METHODS IN MOLECULAR BIOLOGY™

Series Editor
John M. Walker
School of Life Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK

For further volumes:

http://www.springer.com/series/7651
wwwwwwwwwwwww
 Immunogenetics

Methods and Applications in Clinical Practice

Edited by

Frank T. Christiansen
Department of Clinical Immunology, PathWest Laboratory Medicine, Royal Perth Hospital;
School of Pathology and Laboratory Medicine, University of Western Australia, Perth, WA, Australia

Brian D. Tait
National Transplant Services, Australian Red Cross Blood Service, The Royal Melbourne Hospital,
Melbourne, VIC, Australia
Editors
Frank T. Christiansen Brian D. Tait
Department of Clinical Immunology National Transplant Services
PathWest Laboratory Medicine Australian Red Cross Blood Service
Royal Perth Hospital The Royal Melbourne Hospital
Perth, WA, Australia Melbourne, VIC, Australia
School of Pathology and Laboratory
Medicine, University of Western Australia
Perth, WA, Australia

ISSN 1064-3745 ISSN 1940-6029 (electronic)

ISBN 978-1-61779-841-2 ISBN 978-1-61779-842-9 (ebook)
DOI 10.1007/978-1-61779-842-9
Springer New York Heidelberg Dordrecht London

Library of Congress Control Number: 2012938198

© Springer Science+Business Media New York 2012

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this
legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifically for
the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.
Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution
under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not
imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and
regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication, neither
the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be
made. The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

Humana Press is a brand of Springer

Springer is part of Springer Science+Business Media (www.springer.com)
Preface

The ability of mice to distinguish “self” from “nonself” and recognize foreign organisms or
tissues (an immune response) was found to be under genetic control by the seminal experi-
ments of Peter Gorer in 1936 and his description of the H-2 system. This was later extended
to the human with the discovery of the HLA system by Jean Dausset in 1958. Since that
time the study and knowledge of the genes controlling such immune responses in the
human has undergone a rapid expansion. These studies have been facilitated by the devel-
opment of a range of molecular DNA techniques and a remarkable degree of international
collaboration.
The HLA molecules, encoded by a highly polymorphic set of genes located with the
major histocompatibility complex (MHC) on Chromosome 6, are important regulators of
the immune response through mediating antigen presentation and interaction between key
immune mediating cells. They are also the major histocompatibility barriers to transplanta-
tion, which is the clinical paradigm of the self vs. nonself concept. Within the MHC, one of
the best characterized regions of the human genome, are a number of other gene clusters
involved in controlling aspects of the immune response. These encode various accessory
molecules such as TNF, components of the classical complement pathway, and heat shock
proteins. In addition outside the MHC, many other genes encode components which can
act as histocompatibility antigens or are involved in the function of the immune system.
These include genes encoding polymorphic tissue-specific peptides presented within the
context of HLA molecules which have been shown to give rise to a series of minor histo-
compatibility antigens which are barriers to allogeneic transplantation but which can also
be exploited as targets for immunotherapy. Those polymorphic genes encoding the immu-
noglobulin and the T cell receptor molecules, important mediators of humoral and cell-
mediated adaptive immune responses, also play a role in regulation of immune responses.
The immune response is also mediated by a complex of cytokines and cell receptors and
polymorphism in the structural and regulatory elements of the genes have been demon-
strated and shown to have functional correlates. More recently a cluster of genes encoding
receptors on NK cells, including the killer immunoglobulin-like receptors (KIR), which
regulate NK cell activation largely through interaction with ligands encoded within the
MHC, have been described and been shown to be important determinants in transplanta-
tion outcome and in conferring disease susceptibility.
It is now recognized that this diverse range of gene systems involved in the control of
the immune response has been shown to be important in many aspects of clinical practice.
As a result many new molecular and cellular methods have been developed for identifying
these genes and their polymorphisms, and immunogenetic laboratories specializing in these
methods have developed to support transplantation and other clinical programs. This volume
focuses on such methods. The scope is exclusively for human clinical practice and the
emphasis is on those assays which are of established or potential clinical utility and are likely
to be included in the repertoire of tests provided by a routine diagnostic and service

v
vi Preface

laboratory. In the tradition of Methods in Molecular Biology series, the methods provide
details of the materials and equipment required and a step-by-step description of the labo-
ratory method. Details of the critical factors in the performance and interpretation of these
assays and other practical tips are provided through as series of Notes which we trust the
reader will find particularly helpful.
The characterization of the HLA and other genes within the MHC is important to the
routine diagnostic immunogenetics laboratory, particularly in supporting solid organ and
hemopoietic stem cell transplant programs. We have included chapters describing HLA typ-
ing by molecular techniques including sequence-specific oligonucleotide hybridization (SSO)
and sequence-specific priming (SSP) methods, solid phase bead-based assays and by direct
DNA sequencing, and the use of specialist software for HLA allele assignment. An additional
chapter focuses on a number of molecular methods adapted for the typing of specific HLA
alleles which are of diagnostic utility in disease diagnosis and in drug hypersensitivity reac-
tions. In addition to typing of the classical HLA molecules, we have included chapters describ-
ing typing of other MHC genes including the nonclassical HLA molecules HLA-E and
HLA-G, the complement component C4, the MICA genes, and a panel of single nucleotide
polymorphisms (SNPs) throughout the MHC. Methods for detecting the many non-MHC
minor histocompatibility antigens which have been described and for detecting the function-
ally important polymorphisms in a range of cytokine genes are also included.
The detection and characterization of HLA antibodies directed at epitopes found on
donor HLA molecules and the more recently described antibodies directed at MICA
epitopes is known to be important in organ transplantation, and the recent development of
solid phase bead-based assays to detect these antibodies has led to a new understanding of
the role of HLA and MICA antibodies in organ transplant rejection. Chapters describing
the methods for the detection of donor-specific antibodies using the traditional, but still
gold standard, complement dependent cytotoxicity and the more sensitive flow cytometry-
based crossmatches and bead-based antibody detection assays are included. Assays for the
detection of T cell-mediated reactivity against alloantigens either directly or indirectly
through the cross-reactivity of viral-specific T cells—so-called heterologous immunity—are
also described.
The highly polymorphic KIR gene cluster, which encodes a family of NK cell receptors
which are important in controlling NK cell function, has generated considerable interest
recently. Methods detecting and characterizing the polymorphism of this complex gene
system including PCR-SSP and direct sequencing and for detecting NK cell alloreactivity
including a flow cytometry-based assay are described. Alloreactive NK cells have the poten-
tial to be effective therapeutic agents, and a detailed method for the clinical production of
such cells is provided.
In addition to the method chapters we have included a series of overview chapters
highlighting aspects of gene function or various methods available for their study. These
include reviews of the MHC complex, the KIR complex, the human immunoglobulin allo-
types, as well as reviews of the methods for the detection of alloreactive NK cells and the
detection of HLA antibodies by solid phase assays. Because of their expertise and under-
standing of the complexity of the HLA system, many immunogenetics laboratories provide
specialist advice to clinicians in their search for suitable unrelated hemopoietic stem cell
donors. We have therefore included a chapter which provides a detailed and practical guide
to cost-effective strategies for undertaking such searches.
Immunogenetic data is complex and the genes are highly polymorphic and have been
extensively studied. These features have meant that extensive databases of such genes and
 Preface vii

the polymorphisms are now available. Included are chapters on methods for the establish-
ment and management of immunogenetic databases and which describe the use of specialist
HLA, immunoglobulin, and T cell receptor polymorphism databases. In addition two
chapters are included which review the analytical methods available for the study and mea-
surement of human population diversity and the identification, quantitation, and mapping
of disease susceptibility genes respectively. These chapters also provide a number of worked
examples and we trust the readers will find them particularly helpful.
We would like to thank all the contributors to what we believe is an outstanding collec-
tion of manuscripts which we are sure will be widely read and used by the immunogenetics
community. These contributions represent many hours of work and the sharing of detailed
methods and helpful tips gained by many years of experience in the field. We are delighted
that they have been prepared to share this information with the wider scientific community.
We would also like to thank John Walker and Humana Press for inviting us to edit this
volume of the very successful Methods in Molecular Biology series and for his excellent edito-
rial guidance. We trust the final product has rewarded his trust in us.
Finally, we would like to express our special thanks to Natalie Caldwell for her outstand-
ing clerical and editorial assistance. Her enthusiasm and skill has been invaluable as we have
worked through the numerous editorial tasks.

Perth, WA, Australia Frank T. Christiansen

Melbourne, VIC, Australia Brian D. Tait
wwwwwwwwwwwww
Contents

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

1 The Major Histocompatibility Complex: A Paradigm for Studies

of the Human Genome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Richard J.N. Allcock
2 HLA Typing by SSO and SSP Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Heather Dunckley
3 Methods for Diagnostic HLA Typing in Disease Association
and Drug Hypersensitivity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
Michael D. Varney, Alison S.L. Castley, Katri Haimila,
and Päivi Saavalainen
4 HLA Typing Using Bead-Based Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
Daniel Trajanoski and Samantha J. Fidler
5 HLA Typing by Direct DNA Sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
Linda K. Smith
6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio . . . 87
Carla Wirtz and David Sayer
7 Simple Methods for the Detection of HLA-G Variants in Coding
and Non-coding Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Holger Nückel, Erick C. Castelli, Philippe Moreau, Crista Ochsenfarth,
Peter A. Horn, and Vera Rebmann
8 Molecular Typing of HLA-E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
Nina Lauterbach, Christina E.M. Voorter, and Marcel G.J. Tilanus
9 Molecular Analysis of Complement Component C4 Gene Copy Number . . . . 159
Alison S.L. Castley and O. Patricia Martinez
10 Genotyping of Single Nucleotide Polymorphisms by 5¢ Nuclease
Allelic Discrimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Mari Malkki and Effie W. Petersdorf
11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) . . . . . . 183
Yizhou Zou and Peter Stastny
12 Standard Methods for the Management of Immunogenetic Data. . . . . . . . . . . 197
Pierre-Antoine Gourraud, Jill A. Hollenbach, Thomas Barnetche,
Richard M. Single, and Steven J. Mack
13 Analytical Methods for Immunogenetic Population Data. . . . . . . . . . . . . . . . . 215
Steven J. Mack, Pierre-Antoine Gourraud, Richard M. Single,
Glenys Thomson, and Jill A. Hollenbach

ix
x Contents

14 Analytical Methods for Disease Association Studies

with Immunogenetic Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
Jill A. Hollenbach, Steven J. Mack, Glenys Thomson,
and Pierre-Antoine Gourraud
15 Impact of HLA Matching and HLA Antibodies in Organ Transplantation:
A Collaborative Transplant Study View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
Caner Süsal and Gerhard Opelz
16 Screening for Antibodies Against MICA by Luminex Flow Cytometry. . . . . . . 279
Yizhou Zou and Peter Stastny
17 HLA Antibody Detection and Characterization by Solid
Phase Immunoassays: Methods and Pitfalls . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Andrea A. Zachary, Renato M. Vega, Donna P. Lucas, and Mary S. Leffell
18 Detection and Characterisation of Alloreactive T Cells. . . . . . . . . . . . . . . . . . . 309
Mandvi Bharadwaj, Nicole A. Mifsud, and James McCluskey
19 Detection of Allo-HLA Cross-Reactivity by Virus-specific
Memory T-Cell Clones Using Single HLA-Transfected K562 Cells . . . . . . . . . 339
Lloyd J. D’Orsogna, Ellen M.W. van der Meer-Prins, Yvonne M. Zoet,
Dave L. Roelen, Ilias I.N. Doxiadis, and Frans H.J. Claas
20 Separation and Cryopreservation of Lymphocytes from Spleen
and Lymph Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Gabriella Tassone and Samantha J. Fidler
21 Crossmatching by Complement-Dependent Lymphocytotoxicity . . . . . . . . . . 359
Samantha J. Fidler
22 The Lymphocyte Crossmatch by Flow Cytometry
for Kidney Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
Jonathan Downing
23 Overview of the Killer Cell Immunoglobulin-Like Receptor System . . . . . . . . 391
Raja Rajalingam
24 KIR Typing by Non-Sequencing Methods: Polymerase-Chain
Reaction with Sequence-Specific Primers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
David Ordóñez, Manuela Moraru, Natalia Gómez-Lozano,
Elisa Cisneros, and Carlos Vilches
25 Killer Cell Immunoglobulin-Like Receptors (KIR)
Typing by DNA Sequencing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
Lihua Hou, Minghua Chen, Noriko Steiner, Kanthi Kariyawasam,
Jennifer Ng, and Carolyn K. Hurley
26 An Overview of Methods Required to Evaluate Donor NK Cell Alloreactivity
for Haploidentical Haemopoietic Stem Cell Transplantation . . . . . . . . . . . . . . 469
Andrea Velardi
27 The Detection of NK Cell Alloreactivity by Flow
Cytometric CD107a Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
Dianne De Santis, Bree Foley, Campbell S. Witt, and Frank T. Christiansen
28 Clinical Production and Therapeutic Applications of Alloreactive
Natural Killer Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
David H. McKenna, Diane M. Kadidlo, Sarah Cooley, and Jeffrey S. Miller
 Contents xi

29 Minor Histocompatibility Antigen Typing by DNA Sequencing

for Clinical Practice in Hematopoietic Stem-Cell Transplantation . . . . . . . . . . 509
Eric Spierings and Els Goulmy
30 Donor Registries and Search Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
Carolyn K. Hurley, Machteld Oudshoorn, and Michelle Setterholm
31 Cytokine Gene Polymorphisms: Methods of Detection and Biological
Significance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549
Gurvinder Kaur and Narinder Mehra
32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin (IG)
and T Cell Receptor (TR) V-(D)-J Repertoires,
Polymorphisms, and IG Mutations: IMGT/V-QUEST
and IMGT/HighV-QUEST for NGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
Eltaf Alamyar, Patrice Duroux, Marie-Paule Lefranc,
and Véronique Giudicelli
33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis of IG,
TR, MH, IgSF, and MhSF Domain Amino Acid Polymorphism . . . . . . . . . . . 605
François Ehrenmann and Marie-Paule Lefranc
34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization:
A Remarkable Demonstration of Polymorphism . . . . . . . . . . . . . . . . . . . . . . . 635
Marie-Paule Lefranc and Gérard Lefranc

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681
wwwwwwwwwwwww
Contributors

ELTAF ALAMYAR • IMGT ®, the International ImMunoGeneTics Information System®,

Université Montpellier 2, Laboratoire d’ImmunoGénétique Moléculaire (LIGM),
Institut de Génétique Humaine (IGH), UPR CNRS 1142, Montpellier, France
RICHARD J.N. ALLCOCK • Lotterywest State Biomedical Facility Genomics,
School of Pathology and Laboratory Medicine, University of Western Australia,
Nedlands; Department of Clinical Immunology, PathWest Laboratory Medicine,
Royal Perth Hospital, Perth, WA, Australia
THOMAS BARNETCHE • Rheumatology Service, Bordeaux University Hospital, Bordeaux,
France
MANDVI BHARADWAJ • Department of Microbiology and Immunology,
University of Melbourne, Parkville, VIC, Australia
ERICK C. CASTELLI • Molecular Genetics and Cytogenetics Laboratory, and Department of
General Biology, Institute of Biological Sciences, Federal University of Goiás, Goiás, Brazil
ALISON S.L. CASTLEY • Department of Clinical Immunology, PathWest Laboratory
Medicine, Royal Perth Hospital, Perth, WA, Australia
MINGHUA CHEN • Department of Pediatrics, CW Bill Young Marrow Donor Recruitment
and Research Program, Georgetown University Medical Centre, Washington, DC, USA
FRANK T. CHRISTIANSEN • Department of Clinical Immunology, PathWest Laboratory
Medicine, Royal Perth and Fremantle Hospitals, Perth, WA, Australia;
School of Pathology and Laboratory Medicine (PaLM), University of
Western Australia, Perth, WA, Australia
ELISA CISNEROS • Inmunogenética—HLA, Hospital Universitario Puerta de Hierro,
Majadahonda, Spain
FRANS H.J. CLAAS • Department of Immunohematology and Blood Transfusion,
Leiden University Medical Centre, Leiden, The Netherlands
SARAH COOLEY • Department of Medicine, University of Minnesota Cancer Centre,
Minneapolis, MN, USA
DIANNE DE SANTIS • Department of Clinical Immunology, PathWest Laboratory Medicine,
Royal Perth Hospital, Perth, WA, Australia
LLOYD J. D’ORSOGNA • Department of Immunohematology and Blood Transfusion,
Leiden University Medical Centre, Leiden, The Netherlands; PathWest Laboratory
Medicine, Perth, WA, Australia
JONATHAN DOWNING • Tissue Typing Laboratory, New Zealand Blood Service, Auckland,
New Zealand
ILIAS I.N. DOXIADIS • Department of Immunohematology and Blood Transfusion, Leiden
University Medical Centre, Leiden, The Netherlands
HEATHER DUNCKLEY • Australian Red Cross Blood Service, Sydney, NSW, Australia
PATRICE DUROUX • IMGT ®, The International ImMunoGeneTics Information System®,
Université Montpellier 2, Laboratoire d’ImmunoGénétique Moléculaire (LIGM), Institut
de Génétique Humaine (IGH), UPR CNRS 1142, Montpellier, France

xiii
xiv Contributors

FRANÇOIS EHRENMANN • IMGT ®, The International ImMunoGeneTics Information System®,

Université Montpellier 2, Laboratoire d’ImmunoGénétique Moléculaire (LIGM),
Institut de Génétique Humaine (IGH), UPR CNRS 1142, Montpellier, France
SAMANTHA J. FIDLER • Department of Clinical Immunology, PathWest Laboratory
Medicine, Royal Perth Hospital, Perth, WA, Australia; School of Pathology
and Laboratory Medicine (PaLM), University of Western Australia,
Nedlands, WA, Australia
BREE FOLEY • Division of Hematology, Oncology and Transplantation,
University of Minnesota Cancer Centre, Minneapolis, MN, USA
VÉRONIQUE GIUDICELLI • IMGT ®, The International ImMunoGeneTics Information
System®, Université Montpellier 2, Laboratoire d’ImmunoGénétique Moléculaire
(LIGM), Institut de Génétique Humaine (IGH), UPR CNRS 1142, Montpellier, France
NATALIA GÓMEZ-LOZANO • Inmunogenética—HLA, Hospital Universitario
Puerta de Hierro, Majadahonda, Spain
ELS GOULMY • Department of Immunohematology and Blood Transfusion,
Leiden University Medical Center, Leiden, The Netherlands
PIERRE-ANTOINE GOURRAUD • Department of Neurology, University of California,
San Francisco, CA, USA
KATRI HAIMILA • Clinical Laboratory, Finnish Red Cross Blood Service, Helsinki, Finland
JILL A. HOLLENBACH • Center for Genetics, Children’s Hospital and Research Center
Oakland, Oakland, CA, USA
PETER A. HORN • Institute for Transfusion Medicine, University Hospital Essen,
Essen, Germany
LIHUA HOU • Department of Pediatrics, CW Bill Young Marrow Donor Recruitment
and Research Program, Georgetown University Medical Centre, Washington, DC, USA
CAROLYN K. HURLEY • Department of Oncology, CW Bill Young Marrow Donor
Recruitment and Research Program, Georgetown University Medical Centre,
Washington, DC, USA
DIANE M. KADIDLO • Molecular and Cellular Therapeutics, University of Minnesota,
St. Paul, MN, USA
KANTHI KARIYAWASAM • Department of Oncology, CW Bill Young Marrow Donor
Recruitment and Research Program, Georgetown University Medical Centre,
Washington, DC, USA
GURVINDER KAUR • Department of Transplant Immunology and Immunogenetics,
All India Institute of Medical Sciences, New Delhi, India
NINA LAUTERBACH • Transplantation Immunology, Tissue Typing Laboratory, Maastricht
University Medical Center, Maastricht, The Netherlands
MARY S. LEFFELL • Immunogenetics Laboratory, Department of Medicine,
Johns Hopkins University School of Medicine, Baltimore, MD, USA
GÉRARD LEFRANC • IMGT ®, The International ImMunoGeneTics Information System®,
Université Montpellier 2, Laboratoire d’ImmunoGénétique Moléculaire (LIGM),
Institut de Génétique Humaine (IGH), UPR CNRS 1142, Montpellier, France
MARIE-PAULE LEFRANC • IMGT ®, The International ImMunoGeneTics Information
System®, Université Montpellier 2, Laboratoire d’ImmunoGénétique Moléculaire
(LIGM), Institut de Génétique Humaine (IGH), UPR CNRS1142, Montpellier, France
DONNA P. LUCAS • Immunogenetics Laboratory, Department of Medicine, Johns Hopkins
University School of Medicine, Baltimore, MD, USA
 Contributors xv

STEVEN J. MACK • Center for Genetics, Children’s Hospital and Research Center Oakland,
Oakland, CA, USA
MARI MALKKI • Fred Hutchinson Cancer Research Centre, Seattle, WA, USA
O. PATRICIA MARTINEZ • Department of Clinical Immunology, PathWest Laboratory
Medicine, Royal Perth and Fremantle Hospitals, Perth, WA, Australia;
School of Pathology and Laboratory Medicine (PaLM), University of
Western Australia, Nedlands, WA, Australia
JAMES MCCLUSKEY • Department of Microbiology and Immunology, University
of Melbourne, Parkville, VIC, Australia
DAVID H. MCKENNA • Department of Laboratory Medicine and Pathology,
Cancer Centre and Molecular and Cellular Therapeutics, University of Minnesota,
Minneapolis, MN, USA
NARINDER MEHRA • Department of Transplant Immunology and Immunogenetics,
All India Institute of Medical Sciences, New Delhi, India
NICOLE A. MIFSUD • Department of Medicine, Monash University, Alfred Hospital,
Melbourne, VIC, Australia
JEFFREY S. MILLER • Department of Medicine, University of Minnesota Cancer Centre,
Minneapolis, MN, USA
MANUELA MORARU • Inmunogenética—HLA, Hospital Universitario Puerta de Hierro,
Majadahonda, Spain
PHILIPPE MOREAU • Commissariat à l’Energie Atomique, Service de Recherches en Hémato-
Immunologie, I2BM, Institut Universitaire d’Hématologie, Hôpital Saint-Louis,
Paris, France
JENNIFER NG • Department of Pediatrics, CW Bill Young Marrow Donor Recruitment
and Research Program, Georgetown University Medical Centre, Washington, DC, USA
HOLGER NÜCKEL • Department of Haematology and Institute of Pharmacogenetics,
University Hospital Essen, Essen, Germany
CRISTA OCHSENFARTH • Institute of Pharmacogenetics, University Hospital Essen,
Essen, Germany
GERHARD OPELZ • Department of Transplantation Immunology, Institute of Immunology,
University of Heidelberg, Heidelberg, Germany
DAVID ORDÓÑEZ • Inmunogenética—HLA, Hospital Universitario Puerta de Hierro,
Majadahonda, Spain
MACHTELD OUDSHOORN • Department of Immunohematology and Blood Transfusion,
Leiden University Medical Centre, Leiden, The Netherlands; Europdonor Foundation,
Leiden, The Netherlands
EFFIE W. PETERSDORF • Fred Hutchinson Cancer Research Centre, Seattle, WA, USA
RAJA RAJALINGAM • UCLA Immunogenetics Center, Department of Pathology
and Laboratory Medicine, David Geffen School of Medicine at UCLA,
University of California, Los Angeles, CA, USA
VERA REBMANN • Institute for Transfusion Medicine, University Hospital Essen, Essen,
Germany
DAVE L. ROELEN • Department of Immunohematology and Blood Transfusion,
Leiden University Medical Centre, Leiden, The Netherlands
PÄIVI SAAVALAINEN • Research Program for Molecular Medicine, Department of Medical
Genetics, Haartman Institute, Biomedicum Helsinki, University of Helsinki,
Helsinki, Finland
DAVID SAYER • Conexio Pty Ltd, Fremantle, WA, Australia
xvi Contributors

MICHELLE SETTERHOLM • National Marrow Donor Program, Minneapolis, MN, USA

RICHARD M. SINGLE • Department of Mathematics and Statistics, University of Vermont,
Burlington, VT, USA
LINDA K. SMITH • Department of Clinical Immunology, PathWest Laboratory Medicine,
Royal Perth Hospital, Perth, WA, Australia
ERIC SPIERINGS • Department of Immunology, University Medical Centre Utrecht, Utrecht,
The Netherlands
PETER STASTNY • Division of Transplant Immunology, Department of Internal Medicine,
UT Southwestern Medical Centre, Dallas, TX, USA
NORIKO STEINER • Department of Oncology, CW Bill Young Marrow Donor Recruitment
and Research Program, Georgetown University Medical Centre, Washington, DC, USA
CANER SÜSAL • Department of Transplantation Immunology, Institute of Immunology,
University of Heidelberg, Heidelberg, Germany
BRIAN D. TAIT • National Transplant Services, Australian Red Cross Blood Service, The
Royal Melbourne Hospital, Melbourne, VIC, Australia
GABRIELLA TASSONE • Department of Clinical Immunology, PathWest Laboratory Medicine,
Royal Perth Hospital, Perth, WA, Australia
GLENYS THOMSON • Department of Integrative Biology, University of California, Berkeley,
CA, USA
MARCEL G.J. TILANUS • Transplantation Immunology, Tissue Typing Laboratory,
Maastricht University Medical Center, Maastricht, The Netherlands
DANIEL TRAJANOSKI • Department of Clinical Immunology, PathWest Laboratory Medicine,
Royal Perth Hospital, Perth, WA, Australia
ELLEN M.W. VAN DER MEER-PRINS • Department of Immunohematology and Blood
Transfusion, Leiden University Medical Centre, Leiden, The Netherlands
MICHAEL D. VARNEY • Victorian Transplantation and Immunogenetics Service,
Australian Red Cross Blood Service, Melbourne, VIC, Australia
RENATO M. VEGA • Immunogenetics Laboratory, Department of Medicine,
Johns Hopkins University School of Medicine, Baltimore, MD, USA
ANDREA VELARDI • Division of Haematology and Clinical Immunology, Department
of Clinical and Experimental Medicine, University of Perugia, Ospedale Santa Maria
della Misericordia, Perugia, Italy
CARLOS VILCHES • Inmunogenética—HLA, Hospital Universitario Puerta de Hierro,
Majadahonda, Spain
CHRISTINA E.M. VOORTER • Transplantation Immunology, Tissue Typing Laboratory,
Maastricht University Medical Center, Maastricht, The Netherlands
CARLA WIRTZ • Technical Support and Training, Conexio Genomics, Longmont, CO, USA
CAMPBELL S. WITT • Department of Clinical Immunology, PathWest Laboratory Medicine,
Royal Perth Hospital, Perth, WA, Australia; School of Pathology and Laboratory
Medicine (PaLM), University of Western Australia, Nedlands, WA, Australia
ANDREA A. ZACHARY • Immunogenetics Laboratory, Department of Medicine,
Johns Hopkins University School of Medicine, Baltimore, MD, USA
YVONNE M. ZOET • Department of Immunohematology and Blood Transfusion,
Leiden University Medical Centre, Leiden, The Netherlands
YIZHOU ZOU • Division of Transplant Immunology, Department of Internal Medicine,
UT Southwestern Medical Centre, Dallas, TX, USA
 Chapter 1

The Major Histocompatibility Complex: A Paradigm

for Studies of the Human Genome
Richard J.N. Allcock

Abstract
The major histocompatibility complex (MHC) on chromosome 6 is one of the most intensively studied
regions of the human genome and has many features which make it unique. It is the source of much
research interest because of its role in autoimmune and infectious disease susceptibility, and of diagnostic
interest because of its role in transplantation and rejection. It is the most gene-dense and SNP-rich region
of the genome, with large number of complex haplotypes and other features which must be taken into
account when analysing the MHC in the laboratory. This article provides a brief overview of the MHC
highlighting some of the issues that must be considered when developing new methods and assays.

Key words: Major histocompatibility complex, Genetic polymorphism, HLA, Immune genes

1. Introduction

The major histocompatibility complex (MHC) is a 4 Mb region

of chromosome 6 which is one of the most intensively studied
regions of the human genome. It has been the subject of count-
less reviews over more than 20 years and was most recently
reviewed in 2011 (1, 2).
The MHC has attracted the interest of researchers in many
different fields, both medical and scientific, and was one of the first
large regions of the human genome to be sequenced in its entirety
in the 1990s (3). Study of this gene complex has led to a number of
observations on genome structure and biology which have proven
to be applicable to much of the rest of genome. The intense level of
study of this region has meant that many genomic laboratory
techniques have been pioneered or refined examining the MHC.

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_1, © Springer Science+Business Media New York 2012

1
2 R.J.N. Allcock

Given their important roles in disease and transplantation, the

HLA genes (HLA-A, -B, -C, -DRA, -DRB1, -DRB3, -DRB4,
DRB5, -DQA1, -DQB1, and -DPB1) have been the focus of many
clinical and diagnostic studies. Many hundreds of thousands of
individuals have been genotyped at high resolution at the six major
HLA loci. The genotyping data from these loci are presented as
numbered alleles, but it is important to note that an individual
HLA allele can be made up of more than 50–60 individual single
nucleotide polymorphisms (SNPs). Hence, individual alleles are in
fact complex multi-SNP haplotypes, collectively named through a
detailed and extensive nomenclature system (4). There are few
comparably complex variable gene loci elsewhere in the genome
and the methods for dealing with such a level of allelic variation
have been to be developed by researchers in the field. Some of
these issues are addressed elsewhere in this volume. As knowledge
has advanced and new important genes within the MHC region
have been identified, the methodological approaches and nomen-
clature system have been refined and revised to accommodate this
new information. Investigators studying HLA and other genes in
the MHC have had to become expert geneticists, genomicists, bio-
informaticians, and statisticians in order to synthesise and integrate
this knowledge. Because of complexity of the MHC and the need
to develop specialised methods and nomenclature, the biology and
genetics of the MHC has been perceived as an arcane niche.
However, it is now increasingly apparent that many features once
thought unique to the MHC apply more broadly to many other
regions of the genome and that the MHC is rich and unique area
for the study of general features of the human genome.
The MHC is known for a number of features including, but
not limited to:
1. A very high gene density (1 locus every ~15 kb over its entire
length), with a high proportion (20–30%) of the genes hav-
ing a direct or indirect role in innate or adaptive immunity
(for an exhaustive gene list and map, see ref. (2) and also see
http://vega.sanger.ac.uk/Homo_sapiens/Location/
Chromosome?r=6:1-171055067 for the current annotation
of the region).
2. Very short intergenic lengths (<1 kb).
3. High numbers of SNPs including point variations and indels
(1 every 200–300 base pairs (bp)), with many found in coding
regions affecting amino acids (see refs. (5, 6) for descriptions
and maps).
4. Whole-MHC haplotypes and smaller haplotype “blocks” of
approximately 40–80 kb.
5. Variations in gene-content between haplotypes (e.g. the
complement genes and the HLA-DRB region pseudogenes).
 1 The Major Histocompatibility Complex… 3

6. Low levels of recombination and, as a result, en bloc inheritance

of highly conserved MHC haplotypes.
Between them this set of features combine to form a region
whose properties are unique and therefore methods for their study
must be adapted to take account of these features. Many diagnostic
laboratory methods or genotyping techniques, which have been
developed over time, have required revision due to, for example,
unrecognised sequence variations which have led to allelic dropout
during the polymerase chain reaction (PCR) and an inability to
detect particular allele combinations. The modern sequence-based
diagnostic tests for the HLA genes are able to identify the presence
of new unrecognised alleles as well as being able to detect more
than 1,500 alleles and 800 alleles of the class I loci and class II loci,
respectively (7). Such progress reflects a long history of research
and diagnostics effectively “crossing over” in MHC.

2. Physical
Structure and
Gene Content
The MHC lies on the short arm of chromosome 6, with the class
I region at the telomeric end and the class II region at the centro-
meric end (reviewed in ref. (2)). Between them lies the class III, or
central MHC, region of approximately 1 Mb. The class II region is
relatively homogeneous, containing genes predominantly encod-
ing HLA class II proteins (alpha and beta subunits), as well as a
small number of genes involved in processing antigens for class I
proteins (the TAP and PSMB genes). In contrast, the class I region
encodes the classical (HLA-A, -B, -C) and non-classical (HLA-E,
-F, -G) proteins, but is littered with the remnants of many evolu-
tionary relics in the form of pseudogenes. In addition, the class I
region encodes a significant number of Tripartite-motif (TRIM)
family and butyrophilin (BTN) family genes, which may well have
immune-related roles, and a large number of genes with functions
outside the immune system.
The gene families found in the MHC are likely the result of
multiple duplication events and individual pairs of genes can be
almost identical (e.g. HLA-B and HLA-C), or highly similar to
one another (e.g. the MIC family for MHC-class I-like genes—see
ref. (8)) or much more divergent, sharing little more than broad
intron–exon structures and having 20–25% sequence identity (e.g.
TNF, LTA, and LTB—see ref. (9)).
The class I and II regions of the MHC are so named primar-
ily because of the HLA genes encoded within them. Between
them lies the class III region which has no such defining feature
and contains a large number of genes of variable function, both
4 R.J.N. Allcock

within and outside the immune system. Some areas have been
extensively studied (e.g. the complement genes, the TNF cluster
and MICA genes), whilst others have been less well studied. The
1 Mb class III region has the highest gene density of the three
major MHC regions.

3. Genetic
Variation and
Haplotypes
The MHC has long been known to harbour large numbers of
SNPs, particularly in the MHC class I and II genes. However, it is
now recognised that almost all MHC genes are more polymorphic
than most genes elsewhere in the genome (2). These variations
take the form of point mutations, small indels, large indels, and
microsatellites. Within the HLA genes, particularly in the exons
encoding the hypervariable regions, there may be 50–100 SNPs
located within regions of 100–200 bp, many of which cause amino
acid changes (5). Primer-design for PCR is fraught with problems
in these regions and therefore many investigators site primers in
introns or further afield where there are lower levels of variation.
This problem is not limited to the HLA genes however, as variants
are found throughout the MHC, sometimes in very close proxim-
ity to one another.
Initially, only a single MHC sequence, assembled from a vari-
ety of sources and individuals, was available (3) and the identification
of sequence variations was based around “sample” sequencing of
small regions of interest to particular investigators (e.g. see ref.
(10)). Subsequently, the “MHC haplotype project” used standard
mapping and sequencing techniques to resequence the entire
MHC from eight hemi- or homozygous cells lines carrying well-
defined and important haplotypes (11). The vast number of cata-
logued MHC variants generated by this project has allowed a
greater insight into the existence of haplotypes blocks, both within
the MHC, but also elsewhere (see refs. (5, 6)). The level of knowl-
edge of block boundaries and number of the haplotypes within a
given block is far greater in the MHC than elsewhere in the
genome, largely as a consequence of the major interest in the
region and the large number of variants identified. Whilst particu-
lar whole MHC haplotypes based on HLA class I and II combina-
tions have been recognised for more than 20 years (12), it is now
recognised that much smaller haplotype blocks of some 20–60 kb
in length are present. The well-studied TNF block contains more
than 100 SNPs but it can be defined by 20–30 SNPs (13), and the
SNPs present in each haplotype appear to be found in many popu-
lations worldwide, and in similar combinations. Whilst populations
differ in the frequency of individual haplotypes, the fundamental
observation of a common set of haplotypes remains (14).
 1 The Major Histocompatibility Complex… 5

Overall genetic variation in the MHC, as elsewhere in the

genome, is most frequent in intergenic and intronic regions. There
are approximately 500–3,000 intronic variants, and a further
500–4,000 intergenic variants in any given haploptype (5). Coding
region variants are less frequent, approximately 300–500 in a
given haplotype, distributed in HLA and non-HLA genes alike.
However, the HLA-genes contain approximately two thirds of the
non-synonymous coding region variants, consistent with their
functional importance (5). When the sequenced MHC haplotypes
are considered as a whole, there are more than 70,000 identified
and catalogued SNPs in the human MHC.

4. Assay Design
and Approaches
To date, diagnostic assays have focussed on the amplification of
small fragments and either hybridisation to labelled probes or
sequencing using standard methods, such as fluorescent dideoxy
chain terminators (Sanger sequencing). Diagnostic assays for other
parts of the MHC have not been developed to the same extent.
Assays for the detection of MICA polymorphisms have been devel-
oped and are described in this volume. Studies of the complement
region, of interest to some groups, are complicated by a duplicated
gene system in which there can be 0–4 genes present on each hap-
lotype, all of which may be polymorphic (15). The recent significant
interest in copy number variation has led to the development of
methods to examine them effectively, and methods incorporating
both CNV and SNP analysis are emerging. A similar situation exists
at the telomeric end of the HLA class II region, where there is a
variable gene content in the HLA-DRB pseudogenes, which once
again can contain multiple alleles (16).
Arguably, the most complex bioinformatic analysis currently in
use in the diagnostic world is the assignment of a pair of HLA
alleles to a given individual. Currently, based on Sanger sequencing
of the peptide binding-encoding exons, bulk heterozygous
sequence is compared to a database of all combinations of known
HLA alleles. Sophisticated analysis programmes have been devel-
oped for allele assignment and resolution of ambiguities. More
recently, next-generation sequencing technologies have been
applied to the study of MHC genes. One advantage of next-gener-
ation sequencing is the ability to sequence individual DNA strands,
thus unambiguously defining the phase of identified SNPs (17).
How these high-throughput methods will be applied in diagnostic
laboratories is not yet clear, as cost and turn-around times remain
significant. However, it is already evident approaches to enrich
(capture) the MHC and sequence it in its entirety in a matter of
6 R.J.N. Allcock

days are technically feasible. Such approaches will result in a

significant additional increase in our knowledge of sequences origi-
nating from the MHC and in turn pose new challenges for analysis
and data handling.

5. The Future

Many autoimmune diseases have been associated with MHC

alleles and haplotypes and the vast variety of functions encoded by
MHC genes has led to a great deal of speculation as to the path-
ways that are implicated and whether immune system or non-
immune system genes are responsible (recently reviewed in ref.
(1)). To date no definitive answers have emerged and it remains
possible than almost any gene within the MHC could be involved
in autoimmune disease susceptibility. Greater understanding of
MHC gene functions, genetic variability and haplotype structure,
together with other information, such as methylation patterns
(18) and how these epigenetic modifications affect expression,
will be required to elucidate the mechanisms involved. In addi-
tion, a greater examination of variation in the HLA genes may
lead to an improved understanding of the effects of the MHC on
acceptance and rejection of allografts and ultimately further
improvement in transplantation outcome.

Acknowledgements

I am grateful to Winthrop Professor Frank Christiansen for his

critical comments and suggestions during the preparation of this
manuscript.

References

1. Trowsdale J (2011) The MHC, disease and 5. Horton R, Gibson R, Coggill P et al (2008)
selection. Immunol Lett. doi:10.1016/j.imlet. Variation analysis and gene annotation of eight
2011.01.002 MHC haplotypes: the MHC Haplotype Project.
2. Shiina T, Hosomichi K, Inoko H et al (2009) Immunogenetics 60:1–18
The HLA genomic loci map: expression, inter- 6. de Bakker PI, McVean G, Sabeti PC et al (2006)
action, diversity and disease. J Hum Genet 54: A high-resolution HLA and SNP haplotype
15–39 map for disease association studies in the
3. The MHC Sequencing Consortium (1999) extended human MHC. Nat Genet
Complete structure and gene map of a major 38:1166–1172
histocompatibility complex (MHC). Nature 7. Smith LK, Sayer DC, Whidborne RS et al
401:921–923 (2007) Sequencing-based typing identifies
4. Marsh SG, Albert ED, Bodmer WF et al (2010) novel alleles due to single nucleotide polymor-
Nomenclature for factors of the HLA system. phisms in ‘conserved’ regions. Tissue Antigens
Tissue Antigens 75:291–455 69(suppl 1):56–57
 1 The Major Histocompatibility Complex… 7

8. Choy MK, Phipps ME (2010) MICA polymor- Caucasian population: implications for haplotype
phism: biology and importance in immunity tagging. Hum Mutat 24:517–525
and disease. Trends Mol Med 16:97–106 14. Valente FP, Tan CR, Temple SE et al (2009) The
9. Posch PE, Cruz I, Bradshaw D et al (2003) evolution and diversity of TNF block haplotypes
Novel polymorphisms and the definition of in European, Asian and Australian Aboriginal
promoter ‘alleles’ of the tumour necrosis factor populations. Genes Immun 10:607–615
and lymphotoxin alpha loci: inclusion in HLA 15. Yang Z, Yu CY (2000) Organisations and gene
haplotypes. Genes Immun 4:547–558 duplications of the human and mouse MHC
10. Smith WP, Vu Q, Li SS et al (2006) Toward complement gene clusters. J Exp Med 191:
understanding MHC disease associations: partial 2183–2196
resequencing of 46 distinct HLA haplotypes. 16. Traherne JA, Horton R, Roberts AN et al
Genomics 87:561–571 (2006) Genetic analysis of completely sequenced
11. Stewart CA, Horton R, Allcock RJ et al (2004) disease-associated MHC haplotypes identifies
Complete MHC haplotype sequencing for shuffling of segments in recent human history.
common disease gene mapping. Genome Res PLoS Genet 2:e9
14:1176–1187 17. Erlich RL, Jia X, Anderson S et al (2011) Next-
12. Degli-Esposti MA, Leaver AL, Christiansen FT generation sequencing for HLA-typing of class
et al (1992) Ancestral haplotypes: conserved I loci. BMC Genomics 12:42–55
population MHC haplotypes. Hum Immunol 18. Tomazou EM, Rakyan VK, Lefebvre G et al
34:242–252 (2008) Generation of a genomic tiling array of
13. Allcock RJN, Windsor L, Gut IG et al (2004) the human major histocompatibility complex
High-density SNP genotyping defines 17 (MHC) and its application for DNA methylation
distinct haplotypes of the TNF block in the analysis. BMC Med Genomics 1:19
 Chapter 2

HLA Typing by SSO and SSP Methods

Heather Dunckley

Abstract
Human leukocyte antigen (HLA) typing, utilising the sequence-specific oligonucleotide (SSO) and
sequence-specific primer (SSP) technologies, has been in routine use in many tissue typing laboratories
worldwide for more than 20 years since the development of the polymerase chain reaction. Both methods
are very useful for clinical and research purposes and can provide generic (low resolution) to allelic (high
resolution) typing results. This chapter provides an overview of the SSO and SSP methods in relation to
HLA typing.

Key words: HLA typing, PCR-SSP, PCR-SSO, Oligotyping, Reverse-SSO typing

1. Introduction

1.1. Sequence-Specific HLA-SSP typing of the HLA-DR genes was initially developed by
Primer Assay Olerup and Zetterquist (1). Further developments of the method
quickly followed which enabled sequence-specific primer (SSP)
typing of other human leukocyte antigen (HLA) class I and II genes
(2–5). In 1995 Bunce et al. developed a comprehensive SSP typing
system for the HLA-A, -B, -C, -DRB1, -DRB3, -DRB4, -DRB5,
and -DQB1 alleles where all the polymerase chain reactions (PCRs)
were designed to work with identical thermocycler conditions (6).
SSP typing involves the use of primers complementary to particular
HLA allele sequences (e.g. HLA*02:03) or a group of similar alleles
(e.g. all the A*02 alleles). Primers are designed so that the polymor-
phism to be detected is at the 3¢ end of the primer. If there is a
mismatch with the HLA allele at the 3¢ end, the primer will not be
extended by Taq DNA polymerase. In the case of HLA-A, -B, and
-C typing, most primer sets detect polymorphisms in exons 2 and 3
of the class I genes as these regions cover most of the known

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_2, © Springer Science+Business Media New York 2012

9
10 H. Dunckley

Fig. 1. Photo of a sequence-specific primer (SSP) gel. The photo shows amplification of the
control gene and human leukocyte antigen (HLA) genes. The control amplicon is present
in all lanes and the HLA amplicon is present in lanes 3, 5, and 6. The control amplicon is
weaker in lanes 3, 5, and 6.

polymorphisms. For the HLA-DR, -DQ, and -DP genes, most

primer sets cover exon 2 polymorphisms where the majority of the
known sequence variation occurs. The SSP method can be used for
either low resolution (generic) typing which resolves alleles to at
least the antigenic level (the level detected by serological typing
methods), or for high resolution (allele level) typing.
For HLA-SSP typing, primer sets are made up in advance and
dispensed into microtitre trays or PCR tubes and stored frozen. If
freeze dried, the primer sets can be stored at 2–8°C. When an indi-
vidual requires HLA genotyping, DNA is extracted, usually from a
blood sample, and added to a pre-made buffer containing dNTPs
and Taq DNA polymerase and the mixture is then aliquotted into
microtitre trays with the primer sets which are then subjected to
the PCR process, (the same PCR conditions are used for all the
primer sets). Amplification is detected by electrophoresing the
PCR products through an agarose gel containing a dye, such as
ethidium bromide, which binds to DNA. A photo taken while the
gel is exposed to ultra violet light reveals the presence (or absence)
of amplified HLA alleles. Positive amplification is recorded, and
depending on which primer mixes have amplified, the HLA type of
the sample can be assigned (see Fig. 1; Table 1).
In the initial years of SSP method development, there was a
requirement by laboratories to purchase primers and prepare their
own SSP typing trays and to maintain an up to date method as new
HLA alleles were discovered. But the technology was quickly
developed commercially and as a result it has been possible for
some time to purchase SSP typing kits to type the HLA class I and
II loci at either low or high resolution. This requires the laboratory
only to extract DNA from clinical samples, add DNA and Taq
DNA polymerase to the commercially pre-made buffer and primer
sets, perform the PCR and run gels to observe amplification. With
the ever increasing number of HLA alleles, result assignment
requires the use of customised software and commercial HLA-SSP
kit suppliers also have software to assist the result interpretation
 2 HLA Typing by SSO and SSP Methods 11

Table 1
SSP result interpretation and HLA typing from the gel in Fig. 1

Well number

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Control amplicon + + + + + + + + + + + + + +
HLA amplicon − − + − + + − − − − − − − −
DRB1* allele of HLA 01 03 04 07 08 04, 07 09 10 11 12 13 14 15 16
amplicon or
09
“+” Means the amplicon is present; “−” means the amplicon is absent. Based on the positive reactions, the HLA type of
the sample is DRB1*04, 08

and HLA genotyping assignment. The SSP method has the major
advantage of being a very fast method of HLA genotyping, the
process (including DNA extraction) being completed within 2–3 h.
The technique therefore is well suited to the typing of urgent cases,
such as patients requiring HLA-matched platelet transfusions, or
genotyping of donors for organ transplantation. However, because
of the number of PCRs required for each sample, the SSP method
is not suited to the typing of large sample numbers, even with the
use of multi-dispensing pipettes or robotics as there is also the con-
sideration of the number of PCR machines required. Commercial
kits are available which can perform low resolution HLA-A, -B,
-DRB1 typing with 96 PCRs/sample. With the continually increas-
ing large number of HLA alleles, it is becoming more difficult to
resolve the major allele groups with a one-step SSP process.
Additional primer mixes can be added to include newly found
alleles, but there is a limit to the practicality of adding more and
more primer mixes to resolve ambiguities.

1.2. Sequence-Specific HLA-SSO typing was one of the first PCR-based HLA typing meth-
Oligonucleotide Assay ods to be developed, initially for the HLA class II loci and then for
the HLA class I genes (7–12). The technique involves extraction of
DNA followed by amplification of a particular HLA gene locus, for
example HLA-A or HLA-B or HLA-DRB1. Primers are selected to
amplify in one PCR tube all known alleles of the particular HLA
locus which requires genotyping. Generally, for HLA class I typing
exons 2 and 3 are co-amplified as one PCR product while for HLA
class II typing only exon 2 is amplified as these are the gene regions
with the majority of the known sequence variation. The amplified
DNA is blotted onto a nylon membrane which is positively charged
permitting the negatively charged DNA to bind strongly. Many sam-
ples can be blotted onto the same membrane and the membrane size
can be increased as sample numbers increase. Multiple membranes
with the amplified DNA samples are prepared, one for each oligo-
nucleotide probe used in the genotyping method.
12 H. Dunckley

The membrane with the bound DNA is placed in a container

and mixed with a hybridisation solution containing an oligonucle-
otide labelled with a marker such as biotin and which has a DNA
sequence that is specific for a particular HLA allele or group of
alleles. When the amplified DNA sequence is complementary to
the sequence of the oligonucleotide, hybridisation occurs. Non-
hybridised oligonucleotides remain unbound in the solution and
are easily removed in a series of stringency washing steps as are any
oligonucleotides which are non-specifically bound to the amplified
DNA on the nylon membrane. Oligonucleotides are designed so
the polymorphism to be detected is in the middle of the oligonu-
cleotide sequence which facilitates the strongest possible binding
between the amplified DNA and the oligonucleotide. If there is
sequence variation between the amplicon and the end of the oligo-
nucleotide, then only one part of the oligonucleotide binds to the
amplified DNA which increases the likelihood of washing off dur-
ing the stringency washing step. A chemiluminescent or colour-
metric reaction can be used to reveal the presence of bound
oligonucleotide. The reaction of each sample with each oligonu-
cleotide is recorded and the HLA type of the sample can be assigned
based on the pattern of positive reactions.
The sequence-specific oligonucleotide (SSO) method is very
well suited to the typing of large sample numbers, e.g. a routine
typing lab using one to two 96-well microtitre trays can easily test
88–184 samples in one assay (with seven positive control samples
of known HLA type and a negative control). New oligonucle-
otides to cover new alleles or resolve ambiguities can be added
which only requires an additional membrane with the bound
amplified DNA to be prepared. However, with the very high
number of possible heterozygous HLA allele combinations, the
SSO method struggles to resolve all ambiguities as the method
does not distinguish between cis and trans polymorphisms. SSP
typing on the other hand only detects alleles which have both the
forward and reverse primer sequences on the same HLA allele
thus linking two polymorphic regions to the same allele which
assists in resolving some ambiguities. SSO typing takes approxi-
mately 2 days from the PCR stage to result assignment, so is not
a rapid typing method, but as 88–184 samples can easily be typed
in this time for one gene locus by one staff member it is a cost-
effective method.
Because the primers used in SSO typing are designed to amplify
all alleles for a particular locus they are selected from a non-variable
part of the gene. Provided a new allele does not have a polymor-
phism in this region it will be amplified so the SSO method is
generally able to detect new HLA alleles better than the SSP
method. If a new allele has a polymorphism in one of the regions
covered by an SSO, it should be detected and appear as a new pat-
tern which does not match the expected patterns of known alleles.
 2 HLA Typing by SSO and SSP Methods 13

But as the SSOs do not cover all regions of the gene, it is still pos-
sible for new alleles to be missed by this method and they would
only be found if the sample was genotyped by another method
such as DNA sequencing.
While the PCR-SSO method is not a suitable technique for
low sample numbers, there is a method variation known as reverse-
SSO (13, 14) which is well suited to the testing of individual sam-
ples or multiple samples concurrently. In the reverse-SSO
technique, oligonucleotides are individually bound to membranes,
each membrane having all the SSOs bound that are required for a
particular HLA locus or allele group. These membranes are pre-
made so they are available for testing at any time. Sample DNA is
amplified using primers specific for the HLA locus to be typed,
incorporating a label such as biotin. The amplified DNA is then
hybridised to the membrane with the pre-bound oligonucleotides.
Non-specifically bound amplicons are removed by washing and
the hybridisation between the amplicon and membrane bound
oligonucleotide is detected with a chemiluminescent or colour-
based detection system.
A recently developed variation of the reverse-SSO system
employs Luminex xMAP® technology. The Luminex xMAP® tech-
nique utilises polystyrene microspheres which are dyed internally
with red and infrared fluorophores. The two dyes are incorporated
in differing proportions on different beads allowing 100 unique
sets of microspheres to be developed. Each of the 100 bead sets
can be individually identified with a Luminex® analyser even when
in the same reaction tube. For HLA genotyping, each of the micro-
sphere sets is coated with a different oligonucleotide. Test DNA is
amplified by PCR and hybridised to the bead sets in solution, the
process is carried out in a 96-well PCR plate with one sample
tested/well (each well contains all the different bead sets), allow-
ing 96 samples to be tested simultaneously. Non-hybridised DNA
is removed by washing and amplified DNA which has bound to the
SSOs on the microspheres is detected by a fluorescent dye with a
Luminex® analyser using flow cytometric technology. The analyser
automates the data acquisition and provides quantitative results as
well as HLA assignments. Because 100 different bead sets can be
included in each reaction tube or tray well, it is easier to include a
larger number of SSOs with Luminex xMAP® technology than
other SSO or reverse-SSO methods, enabling a better level of HLA
typing resolution.
While reverse-SSO typing still has the drawbacks of ambiguity
resolution that the original SSO method has, it is well suited to
genotyping of small sample numbers and there are several com-
mercial kits available utilising this technology. Some commercial
kits incorporate robotics for the hybridisation, washing and detec-
tion steps making the method also suitable for the typing of larger
sample numbers.
14 H. Dunckley

1.3. Interpretation Because of the large and ever increasing number of HLA alleles,
of Results data analysis and HLA genotyping assignment now requires the
use of software, commercial HLA typing kits generally provide
software programs for this purpose. For the SSP, SSO and reverse-
SSO methods, most analysis software follows the same principle.
The reaction patterns of known alleles are established for the
primer sets used in the SSP kit or the oligonucleotides in the SSO
or reverse-SSO methods based on known HLA sequence data
from the IMGT/HLA database (15, 16). When a sample is tested,
the reaction pattern obtained is compared to the reaction patterns
of known alleles. The software then provides a list of possible HLA
types based on the patterns of known alleles which match the sam-
ple being tested. With the SSP and SSO methods, multiple results
will sometimes be obtained as it is difficult for these methods to
completely resolve all possible heterozygous allele combinations.
However, provided the alternative results are within the same
allele group and do not cross antigen specificities, this is not usu-
ally an issue particularly if a low resolution typing result is satisfac-
tory. The SSO and SSP methods can also be used in conjunction
with sequencing-based typing (SBT) by providing a generic result
which determines the primers to be used in allele level SBT. In
some cases, ambiguities can arise with SBT (with heterozygous
sequences) in which case the SSP method can be used to help
resolve an allele level typing. Because of the continual increase in
the identification of new alleles, laboratories must keep all typing
methods up to date which requires the databases used for SSO
and SSP result interpretation to be updated at least every 6 months.
The IMGT/HLA database of officially named HLA alleles is
updated every month (15, 16).
While molecular HLA typing is more accurate and provides a
higher level of resolution than serological tissue typing (17–20),
molecular methods will not detect new alleles if a new polymor-
phism is not in the region where an amplification primer or SSO
matches the gene. In addition, null or non-expressed alleles are
usually detected when molecular typing is used in conjunction with
serological methods. When only one method is used both serology
and molecular techniques sometimes fail to detect null alleles.
While null alleles are not common, tissue typing protocols for
transplantation should be developed to avoid missing these non-
expressed alleles, especially in matching for haematopoietic stem
cell transplantation.

1.4. Clinical The SSP, SSO and reverse-SSO methods are all suitable for use in
Applications routine tissue typing laboratories as well as for research purposes.
With a variety of commercial kits now available, laboratories can
choose which method is most suitable for their work, and a combi-
nation of kits can be used for different requirements. In laborato-
ries testing for clinical purposes like transplantation it is useful to
 2 HLA Typing by SSO and SSP Methods 15

have more than one method available so that any unusual results
can be verified. If dealing with low sample numbers or urgent
testing, then SSP is very practical. With low to large sample num-
bers reverse-SSO can be used, this method also gives a reasonable
turnaround time for testing. For large sample batches and when
the test time is less urgent, then SSO typing is a very cost-effective
method.
As it is now easier to buy commercial HLA-SSP and reverse-
SSO kits rather than developing in-house methods, the protocols
described below summarise the steps that would be followed when
using a commercial kit, some specific details being provided by the
kit manufacturer supplier. When using commercial kits, each ship-
ment and batch of kits needs to be checked before routine use.
This can be accomplished using samples of known HLA type to
verify the typing results obtained with the kit.

2. Materials

2.1. HLA-SSP Typing 1. Commercial kits will usually supply the primer sets (both HLA
and control primers) already pre-dispensed into microtitre
trays or tubes ready for use in a thermocycler. Some trays will
also include the dNTPs and PCR buffer with the primers.
Sometimes these mixes will be lyophilised, other times they
will be in solution but covered with a layer of paraffin oil to
avoid evaporation.
2. PCR buffer and dNTPs, if not included with the pre-dispensed
primer sets, are usually included with the commercial kits as is
the gel loading buffer.
3. Taq DNA Polymerase—may or may not be included with the
commercial kit.
4. Tube caps or tray sealers for the PCR trays are usually included
with the commercial kits.
5. Sterile molecular grade water (dH2O).
6. Pipettes and sterile tips to dispense 1–10 μL, 10–200 μL, and
200–1,000 μL volume ranges. Electronic multi-dispensing
pipettes are better if the sample numbers are high because of
the large amount of pipetting involved.
7. Thermocycler with a heated lid.
8. Agarose horizontal gel tank equipment, gel casting trays, gel
combs, and power supply.
9. Agarose (DNA grade).
10. Electrophoresis buffer as specified by the kit supplier.
11. Microwave or heating apparatus to dissolve the agarose.
16 H. Dunckley

12. DNA molecular weight marker to cover a range of

50–2,000 bp.
13. Ethidium bromide 10 mg/mL (see Note 1).
14. UV transilluminator and photographic equipment—a digital
photo allows the photo to be stored electronically.
15. Gel documentation forms and worksheets if not included with
the commercial kit.
16. Data analysis and HLA assignment software (from the kit sup-
plier) and a computer.

2.2. HLA Reverse-SSO 1. A master mix containing biotinylated primers, dNTPs and buf-
Typing fer is usually supplied with commercial kits. MgCl2 may be pro-
vided in a separate tube. These are normally supplied frozen.
2. Taq DNA polymerase may be included in the master mix, or
may be provided in a separate tube, or may not be included
with a commercial kit.
3. Strips with pre-bound oligonucleotides are provided with com-
mercial kits.
4. Control DNA is usually supplied with a commercial kit.
5. Sterile molecular grade water (dH2O).
6. Pipettes and sterile tips to dispense 1–10 μL, 10–200 μL, and
200–1,000 μL volume ranges. Electronic multi-dispensing
pipettes are better if the sample numbers are high because of
the large amount of pipetting involved.
7. PCR plastic ware—tubes or microtitre trays, caps or tray sealers.
8. Thermocycler with a heated lid.
9. Agarose horizontal gel tank equipment, gel casting trays, gel
combs, and power supply.
10. Agarose (DNA grade).
11. Electrophoresis buffer as specified by the kit supplier.
12. Microwave or heating apparatus to dissolve the agarose.
13. Ethidium bromide 10 mg/mL (see Note 1).
14. UV transilluminator and photographic equipment—a digital
photo allows the photo to be stored electronically.
15. Solutions required for denaturation, hybridisation, washing,
and probe detection are included with commercial kits (see
Note 2).
16. Some reverse-SSO kit suppliers provide automated assay equip-
ment. Otherwise, the process can be carried out manually and
requires a rocking platform, shaking water bath and containers
for hybridisation, washing and probe detection steps as well as
sterile pipettes (2–50 mL), and adjustable volume dispensers.
 2 HLA Typing by SSO and SSP Methods 17

17. Water bath and filter forceps.

18. Ethanol.
19. Gel documentation forms and worksheets for recording oligo-
nucleotide results if not included with the commercial kit.
20. Data analysis and HLA assignment software (from the kit sup-
plier) and a computer.

3. Methods

3.1. HLA-SSP Method 1. Remove the PCR tray containing the primer sets and the kit
components from the freezer or fridge in the pre-PCR area
(see Note 3).
2. Following the manufacturer’s instructions, add the required
amount of dH2O to the PCR buffer provided in the kit, mix
well and then add Taq DNA polymerase. Add the amount
specified by the manufacturer to the negative control well of
the PCR tray (see Notes 4 and 5).
3. Add the required amount of DNA to the remaining buffer/
enzyme solution, mix well, and add the amount specified by
the manufacturer to each of the test wells of the PCR tray. Seal
or cap each tube tightly (see Notes 6–8).
4. Ensure the liquid of all the reaction mixes is at the base of the
tray—hand flick the tray, or tap the tray gently on the bench,
or pulse centrifuge the tray at a low speed (see Note 9).
5. In the PCR or post-PCR area, place the tray in the thermocy-
cler and perform the PCR according to the kit supplier’s
specifications (see Notes 10–13).
6. When the PCR program is complete, remove the tray from the
thermocycler.
7. In the post-PCR area, electrophorese the amplified samples and
DNA molecular weight marker on an agarose gel containing
ethidium bromide in the buffer recommended by the kit sup-
plier. The supplier will provide details on the agarose concentra-
tion, gel size, well size, and voltage required (see Notes 14–17).
8. Take a photo of the gel under a UV light (see Note 18).
9. Check that the negative control is negative (no amplification
product present). Score the positive PCRs (checking that the PCR
product is the correct size) and note any wells where both the
control PCR and HLA PCR failed as these cannot be scored as
negative but as an unknown result (see Notes 19–22; Table 2).
10. Input the results of the PCRs (positive, negative, or unknown/
no result) into the data analysis software program and obtain
the HLA typing from the software (see Notes 23–28).
18 H. Dunckley

Table 2
Interpretation of SSP gel results

HLA amplicon Control amplicon Result

Present Present Positive for the HLA amplicon
Present Absent Positive for the HLA amplicon
Absent Present Negative for the HLA amplicon
Absent Absent No result

3.2. HLA Reverse-SSO 1. In the pre-PCR area, make up the reaction mix as per the man-
Method ufacturer’s recommendations (buffer, water, dNTPs, primers,
and Taq DNA polymerase). The amount of reaction mix made
will depend on the number of samples to be amplified. Aliquot
the reaction mix into the PCR tubes or tray (see Notes 3, 5, 7,
and 29).
2. Add one DNA sample (amount as per the kit supplier
specifications) to each reaction mix on the tray checking that
the correct sample is added to each well or tube (see Note
30).
3. Seal or cap each tube tightly, ensure the liquid of all the reac-
tion mixes is at the base of the tray—hand flick the tray, or tap
the tray gently on the bench, or pulse centrifuge the tray at a
low speed (see Notes 8 and 9).
4. In the PCR or post-PCR area, immediately place the tray in
the thermocycler and perform the PCR according to the kit
supplier’s specifications (see Note 31).
5. When the PCR program is complete, remove the tray from the
thermocycler.
6. In the post-PCR area, electrophorese the amplified samples on
an agarose gel containing ethidium bromide in the buffer rec-
ommended by the kit supplier. The supplier will provide details
on the agarose concentration, gel size, well size, and voltage
required. Note: the agarose gel step is optional, some commer-
cial kits do not include this step or steps 7 and 8 below.
7. Photograph the gel under a UV light (see Note 18).
8. Check that the negative control is negative (no amplification
band present) and the PCR has worked for the test samples
(amplification band evident on the gel).
9. Denature the amplified DNA sample as per the instructions in
the kit and add the denatured DNA to the container with the
 2 HLA Typing by SSO and SSP Methods 19

membrane strip which has the pre-bound oligonucleotides (see

Note 32).
10. Incubate the amplicon DNA with the membrane for the
hybridisation step and then remove unbound or non-specifically
bound DNA by washing following the instructions in the kit
(see Notes 33–35).
11. Add the conjugate to the membrane which binds to the
labelled amplicons captured by the oligonucleotides. Remove
unbound conjugate by washing following the kit instructions
(see Note 36).
12. Incubate the membranes with the substrate solution from the
kit for the colourimetric reaction. When the colour reaction
has developed sufficiently, remove excess substrate by washing
as per the kit instructions (see Note 37).
13. Add the short-term storage buffer from the kit to the
membranes. A photo or scan (with a normal flat bed scanner)
can be taken of the membrane strips for long-term digital
record storage.
14. Use the strip template overlay provided with the kit to identify
the probes correctly.
15. Check that the positive and negative controls give the correct
results. Some kits also include an intensity reference probe.
The reaction strength of the intensity reference probe is
designed to be less than the reaction with all the HLA oligo-
nucleotides on the strip. If the intensity reference probe signal
is absent, there is a possibility that hybridisation with the HLA
oligonucleotides has not been detected so the whole test will
need to be repeated. Any HLA oligonucleotide signals that are
equal to or greater than the intensity of the control probe
should be scored as positive (see Note 38).
16. Score the hybridisation reactions with each oligonucleotide
and input the results into the data analysis software program
and obtain the HLA typing from the software (see Notes 24
and 27).

4. Notes

1. Ethidium bromide is a mutagen. Handle with appropriate

personal protective equipment including gloves and eye
protection.
2. The denaturation solution normally contains sodium hydroxide.
The hybridisation and washing solutions usually contain a buffer
20 H. Dunckley

with sodium chloride, sodium phosphate and EDTA (SSPE),

and sodium dodecyl sulphate (SDS). When using biotinylated
primers, the conjugate is generally streptavidin-horseradish
peroxide (HRP). The substrate for streptavidin-HRP which is
used for probe detection depends on the kit supplier.
3. All work involving PCR requires the laboratory to have
separate areas for pre- and post-PCR work so there is no risk
of contaminating pre-PCR samples with PCR amplified
product.
4. If an error is made during the PCR setup, this can lead to
weak or no amplification of the internal control or test sample
(one or multiple PCRs which may be incorrectly scored as
negative reactions). The PCR must be set up exactly as
specified in the kit instructions in relation to reagent volumes
and concentrations. Ensure that pipettes are calibrated.
5. If the Taq DNA polymerase is not performing correctly, there
will be weak or no amplification. Check the Taq DNA poly-
merase has been validated before use, re-verify if required with
a known DNA sample.
6. If insufficient test DNA, poor quality or degraded DNA is used
or PCR inhibitors are present in the DNA (e.g. too much
EDTA), there may be no amplification of the internal control
or the HLA alleles, or only weak amplification (one or multiple
PCRs which may be incorrectly scored as negative reactions).
If this occurs, the volume, concentration, and purity of the test
DNA should be checked. If required re-precipitate the DNA
(ethanol and salt step) and resuspend the DNA in a smaller
volume of dH2O or repeat the whole DNA extraction with
new reagents. DNA can be extracted from nucleated cells by
any preferred method (in-house or commercial kit, manual or
robotic system). If blood samples are used, the anticoagulant
should be EDTA or ACD, not heparin which can inhibit the
PCR. DNA should be kept in TE buffer for long-term storage
but the aliquots for SSP work should be diluted in dH2O, the
final concentration of the DNA prior to amplification should
be according to the recommendations of the kit supplier as this
varies depending on the kit used. The OD260/280 of the DNA
should be between 1.6 and 1.9.
7. If the test DNA is not resuspended uniformly or the Taq DNA
polymerase or PCR reagents are not mixed properly before
use, there may be random PCR failures. Pipette the test DNA
up and down several times to aid re-suspension; incubate test
DNA on a heat block or in an incubator for 10 min at 70°C.
Mix individual reagents briefly by vortexing (pulse only) before
use. Reagents should also be centrifuged briefly (pulse only) so
the reagent solutions are all at the base of the tube before use.
 2 HLA Typing by SSO and SSP Methods 21

Ensure the Taq DNA polymerase/buffer/DNA mixture is also

mixed well before dispensing into the PCR tray.
8. If the PCR tubes or tray are not capped or sealed tightly, there
may be evaporation during the PCR leading to random
amplification failure across the PCR plate. Make sure the caps
or tray sealer are fixed properly (use a capping tool). Check the
heated lid of the thermocycler is firmly in place.
9. If the PCR mix is not at the base of the tube in the PCR tray
(some of the mix is on the side or lid of the tube), there may
be random amplification failure leading to false negative results.
Ensure all the PCR mix is under the paraffin oil if it is used.
10. A delay between the PCR setup and start of the PCR program
can lead to false positive SSP reactions. Do not allow more
than a 5 min delay between the PCR setup and commencing
the PCR program.
11. Incorrect PCR conditions, for example an annealing tempera-
ture which is too high, can result in a low level of amplification
or no amplification of the test DNA. One or multiple PCRs
may be affected and incorrectly identified as negative. If this
occurs the PCR program should be checked.
12. Temperature variation across the heat block in the thermocy-
cler can cause random amplification failure across the PCR
plate which may be falsely interpreted as negative reactions.
Make sure the thermocycler is calibrated and re-calibrate if
necessary to ensure the temperatures of the wells in the heat
block are correct during thermocycling. Calibration verification
is carried out by running a thermocycler program and measur-
ing the temperature of the wells with a temperature probe to
check the observed temperatures are within specification. If
the observed temperatures are not within specification, the
thermocycler will need to be re-calibrated following the equip-
ment manufacturer’s instructions.
13. Random amplification failure can be caused by inadequate
contact between the thermocycler heat block and PCR tray or
tubes. Ensure that the correct trays or PCR tubes are used for
the thermocycler.
14. A DNA molecular weight marker should be run on the gel as
the SSP method amplifies DNA of many different sizes so
results need to checked against the size marker to ensure that
a positive amplification result is of the correct size and not a
false positive or primer dimer.
15. The presence of blurry amplification bands on the gel may be
caused by the electrophoresis buffer becoming too hot due to
excess voltage when running the gel; the wrong electrophore-
sis buffer composition; old electrophoresis buffer being used;
or the agarose not being dissolved properly before pouring the
gel. To avoid these problems lower the voltage during gel elec-
22 H. Dunckley

trophoresis, use fresh electrophoresis buffer and ensure the

agarose is dissolved properly before pouring the gel.
16. Heavy streaking in random wells can be caused by uneven
suspensions of DNA. Mix the PCR product up and down
several times before loading the gel.
17. Occasional faint amplification bands may be due to a gel
loading problem. Use slow, steady pipetting when loading the
gel so the sample does not float away.
18. A photo can be taken of the gel on a UV transilluminator using
a digital camera with an orange filter.
19. The negative control well contains the control primer set but
no DNA. Any amplification seen when the negative control is
run on the gel is evidence of DNA contamination at some
stage of the test process and results of the test are invalid.
20. To avoid the risk of false negative reactions in the SSP method,
internal controls must be used for every SSP PCR mix to ver-
ify the PCR conditions. Typically these controls are primers
for a non-variable or conserved gene present in all samples
and the control primer set is selected to amplify a DNA frag-
ment of a different size to that expected from the HLA primer
sets. In order to score a reaction with an HLA primer set as
negative, the control amplification must be positive (amplified
DNA detected) and the HLA amplicon absent. If the control
amplicon is absent and the HLA amplicon is also absent, it
cannot be assumed that this is a true negative result as there
may be inhibition of the PCR in that tray well or tube.
However, the presence of a specific HLA amplicon with no
control amplification can be scored as a positive result. This
conclusion is based on the fact that the PCR of the HLA gene
usually amplifies better than the control PCR as the control
primers are present in a lower concentration in order to favour
the HLA specific reaction (see Fig. 1). Table 2 summarises the
interpretation of SSP results.
21. Care must be taken when assigning homozygous results by the
SSP method, particularly if any of the control amplicons are
absent. The rates of homozygosity should be monitored, espe-
cially if there is a high rate of control amplification failure.
22. Interpretation of primer dimer as specific amplification bands
can lead to false positive results. Always check the amplification
band is the correct size.
23. Non-specific amplification (ladders or smears) may be seen if
there are impurities in the test DNA or excess test DNA is
used. Check the DNA quality and concentration. If required
re-precipitate the DNA (ethanol and salt step) and resuspend
the DNA in dH2O or repeat the whole DNA extraction with
new reagents.
 2 HLA Typing by SSO and SSP Methods 23

24. The presence of more than two specific HLA alleles indicates
contamination between PCR products or between DNA
samples during the DNA extraction or PCR preparation
stages. Make sure the pre- and post-PCR work is kept sepa-
rate (different rooms, lab coats, gloves, equipment and
reagents); clean the working area with 10% bleach and ensure
there is no mixing of PCR amplicons. Check there has been
no mix up at the DNA extraction stage—check other test
samples for a possible mix up.
25. False positive SSP reactions can be caused by contamination
when dispensing reagents to set up the PCR or when loading
the gel with the amplified product. Take more care when
pipetting during the PCR setup or when loading the gel.
26. A false positive or false negative reaction may be due to an
error in loading the gel so that the PCRs have been loaded in
the wrong order. Check the alignment of the PCR mixes and
gel lanes.
27. An unexpected positive or negative reaction may sometimes be
due to the presence of a new allele. Repeat the typing and if the
same reaction is seen verify the typing by SBT.
28. Persistent false negative or positive reactions with the same
PCR mix may be due to a kit problem. Double check the kit
with a control sample and advise the supplier if the control
sample also gives the same false negative or positive result.
29. If an error is made during the PCR setup, this can lead to inefficient
amplification and weak probe signals. The PCR must be set up
exactly as specified in the kit instructions in relation to reagent
volumes and concentrations. Ensure that pipettes are calibrated.
30. If insufficient test DNA, poor quality or degraded DNA is used,
there may be inefficient amplification and weak probe signals. If
this occurs, the volume, concentration, and purity of the test
DNA should be checked. If required re-precipitate the DNA
(ethanol and salt step) and resuspend the DNA in dH2O or repeat
the whole DNA extraction with new reagents. DNA can be
extracted from nucleated cells by any preferred method (in-house
or commercial kit, manual or robotic system). If blood samples
are used the anticoagulant should be EDTA or ACD, not hepa-
rin which can inhibit the PCR. DNA should be kept in TE buffer
for long-term storage, the final concentration of the DNA prior
to amplification should be according to the recommendations of
the kit supplier as this varies depending on the kit used. The
OD260/280 of the DNA should be between 1.6 and 1.9.
31. Incorrect PCR conditions can result in a low level of
amplification or no amplification of the test DNA (and conse-
quently weak or no hybridisation signals). If this occurs, the
PCR program should be checked.
24 H. Dunckley

32. Always handle the membranes at the end of the strips using
forceps.
33. If the water bath, incubator, or wash solution temperature is
too high when removing non-specifically bound probe (strin-
gency wash step), it can result in weak hybridisation signals.
If the wash temperature is too low, there can be false positive
probe signals or high background on the probe support
membrane. It is important to set the temperature with a cali-
brated thermometer and to monitor the temperature during
the stringency wash step.
34. If the wash buffer solution is made up incorrectly, this can
cause either weak hybridisation signals, high background on
the probe support membrane or false positive probe signals. It
is critical that wash buffer solutions are prepared carefully and
that salt concentrations are correct.
35. High background on the probe support membrane can be
caused by detergent residue in the hybridisation or wash con-
tainers. Do not use detergent when washing the containers or
beakers used for making solutions but rinse the containers very
well with water.
36. High background on the probe support membrane can be due
to the conjugate step being performed at an incorrect tem-
perature. Monitor this step with a calibrated thermometer.
37. Ensure the substrate incubation time is correct. If the time is
too short, this can lead to weak hybridisation signals and if it is
too long it can result in high background on the probe support
membrane. Also check that the substrate is at room tempera-
ture before it is used.
38. If the positive control result is strong but the probe intensity of
the test samples is weak, this is most likely due to a weak PCR
because of insufficient test DNA or poor quality DNA. If this
occurs, the volume, concentration, and purity of the test DNA
should be checked. If required re-precipitate the DNA (etha-
nol and salt step) and resuspend the DNA in dH2O or repeat
the whole DNA extraction with new solutions. Use non-hepa-
rinised blood to extract DNA.

References

1. Olerup O, Zetterquist H (1992) HLA-DR typ- 2. Olerup O, Aldener A, Fogdell A (1993) HLA-
ing by PCR amplification with sequence-specific DQB1 and -DQA1 typing by PCR amplification
primers (PCR-SSP) in 2 hours: an alternative to with sequence-specific primers (PCR-SSP) in 2
serological DR typing in clinical practice includ- hours. Tissue Antigens 41:119–134
ing donor-recipient matching in cadaveric 3. Bunce M, Taylor CJ, Welsh KI (1993) Rapid
transplantation. Tissue Antigens 39:225–235 HLA-DQB typing by eight polymerase chain
 2 HLA Typing by SSO and SSP Methods 25

reaction amplifications with sequence-specific 11. Gao X, Jakobsen IB, Serjeantson SW (1994)
primers (PCR-SSP). Hum Immunol 37: Characterization of the HLA-A polymorphism
201–206 by locus-specific polymerase chain reaction
4. Browning MJ, Krausa P, Rowan A et al (1993) amplification and oligonucleotide hybridiza-
Tissue typing the HLA-A locus from genomic tion. Hum Immunol 41:267–279
DNA by sequence-specific PCR: comparison of 12. Cereb N, Yang SY (1997) Dimorphic primers
HLA genotype and surface expression on col- derived from intron 1 for use in the molecular typ-
orectal tumor cell lines. Proc Natl Acad Sci ing of HLA-B alleles. Tissue Antigens 50:74–76
USA 90:2842–2845 13. Erlich H, Bugawan T, Begovich AB et al (1991)
5. Sadler AM, Petronzelli F, Krausa P et al (1994) HLA-DR, DQ and DP typing using PCR
Low-resolution DNA typing for HLA-B amplification and immobilized probes. Eur J
using sequence-specific primers in allele- or Immunogenet 18:33–55
group-specific ARMS/PCR. Tissue Antigens 14. Bugawan TL, Apple R, Erlich HA (1994) A
44:148–154 method for typing polymorphism at the HLA-A
6. Bunce M, O’Neill CM, Barnardo MCNM et al locus using PCR amplification and immobilized
(1995) Phototyping: comprehensive DNA typ- oligonucleotide probes. Tissue Antigens
ing for HLA-A, B, C, DRB1, DRB3, DRB4, 44:137–147
DRB5 & DQB1 by PCR with 144 primer mixes 15. Robinson J, Malik A, Parham P et al (2000)
utilising sequence-specific primers (PCR-SSP). IMGT/HLA—a sequence database for the
Tissue Antigens 46:355–367 human major histocompatibility complex.
7. Wordsworth BP, Allsopp CE, Young RP et al Tissue Antigens 55:280–287. http://www.ebi.
(1990) HLA-DR typing using DNA ac.uk/imgt/hla/
amplification by the polymerase chain reaction 16. Robinson J, Mistry K, McWilliam H et al
and sequential hybridization to sequence- (2011) The IMGT/HLA database. Nucleic
specific oligonucleotide probes. Immunogenetics Acids Res 39(suppl 1):D1171–D1176
32:413–418 17. Mytilineos J, Scherer S, Dunckley H et al (1993)
8. Morel C, Zwahlen F, Jeannet M et al (1990) DNA HLA-DR typing results of 4000 kidney
Complete analysis of HLA-DQB1 polymorphism transplants. Transplantation 55:778–781
and DR-DQ linkage disequilibrium by oligonu- 18. Mytilineos J, Lempert M, Middleton D et al
cleotide typing. Hum Immunol 29:64–77 (1997) HLA class I DNA typing of 215 “HLA-
9. Scharf SJ, Griffith RL, Erlich HA (1991) Rapid A, -B, -DR zero mismatched” kidney trans-
typing of DNA sequence polymorphism at the plants. Tissue Antigens 50:355–358
HLA-DRB1 locus using the polymerase chain 19. Lorentzen DF, Iwanaga KK, Meuer KJ et al
reaction and nonradioactive oligonucleotide (1997) 25% Error rate in serologic typing of
probes. Hum Immunol 30:190–201 HLA-B homozygotes. Tissue Antigens
10. Kimura A, Sasazuki T (1992) Eleventh inter- 50:359–365
national histocompatibility workshop refer- 20. Mytilineos J, Christ U, Lempert M et al (1997)
ence protocol for the HLA-DNA typing Comparison of typing results by serology and
technique. In: Tsuji K, Aizawa M, Sasazuki T polymerase chain reaction with sequence-
(eds) HLA 1991. Oxford University Press, specific primers for HLA-Cw in 650 individuals.
Oxford, p 397 Tissue Antigens 50:395–400
 Chapter 3

Methods for Diagnostic HLA Typing in Disease Association

and Drug Hypersensitivity
Michael D. Varney, Alison S.L. Castley, Katri Haimila,
and Päivi Saavalainen

Abstract
This chapter describes the application of diagnostic HLA typing for disease association and five methods
used for specific HLA genotypes. The methods utilise a combination of polymerase chain reaction (PCR)
amplification to detect sequence polymorphism by the presence or absence of amplification, nucleotide
sequencing of the PCR product, and hybridisation of the PCR product with labelled probes. The probes
are specific for sequence polymorphism associated with the genotype and are attached to either a Micro
Bead or a Solid Phase. In addition, the detection of single nucleotide polymorphism(s) which “tag” for the
genotype using a real-time PCR is described.

Key words: HLA, Disease, Association, Hypersensitivity, Methods, B*27, B*57, Coeliac, B*1502,
Narcolepsy, DQ

1. Introduction

Many diseases have been shown to be associated with antigens,

alleles, epitopes, or haplotypes at different loci within the HLA
region and yet the diagnostic value of these associations, with a
number of exceptions, remains unclear. This chapter briefly out-
lines the challenges of disease association studies and their role in a
diagnostic HLA typing laboratory and describes a combination of
locally designed and commercially available HLA typing methods
utilised for diagnostic HLA typing requests.
Typically in disease association studies, the frequency of a par-
ticular HLA polymorphism in a patient population is compared to
a control population by constructing a 2 × 2 contingency table and

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_3, © Springer Science+Business Media New York 2012

27
28 M.D. Varney et al.

testing the null hypothesis that no difference exists using the

chi-squared test with Yates’ correction or Fishers exact test. The
strength of the association is measured by calculating the relative risk
or odds ratio in the presence of a statistically significant difference
(typically p < 0.05). To avoid a type 1 error (false positive finding)
when many HLA polymorphisms are compared, the p value may
be conservatively corrected by multiplying it by the number of
comparisons made (Bonferroni correction). Association studies in
the HLA region are further complicated by the presence of linkage
disequilibrium in that alleles at different loci are not always ran-
domly associated. In association studies this makes the delineation
of the causative loci problematic. Whilst appropriate statistical
consideration is important in association studies, the selection of a
clearly defined patient group and matched controls remains pivotal.
The use of HLA genotyping as a diagnostic test is dependent
on the test sensitivity (proportion of disease positive patients
correctly identified by the genotype i.e. true positive (TP) vs. false
positive (FP)) and the test specificity (proportion of disease negative
patients correctly identified by the genotype i.e. true negative (TN)
vs. FP). However, the diagnostic relevance of the genotype result
is the probability that it will positively predict disease (the proba-
bility that a genotype positive person will have the disease i.e. TP/
TP + FP) or the absence of disease (the probability that a genotype
negative person will not have the disease i.e. TN/TN + FN)
(reviewed at (1, 2)). Whilst a positive predictive value may be
requested of HLA genotyping, its value often lies in the negative
predictive value. For example, the HLA genotypes associated with
coeliac disease (Table 1) have a negative predictive value greater
than 99%.
A large number of methods for HLA typing utilising the poly-
merase chain reaction (PCR) amplification of DNA have been pub-
lished and described in collaborative workshops and this volume.
This chapter has selected five methods and described their use in
determining the genotypes associated with disease and drug hyper-
sensitivity described in Table 1. However, it is recognised that each
method has the potential to determine other genotypes and that
the methods may be used in conjunction with one another.
Sequence Specific Priming (SSP) utilises the inefficient PCR
amplification of a primer that contains a 3¢ nucleotide mismatch.
This inefficiency permits the design of PCR primers that, while
specific for their target sequence, will not amplify closely related
sequences (4). By standardising the PCR cycling conditions for a
range of primers, multiple nucleotide differences in HLA genes
can be detected using a pattern of amplification that corresponds
to HLA alleles (5). The methods described below for HLA-DR
and DQ are used locally to determine the genotypes associated
with coeliac disease, narcolepsy, Type 1 diabetes, and carbam-
azepine hypersensitivity in conjunction with sequence-based
typing (SBT) to resolve the DQB1*0302, 0602, and B*15 alleles.
3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 29

Table 1
Diagnostic HLA disease and hypersensitivity associations

Disease/hypersensitivity HLA association Notes

Ankylosing spondylitis B-27 High negative predictor B27 allele type

may provide further discrimination;
B*2702-08,10/14/15/19 positively
associated, whereas B*2709 appears
protective (3)
Narcolepsy DQB1*0602 DRB1*15 is in strong linkage disequilibrium
with DQB1*0602
Coeliac disease DQA1*05, DQB1*02, High negative predictor
DQB1*0302
Rheumatoid arthritis DRB1*0401/04/05/13, Positive DRB1 allele association reflects a
0101/02, 1402, 1001 “shared epitope” (QKRAA) at amino
acids 70–75
Type 1 diabetes DRB1*03-DQB102, DRB1* May be used to assess familial risk
04-DQB1*0302 (DRB1*
15-DQB1*0602 protective)
Abacavir hypersensitivity B*5701 High negative predictor
Carbamazepine hyper- B*1502 Restricted to Asian genetic background
sensitivity (Stevens–
Johnson syndrome)

In addition, methods for the detection of HLA-B27 and B-57

using whole blood rather than DNA as a template are described.
SBT typically uses the Sanger approach with fluorescently
labelled dideoxynucleotides in a cycle sequencing reaction using a
purified PCR product. The subsequent reaction is “cleaned” of
unincorporated dyes, dNTPs, and primers by precipitation or mag-
netic bead separation and the resulting fragments detected in an
automated analyser. Nucleotide sequence is interpreted from the
fragment analysis using an automated analysis program and com-
pared to the known sequence of HLA alleles. The methods
described below use PCR primer pairs specific for groups of alleles
to allow sequencing of a single allele e.g. DQB1*02, *0302,*0602,
DRB1*01,* 04, *14, *10, B*57, and B*15 allele combinations.
Other approaches include sequencing of a single PCR product at
each loci (6) using Sanger sequencing or the application of next
generation sequencing (7). The reagents for both SSP and SBT at
most HLA loci and analysis software packages are available
commercially.
Micro Bead hybridisation assays utilise a series microsphere
beads containing immobilised biotinylated oligonucleotide probes
that are hybridised to PCR amplified DNA in a single well of a
96-well microtitre plate. Individual beads are recognised by a
30 M.D. Varney et al.

Table 2
Coeliac disease associated DQ2 and DQ8 haplotypes and their prediction by the six
tagging single nucleotide polymorphisms (SNPs)

Positive Negative
predicting predicting Applied
DQ allele (other allele (other biosystems
haplotype DRB1 DQA1 DQB1 SNP rs number allele) allele) assay number

DQ2.5 03 0501 0201 rs2187668 T(C) C_58662585_10

DQ8 04 0301 0302 rs7454108 C(T) C_29817179_10
DQ7 11/12/13 0505 0301 rs4639334 A(G) C_42975350_10
(rs79616158a)
DQ2.2 07 0201 0202 rs2395182 T(G) C_11409965_10
rs4713586 G(A) C_27950246_10
rs7775228 C(T) C_29315313_10
The CD associated DQA1*05 and DQB1*02 alleles (marked bold) are located either in cis in the same DQ2.5 (DR-
DQ2) haplotype or in trans in patients carrying both DQ7 and DQ2.2 haplotypes
a
The TaqMan assay number C_42975350_10 for DQ7 refers currently to a changed SNP rs number (rs79616158)
which is different to that in the referenced tables (8, 9)

combination of two internal fluorescent dyes and positive hybridisation

by fluorescently labelled (Streptavidin-PE) captured PCR products in
a flow cytometric analysis system. A number of commercial compa-
nies e.g. One Lambda (http://www.onelambda.com) and Lifecodes
(http://www.gen-probe.com) provide complete HLA typing kits.
The method described below is a summary of the commercial
Luminex/Labtype (http://www.OneLambda.com) protocol with
minor modifications.
A large number of solid phase hybridisation assays have been
described which include the binding of either PCR product or oli-
gonucleotide probe to a nylon membrane support and the colour
or chemiluminescent detection of probes. The method described
below utilises a biotinylated PCR product which is immobilised to
streptavidin-coated microtitre plate and hybridised with a series of
fluorescent probes in a single well of a microtitre plate. The probes
are labelled with chelates of lanthanide metals which have distinct
fluorescent profiles allowing the detection of several probes in the
same assay. The method is a summary of the commercial DELFIA
(http://www.las.perkinelmer.com) Coeliac Disease assay.
An alternative approach to HLA genotyping is the typing of
single nucleotide polymorphisms (SNPs) which tag for HLA hap-
lotypes. For coeliac disease, six tagging SNPs have been identified
and validated in Caucasian populations only for the associated
HLA-DQ2 and DQ8 haplotypes (Table 2). Whilst the specificities
and sensitivities are high (8, 9) (97–100%), they are not complete
 3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 31

due to the presence of rare haplotypes and so the method is

not recommended for diagnostic testing but is cost effective and
suitable for high-throughput screening assays. The method uses
TaqMan® chemistry (http://www.AppliedBiosystems.com) and
two allele-specific dual labelled probes with alternative 5¢ dyes and
a 3¢ Quencher dye which bind to the SNP complementary target
sequence in a real-time PCR. Positive hybridisation results in the
release of a unique reporter dye detected by a fluorimeter allowing
allelic discrimination.

2. Materials

2.1. Sequence 1. HLA-B 10× PCR Buffer: To prepare dissolve 8.11 g Trisma-
Specific Priming Base in 80 mL H2O on a magnetic stirrer and adjust to pH = 8.8
and Sequence-Based using concentrated HCl. Add 2.19 g (NH4)2 SO4 and dissolve.
Typing Make up to 95 mL, filter through a 0.22-mm filter, and add
1 mL Tween 20. Aliquot and store frozen.
2. SSP Buffer: To prepare 1,500 mL add 500 mL 10× 1.5 mM
MgCl2 PCR Buffer (as supplied with Taq DNA Polymerase),
500 mL 10× dNTPS, 250 mL Glycerol, 200 mL H2O, 50 mL
Cresol Red Solution (10 mg/mL diluted in H2O).
3. 10× dNTPS: A stock solution is prepared containing a final
concentration of 2 mM of dCTP, dGTP, dATP, and dTTP
diluted in H2O.
4. B*1502 Master Mix: For each sample a 39-mL Master Mix is
prepared containing 5 mL HLA-B 10× PCR buffer, 2 mL
dNTPs (40 mM), 2 mL 50 mM MgCl2, 3.25 mL B*1502
primer mix, 35.35 mL H2O.
5. B*27 Master Mix: For each sample a 39-mL Master Mix is pre-
pared containing 5 mL HLA-B 10× PCR buffer, 0.3 mL dNTPs
(40 mM), 6 mL B*27 primer mix, 27.7 mL H2O.
6. B*57 Master Mix: For each sample a 39-mL Master Mix is pre-
pared containing 5 mL HLA-B 10× PCR buffer, 2 mL dNTPs
(40 mM), 3 mL 50 mM MgCl2, 15.75 mL B*57 primer mix,
13.25 mL H2O.
7. Formamide Master Mix: For each sample prepare a formamide
mix containing 5 mL Deionised Formamide and 4 mL deionised
H2O.
8. 2.5 mM MgCL2 DQB PCR Buffer: To prepare sufficient for
200 reactions (4,310 mL) add 500 mL 10× PCR Buffer (as sup-
plied with Taq DNA Polymerase, no MgCl2), 250 mL 50 mM
MgCl2, 500 mL 10× dNTPS, 250 mL DiMethyl Sulphoxide,
3,010 mL H2O; 2.0 and 1.2 mM MgCL2 DQB PCR Buffers
are prepared by adjusting the proportion of 50 mM MgCl2 and
H2O appropriately.
32 M.D. Varney et al.

9. 1.5 mM MgCL2 DRB PCR Buffer: To prepare sufficient for

200 reactions (4,560 mL) add 500 mL 10× PCR Buffer (as
supplied with AmpliTaq DNA Polymerase, no MgCl2), 150 mL
50 mM MgCl2, 500 mL 10× dNTPS, 3,410 mL H2O; 2.0 mM
MgCl2 DRB PCR Buffer is prepared by adjusting the propor-
tion of 50 mM MgCl2 and H2O appropriately.
10. SBT HLA-B Master Mix: To prepare sufficient for 200 reac-
tions (4,960 mL) add 500 mL HLA-B 10× PCR Buffer, 180 mL
50 mM MgCl2, 500 mL 10× dNTPS, 150 mL Dimethyl
Suphoxide, and HLA-B gen Primer Mix 600 mL, 2,670 mL
H2O or HLA-B TA Primer Mix 600 mL, 500 mL 0.1 M b-Mer-
captoethanol, 2,170 mL H2O or HLA-B CG Primer Mix
400 mL, 500 mL 0.1 M b-Mercaptoethanol, 2,370 mL H2O.
11. Specialist equipment: Applied Biosytems 3730 DNA Analyser
or equivalent.

2.2. Micro Bead 1. Denaturing Buffer, Neutralising Buffer, Hybridisation Buffer,

Hybridisation Typing Bead Mixtures, Wash Buffer, 100× SAPE, SAPE buffer as
Assay supplied by the manufacturer (OneLambda).
2. Specialist equipment: Luminex® LABScan™ Analyser or
equivalent.

2.3. DELFIA 1. Wash Concentrate, Bio-Hybridization Control, Hybridization

Hybridisation Buffer, labelled probes, Enhancer Solution, Enhancement
Solution, streptavidin coated microtitration strips as supplied
by the manufacturer.
2. dNTP mix (10 mM).
3. Sodium Hydroxide solution (20 mmol/L): Prepare the solu-
tion: 2.0 mL NaOH Titrisol 1 mol/L, add ultrapure water to
a volume of 100 mL, store maximum 2 months, the solution is
corrosive.
4. Tris buffer 10 mM, pH 8.0, Tris, p.a. 0.61 g, ultrapure water
add 500 mL, store maximum 1 year, the solution is irritating.
5. Specialist equipment: automatic microplate washer, e.g.
DELFIA Platewasher, plateshaker, a time-resolved fluorometer
e.g. VICTOR.

2.4. TaqMan® Typing 1. TaqMan® (2×) Genotyping Master Mix (Applied Biosystems).
of HLA-Tagging SNPs 2. TaqMan® (20×) SNP Genotyping Assay Mix (Applied
in Coeliac Disease Biosystems).
3. PCR-quality H2O.
4. Specialist equipment: ABI 7900HT real-time PCR system
(Applied Biosystems) or equivalent with 84- or 96-well optical
PCR plates and lids.
 3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 33

3. Methods

3.1. Sequence Specific 1. The identification of an individual or group of alleles is

Priming determined by the presence or absence of the correctly sized
test PCR product (Table 3) using standard agarose or poly-
acrylamide gel electrophoresis. In addition, a control PCR
product indicates the success of the assay which may or may
not be present when the test PCR is positive.
2. DRB/DQ. For each primer mix (see Note 1) being tested,
prepare an SSP Master Mix containing 5 mL of the primer mix,
3 mL SSP Buffer, and 0.1 mL Taq DNA Polymerase (5 U/mL).
3. Aliquot 8 mL of SSP Master Mix into a 200-mL PCR strip tube
or plate and add 2 mL DNA (50 ng/mL) (see Note 2).
4. Seal the tubes/plates and cycle in a PCR machine for 96°C-
120 s, 96°C-10 s, 65°C-60 s ×10 cycles, 96°C-10 s, 61°C-
50 s, 72°C-30 s ×20 cycles, 20°C hold.
5. HLA-B*1502. For each sample add 0.4 mL of AmpliTAQ
DNA polymerase to 47.6 mL of B*1502 Master Mix (see Note 1),
mix by gently flicking and add 2 mL DNA (30 ng/mL).
6. Seal the tubes and cycle in a PCR machine for 96°C-60 s ×1
cycle, 96°C-25 s, 70°C-45 s, 72°C-45 s ×4 cycles, 96°C-625 s,
65°C-50 s, 72°C-45 s ×24 cycles, 96°C-25 s, 55°C-60 s, 72°C-
80 s ×10 cycles, 15°C hold.
7. HLA-B*27 and B*57 (whole blood, see Note 2). Add 9.0 mL
of formamide mix to a PCR microtube and 1 mL of whole
blood, cap the tube, and mix by gently flicking.
8. Incubate at 95°C in a thermal cycler for 5–10 min, remove the
tubes and immediately place them on ice for at least 2 min.
9. For each sample add 1.0 mL of AmpliTAQ DNA polymerase to
39 mL of B*27 or B*57 Master Mix (see Note 1) and add to the
denatured blood sample. Mix by gently flicking.
10. Seal the tubes and for B*27 cycle in a PCR machine for 85°C-
120 s, 55°C-45 s, 60°C-60 s ×7 cycles, 85°C-60 s, 50°C-45 s,
60°C-90 s ×15 cycles, 85°C-60 s, 45°C-45 s, 60°C-120 s ×10
cycles, 15°C hold. For B*57 cycle for 85°C-120 s, 86°C-25 s,
60°C-45 s, 62°C-45 s ×4 cycles, 86°C-25 s, 55°C-50 s, 62°C-
45 s ×24 cycles, 86°C-25 s, 45°C-45 s, 62°C-45 s ×10 cycles,
15°C hold.

3.2. Sequence-Based 1. Select the PCR reaction(s) for the association to be tested
Typing using the primer pair specificities listed in Table 4. Additional
primer pairs e.g. DQB1*0201/02/03, B CG are described to
assist in the resolution of ambiguous results.
 34

Table 3
SSP primer pair sequences and specificity used in diagnostic HLA disease and hypersensitivity association testing

Primer pair specificitya Primer mix final

(PCR product size) Primer sequences (direction) concentration (mM)
DQA1*05 (186 bp) GGTCCCTCTGGCCAGTTC (f) 1.5
M.D. Varney et al.

GTTGGAGCGTTTAATCAGAC (r) 1.5

DQB1*0201-05 (200 bp) TGCGTCTTGTGAGCAGAAG (f) 1.5
CGTGCGGAGCTCCAACTG (r) 1.5
DQB1*0302/03/07/11/12/15/18/ GACGGAGCGCGTGCGTCT (f) 1.5
20/23/26,*0604/05/07/09/25/27/
AACTCCGCCCGGGTCCT (r) 1.5
32/34-39 (170 bp)
DQB1*0501-05, 0601-39 (285 bp) TTCCCCGCAGAGGATTTCGT (f) 1.5
CTCTCCTCTGCARGATCCC (r) 1.5
DRB1*1501-43 except 150107,1521/27/34 TTCCTGTGGCAGCCTAAGAGG (f) 1.0
(215 bp)
TCCACCGTGGCCCGCGC (r) 1.0
DQB1*0601-39 except 0623 (220 bp) GACGGAGCGCGTGCGTCT (f) 0.5
GACGGAGCGCGTGCGTTA (f) 0.5
GCAAGATCCCGCGGAACG (r) 0.5
GCAGGATCCCGCGGTACC (r) 0.5
PCR amplification control used in all DQ TGCCAAGTGGAGCCCCAA (f) 0.2–0.4
primer pairs (693 bp)
GCATCTTGCTCTGTGCAGAT (r) 0.2–0.4
b
B*1502 (357 bp) CGGAACACACAGATCTCCAAGACC (f) 3.57
GCCATACATCCTCTGGATGAT (r) 3.57
 3

B*1502 amplification control (439 bp) CAGTGCCTTCCCAACCATTCCCTTA (f) 0.71

ATCCACTCACGGATTTCTGTTGTGTTTC (r) 0.71
B*27 (141 bp) AGT CTG TGC CTT GGC CTT GC (f) 6.67
GCT ACG TGG ACG ACA CGC (r) 6.67
B*27 amplification control (439 bp) CAGTGCCTTCCCAACCATTCCCTTA (f) 0.42
ATCCACTCACGGATTTCTGTTGTGTTTC (r) 0.42
B*57 (130 bp) TGTAAAACGACGGCCAGTGTCTCACATCATCCAGGT (f) 1.82
B*57 amplification control (439 bp) CAGGAAACAGCTATGACCATCCTTGCCGTCGTAGGCGG (r) 1.82
CAGGAAACAGCTATGACCATCCTTGCCGTCGTAGGCAG (r) 1.82
CAGTGCCTTCCCAACCATTCCCTTA (f) 1.09
ATCCACTCACGGATTTCTGTTGTGTTTC (r) 1.09
a
IMGT database 2.28
b
The HLA-B*1502 primers also amplify some rarer HLA-B alleles, all positive results must be confirmed by an alternative HLA-B PCR/SBT assay
Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity
35
 36

Table 4
SBT primer pair sequences and specificity used in diagnostic HLA disease and hypersensitivity association testing

Primer pair Id and specificitya Primer mix

(sequencing primer, PCR product size) Primer sequence (direction) concentration
M.D. Varney et al.

DQB2,3,4 TGTAAAACGACGGCCAGTTATTCCCCGCAGAGGATTTCG (f) 6.25 mM

DQB1*0201-05,*0301-26,*0401-04 (M13, 285 bp) TGTAAAACGACGGCCAGTTATTCCTCGCAGAGGATTTCG (f) 6.25 mM
TGTAAAACGACGGCCAGTCTCGCCGCTGCAAGGTCGT (r) 12.5 mM
DQB5,6 TGTAAAACGACGGCCAGTTATTCCCCGCAGAGGATTTCG (f) 12.5 mM
DQB1*0501-05, 0601-39 (M13, 285 bp) CAGGAAACAGCTATGACCCTCTCCTCTGCAAGATCCC (r) 6.25 mM
CAGGAAACAGCTATGACCCTCTCCTCTGCAGGATCCC (r) 6.25 mM
PCR amplification control (693 bp) TGCCAAGTGGAGCCCCAA (r) 2.5 mM
GCATCTTGCTCTGTGCAGAT (r) 2.5 mM
DQB2 TGTAAAACGACGGCCAGTAGGAAGGCGGATTCCCGAAGA (f) 12.5 mM
DQB1*0201-03 (M13, 584 bp) CAGGAAACAGCTATGACCCTGTCCTCTGGGGTGGAACA (r) 12.5 mM
DQB7 TGTAAAACGACGGCCAGTAGGAAGGCGGATTCCCGAAGA (f) 12.5 mM
DQB1*0301/04 (M13, 591 bp) CAGGAAACAGCTATGACCTGTCCCCTGGGGTGGAACG (r) 12.5 mM
DR1 (M13, 463 bp) TGTAAAACGACGGCCAGTTCCCAGTGCCCGCTCCCT (f) 12.5 mM
CAGGAAACAGCTATGACCACACACTCAGATTCTCCGCTT (r) 12.5 mM
DR3/11/13/14 (M13, 501 bp) TGTAAAACGACGGCCAGTTGGTGGGCGTTGGGGCG (f) 12.5 mM
CAGGAAACAGCTATGACCACACACACACTCAGATTCCCA (r) 12.5 mM
DR4 (M13, 528 bp) TGTAAAACGACGGCCAGTAGCCACTGGGTGCGTGTT (f) 12.5 mM
CAGGAAACAGCTATGACCACACACACACTCAGATTCTCC (r) 12.5 mM
 3

DR10 (M13, 495 bp) TGTAAAACGACGGCCAGTGGCGTTGCGGGTCGGCG (f) 12.5 mM

CAGGAAACAGCTATGACCACACACAGAGTCAGATTCCCA (r) 12.5 mM
DR15/16 (M13, 756 bp) TGTAAAACGACGGCCAGTGGTGGGTGCTGTTGAAGGT (f) 12.5 mM
CAGGAAACAGCTATGACCACACACACACTCAGATTCCCA (r) 12.5 mM
B CG TGTAAAACGACGGCCAGTCGGGGGCGCAGGACCCGG (f) 10 mM
B*13,15,18,27,37,38,39,40,41,44, 45,46,47,49,50,54,55,56, CAGGAAACAGCTATGACCCCATCCCSGGCGAYCTAT (r) 10 mM
57,59,82 (M13, ~1 kb)
B TA TGTAAAACGACGGCCAGTGGCGGGGGCGCAGGACCTGA (f) 5 mM
B*07,08,14,1502,35,42,48,51,52, 53,58,67,78,81 (M13, ~1 kb) CAGGAAACAGCTATGACCCCATCCCSGGCGAYCTAT (r) 5 mM
M13 sequencing primers TGTAAAACGACGGCCAGT (f) 3.2pM
CAGGAAACAGCTATGACC (r) 3.2pM
B gen (B gen, ~1 kb size) GGGTCCCAGTTCTAAAGTCCCCACG (f) 3.33 mM
CCATCCCCGGCGACCTATAGGAGATG (r) 3.33 mM
AGGCCATCCCGGGCGATCTAT (r) 3.33 mM
B gen sequencing primers GGGCGCAGGACCYGRGGA (f) 3.2 pM
GGAGGCCATCCCCGGCGACCTAT (r) 3.2 pM
a
IMGT database 2.28
Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity
37
38 M.D. Varney et al.

2. DQB1 PCR: A 25-mL PCR is prepared for each sample con-

taining 21.5 mL DQB PCR Buffer (DQB234 = 2.5 mM MgCl2,
DQB56 and DQB7 = 1.5 mM MgCl2, DQB2 = 1.2 mM
MgCl2), 1 mL DQB Primer Mix (see Note 3), 1.25 mL Glycerol,
0.2 mL Taq DNA Polymerase (5 U/mL), 1 mL DNA
(100 ng/mL) in a 200-mL PCR strip tube or plate.
3. Seal the tubes/plates and cycle in a PCR machine for 96°C-120 s
×1 cycle, 96°C-10 s, 65°C-60 s ×10 cycles, 96°C-10 s,61°C-50 s,
72°C-30 s ×20 cycles, 20°C hold. Proceed to step 7.
4. DRB1 PCR: A 25-mL PCR is prepared for each sample con-
taining 22.5 mL DRB PCR Buffer (DR4 = 1.5 mM MgCl2,
DR1, DR3/11/13/14, DR10 = 2.0 mM MgCl2), 1 mL DRB
Primer Mix (see Note 3), 0.2 mL AmpliTaq DNA Polymerase
(5 U/mL), 1 mL DNA (100 ng/mL) in a 200-mL PCR strip
tube or plate.
5. Seal the tubes/plates and cycle in a PCR machine for (DR1,
DR15/16, DR4, DR3/11/13/14) 96°C-120 s ×1 cycle, 96°C-
10 s, 71°C-60 s ×15 cycles, 96°C-30 s, 61°C-40 s, 72°C-30 s
×20 cycles, 72°C-10 min, 20°C hold or (DR10) 96°C-120 s ×1
cycle, 96°C-10 s, 65°C-60 s ×10 cycles, 96°C-10 s, 59°C-50 s,
72°C-30 s ×20 cycles, 20°C hold. Proceed to step 8.
6. HLA-B PCR: A 25-mL PCR is prepared for each sample con-
taining 23.9 mL HLA-B PCR Master Mix (see Note 3), 0.1 mL
AmpliTaq DNA Polymerase (5 U/mL), 1 mL DNA (100 ng/mL)
in a 200-mL PCR strip tube or plate.
7. Seal the tubes/plates and cycle in a PCR machine for
96°C-6 min ×1 cycle, 96°C-30 s, 70°C-60 s, 72°C-30 s ×10
cycles, 96°C-30 s, 68°C-60 s, 72°C-30 s ×15 cycles, 96°C-
30 s, 66°C-60 s, 72°C-30 s ×6 cycles, 96°C-25 s, 55°C-60 s,
72°C-120 s ×4 cycles, 72°C-10 min, 1 cycle, 20°C hold.
8. The PCR products may be visualised and approximate size
determined using standard agarose or polyacrylamide gel
electrophoresis.
9. For each PCR product, remove excess dNTPS and primers by
adding 5 mL of PCR product to 5 mL of Exo SAP-IT and incu-
bating at 37°C-15 min, 80°C-15 min, 4°C hold using a PCR
machine. Dilute the cleaned PCR product by adding 30 mL of
sterile water.
10. Cycle sequencing: Use a 96-well PCR plate to aliquot per
sample; 4 mL cleaned PCR product, 10 mL H2O, 4 mL 5×
Sequencing Buffer (Applied Biosystems), 1 mL M13 or B gen
Sequencing Primer, 1 mL Big Dye Terminator (version 3.1,
Applied Biosystems) and cycle for 96°C-10 s ×1 cycle, 96°C-
10 s, 50°C-5 s, 60°C-4 min ×25 cycles, 4°C hold. Pulse centri-
fuge the PCR plates prior to amplification and store the product
at 2–8°C in the dark.
 3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 39

11. Dye Terminator Removal: Prepare a Master Mix containing

50 mL 95% Ethanol and 2 mL 3 M Sodium Acetate per sample
and aliquot 52 mL into each sequencing reaction. Replace the
PCR caps in their original order and invert the plate at least ten
times to mix.
12. Centrifuge the plate for 1 min at approximately 800 × g and
leave at room temperature in the dark for 15 min.
13. Spin the plate at approximately 2,060 × g for 45 min, carefully
remove from the centrifuge, remove the lids, and immediately
discard the supernatant by inverting and flicking the plate onto
an absorbent towel (see Note 4).
14. Place the inverted plate on new absorbent towel in the centri-
fuge centrifuging at approximately 800 × g for 1 min.
15. Remove the plate from the centrifuge and add 200 mL of 70%
ethanol to each pellet. Discard the absorbent towel from the
centrifuge.
16. Recap the plate, invert two to three times to mix and centrifuge
at approximately 2,060 × g for 10 min.
17. Discard the supernatant by removing the caps, inverting and
flicking the plate, and blotting onto an absorbent towel and
centrifuging at approximately 800 × g for 1 min (see Note 4).
18. Repeat steps 14–16.
19. Dry the upright plate in the 70°C oven for at least 30 min and
discard the absorbent towel from the centrifuge.
20. Resuspend in 12 mL Hi-Di formamide, denature at 95°C, cool
on an ice bath prior to loading on the Applied Biosytems 3730
DNA Analyser according to the manufacturers’ instructions.
The analyser separates the dye labelled DNA fragments in a
polymer-filled capillary array by electrophoresis. The fragments
are detected by excitation of the fluorescent dye which is dis-
played as an electropherogram by the data collection software.
Base calling software is then used to assign a sequence.
21. HLA alleles are assigned by comparing the sequence obtained
to a library of reference sequences (http://www.ebi.ac.uk/
imgt/hla/). A number of analysis programs have been
described (10–12) and detailed elsewhere in this volume to
accommodate the HLA reference library and the analysis of
both multiple and heterozygote templates.

3.3. Micro Bead 1. A 10-mL PCR is prepared for each sample containing 6.9 mL
Hybridisation Typing D-mix, 2.0 mL Primer Set, 0.1 mL Taq DNA Polymerase (not
Assay (Luminex/ supplied) and 1 mL DNA (20 ng/mL). Cycle in a Perkin Elmer
Labtype) PCR machine with ramp setting at 9600 for 96°C-180 s ×1
cycle, 96°C-20 s, 60°C-20 s, 72°C-20 s ×5 cycles, 96°C-10 s,
60°C-15 s, 72°C-20 s ×30 cycles, 72°C-10 min, 4°C hold.
40 M.D. Varney et al.

2. The PCR products may be visualised and approximate size

determined using standard agarose or polyacrylamide gel
electrophoresis.
3. Start the Luminex® LABScan™ Analyser according to the
manufacturers’ instructions and set a PCR machine to a 60°C
hold at least 30 min before use.
4. For each sample aliquot 2.5 mL of PCR product and 1.3 mL of
Denaturing Buffer into a 96-well PCR microtitre plate, seal
the plate, pulse centrifuge, vortex gently, and incubate for
10 min at room temperature.
5. Carefully remove the tray seal and add 2.5 mL of Neutralising
Buffer, seal the plate, pulse centrifuge, vortex gently, and place
the tray on crushed ice or a tray cooler. Note the colour change
from pink to yellow (see Note 5).
6. Prepare the hybridisation mixtures by vortexing the required
bead mixtures for 1 min and adding 2 mL of bead mixture to
17 mL of Hybridisation Buffer for each sample being tested.
7. Carefully remove the tray seal, add 19 mL of hybridisation mix-
ture per sample, remove from the ice/cooler, place in a tray
holder, seal, and vortex thoroughly.
8. Remove the tray from the holder and place in a thermocycler
held at 60°C for exactly 15 min.
9. Immediately place in tray holder, remove the seal and add
50 mL Wash Buffer, seal and centrifuge for 5 min at approxi-
mately 1,200 × g.
10. Retaining the plate in the holder, remove the seal and remove
the Wash Buffer by firmly flicking the tray over a sink or wet
waste bin. Maintaining the tray in an inverted position, firmly
blot the tray on an absorbent towel immediately followed by
five gentle taps of the tray on the towel. Return the tray to an
upright position and gently vortex.
11. Repeat steps 9–10 twice for a total of three washes.
12. Prepare sufficient SAPE solution (0.25 mL 100× SAPE stock
and 24.75 mL SAPE buffer per sample). Keep refrigerated and
protected from light until ready to use.
13. Add 25 mL of SAPE solution to each well, seal the tray, vortex,
remove from the holder and place in the thermocycler main-
tained at 60°C for exactly 5 min.
14. Immediately place in tray holder, remove the seal and add
50 mL Wash Buffer, seal and centrifuge for 5 min at approxi-
mately 1,200 × g.
15. Retaining the plate in the holder, remove the seal and remove
the Wash Buffer by firmly flicking the tray and whilst inverted
firmly blot the tray on an absorbent towel followed by five
 3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 41

gentle taps on a clean towel. Return the tray to an upright

position and gently vortex.
16. Add 70 mL of Wash Buffer to each well, gently mix using a
pipette and transfer entire well contents to a 96-well round
bottom reading plate. Store at 2–8°C in the dark until ready
to read.
17. Refer to the manufacturers’ instructions for the calibration,
operation, and data acquisition by the Luminex® LABScan™
Analyser and analysis by the HLA Fusion software.

3.4. Solid Phase 1. Prepare a 45-mL PCR Master Mix for each sample and a nega-
Hybridisation Typing tive control containing 19 mL ultrapure H2O, 5 mL AmpliTaq
(DELFIA Coeliac Gold 10× PCR Buffer, 5 mL dNTP (2 mM), 5 mL DQA1fo
Disease) (10 pml/mL), 5 mL DQA1re (10 pml/mL), 2.5 mL DQB1fo
(10 pml/mL), 2.5 mL DQB1re (10 pml/mL), 1 mL AmpliTaq
Gold per sample (see Note 6).
2. Pipette 45 mL of the Master Mix to each PCR well/tube and
add 5 mL of DNA (optimally 20 ng/mL) and cycle at 95°C-
15 min ×1 cycle, 95°C-50 s, 54°C-60 s, 72°C-60 s ×40 cycles,
72°C-5 min, 4°C hold. The PCR products may be stored at
2–8°C for 2 days or at approximately −20°C for 2–3 weeks
prior to assay.
3. The PCR products may be visualised and approximate size
determined using standard agarose or polyacrylamide gel
electrophoresis.
4. Binding the PCR products: Dilute the Wash Concentrate
(50 + 1,200 mL) and warm half of it in an incubator at 43°C
(see Note 7). Dilute the Bio-Hybridization Control to DELFIA
Hybridization Buffer (0.5 + 49.5 mL).
5. Pipette 10 mL of water to the wells 1–4 (background control
for hybridization), 10 mL of diluted Bio-Hybridization Control
to the wells 5–8, 10 mL of negative control to the wells
9–12,10 mL of each PCR product to 4 wells, 50 mL Hybridization
Buffer to each well mentioned above.
6. Incubate 30 min at room temperature on a plateshaker.
7. Dilute the probes with the Hybridization Buffer, 100 mL is
required per well. Prepare two solutions: A including Europium
(Eu)-DQB1*0302 (1 ng/well) and Terbium(Tb)-DQB1*02
(1 ng/well), and B including Samarium (Sm)-DQA1*05
(3 ng/well), Tb-DQB1 Control (3 ng/well), and Eu-DQA1
Control (0.5 ng/well). The dilutions must be kept on ice and
used within 30 min. The probes must be on ice and returned
to a freezer as quick as possible (see Note 8).
8. Denaturation and hybridization: Wash the plate four times
with 400 mL of prewarmed washing dilution using the auto-
matic washer.
42 M.D. Varney et al.

9. Add 150 mL of NaOH solution to each well and incubate for

5 min at room temperature on a plateshaker.
10. Wash the plate four times with 400 mL of prewarmed washing
dilution using the automatic washer.
11. Pipette 100 mL of solution A to the wells 1–2, 5–6, 9–10 and
100 mL of solution B to wells 3–4, 7–8, 11–12, etc. There are
four replicates for each sample on a plate.
12. Seal the plate and incubate for 2 h (±10 min) at 37°C.
13. Wash the plate six times with 400 mL of prewarmed washing
dilution using the automatic washer.
14. Quantitation: Add 200 mL of Enhancement Solution at room
temperature (see Note 9) and incubate for 30 min at room
temperature on a plateshaker (see Note 10).
15. Measure the Eu and Sm labels by a time-resolved fluorometer.
16. Add 50 mL of Enhancement Solution to each well (see Note 9)
and incubate for 5 min at room temperature on a plateshaker.
17. Measure the Tb label using a Time-resolved fluorometer (see
Note 11).
18. The probes detect alleles DQB1*0302 (also *0304-5 and
*0307-8), DQB1*02 (specific for *0201-03), and DQA1*05
(specific for *0501-05). Homozygotes and heterozygotes
cannot be separated.

3.5. TaqMan® Typing 1. For a 384-well microplate, a 4-mL PCR is prepared for each
of HLA-Tagging SNPs sample containing 2.0 mL TaqMan® (2×) Genotyping Master
in Coeliac Disease Mix (Applied Biosystems), 0.2 mL TaqMan® (20×) SNP
Genotyping Assay mix and 1.8 mL DNA (10 ng/reaction)
suspended in PCR-quality H2O. Prepare a pre-mix of the
Genotyping Master Mix and SNP Genotyping Assay mix for
the required number of samples and aliquot 2.2 mL to each
well of an empty plate followed by 1.8 mL of DNA using a
multichannel pipette. Alternatively, first transfer the DNA
using a pipette robot into the wells of six identical 384 plates
and let the DNA dry into the bottom of the wells at room
temperature or 37°C. After drying, the plates can be stored
covered with lid at room temperature. To the plates add the
pre-mix, described above but including 1.8 mL PCR-quality
H2O instead of DNA, using a multichannel or stepper pipette.
To reduce reagent use, the pre-mix may be modified to 1.5 mL
Genotyping Master Mix, 0.12 mL SNP Genotyping Assay mix,
and 0.58 mL H2O (or 2.38 mL H2O if plate dried DNA is used).
This may also improve the amplification of some of the SNPs
(see Note 12).
3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 43

2. Leave at least two wells as negative controls with pre-mix but

no DNA. The computer program is designed to use these two
values to exclude background. They also act as a quality con-
trol for the Master Mix, SNP Assay mix and water for contami-
nation. Seal the plate and pulse centrifuge to bring the fluid
down. Start ABI 7900HT real-time PCR cycler and complete the
assay and well information data according to the instructions
of the system.
3. Perform a PCR run of 10 min at 95°C (on the first cycle only),
followed by at least 50 cycles of 15 s at 92°C and 1 min at
60°C (see Note 13). Perform these amplification steps with
program SDS Standard Curve (Absolute Quantification, AQ).
After the plate is ready, run an end-point analysis using pro-
gram SDS Allelic Discrimination (AD) to obtain allele assign-
ment. The results for each sample are listed as “allele 1”
(homozygous FAM), “allele 2” (homozygous VIC), or “both”
(heterozygote). Check from the SNP assay data sheet which
allele of the SNP is labelled with FAM, and which one with VIC.
4. Check also the X-Y-plot of the signal intensities from all 384
samples together to double check that three well separated
genotype clusters and the negative control cluster are seen.
Samples with unclear clustering are removed from the analysis
and may require re-examination of the real-time amplification
curve and/or classical HLA typing. For the DQ7 tagging, SNP
rs4639334 four genotype clusters instead of three are some-
times seen. This is likely to be caused by additional SNPs within
the probe or primer target sequence carried by some donors.
Attention should be paid to allele calling in addition to checking
Hardy-Weinberg equilibrium of the observed genotypes.
5. Export the results from Allelic Discrimination (AD) file into
text file. Once all six SNP results are complete, order the
detected genotypes of each SNP into Excel for interpretation
of the predicted HLA-haplotypes (Table 2).
6. For DQ2.5, DQ8, and DQ7, there is only one SNP for each,
and hetero/homozygosity of the predicted SNP alleles usually
predicts hetero/homozygosity of these HLA haplotypes. For
DQ2.2, one needs the predictive alleles of all three tagging
SNP alleles for best prediction value. Report the interpreted
HLA types e.g. DQ2.5/DQ8, DQ2.2/DQ7, DQ2.5/DQ2.5,
DQ2.5/DQX, DQX/DQX, etc. where DQX means any other
allele than the detected DQ2.5, DQ8, DQ7, or DQ2.2.
7. For some rare haplotypes, there might be more than two DQ
alleles suggested by the tagging SNPs, and in such cases classical
DQ (and possibly also DR typing) is recommended. Exclusion
of false negative results by traditional HLA typing is also
recommended for confirmed celiac disease patients for which
44 M.D. Varney et al.

the tagging SNP method does not predict any DQ2/DQ8

risk haplotypes. Discrepancies between the tagging SNP method
and classical HLA-DQ typing may be resolved with reference
to the published reference tables (8, 9).

4. Notes

1. SSP primer mixes (Table 3) have been designed locally or are

drawn from existing literature (13–17).
2. DNA. With the exception of the whole blood SSP method,
genomic DNA is used in each method described. A number of
published and commercial DNA extractions methods have
been described. Heparin is typically avoided as an anti-coagu-
lant as it may inhibit the PCR and must not be used in the
HLA-B*27 and B*57 whole blood protocols. The purity and
concentration of the DNA can be critical in PCR-based
assays.
3. Primer sequences (Table 4) have been selected from previous
publications with minor modifications or derived locally based
on published sequences (18–23). The DQB*02/03/04 and
*05/06 primer mixes contain a PCR amplification control.
4. Be careful not to discard the pellets, if the supernatant cannot
be discarded immediately re-spin the plate for 2 min before
discarding.
5. The colour must be yellow to proceed; if it remains pink add
an additional 1 mL of Neutralising Buffer until yellow.
6. The PCR primer sequences for the celiac disease assay are:
DQA1fo Bio TATGGTGTAAACTTGTACCAGT, DQA1re
GGTAGCAGCGGTAGAGTTG, DQB1fo GCATGTGCTA
CTTCACCAACG, DQB1re Bio-CCTTCTGGCTGTTCCA
GTACT. The Biotin labelled primers must be HPLC purified,
the unlabelled primers may be desalt purified. Dilute the
primers to 100 pmol/mL in Tris buffer to store at −20°C or
to 10 pmol/mL in Tris buffer for use (maximum storage is 3
months at −20°C).
7. Diluted Wash Concentrate can be stored for a maximum of
2 weeks.
8. Empty tubes for the dilutions should be cooled down on ice
beforehand.
9. It is recommended to use a separate pipette because of high
contamination risk.
10. It is recommended not to seal the plate because a plastic seal
easily spreads liquids from well to well cross-contaminating
samples.
 3 Methods for Diagnostic HLA Typing in Disease Association and Drug Hypersensitivity 45

11. The hybridization control must be over 100 cps, the DQA1
control over 700 cps, and the DQB1 control over 60 cps.
Washes must be done very carefully as background signal can
easily interfere with quantification.
12. In many populations, over 90% of patients with coeliac disease
are positive for the DQA1*05-DQB1*02 haplotype. Whenever
possible, it is therefore advisable to have both patients and
controls within each genotyped 384-well plate to provide a
good cluster definition of the three genotypes in the analysis
step.
13. A long number of cycles may be required to see a clear plateau
phase when the SNPs are not working optimally due to the
difficult template sequence.

Acknowledgements

The DR/DQ SSP, SBT, and micro bead hybridisation methods are
a culmination of input by the molecular typing staff of the
Transplantation and Immunogenetics Service of the Australian
Red Cross Blood Service.

References

1. Warrens AN (2000) Statistical considerations in 7. Lind C, Ferriola D, Mackiewicz K, Heron S,

analyzing HLA and disease associations. In: Rogers M, Slavich L, Walker R, Hsiao T,
Lechler R, Warrens A (eds) HLA in health and McLaughlin L, D’Arcy M, Gai X, Goodridge
disease, 2nd edn. Academic, London D, Sayer D, Monos D (2010) Next-generation
2. http://www.dorak.info/hla/stat.html sequencing: the solution for high-resolution,
3. Brown MA (2009) Genetics and the pathogen- unambiguous human leukocyte antigen typing.
esis of ankylosing spondylitis. Curr Opin Hum Immunol 71:1033–1042
Rheumatol 21:318–323 8. Monsuur AJ, de Bakker PI, Zhernakova A,
4. Wu DY, Ugozzoli L, Pal BK, Wallace RB Pinto D, Verduijn W, Romanos J, Auricchio R,
(1989) Allele-specific enzymatic amplification Lopez A, van Heel DA, Crusius JB, Wijmenga
of beta-globin genomic DNA for diagnosis of C (2008) Effective detection of human leuko-
sickle cell anemia. Proc Natl Acad Sci USA cyte antigen risk alleles in celiac disease using
86:2757–2760 tag single nucleotide polymorphisms. PLoS
5. Olerup O, Zetterquist H (1992) HLA-DR One 3:e2270
typing by PCR amplification with sequence- 9. Koskinen L, Romanos J, Kaukinen K, Mustalahti
specific primers (PCR-SSP) in 2 hours: an K, Korponay-Szabo I, Barisani D, Bardella MT,
alternative to serological DR typing in clinical Ziberna F, Vatta S, Széles G, Pocsai Z, Karell K,
practice including donor-recipient matching Haimila K, Adány R, Not T, Ventura A, Mäki M,
in cadaveric transplantation. Tissue Antigens Partanen J, Wijmenga C, Saavalainen P (2009)
39:225–235 Cost-effective HLA typing with tagging SNPs
6. Sayer DC, Whidborne R, De Santis D, predicts celiac disease risk haplotypes in the
Rozemuller EH, Christiansen FT, Tilanus MG Finnish, Hungarian, and Italian populations.
(2004) A multicenter international evaluation of Immunogenetics 61:247–256
single-tube amplification protocols for sequenc- 10. Helmberg W, Lanzer G, Zahn R, Weinmayr B,
ing-based typing of HLA-DRB1 and HLA- Wagner T, Albert E (1998) Virtual DNA
DRB3,4,5. Tissue Antigens 63:412–423 analysis—a new tool for combination and
46 M.D. Varney et al.

standardised evaluation of SSO, SSP and induced cutaneous reactions in Han Chinese
sequencing-based typing results. Tissue [Erratum in: epilepsia 2008]. Epilepsia
Antigens 51:587–592 48:1015–1018
11. Sayer DC, Goodridge DM, Christiansen FT 18. Kotsch K, Wehling J, Blasczyk R (1999)
(2004) Assign 2.0: software for the analysis of Sequencing of HLA class II genes based on the
Phred quality values for quality control of HLA conserved diversity of the non-coding regions:
sequencing-based typing. Tissue Antigens sequencing based typing of HLA-DRB genes.
64:556–565 Tissue Antigens 53:486–497
12. Rozemuller EH, Ploeger LS, Mulder W (2008) 19. Voorter CE, Kik MC, van den Berg-Loonen
SBTengine; the ultimate solution for high reso- EM (1998) High-resolution HLA typing for
lution HLA typing. Hum Immunol 69:131 the DQB1 gene by sequence-based typing.
13. Olerup O, Aldener A, Fogdell A (1993) HLA- Tissue Antigens 51:80–87
DQB1 and -DQA1 typing by PCR amplification 20. Van Dijk A, Melchers R, Tilanus M, Rozemuller
with sequence-specific primers (PCR-SSP) in 2 E (2007) HLA-DQB1 sequencing-based typing
hours. Tissue Antigens 41:119–134 updated. Tissue Antigens 69(Suppl 1):
14. Olerup O (1994) HLA-B*27 typing by group 64–65
specific PCR amplification. Tissue Antigens 21. Dunn PP, Day S, Williams S, Bendukidze N
43:253–256 (2005) HLA-DQB1 sequencing-based typing
15. Sayer DC, Cassell HS, Christiansen FT (1999) using newly identified conserved nucleotide
HLA-B*27 typing by sequence specific sequences in introns 1 and 2. Tissue Antigens
amplification without DNA extraction. Mol 66:99–106
Pathol 52(5):300–301 22. Tilanus MGJ (ed) (2000) IHWG technical
16. Martin AM, Nolan D, Mallal S (2005) HLA- manual. Genomic analysis of the human MHC.
B*5701 typing by sequence-specific DNA based typing for HLA Alleles & linked
amplification: validation and comparison polymorphisms. Fred Hutchinson Cancer
with sequence-based typing. Tissue Antigens Research Centre, Seattle, USA
65:571–574 23. Cereb N, Yang SY (1997) Dimorphic primers
17. Man CB, Kwan P, Baum L, Yu E, Lau KM, derived from intron 1 for use in the molecular
Cheng AS, Ng MH (2007) Association between typing of HLA-B alleles. Tissue Antigens
HLA-B*1502 allele and antiepileptic drug- 50:74–76
 Chapter 4

HLA Typing Using Bead-Based Methods

Daniel Trajanoski and Samantha J. Fidler

Abstract
The LABType® SSO (One Lambda, Inc) and Gen-Probe LIFECODES HLA-SSO HLA Typing tests are
rapid and efficient assays for determining human leukocyte antigens (HLAs). The principle of these assays
involves the hybridization of reverse sequence-specific oligonucleotide probes each attached to a unique
colour coded microsphere to identify HLA class I and class II alleles. Target DNA is polymerase chain
reaction (PCR) amplified using group-specific primers and then biotinylated which allows it to be detected
using R-Phycoerythrin-conjugated Streptavidin. The PCR product is then denatured and allowed to
hybridise to complementary DNA probes conjugated to fluorescently code microsperes. The Luminex®
Flow Analyser achieves detection by determining the fluorescent intensity of PE on each microsphere. The
assignment of HLA alleles is based on the reaction pattern of the various beads compared to patterns with
known HLA alleles.

Key words: LABType® SSO, Luminex®, Denaturation, Gen-Probe LIFECODES HLA-SSO,

Hybridisation, R-Phycoerythrin-conjugated streptavidin, Polymerase chain reaction, Human leuko-
cyte antigen, Microsphere

1. Introduction

Over the past few decades, human leukocyte antigen (HLA) typing
methods have utilized advances in technology to provide assays
with higher resolution whilst reducing cost and time. Historically,
HLA typing was first performed by serological typing of HLA anti-
gens using antisera with defined specificities (1, 2), but this proce-
dure has largely been replaced by DNA typing methods. The
polymerase chain reaction (PCR) assay developed by Saiki et al. in
1985 (3) was a major breakthrough technology for HLA genotyp-
ing, and methods using PCR-amplified DNA are now widely used
by HLA laboratories throughout the world. Currently there are
numerous DNA-based methods available that can detect HLA
alleles from PCR-amplified DNA, such as restriction fragment

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_4, © Springer Science+Business Media New York 2012

47
48 D. Trajanoski and S.J. Fidler

length polymorphism (RFLPs) (4), direct sequencing (5), sequence-

specific primers (SSPs) (6), and many others. Several of these are
described elsewhere in this volume. However, all these methods
have their limitations in one way or another in terms of speed, lack
of resolution, low throughput of sample number, and high cost.
A novel method has been developed by combining PCR-
amplified DNA and sequence-specific oligonucleotide probe (SSO)
protocols with the Luminex® xMAP® technology.
The LABType® SSO (One Lambda, Inc) and Gen-Probe
LIFECODES SSO HLA Typing Assays use SSOs bound to
fluorescently coded microspheres to identify alleles encoded by the
sample DNA. The principle involves adapting Luminex® xMAP® tech-
nology to the reverse SSO DNA typing method. Initially target
genomic DNA is PCR amplified using a group-specific primer. The
PCR product is biotinylated, which allows it to be detected using
R-Phycoerythrin-conjugated Streptavidin (SAPE). The PCR product
is denatured and allowed to rehybridise to complementary DNA
probes conjugated to fluorescently coded microspheres. After wash-
ing the beads, bound amplified DNA from the test sample is tagged
with SAPE. Using the Luminex® Flow Analyser, the amount of PE on
each specifically identified microsphere is detected. The assignment of
the HLA typing is based on the pattern of reactivity of each bead
compared to patterns associated with published gene sequences.

2. Materials

2.1. LABType SSO (One 1. Luminex®100 or 200 Flow Analyser.

Lambda) 2. Luminex® XY Platform.
2.1.1. Instrument 3. Centrifuge
Requirements – Mikro 22 Hettich Zentrifugen or equivalent for 1.5-mL
microtube.
– Sigma Laboratory Centrifuge 4–15 or equivalent for
96-well microplate.
4. Vortex mixer with adjustable speed.
5. Thermal cycler—ABI 9700 or Equivalent (see Note 1).
6. Electrophoresis apparatus.
7. UV transiluminator with photographic image software.
8. Freezer (−20 to −80°C).
9. Refrigerator (4–8°C).

2.1.2. Materials and 1. Genomic DNA—patient samples and controls (see Note 2).
Reagent Requirements
2. Deionised water.
3. 70% Ethanol.
 4 HLA Typing Using Bead-Based Methods 49

4. 20% Chlorine Bleach solution or equivalent.

5. Luminex Sheath Fluid.
6. Luminex Calibrator beads.
7. Luminex Control beads.
8. AmpliTaq polymerase 1,000 U, 5 U/μL (ABI Part# N808-
0101) (see Note 3).
9. Isorack (Eppendorf).
10. Crushed Ice bath.
11. 1.5-mL Microtube (Eppendorf or equivalent).
12. 5-, 20-, and 50-mL Disposable tubes (for making dilutions).
13. Disposable pipette tips (10, 20, 100, 200, 1,000 μL).
14. PCR pad.
15. Timers.
16. 10-, 20-, 100-, 200-, and 1,000-μL pipette–Gilson or equivalent.
17. PCR 96-well microplates.
18. 96-well, 250-μL V-bottom, white polystryene microplates.
19. Aluminium foil.
20. Adhesive Tray seal (OLI Cat# SSPSEA300) or equivalent.
21. PCR plate cap strips.
22. Wash Troughs/Trays.
23. 1% Seakem ME agarose.
24. Electrophoresis loading Dye—6× DNA loading dye (Fermentas)
or equivalent.
25. DNA lambda ladder—Generuler 100 bp DNA Ladder
(Fermentas) or equivalent (see Note 4).
26. 1× TBE buffer.
27. 5 mg/mL Ethidium Bromide.
28. SAPE (One Lambda, Inc).
29. One Lambda SSO Genotyping Kit (see Note 5) containing
– LABType® SSO Bead Mixture (Pre-optimised and tested
mixture of microspheres with covalently attached probes).
– LABType® SSO Hybridisation Buffer.
– LABType® SSO Wash buffer.
– LABType® SSO Denaturation Buffer.
– LABType® SSO Neutralisation Buffer.
– LABType® SSO Primer Set (Pre-optimised HLA loci-specific
primer mix).
– LABType® SSO D-Mix.
– LABType® SAPE Buffer.
50 D. Trajanoski and S.J. Fidler

2.2. Gen-Probe Refer to Subheading 2.1.1.

LIFECODES SSO
2.2.1. Instrument
Requirements

2.2.2. Materials and The same as Subheading 2.1.2 except replace One Lambda SSO
Reagent Requirements Genotyping kit with:
● Gen-Probe LIFECODES HLA-SSO Typing Kit (see Note 6)
containing
– LifeCodes HLA master mix.
– LifeCodes Probe Mix.
– Dilution Solution.
– SAPE.

3. Methods

The methods in this chapter have been written according to cur-

rent manufacturer’s protocols. Any changes to these methods such
as volumes of reagents and thermal cycler conditions can be found
in the kit insert and the appropriate websites (http://www.
onelambda.com and http://www.gen-probe.com). These should
also be checked following any lot change.

3.1. LABType® SSO 1. Create a LABType® PCR program on your thermal cycler as
HLA Typing Method shown in Table 1.
3.1.1. DNA Amplification

Table 1
LABType® SSO polymerase chain reaction (PCR) program

Step Temperature and incubation time Number of cycles

Step 1 96°C for 03:00 min 1
Step 2 96°C for 00:20 min 5
60°C for 00:20 min
72°C for 00:20 min
Step 3 96°C for 00:10 min 30
60°C for 00:15 min
72°C for 00:20 min
Step 4 72°C for 10:00 min 1
Step 5 4°C forever 1
 4 HLA Typing Using Bead-Based Methods 51

Table 2
LABType® amplification mixture

Amplification AmpliTaq
Number of reactions D-mix (mL)a primer (mL)a polymerase (mL)
1 13.8 4 0.2
10 138.0 40 2.0
50 690.0 200 10.0
96 1,491.0 432 21.6
a
As provided in the One Lambda SSO genotyping kit

2. Turn on thermal cycler to keep heated lid warm.

3. Select the appropriate HLA typing kit (locus and resolution)
and thaw out amplification primers and D-mix (see Note 7).
Make sure to keep on ice until ready to use (see Note 8).
4. Vortex D-mix and amplification primer for at least 15 s and
perform a quick centrifuge (5 s) to spin down any liquid trapped
under lid.
5. Create the amplification mixture by adding correct volumes of
D-mix and amplification primers in a 1.5-mL Eppendorf tube
according to Table 2 (see Note 9).
6. Pipette 2 μL of each DNA template (see Note 10) into a
96-well PCR microplate.
7. Add the correct amount of AmpliTaq Polymerase (see Table 2)
to the amplification mixture and vortex for 5s (see Note 11).
8. Aliquot 18 μL of amplification mixture into each well contain-
ing DNA (see Note 12).
9. Cap or seal the plate and give a quick vortex to mix contents in
plate.
10. Place PCR plate in thermal cycler and start the “LABType®
PCR program” shown in Table 1.
11. After the “LABType® PCR program” is finished, remove PCR
plate and store in a 4°C fridge.
12. Pour agarose in a gel tray with the correct numbers of combs
needed to electrophorese all the PCR products. (It is recom-
52 D. Trajanoski and S.J. Fidler

mended you electrophorese all samples to confirm an

amplification product band is present.)
13. Once the agarose gel is set, place gel tray in electrophoresis
tank and pour 1× TBE buffer until agarose is fully immersed in
the 1× TBE buffer.
14. Mix 5 μL of DNA lambda ladder and 5 μL of loading dye
and pipette into first well of each row of the agarose gel (see
Note 13).
15. In a separate clean PCR plate, mix 5 μL of PCR product and
5 μL of loading dye and pipette into remaining empty wells.
Repeat this for all the samples.
16. Place the cover on the electrophoresis tank and run at 150 V
for 20 min.
17. Take out gel and place on UV transilluminator to inspect PCR
products. Save photo and check all samples have correctly
amplified (see Note 14).

3.1.2. Denaturation/ 1. Turn on thermal cycler and run program to 60°C hold. Make
Neutralisation sure lid is heated to optimum temperature before use.
2. Remove reagents listed in Tables 3 and 4 and supplied in the
SSO genotyping kit from −80°C freezer to room temperature
to thaw and then prepare all reagent volumes in an appropri-

Table 3
LABType® reagent volumes

Denaturation Neutralisation Hybridisation Wash Bead mixture

Number of testsa buffer (mL)b buffer (mL)b buffer (mL)b buffer (mL)b (mL)b

1 2.5 5 34 480 4
10 25 50 340 4,800 40
20 50 100 680 9,600 80
50 125 250 1,700 24,000 200
96 240 480 3,264 46,080 384
a
Number of tests includes all samples to be tested including appropriate controls and extra volume (see Notes 2 and 9
for comments)
b
As provided in the One Lambda SSO genotyping kit
 4 HLA Typing Using Bead-Based Methods 53

Table 4
LABType® R-Phycoerythrin-conjugated Streptavidin
(SAPE) and SAPE buffer volumes

Number of tests SAPE stock volume (mL) SAPE buffer volume (mL)a

1 0.5 49.5
10 5.0 495.0
20 10.0 990.0
50 25.0 2,475.0
96 48.0 4,752.0
a
As provided in the One Lambda SSO genotyping kit

ately sized disposable tube as necessary according to Tables 3

and 4 (see Notes 8 and 15).
3. Return any unused reagents to 2–8°C (do not re-freeze bead
mixture after thawing).
4. Prepare a crushed ice bath.
5. Transfer 5 μL of each amplified DNA sample into the corre-
sponding well of a clean 96-well PCR plate (see Note 16).
6. Add 2.5 μL Denaturation Buffer to each well of the PCR plate
and mix thoroughly by pipetting up and down. Incubate at
room temperature (20–24°C) for 10 min.
7. Add 5 μL Neutralisation Buffer with pipette to each well of the
PCR plate and mix thoroughly (see Note 17).
8. Place PCR plate on ice bath or 96-well ice rack.

3.1.3. Hybridisation 1. Prepare Hybridisation Mixture by combining appropriate vol-

umes of Bead Mixture and Hybridisation buffer according to
Table 3 (see Note 18).
2. Add 38 μL hybridisation mixture to each well.
3. Cover with tray seal and give the PCR plate a quick mix using
the vortex (see Note 19).
4. Place PCR plate into the pre-heated (60°C) thermal cycler,
cover with PCR pad, and close and tighten lid. Incubate for
15 min.
5. Remove the PCR plate seal and quickly add 100 μL wash buf-
fer to each well (see Note 20).
6. Cover PCR plate with plate seal and centrifuge plate for 5 min
at 1,000–1,300 × g.
54 D. Trajanoski and S.J. Fidler

7. Remove the tray seal, then remove the wash buffer by

dumping/flicking plate upside down (see Note 21). Gently pat
on absorbant paper three times to remove any residual liquid.
8. Repeat steps 4–6 two more times for a total of three wash
steps.
9. Prepare the 1× SAPE solution at the third centrifugation stage
according to Table 4.
3.1.4. Labelling
1. Add 50 μL of 1× SAPE solution to each well of the PCR
plate.
2. Cover plate with seal and vortex thoroughly at low speed.
3. Place PCR plate into the pre-heated (60°C) thermal cycler,
cover with PCR pad, and close lid. Incubate for 5 min.
4. Remove the PCR plate seal and quickly add 100 μL wash buf-
fer to each well (see Note 20).
5. Cover PCR plate with plate seal and centrifuge plate for 5 min
at 1,000–1,300 × g.
6. Remove wash buffer by dumping/flicking plate upside down
(see Note 12).
7. Add 70 μL wash buffer to each well and gently mix by
pipetting.
8. Transfer contents of each well to a 96-well V-bottom microplate
(reading plate) using an 8-channel pipette. To avoid contami-
nation use fresh pipette tips each time.
9. Cover reading plate with seal and aluminium foil. Keep in dark
and at 4°C until ready to read using LABScan™ Flow Analyser
(see Note 22).

3.1.5. Sample Acquisition 1. Turn on the system and set up the Luminex® Flow Analyser for
on Luminex® Flow sample acquisition and calibration. (Further details of how to
Analyser use the Luminex® Flow Analyser can be found in the Luminex®
user’s manual.)
2. Verify the levels of sheath fluid and waste fluid (see Note 23).
3. Perform Warm up by selecting “Warm Up” button in the
Luminex opening system. The warm up requires no plate or
solution.
4. Prime the instrument by selecting “Prime”. Again this step
requires no plate or solution.
5. Perform a 70% alcohol flush, followed by four washes with
sheath fluid.
6. A calibration should be performed every time the instrument
is turned on or if the temperature of the instrument has
increased or decreased by more than 3°C from the time of
calibration.
 4 HLA Typing Using Bead-Based Methods 55

7. To calibrate instrument, vortex calibration (CAL1 and 2) and

control (CON1 and 2) beads thoroughly (see Note 24).
8. Place four drops of the bead solutions in each of four wells
(e.g. A1, B1, C1, and D1) in the following order: CAL1,
CAL2, CON1, and CON2.
9. In the acquisition software, select “CAL1” indicating the location
of the bead solution. The instrument will then perform the
calibration for CAL1. If the calibration is successful, this will
be indicated in the “diagnostic” tab.
10. Repeat steps 8 and 9 for CAL2, CON1, and CON2 (see
Note 25).
11. Once calibration has been performed, perform four washes.
12. Select the correct template according to the product catalogue
and lot number being used (see Note 26).
13. Create a filename for the samples to be run, for example
DATE_LOCUS_USERINITIALS.
14. Check all template settings are correct (see Note 27).
15. Load the 96-well reading plate from Subheading 3.1.4 step 9
into instrument and select “Start”.
16. Once run is finished, wash three times with sheath fluid, sani-
tise once with 20% bleach, wash three times with distilled water
and soak with distilled water twice.
17. Shut down Luminex® instrument.

3.1.6. Data Analysis: One 1. Data analysis for the HLA Class I and II assays are performed
Lambda HLA Fusion™ with the software package HLA Fusion™. The LABType®
analysis feature of the program analyses Luminex .csv output
files and is based on catalogue specifications provided with the
software. Import the data files as specified by the manufacturer
(see Note 28).
2. The analysis software establishes the HLA type by firstly cate-
gorising each of the beads via its internal fluorescence, then
determining whether a probe is bound or not by the strength of
external fluorescence on that bead produced by the SAPE. For a
particular sample, the bead/probe fluorescence for each of the
beads bearing a particular probe is visualised in the lower left-
hand quadrant of the software (Fig. 1) as a bar chart. Any probe
can be selected from this bar chart for further scrutiny. The
upper left quadrant shows the manufacturer’s QC panel reactiv-
ity with the selected probe. In this example, the QC for probe
A107 (Bead 512) is shown (see Note 29). The upper right
quadrant shows the probe reactivity for all samples in the run
being analysed (see Note 30). The lower right quadrant shows
the software analysis derived HLA type of the sample based on
the positive and negative probe reactions (see Note 31).
56 D. Trajanoski and S.J. Fidler

Fig. 1. HLA-Fusion® analysis software. Quadrant 1 shows the bead/probe fluorescence for each of the beads bearing a
particular probe. Any probe can be selected from this bar chart for further scrutiny. Quadrant 2 shows the manufacturer’s
QC panel reactivity with the selected probe (see Note 29). Quadrant 3 shows the probe reactivity for all samples in the run
being analysed (see Note 30). Quadrant 4 shows the software analysis derived HLA type of the sample based on the
positive and negative probe reactions (see Note 31). Courtesy of One Lambda, Inc. Used with permission.

Table 5
Gen-Probe LIFECODES components for amplification mix

Component Volume per PCR sample reaction

LIFECODES master mixa 15 μL

Genomic DNA 4 μL (see Note 34)
Taq polymerase 0.5 μL (2.5 U)
Water 30.5 μL
a
As provided in the Gen-Probe SSO genotyping kit

3. The reporting function in the Fusion™ software allows result-

ing via a printed hard-copy report, or export of results via a .
csv file (see Note 32).

3.2. Gen-Probe 1. Select the appropriate SSO kit for the HLA locus to be typed.
LIFECODES HLA-SSO 2. Allow the master mix (containing amplification primers) to
Typing Method thaw to room temperature (see Note 8).
3.2.1. DNA Amplification 3. Using Table 5, prepare the components for amplification for
n + 2 reactions using the indicated amount of each component
per reaction (except for DNA) (see Note 33).
 4 HLA Typing Using Bead-Based Methods 57

Table 6
Gen-Probe LIFECODES thermal cycler conditions
for amplification

Step Temperature and incubation time Number of cycles

Step 1 95°C for 05:00 min 1

Step 2 95°C for 00:30 min 8
60°C for 00:45 min
72°C for 00:45 min
Step 3 95°C for 00:30 min 32
63°C for 00:45 min
72°C for 00:45 min
Step 4 72°C for 15:00 min 1
Step 5 4°C forever 1

4. Bring to a final volume of 50 μL (minus the volume of DNA

template) per reaction with nuclease free water. Gently vortex,
spin down, and place on ice (see Note 9).
5. Pipette the appropriate amount of DNA template into the
PCR tubes (see Note 34).
6. Aliquot the amplification mix into the PCR tubes containing
the DNA. (The total volume of amplification mix and DNA
template should equal 50 μL for each sample reaction.)
7. Cap or seal wells tightly to prevent evaporation during PCR.
8. Place samples in the thermal cycler and run program (see
Table 6).
9. After the program is finished, remove PCR plate and store in a
4-C fridge.
10. Pour agarose in a gel tray with the correct number of combs
needed to electrophorese all the PCR products. (It is recom-
mended you electrophorese all samples to confirm amplification
product band is present).
11. Once the agarose gel is set, place gel tray in electrophoresis
tank and pour 1× TBE buffer until agarose is fully immersed in
the 1× TBE buffer.
12. Mix 5 μL of DNA lambda ladder and 5 μL of loading dye
and pipette into first well of each row of the agarose gel (see
Note 12).
13. In a clean PCR plate, mix 5 μL of PCR product and 5 μL of
loading dye and pipette into the remaining empty wells of the
agarose gel. Repeat this for all the samples.
58 D. Trajanoski and S.J. Fidler

14. Place the cover on the electrophoresis tank and run at 150 V
for 20 min. Take out gel and place on UV transilluminator to
inspect PCR products. Save photo and check all samples have
correctly amplified (see Note 35).

3.2.2. Hybridisation 1. Warm probe mix in a 55–60°C heat block for at least 5–10 min
to thoroughly solubilise components in probe mixture.
2. Vortex briefly for about 15 s to thoroughly suspend the beads
(see Note 18).
3. Aliquot 5 μL of locus-specific PCR product from Subheading
3.2.1 step 7 into a 96-well PCR plate (see Note 12).
4. Aliquot 15 μL of corresponding probe mix for the locus-
specific product into each well. When aliquoting probe mix to
more than 10 wells, gently vortex probe mix after each set of
ten. Seal plate with film.
5. Hybridize samples under the incubation conditions indicated
in Table 7.
6. While the samples are hybridizing, prepare a 1:200 Dilution
Solution/SAPE mixture. Combine 170 μL Dilution Solution
(DS) and 0.85 μL 1 mg/mL SAPE per sample (follow Table 8
and see Notes 15, 36, and 37).

Table 7
Gen-Probe LIFECODES thermal cycler conditions |for
hybridization

Step Temperature and incubation time Number of cycles

Step 1 95°C for 05:00 min 1

Step 2 47°C for 30:00 min 1
Step 3 56°C for 10:00 min 1
Step 4 56°C HOLD

Table 8
Gen-Probe LIFECODES SAPE and dilution solution volumes

Number of samples Dilution solution (DS) mLa SAPE stock volume (mL)b

1 170 0.85
5 850 4.25
10 1,700 8.5
20 3,400 17
50 8,500 42.5
a
As provided in the Gen-Probe SSO genotyping kit
 4 HLA Typing Using Bead-Based Methods 59

7. At the 56°C hold step (i.e. step 4, Table 7), while the tray is on
the thermal cycler, dilute each sample with 170 μL of the
prepared DS/SAPE mixture. It is critical to dilute all samples
within 5 min (following the 10 min 56°C step i.e. step 3,
Table 7) (see Note 15).
8. Transfer contents of each well to a 96-well V-bottom microplate
(reading plate) using an 8-channel pipette. To avoid contami-
nation use fresh pipette tips each time.
NOTE: The Gen-Probe LIFECODES HLA-SSO assay requires
no centrifugation or wash steps.
9. Cover reading plate with seal and aluminium foil. Keep in dark
and at 4°C until ready to read using LABScan™ 100 Flow
Analyser (see Note 22).

3.2.3. Sample Acquisition The procedure for sample acquisition for the Gen-Probe
on Luminex® Flow LIFECODES SSO method is the same as that for the ONE
Analyser LAMBDA method. Follow Subheading 3.1.5 steps 1–17.

3.2.4. Data Analysis-Gen- 1. Data analysis for the HLA Class I and II assay is performed
Probe LifeMatch Quick- with the software package LifeMatch™. The Quick-Type™
Type™ analysis feature of the program analyses Luminex .csv output
files and is based on catalogue specifications provided with the
software. Import the data files as specified by the manufacturer
(see Note 38).
2. The analysis software establishes the HLA type by firstly cate-
gorising each of the beads via its internal fluorescence, then
determining whether a probe is bound or not by the strength
of external fluorescence on that bead produced by the SAPE.
The probe reactivity thresholds and probe assignments for an
individual sample are shown in the table in section A (Fig. 2).
In the example shown in Fig. 2, for probe 331 the positive and
negative thresholds are 0.624 and 0.511, respectively. For the
sample tested, for probe 331 the fluorescence observed is 0.992
which is greater than the positive threshold. This is scored
positive. An assignment can be changed from positive to nega-
tive or vice versa in this table (see Notes 30 and 39). The drop-
down in section B allows selection of an individual probe. In
this section, the probe reactivity for all samples in the run under
analysis can be viewed. In the example in Fig. 2, probe 202 gave
a fluorescence above the positive threshold for three of the sam-
ples displayed. Section C shows the selected probe reactivity for
all historical samples. These local data can be used to allow the
user to modify the cut-offs as required. The final HLA type is
displayed in section D, including any results from other analyses
of the same sample. In this example, the final HLA type is A*01,
A*02. The suggested allelic combination for an individual sample
is shown in section E. In this example, possible combinations
60 D. Trajanoski and S.J. Fidler

Fig. 2. LIFECODES® Quick-Type™ analysis software. Section A shows the probe reactivity thresholds and probe
assignments for an individual sample. An assignment can be changed from positive to negative or vice versa in this
table (see Notes 30 and 39). The drop-down in section B allows selection of an individual probe. In this section, the
probe reactivity for all samples in the run under analysis can be viewed. Section C shows the selected probe reactivity
for all historical samples. The final HLA type is displayed in section D, including any results from other analyses of the
same sample. The suggested allelic combination for an individual sample is shown in section E. Courtesy of Gen-
Probe, Inc. Used with permission.

are A*01:01:01:01, A*02:01:01:01 or A*01:01:01:01, A*02:

01:01:02L or A*01:01:01:01, A*02:01:02, etc.
3. The reporting function in the LIFECODES™ software allows
resulting via a printed hard-copy report or export of results via
a .csv file (see Note 23).

4. Notes

1. Ramping speed is critical for this PCR reaction. If an ABI 9700

thermal cycler is not available, a detailed work-up of the assay
may be required adjusting ramping speed as appropriate.
2. At least one control DNA of well-established HLA type should
be included in each typing run. This control can be used as an
amplification control, hybridisation control, and as a run con-
trol (sample order control). The concentration of DNA
required is defined by the manufacturer in the kit insert.
 4 HLA Typing Using Bead-Based Methods 61

3. AmpliTaq is recommended by the manufacturer. If an alternative

is used, it will need careful QC to ensure appropriate amplification
of DNA occurs.
4. A DNA ladder from 100 to 1,000 bp is sufficient to size any
product from the LABType SSO range.
5. The LABType SSO range includes HLA typing kits specific for
HLA-A, -B, -C, -DRB1, -DRB3,4,5, -DQB1, -DQA1, -DPB1.
Intermediate resolution kits are available for all loci and high
resolution kits are available for some. For the full range refer to
http://www.onelambda.com.
6. The LifeCodes SSO range includes HLA typing kits specific for
HLA-A, -B, -C, -DRB1,-DRB3,4,5, -DQB1, -DQA1, -DPB1.
Intermediate resolution kits are available for all loci. For the
full range refer to http://www.gen-probe.com.
7. D-mix is generic and may be used with any HLA typing kit.
However, the amplification primers are specific for each kit and
lot specific and therefore must not be interchanged.
8. Once the reagents are thawed at room temperature, vortex the
vials at medium speed for 20 s followed by a quick centrifuge
spin. This is done to spin down any liquid present in the lid.
After centrifuging, place them in a cold isorack or crushed ice
to keep cool.
9. When making up the amplification mixture always make extra
volume for the required number of tests needed. The extra
volume may be needed depending on pipetting techniques and
calibration status of equipment.
10. ONE LAMBDA recommend DNA at 20 ng/μL concentration
with A260/A280 ratio 1.65:1.8. DNA should not be resus-
pended in EDTA. Where DNA concentration is less than
20 ng/μL, the assay may be modified by adding a greater volume
of DNA, and reducing the volume of D-mix as appropriate.
11. When adding Taq polymerase, gently mix using vortex on low
speed or by flicking tube with fingers. Excessive fast mixing
may denature the Taq polymerase resulting in poor PCR
amplification of the template DNA.
12. Take care not to crosscontaminate samples with one another.
Either use a fresh tip each time or if you are multidispensing,
place the tip on the top left side of each well and dispense. This
should not cause contamination as the tip touches a clean part
of the well each time.
13. Make sure to use an 8-channel pipette as this will reduce
pipetting time, and always use fresh clean tips for each dis-
pense. Never re-use tips, as this will cause contamination.
14. The size of the PCR amplification product differs according to
the HLA typing kit used. The expected product size can be found
62 D. Trajanoski and S.J. Fidler

in the manufacturer’s kit insert or at http://www.onelambda.

com. If no band is present, the DNA concentration and quality
should be rechecked and the PCR repeated. If the band is of the
wrong size, it is possible the amplification primers used for that
particular loci were incorrect. The PCR should be repeated care-
fully checking the amplification primers before use.
15. Once thawed, the 100× SAPE solution should be kept refrig-
erated. Do not remove the 100× SAPE solution from the fridge
until needed. The SAPE is light sensitive and must always be
kept in the dark. Prepare the SAPE solution during the third
wash step and immediately use after.
16. Make sure sample location and sample identification are cor-
rectly noted by following a proforma or spreadsheet. It is easy
to mix up sample locations, especially if you are testing 96
samples.
17. Note a colour change from clear to pale yellow. If there is no
colour change, add a further 1 μL of neutralisation buffer,
which should trigger a colour change. If there is still no colour
change, then there is a possibility that the sample did not dena-
ture properly. It is good practice to note down which wells did
not change colour just in case they fail during the reading of
the samples. This way you can identify where a problem has
occurred.
18. Make sure to vortex Bead Mixture vigorously to give a homog-
enous bead mixture. This is best done by vortexing the vial
three times for 10 s each time.
19. Hold the PCR plate firmly and flat on the vortex to avoid
splashing of liquid onto the tray seal.
20. Use an 8-channel pipette and a wash trough. Or alternatively
you can use an electronic pipette set to multidispense. This will
enable you to quickly add wash buffer to all wells. To eliminate
contamination, make sure not to dispense the wash buffer to
the bottom of the well, but the top side of the well. If, how-
ever, you do dispense to the bottom of the well, make sure to
use fresh tips each time.
21. Make sure to do this in one very quick flick in order to avoid
loosing beads. If done slowly, the wash buffer can wash away
the beads that are in the bottom of the well as it runs on the
side of the well walls. Speed and correct techniques are vital at
this step. A good flick will get rid of any excess wash buffer
and/or reagents and produce a good strong signal. Once the
tray has been flicked and returned to the upright position, do
not invert the tray for a second time before blotting, as this
will cause the beads to be flushed out onto the paper towel. If
 4 HLA Typing Using Bead-Based Methods 63

this has occurred, repeat steps 4 and 5 in Subheading 3.1.4

before blotting.
22. For best results, read the plate as soon as possible. Prolonged
storage of samples (more than 4 h) may result in signal loss. If
samples cannot be read immediately, cover in aluminium foil
and store at 4°C. Mix the samples thoroughly with gentle vor-
texing before reading as the beads may have set over time.
23. Waste container levels must always be monitored. The run 0.2
and 0.3 batch tab displays a warning when the sheath fluid
container needs to be refilled. Check all tubes are running
freely and there is no line blockage or breakage.
24. The calibration beads are in larger 10 mL vials and vigorous
mixing is needed. Vortex for at least 1 min at high speed to get
a homogenous bead mixture.
25. The calibrator microspheres are used to normalise the settings
for the reporter, classification, and the doublet discriminator
channels. If calibration of any of the calibration or control
beads fails, then the calibration must be performed again.
Failure could be due to either air in the lines of the instrument
or bead mix was insufficiently vortexed. Performing a couple
of wash steps will eliminate air in the lines of the instrument.
26. Templates can be found and downloaded from ONE LAMBDA
via its website http://www.download.onelambda.com or from
Gen-Probe via its website http://www.gen-probe.com.
27. It is extremely important that the correct lot number is selected.
Incorrectly assigning the lot will result in invalid data.
28. The HLA Fusion™ software has many import features and can
be adapted to suit the laboratories’ reporting systems. Full
training on the software is commonly provided by ONE
LAMBDA or its distributors, but basic functionality can be
easily followed in the user manual.
29. The manufacturer’s QC shows the probe reactivity under ideal
conditions. The default probe cut-off is set in the software by
the manufacturer using the QC panel.
30. Different DNA extraction methods (quality of DNA) and dif-
ferences in thermal cyclers may contribute to differences in
probe reactivity from the manufacturer’s QC. The probe cut-
off can therefore be adjusted according to the individual labo-
ratory QC.
31. The “Force” tab shows the antigen assignment possible if an
individual probe has false positive or false negative reactivity.
The “Close Bead Reaction” section shows any probes which
are close to the set cut-off for positivity. Adjusting any of the
64 D. Trajanoski and S.J. Fidler

cut-offs may change the antigen assignment and should be

checked carefully.
32. The .csv file export format allows import directly into the labora-
tory reporting system, thereby minimising transcription errors.
33. The master mix contains locus-specific amplification primers
and therefore cannot be interchanged for amplification of other
loci.
34. Gen-Probe recommends DNA at 10–200 ng/μL concentra-
tion with A260/A280 ratio 1.65:1.8. Two hundred nanogram
of DNA template is required per reaction. Based on an average
yield of 50 ng/μL (for robotic DNA extraction systems), 4 μL
of template is required per test. This volume will be altered
according to the DNA concentration of an individual sample,
alternatively all samples could be standardised to the same con-
centration by dilution.
35. The size of the PCR amplification product differs according to
the HLA typing kit used. The expected product size can be
found in the manufacturer’s kit insert or at http://www.gen-
probe.com.
36. It is recommended to make enough Dilution Solution Mixture
for n + 2 samples to account for pipetting loss.
37. The Dilution Solution may, if required, be warmed at 45°C for
5 min and vortexed to ensure all components are in solution.
Dilution Solution must be at room temperature (18–30°C)
before making the SAPE/DS mixture.
38. The LIFECODES™ software has many import features and
can be adapted to suit the laboratories’ reporting systems. Full
training on the software is commonly provided by Gen-Probe
or its distributors, but basic functionality can be easily followed
in the user manual.
39. Adjusting any of the cut-offs may change the antigen assign-
ment and should be checked carefully.

Acknowledgments

HLA Fusion screenshots reproduced with kind permission from

One Lambda, Inc; LIFECODES Quick-Type™ screenshots repro-
duced with kind permission from Gen-Probe.
We wish to acknowledge the following for assistance with
writing this manuscript: Peter Brescia (One Lambda, Inc) and Rob
Vorhaban (Gen-Probe, Inc).
 4 HLA Typing Using Bead-Based Methods 65

References

1. Claas FH, van der Poel JJ, Castelli-Visser R, Pool J, 4. Olerup O (1990) HLA class I typing by digestion
Chen RB, Xu KY, van Rood JJ (1985) Interaction of PCR-amplified DNA with allele specific restric-
between des-Tyr1-gamma-endorphin and HLA tion endonucleases. Tissue Antigens 36:83–87
class I molecules: serological detection of HLA- 5. Sayer D, Whidborne R, Brestovac B et al (2001)
A2 subtype. Immunogenetics 22:309–314 HLA-DRB1 DNA sequencing based typing: an
2. Kojima S, Inaba J, Yoshimi A et al (2001) approach suitable for high throughput typing
Unrelated donor marrow transplantation in chil- including unrelated bone marrow registry
dren with severe aplastic anaemia using cyclo- donors. Tissue Antigens 57:46–54
phosohamide, anti-thymocyte globulin and total 6. Olerup O, Zetterquist H (1992) HLA-DR typ-
body irradiation. Br J Haematol 114:706–711 ing by PCR amplification with sequence-
3. Saiki RK, Scharf S, Faloona F et al (1985) specific primers (PCR-SSP) in 2 hours: an
Enzymatic amplification of beta-globulin alternative to serological DR typing in clinical
genomic sequences and restriction site analysis practice including donor-recipient matching in
for diagnosis of sickle cell anaemia. Science cadaveric transplantation. Tissue Antigens
230:1350–1354 39:225–235
 Chapter 5

HLA Typing by Direct DNA Sequencing

Linda K. Smith

Abstract
Sequencing-based typing is a high resolution method for the identification of HLA polymorphisms. The
majority of HLA Class I alleles can be discriminated by their exon 2 and 3 sequence, and for Class II alleles,
exon 2 is generally sufficient. There are polymorphic positions in other exons which may require additional
sequencing to exclude certain alleles with differences outside exon 2 and 3, depending on the clinical
requirement and relevant accredition guidelines. The process involves selective amplification of target
alleles by PCR, agarose gel electrophoresis of the PCR products to assess the quantity and quality, followed
by purification of PCR amplicons to remove excess primer and dNTPs. Cycle sequencing reactions using
Applied Biosystems™ BigDye® Terminator Ready Reaction v1.1 or v3.1 Kit are performed, then purification
of sequence reactions before electrophoresing using Applied Biosystems™ 3730 or 3730XL Genetic
Analyser (or similar). Data is processed by specialised software packages, which compare the sample
sequence to the sequences of all possible theoretical allele combinations to assign an accurate genotype.
Examination of all nucleotides, both at conserved and polymorphic positions enables the direct identification
of new alleles, which may not be possible with techniques such as SSP and SSO typing.

Key words: PCR, Sequencing, HLA, Sequencing-based typing

1. Introduction

1.1. Amplification Sequencing is carried out using a PCR product as the template.
of Target Region The PCR amplifies a specific region of the gene/locus of interest,
Using PCR which is then sequenced from both directions to generate a high-
quality DNA sequence. Bi-directional sequencing is required to
achieve the desired level of accuracy and confidence in the base
calls at critical positions throughout the length of the sequence
(see Fig. 1 for overview).
Two of the most common strategies for designing PCR methods
for sequencing-based typing (SBT) include:
(a) Designing primers with a sequencing “tail” or “tag” (Fig. 2).
The most common tail (or tag) sequence is M13, which was

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_5, © Springer Science+Business Media New York 2012

67
68 L.K. Smith

Fig. 1. Steps for sequencing of PCR templates.

Fig. 2. Sequencing with tailed PCR primers.

 5 HLA Typing by Direct DNA Sequencing 69

Fig. 3. Sequencing with internal sequencing primers and tailed PCR primers.

initially used for sequencing clones constructed in the M13

bacteriophage. The potential disadvantage of using tailed
primers is the increase in primer length, which requires higher
quality oligonucleotides, and to a lesser extent, challenges in
designing primers with a tail.
(b) Using nested (internal) sequencing primers (Fig. 3). Internal
primers may be required for sequencing of long PCR frag-
ments, where the length of good quality sequence from each
end of the PCR template is insufficient to cover the entire
region required. This applies mainly to HLA Class I loci, where
data from more than one exon is required. This approach may
also be used if the entire gene is amplified in the PCR, but only
certain exons are intended to be sequenced (1).
There are now commercial kits available from several sources
which include both PCR and sequencing primers for the majority
of HLA loci. Alternatively, there are a number of published proto-
cols (2–15), which can be reproduced in a laboratory planning to
implement HLA typing using direct sequencing. This latter
approach is likely to require a degree of optimisation, as the instru-
mentation and other operating conditions may vary from the orig-
inal reporting laboratory. If there are no suitable commercial kits,
or published protocols suitable for the specific target of interest,
then the available sequence data for the gene can be examined to
determine whether primers can be designed in conserved regions
70 L.K. Smith

to specifically amplify the target region. PCR conditions are then

optimised to determine the appropriate buffer conditions, and
thermal cycling settings. The limiting factor to such an approach in
many cases is the availability of sequence data.

1.2. Principles Fluorescence-based cycle sequencing requires a DNA template, a

of Sequencing sequencing primer, thermal stable DNA polymerase, nucleotides
(dNTPs), dideoxynucleotides (ddNTPs) where each of the four
dNTPs are labelled with one of the four specific fluorescent dyes,
and buffer . Thermal cycling of the reactions creates and amplifies a
range of extension products that are terminated when one of four
fluorescently labelled dideoxynucleotides is incorporated. Each dye
emits a unique wavelength when excited by a laser, and the fluorescent
dye on the extension product identifies the 3¢ terminal dideoxynu-
cleotide as A, C, G, or T. The main advantages of dye terminator
chemistry are that it enables the use of unlabelled sequencing
primers, the reactions are performed in one tube, and there are
fewer pipetting steps compared with dye primer chemistry (1).
Historically, DNA sequencing products were separated using
polyacrylamide gels poured between two glass plates. Capillary
electrophoresis using a denaturing liquid polymer has largely
replaced the use of gel separation, which simplifies workflow,
increases throughput, and is easier to use. During capillary electro-
phoresis, the extension products of the sequencing reaction enter
the capillary via electrokinetic injection. High voltage is applied
to the buffered sequence reaction, which forces the negatively
charges fragments into the capillaries. Products are separated by
size based on their total charge. As the fragments reach the positive
electrode, they pass through the path of a laser beam, which causes
the dyes to fluoresce. An optical detection device captures the
signal, and the software converts this to digital data (16).
The data is then examined using specialised software which
will align the resulting sequences, allow the operators to edit the
base calls if required, and compare the final consensus sequence
to a reference library to determine the allele combination present
(see Chapter 6).
The method included in this chapter is based on reagents and
software developed by Conexio Genomics Pty Ltd (CG-HLA
Sequence Based Typing Kits™) (17). Note that the kits are designed
to amplify a larger region of many loci than what is required by
many tissue typing accrediting bodies, but they allow resolution of
certain Class I null alleles as required by the National Marrow
Donor Program (NMDP). The regions amplified for each locus
are as follows:

HLA-A: exons 1–4 DRB1: exons 2 + 3

HLA-B: exons 1–4 DQB1: exons 2 + 3
HLA-C: exons 1–8
 5 HLA Typing by Direct DNA Sequencing 71

In a medium to high throughput situation, it is advisable, where

possible, to use robotic workstations (automated liquid handling
platforms) to reduce the incidence of repetitive pipetting injury.

2. Materials

2.1. Specimen 1. Sterilised tissue culture grade water (negative/no template

Requirements control).
(See Note 1) 2. High molecular weight human genomic DNA (concentration
range 20–100 ng/mL and OD 260/280 ratio >1.8) extracted
from ACD or EDTA anticoagulated whole blood specimens.
Do not use samples containing Heparin (see Note 2).

2.2. PCR Amplification 1. CG-HLA Sequence Based Typing Kits™ (see Table 1 for
of Target DNA components).
and Agarose Gel 2. Sterile molecular grade or tissue culture grade deionised water.
Electrophoresis
3. Adjustable pipettes (manual or electronic, various volumes)
and aerosol resistant pipette tips.
4. Thermal Cycler with heated lid.
Conexio Genomics kits have been tested using the following
thermal cyclers: Applied Biosystems™ (by Life Technologies™)
GeneAmp PCR system 9700, Biorad DYAD™ (previously MJ
Research DNA Engine DYAD), and Eppendorf Mastercycler
Pro™. Use of other thermal cyclers will require validation by
the user.
5. 0.2-mL thin walled reaction tubes either in 8-tube strips or
96-well plates. Use those to suit your thermal cycler as not all
plates fit all thermal cyclers.
6. Strip lids to fit 96-well PCR plates.
7. Sterile 1.5-mL tubes.
8. Vortex mixer.
9. Sterile biological safety cabinet or hood.
10. Agarose gel electrophoresis apparatus.
11. 1% Agarose gel (molecular biology grade) in 0.5× TBE, con-
taining 0.1 g/mL ethidium bromide.
12. 0.5× TBE electrophoresis buffer, diluted from 10× TBE which
is prepared as follows:

10× TBE (for 2 L)

Trizma base 215.6 g
Boric acid 110 g
EDTA (di-sodium salt) 16.4 g
 72
L.K. Smith

Table 1
Conexio genomics HLA sequencing-based typing kit details

Pre-PCR contents Post PCR contents

Kit Catalogue number No of tests Reagent Volume (mL) Primer name Volume (mL)
HLA-A XH-PD1.1-2(50) 50 DNA Pol 1 × 55 AEX1F, AEX1R, AEX2F, AEX2R, AEX3F, 1 × 110 of each
AEX3R, AEX4F, AEX4R
CG-HLA-A mix 1 × 880
XH-PD1.1-2(100) 100 DNA Pol 1 × 110 AEX1F, AEX1R, AEX2F, AEX2R, AEX3F, 1 × 220 of each
AEX3R, AEX4F, AEX4R
CG-HLA-A mix 2 × 880
HLA-B BS-PD2.1-2(50) 50 DNA Pol 1 × 55 BEX1F, BEX2F, BEX2R, BEX3F, BEX3R, 1 × 110 of each
BEX4F, BEX4R
CG-HLA-B mix 1 × 880
BS-PD2.1-2(100) 100 DNA Pol 1 × 110 BEX1F, BEX2F, BEX2R, BEX3F, BEX3R, 1 × 220 of each
BEX4F, BEX4R
CG-HLA-B mix 2 × 880
HLA-C HH-PD 3.2-2(50) 50 DNA Pol 1 × 55 CEX1F, CEX1R, CEX2F, CEX2R, CEX3F, 1 × 110 of each
CEX3R, CEX4F, CEX4R, CEX5F, CEX5R,
CG-HLA-C mix 1 × 880
CEX6F, CEX6R, CEX7F, CEX7R, CEX8F
HH-PD 3.2-2(100) 100 DNA Pol 1 × 1101 CEX1F, CEX1R, CEX2F, CEX2R, CEX3F, 1 × 220 of each
CEX3R, CEX4F, CEX4R, CEX5F, CEX5R,
CG-HLA-C mix 2 × 880
CEX6F, CEX6R, CEX7F, CEX7R, CEX8F
 DRB1 HH-PD5.2-4(50) 50 DNA Pol 1 × 18 DRB1EX2F, DRB1EX3R, DRB1EX3R, 1 × 110 of each
RB-TG344-R
CG-HLA-DRB1 mix 1 × 920
HH-PD5.2-4(100) 100 DNA Pol 1 × 36 DRB1EX2F, DRB1EX3R, DRB1EX3R, 1 × 220 each
RB-TG344-R
CG-HLA-DRB1 mix 2 × 920
DQB1 PQ-PD6.2-1(50) 50 DNA Pol 1 × 18 DQB1EX2F, DQB1EX2R, DQB1EX3F. 1 × 110 of each
DQB1EX3R
CG-HLA-DQB1 mix 1 × 920
PQ-PD6.2-1(100) 100 DNA Pol 1 × 36 DQB1EX2F, DQB1EX2R, DQB1EX3F. 1 × 220 of each
DQB1EX3R
CG-HLA-DQB1 mix 2 × 920
The PRE-PCR component of each kit consists of a vial/s of a locus-specific PCR mix (e.g. CG-HLA-A mix) consisting of PCR buffer, dNTPs, MgCl2, locus-specific PCR
primers, and a single vial of DNA polymerase (DNA Pol). The POST-PCR kit contains sequencing primers (e.g. AEX1F)
5
HLA Typing by Direct DNA Sequencing
73
74 L.K. Smith

Weigh each of the reagents and place into a clean sterile 2-L
conical flask. Dissolve in 1.2 L of Milli-Q water with continuous
mixing on a magnetic stirrer. When dissolved make up to 2 L
with Milli-Q water. Place buffer into 2 × 2 L clean sterile auto-
clavable containers. The buffer is divided into two containers
so as not to spill over during autoclaving. Autoclave to sterilise.
Store in the dark.
13. Electrophoresis loading buffer. Prepare as follows:
Place 20 mL of Milli-Q water into a clean sterilised glass
beaker; add 8 g of sucrose and 0.05 g Bromophenol Blue. Mix
using a magnetic stirrer. Once dissolved, make the solution up
to 160 mL with Milli-Q water. Store at 4°C for immediate use,
or for long term storage at −20°C.
14. PCR Marker (“Lambda ladder”) suitable to cover range of
300–1,300 bp (e.g. TrackIt 100 bp DNA ladder or TrackIt
1 Kb Plus DNA ladder, from Invitrogen™).
15. BIO-RAD™ GEL DOC 2000 Transilluminator or similar.

2.3. PCR Purification 1. ExoSAP-IT® (USB Products).

2. 2 mM MgCl2.
3. Shaker.

2.4. Sequencing, 1. BigDye® Terminator (BDT) Sequencing Kits v3.1 or v1.1

Purification, Ready Reaction mix (from Applied Biosystems™ by Life
and Denaturation Technologies™).
2. BigDye® Terminator v1.1/3.1 5× Sequencing Buffer.
3. 96-Well half skirt PCR plates to fit Applied Biosystems™
sequencers (e.g. Axygen™: PCR-96-ABC).
4. CR cooler 200-mL starter kit (Eppendorf™) or ice bath to fit
96-well plate.
5. Table top centrifuge with plate adapters and ability to reach
2,500 × g (e.g. Heraeus™ Megafuge 2.0).
6. 125 mM EDTA, pH 8.0.
7. Absolute ethanol. Each run requires freshly prepared 80%
ethanol consisting of Absolute ethanol and tissue culture grade
water. DO NOT USE DENATURED ALCOHOL.
8. Hi-Di™ Formamide (Applied Biosystems™).
9. Vacuum pump and vacuum apparatus.
10. Automated DNA sequencer and accessories (e.g. Applied
Biosystems™ ABI Prism® 3730 or 3730-XL Genetic Analyser)
including Data Collection software.
Conexio Genomics kits have been tested and validated on
Applied Biosystems™ ABI Prism® 3100, 3730 or 3730-XL
 5 HLA Typing by Direct DNA Sequencing 75

capillary sequencers and software. The use of other sequencing

platforms requires validation by the user prior to use.
11. HLA Sequencing analysis software, e.g. Assign SBT™, version
3.5 or higher, or Assign ATF™ (Conexio Genomics Pty Ltd) as
described in Chapter 6.

3. Methods

3.1. PCR Amplification 1. For each locus set up one reaction for each sample being
of Target DNA (17) amplified (see Note 3). Include appropriate positive controls
of known genotype (see Note 4) and at least one negative con-
trol (see Note 5) for each group of samples being amplified.
2. CG-HLA Sequence Based Typing Kits™ are supplied as sepa-
rate aliquots of primer mix and polymerase. The polymerase
needs to be added to the primer mix just prior to setting up
the PCR reactions. Thaw the required number of vials of the
appropriate PCR Mix. For example, for a run of 15 samples
including controls, make up enough primer/polymerase mix
for 16 samples. This allows for slight pipetting inaccuracies
which can be due to a small amount of the solution adhering
to the outside of the tips. Once thawed, vortex briefly to mix
components.
3. Calculate the number of samples to be tested and dispense the
required amount of PCR mix and DNA polymerase into a
sterile tube. Refer to Table 2 for volumes required per sample.
Vortex the solution three to four times, for approximately 1 s
each time.
4. Dispense 17 mL of the primer/polymerase mix prepared in
step 3 above into each reaction well of either a 96-well plate or
8-tube strips.
5. Add 3 mL of sample DNA or appropriate positive control DNA
to each reaction well. Add 3 mL of sterile water to the negative
control reaction well.

Table 2
PCR mastermix volumes—per sample

Locus HLA-A (mL) HLA-B (mL) HLA-C (mL) HLA-DRB1 (mL) HLA-DQB1 (mL)

Locus-specific PCR mix 16 16 16 16.7 16.7

(e.g. CG-HLA-A mix)
DNA-Pol 1 1 1 0.3 0.3
76 L.K. Smith

6. Seal the reaction wells or tubes. Mix thoroughly but gently by

vortexing, then centrifuge briefly to ensure all reaction compo-
nents are at the bottom of the wells (i.e. turn centrifuge on,
allow it to reach a speed of approximately 185 × g, and then
turn off).
7. Place the 96-well plate or tube strips into a thermal cycler and
amplify the target sequence according to the thermal cycling
conditions below.
Thermal cycling program
95°C 10 min
96°C 20 s
60°C 30 s 33 cycles
72°C 3 min
15°C Hold

NB: Amplification takes approximately 2.5 h

to complete.

8. When the PCR is completed, remove the plate from the ther-
mal cycler and either proceed directly to gel electrophoresis or
store at 4°C until required. Purification of amplicons by
ExoSap-IT® treatment (see Subheading 3.3) should occur
within 24 h of completion of PCR.

3.2. Agarose Gel 1. Confirm successful amplification of the template DNA by

Electrophoresis agarose gel electrophoresis using 2 mL of each PCR product
combined with 5 mL loading buffer (If different volumes of
loading buffer are to be used, this should be validated prior to
use). The use of 1% agarose gel is recommended (17). See
Note 6 for voltage and running time recommendations.
2. Photograph the gel using a BIO-RAD GEL DOC 2000
Transilluminator or similar. The ethidium bromide-labelled
DNA bands are illuminated by the UV light.

Table 3
Expected number of PCR products and respective sizes for each locus

Locus Expected band sizes

HLA-A »2 kb
HLA-B »2 kb
HLA-C »1 and 1.4 kb
HLA-DRB1 »450–650 bp (banding pattern will vary depending on which of the specific allele
groups are present, maximum of two bands expected)
HLA-DQB1 »300 and 350 bp
 5 HLA Typing by Direct DNA Sequencing 77

3. The number and expected sizes of the resultant amplicons will

vary according to the locus and sample genotype. Expected
PCR amplicon sizes are indicated in Table 3. Band intensities
should be reproducible within each assay if DNA quality is
consistent. A sample with a slightly weaker band can still be
processed for sequencing using a lower post-ExoSAP-IT® dilution
volume. Very weak products should not be sequenced (17).
See Note 7 for interpretation of positive and negative PCR
controls.

3.3. Purification PCR products contain excess primers and unconsumed dNTPs
of PCR Products which require removal before sequencing reactions can be per-
for Sequencing formed. They may interfere with the efficiency of the sequencing
reaction, or can result in background “noise” in the sequence.
There are a number of options for purifying PCR products: Spin
columns, enzyme digestion, or magnetic bead-based technology.
Purification systems other than ExoSAP-IT® (e.g. Agencourt®
AMPure® XP or column-based systems) can be used to purify these
PCR products. It is strongly recommended that users validate these
procedures before proceeding. If ExoSAP-IT® is used; it is recom-
mended that users follow the procedure described below (17).
1. Prepare a mastermix consisting of 4 mL of ExoSAP-IT® and
8 mL of 2 mM MgCl2 per sample. Dispense 12 mL of the
mastermix into each PCR sample to be sequenced. (This is
determined by the band intensities seen following agarose gel
electrophoresis as described in Subheading 3.2. Very weak
products are not suitable for sequencing and therefore do not
need to be treated with ExoSAP-IT®.) Seal the tubes or plates,
vortex or place on shaker for 2 min, and centrifuge as described
in Subheading 3.1, step 6 to ensure all reagents are at the
bottom of the wells, before placing into the thermal cycler.
2. Run the thermal cycler according to the following protocol:
37°C-30 min
80°C-15 min
4°C-hold
3. Upon completion, dilute the purified product 1:8 with sterile
water. This dilution step will ensure that there is sufficient
template to perform the sequencing reactions and ensure that
the concentration of the template is sufficient to produce good
quality sequence data. Note that weaker PCR products may
require a lower dilution factor than 1:8.
4. ExoSAP-IT® treated samples may be stored at 4°C until ready
for use.
78 L.K. Smith

3.4. Sequencing All sequencing reactions must be set up in a designated (“post-

of PCR Amplified PCR”) area. This cannot be the same area as used to set up the
Template PCR reactions.
In instances where heterozygous ambiguities are to be resolved
with hemizygous sequencing primers, such as HARPs®
(Heterozygous Ambiguity Resolving Primer), the procedure is the
same as described in this method.
The RB-TG344-R included in the CD-HLA-DRB1 kit
(Table 4) is a HARP® directed towards the codon 86 dimorphism.
Its use is optional (17).
1. Allow the vial of BDT and other reagents to thaw at room
temperature. Expose BDT to light as little as possible. Once
thawed, transfer reagents to ice. Reagents should be kept on
ice and the sequencing reactions should be prepared on a cool-
ing block or on an ice bath (see Note 8 for Quality testing of
BDT and buffer batches).
2. For each reaction, prepare a fresh sequencing primer reaction
mix as follows (3).
Reagent Volume required (mL)

Sequencing primer 2.0

BDT v3.1 ready reaction mix 1.0
5× Sequencing buffer 3.5
Water 11.5

For multiple samples requiring the same sequencing primer,

make a large mix of the components above, allowing for
pipetting error (e.g. for nine samples, make enough mix for ten
reactions). See Note 9 regarding optimisation of these reagent
ratios for different instruments.
3. Mix each sequencing primer reaction mix gently by pulse
(2–3 s) vortexing.

Table 4
Sequencing primers provided for use with each locus

Locus Sequencing primers

HLA-A AEX1F, AEX1R, AEX2F, AEX2R, AEX3F, AEX3R, AEX4F, AEX4R

HLA-B BEX1F, BEX2F, BEX2R, BEX3F, BEX3R, BEX4F, BEX4R
HLA-C CEX1F, CEX1R, CEX2F, CEX2R, CEX3F, CEX3R, CEX4F, CEX4R, CEX5F,
CEX5R, CEX6F, CEX6R, CEX7F, CEX7R, CEX8F
HLA-DRB1 DRB1EX2F, DRB1EX3R, DRB1EX3R, RB-TG344-R
HLA-DQB1 DQB1EX2F, DQB1EX2R, DQB1EX3F, DQB1EX3R
 5 HLA Typing by Direct DNA Sequencing 79

4. Dispense 18 mL of the sequencing reaction mix to each

appropriate reaction tube or well of a 96-well plate. If 8-tube
strips or individual tubes are used, they should be placed in a
carrier so as to avoid mixing up the positions during the
procedure.
5. Add 2 mL of purified PCR product (from Subheading 3.3, step
3 and 4) to each appropriate well (see Note 10).
6. Cap or seal wells or plate tightly, mix gently then centrifuge
briefly as previously described in Subheading 3.1, step 6 to
ensure all reagents are at the bottom of the tubes.
7. Place the reaction tubes or plate into a thermal cycler and run
according to the following program (17) (see Note 11):

96°C for 10 s
50°C for 5 s 25 cycles

60°C for 2 min

Then 4°C indefinitely

8. Once the program is complete, remove the reaction tubes or

plate from the thermal cycler and either proceed directly to
purification of the sequencing reaction products or store at
4°C until required. It is recommended that sequence reaction
samples are purified and run on the DNA sequencer within
24 h (17).

3.5. Purification See Note 12 for additional information.

and Concentration
1. Remove the reaction tubes or 96-well plate from the thermal
of Big Dye Terminator
cycler (or 4°C storage) and centrifuge briefly as previously
Sequencing Products described in Subheading 3.1, step 6. If reusable lids/caps have
Using Ethanol been used during thermal cycling, label the lids/caps to avoid
Precipitation (17, 18) cross-contamination.
2. Each reaction requires 5 mL of 125 mM EDTA and 60 mL
ethanol (see Note 13). Make a mix of the following for a full
96-well plate:
550 mL of 125 mM EDTA.
6.6 mL of 100% ethanol.
Add 65 mL of the above mix to each sample.
The final ethanol concentration is 70%.
3. Replace caps and invert tubes or plate four times to ensure
thorough mixing.
4. Centrifuge briefly as previously described in Subheading 3.1,
step 6.
80 L.K. Smith

5. Leave at room temperature, protected from light, for 15 min

to precipitate the extension products. Slightly longer precipita-
tion times are acceptable, preferably no longer than 1 h (18)
(see Note 14).
6. Place the tubes or plate in a centrifuge and spin according to
the following guidelines:
Centrifuge speed 1,400–2,000 × g: 45 min.
Centrifuge speed 2,000–3,000 × g: 30 min.
Important: Proceed to the next step immediately. If this is not
possible, then spin the tray for an additional 10 min immedi-
ately before performing the next step.
7. Without disturbing the precipitates, remove the strip lids and
discard the supernatant, by inverting the tubes or plate onto
absorbent tissues (e.g. facial tissues).
8. Place the inverted tubes or plate with the tissues into the cen-
trifuge and spin briefly as described previously in Subheading 3.1,
step 6.
9. Remove the plate from the centrifuge, discard tissues, and then
add 60 mL of freshly prepared 80% ethanol to each pellet. Cap
the tubes and then invert the tubes or plate four times to mix.
10. Place the tubes or plate in centrifuge and spin at approximately
1,700 × g for 15 min.
11. Repeat steps 8–10, then remove from centrifuge, and discard
tissues. Vacuum dry for 15 min, or air dry for 1 h, in the dark.
Reseal plate if not proceeding immediately to denaturation step.
12. Resolubilise pellets by adding 12 mL of Hi-Di™ formamide.
Thoroughly vortex tubes or plate for approximately 15 s.
13. Denature samples by placing in a thermal cycler which is already
at 98°C, for 5 min.
14. At the end of 5 min, remove from thermal cycler and immedi-
ately place on ice for at least 5 min. Centrifuge briefly as
described previously in Subheading 3.1, step 6 to bring down
any condensation.
ENSURE THAT THERE ARE NO AIR BUBBLES IN THE
WELLS. THESE CAN ENTER AND DAMAGE THE
CAPILLARY.
15. Load the reaction plate onto the automated sequencer and
prepare the data collection file according to the sequencer
manufacturer specifications. Samples should be run within
24–36 h; otherwise, the Hi-Di™ formamide will begin to break
down to formic acid and there will be significant loss of resolu-
tion. If it is not possible to run the plate within this time frame,
it should be stored at 4°C (17, 18).
 5 HLA Typing by Direct DNA Sequencing 81

Table 5
Data collection settings for ABI 3730 or 3730xl

Parameter Setting

Dye set Z-BigDye V3

Mobility file KB_3730_POP7_BDTV3
Basecaller KB.bcp
Run module Regular FastSeq50_POP7
Injection time 15 s
Run time 2.5 h

3.6. Electrophoresis 1. Run the sequencing product plates according to the

of Sequencing instrument parameters in Table 5. These have been validated
Products on 3730 by the manufacturer using POP-7™. These parameters may
Sequencer require user validation for other polymers and instruments.
Please refer to the appropriate instrument user’s manual for
detailed instructions and guidance (17).
2. Use the instrument’s data collection software to process the
raw collected data and create the sequence files. Please refer to
the appropriate instrument user’s manual for detailed instruc-
tions and guidance (19).
3. See Note 15 for brief sequence quality troubleshooting guide.

3.7. Allele Assignment Sequence analysis may be performed using Assign SBT™ or Assign
ATF™ software. For more details, please refer Chapter 6 or the
Conexio Genomics Pty Ltd. Web site (http://www.conexio-
genomics.com). See Notes 16–19 for some critical issues in allele
assignment.

4. Notes

1. Care should be taken to ensure reagents are sterile and the

water is of high quality for PCR and sequencing. DNases and
RNases can degrade DNA and PCR products, resulting in poor
quality sequence. The presence of salts or other contaminants
can inhibit PCR and sequencing reactions and can compete
with DNA product in electrochemical injection on capillary-
based sequencers.
2. Genomic DNA from anticoagulated whole blood, buffy coat
or buccal mucosa swabs is suitable for HLA typing by DNA
sequencing. ACD or EDTA is the preferred anticoagulant,
82 L.K. Smith

whereas heparin inhibits PCR reactions and is therefore not

recommended. There are many commercial kits for DNA
extraction, both manual and automated. The concentration
and purity of the isolated DNA should be determined using a
spectrophotometer before proceeding to the PCR. Some
commercial kits recommend a specific concentration to which
the DNA is diluted. Others may use a particular volume of
DNA within a range of concentrations. Very low DNA con-
centrations (<15 ng/mL) should be avoided as the failure rate
will be increased, and there is some potential for uneven
amplification of heterozygous samples.
3. PCR set-up is to be performed on ice or a cold-brick in a
dedicated (“pre-PCR”) area (e.g. biological safety cabinet).
Mastermix preparation should be performed in a separate area
to that used for addition of DNA.
4. DNA samples with known HLA types should be included in
each PCR assay to ensure the primers are amplifying the
target region as expected. For example, it should be possible to
cover all of the broad DRB1 allele groups using four controls,
such as DRB1*01 + 03, DRB1*04 + 07, DRB1*09 + 15, and
DRB1*01 + 10. For DQB1, a combination of control samples
to cover DQB1*02, 03, 04, and 05/06 is suggested. For
Class I loci, it is sufficient to use a single heterozygous sample.
These controls should be sequenced to confirm the identity of
the sample.
5. A negative or blank control (water) should be included on
each PCR run to detect contamination. This sample should
have all reaction components except DNA. Any PCR run
where there is amplification evident in the negative control
should be repeated, as there is likely to be contamination of
one or more reagents.
6. The maximum voltage for electrophoresis is dependent upon
the size of the tank. For a large tank (42 cm), the maximum
voltage is approximately 150 V for 20–30 min. The maximum
voltage for a medium and small size tank is 130 V for 20–30 min.
The smaller the gel, the lower the voltage required. Good
resolution of PCR products are obtained if gels are run at a
lower voltage and for a longer period of time.
7. If the positive control has failed (no band visible on gel, or a
band of the incorrect size), or there is a band visible in the
negative control, the PCR run must be repeated. If desired, all
samples can be visualised in this manner. It is advisable to do
this until you are satisfied with the reproducibility of the assay.
It should also be considered when starting a new batch of
primers or master mix.
 5 HLA Typing by Direct DNA Sequencing 83

8. The Big Dye Terminator and Sequencing Buffer must be

tested together and the two test lots should be used together
routinely. If ordering large-scale BDT kits, extra bottles of
buffer may need to be ordered in order to keep the batches
coordinated.
9. These volumes apply only to the Applied Biosystems™
3730/3730xl sequencer referred to in this chapter. Other
sequencers may require different combinations. Less sensitive
instruments may require a higher concentration of BDT, or
more template. It is advisable to carry out a titration of BDT
and template to determine the optimum combination for your
laboratory. When increasing the volume of BDT, the volume
of 5× sequencing buffer should be decreased proportionately.
If increasing primer or template volume, the volume of water
should be adjusted accordingly so that total volume is still
20 mL.
10. Care must be taken to prevent cross-contamination of sequence
reactions. Fresh tips must be used for each sequencing primer
reaction mix and each PCR product. PCR plates should be
briefly centrifuged, as previously described in Subheading 3.1,
step 6, before lids or seals are removed so that any condensa-
tion that may have formed does not flick into another well
during removal.
11. Note that the sequencing thermal cycling conditions are
slightly different from Applied Biosystems™ recommended
cycling conditions, which are as follows:

96°C for 10 s
50°C for 5 s 25 cycles
60°C for 4 min
Then 4°C indefinitely

The use of these standard conditions in place of Conexio

Genomics’ recommended protocol would require validation by
the user. This may be necessary if using an outside sequencing
service, as opposed to having access to your own instrument. In
a sequencing service with multiple sample types from different
“customers”, it may not be possible to have specific sequence
cycling conditions for a particular assay. An initial comparison
between the two cycling conditions is suggested, for all loci
being tested, as the extent to which they are affected may vary.
12. Purification of the sequencing reaction products may be
carried out by procedures other than the ethanol precipitation
method described here. It is strongly recommended that users
validate these procedures before proceeding.
84 L.K. Smith

13. Absolute ethanol absorbs water from the atmosphere, gradu-

ally decreasing its concentration. This can lead to inaccurate
final concentrations of ethanol, which can affect sequencing
protocols. Ethanol should be stored in small aliquots so that
there is limited opportunity for repeated use of each aliquot,
which reduced exposure to the atmosphere.
14. Precipitation times <15 min will result in the loss of very short
extension products. Precipitation times >24 h will increase the
precipitation of unincorporated dye terminators (1).
15. Poor sizing precision or allele calling, poor sequence resolution
and/or decreased signal intensity may be an indication that the
capillary array on the sequencer requires renewal. Capillary
arrays degrade over time. They should last at least 300 runs
(will probably last 600 runs). However, these symptoms are
similar to those seen when poor grade Hi-Di™ formamide is
used, or has been exposed to air for too long (19).
16. It should be noted that mutations at amplification or sequenc-
ing primer sites are possible and may result in allele drop-out.
Samples that suggest a homozygous typing result should be
confirmed by alternative procedures. These may include serol-
ogy, or other PCR-based assays, such as SSP or SSO typing.
17. The nature of HLA sequence-based typing is such that factors
other than the PCR mix may result in preferential amplification
or allele drop out. As a consequence, apparent homozygous
typing results should be confirmed using alternative methods as
mentioned above in Note 16, or family genotyping.
18. If sequence electropherograms have high background, there is
potential for errors in base calling which may result in incorrect
allele assignment. Laboratory personnel should be capable
(through experience and training) of assessing the quality of
sequence data and excluding those samples which are not of
sufficient quality to interpret reliably. If using the Assign software
package described in Chapter 6, there will be a visual indication
of sequence quality (the degree or intensity of “redness”) of each
base. The user will be prompted by the software to confirm the
base at each position with low quality, if that function is selected
in the Editing Tool. See Chapter 6 for further details.
19. SBT generally results in high resolution typing, however not
all allele combinations can be resolved. The location and
specificity of the PCR and sequencing primers will determine
which ambiguous allele combination can occur. Three differ-
ent categories of ambiguity are recognised (20):
– Single alleles that cannot be distinguished because the
resolving positions are outside the region amplified. For
example, where exons 2 and 3 are amplified for Class I loci,
and the difference between the alleles is in exon 4.
 5 HLA Typing by Direct DNA Sequencing 85

– Several allele pairs that result in the same heterozygous

sequence. This occurs when the alleles share sequence
motifs, but in different combination, so it is impossible to
determine whether the motifs are in cis or trans linkage.
– Combinations of alleles where one of the alleles has incom-
plete sequence for the region analysed.
The only way to resolve ambiguities of the first type is to
employ an assay that amplifies the additional region required.
Separation of alleles followed by additional sequencing can be
employed to resolve the other kinds of heterozygous ambigui-
ties. This may be by means of additional allele or group-specific
PCR amplification or by the use of allele or group-specific
sequence primers to separate the alleles in the primary heterozy-
gous PCR product. If using the Assign software package, the
user can produce a list of suggested resolving primers from
the “Reports” menu.

References

1. Applied Biosystems (2000) Automated DNA 8. Voorter CE, Kik MC, van den Berg-Loonen
sequencing chemistry guide. © Copyright EM (1998) High-resolution HAL typing for
2000, Applied Biosystems the DQB1 gene by sequence-based typing.
2. Cereb N, Maye P, Lee S, Kong Y, Yang SY Tissue Antigens 51(1):80–87
(1995) Locus-specific amplification of HLA 9. Van der Vlies SA, Voorter CE, van den Ber-
class I genes from genomic DNA: locus-specific Loonen EM (1998) A reliable and efficient
sequences in the first and third introns of high resolution typing method for HLA-C
HLA-A, -B and -C alleles. Tissue Antigens using sequence-based typing. Tissue Antigens
45(1):1–11 52(6):558–568
3. Petersdorf EW, Hansen JA (1995) A compre- 10. Kurz B, Steiert I, Heuchert G, Muller CA
hensive approach for typing the alleles of the (1999) New high resolution typing strategy for
HLA-B locus by automated sequencing. Tissue HLA-A locus alleles based on dye terminator
Antigens 46(2):73–85 sequencing of haplotypic group-specific PCR-
4. Blasczyk R, Wehling J, Weber M, Salama A amplicons of exon 2 and exon 3. Tissue Antigens
(1996) Sequence analysis of the 2nd intron 53(1):81–96
revealed common sequence motifs providing 11. Kotsch K, Wehling J, Blasczyk R (1999)
the means for a unique sequencing based Sequencing of HLA class II genes based on the
typing protocol of the HLA-A locus. Tissue conserved diversity of the non-coding regions:
Antigens 47(2):102–110 sequencing based typing of HLA-DRB genes.
5. Cereb N, Yang SY (1997) Dimorphic primers Tissue Antigens 53(5):486–497
derived from intron 1 for use in the molecular 12. Sayer D, Whidborne R, Brestovac B, Trimboli
typing of HLA-B alleles. Tissue Antigens F, Witt C, Christiansen F (2001) HLA-DRB1
50(1):74–76 DNA sequencing based typing: an approach
6. Kotsch K, Wehling J, Kohler S, Blasczyk R suitable for high throughput typing including
(1997) Sequencing of HLA class I genes based unrelated bone marrow registry donors. Tissue
on the conserved diversity of the noncoding Antigens 57(1):46–54
regions: sequencing-based typing of the HLA-A 13. Sayer DC, Whidborne R, De Santis D,
gene. Tissue Antigens 50(2):178–191 Rozemuller EH, Christiansen FT, Tilanus MG
7. Voorter CE, de Bruyn-Geraets D, van den (2004) A multicenter international evaluation
Berg-Loonen EM (1997) High-resolution of the single-tube amplification protocols for
HLA typing for the DRB3/4/5 genes by sequencing-based typing of HLA-DRB1 and
sequence-based typing. Tissue Antigens 50(3): HLA-DRB3,4,5. Tissue Antigens 63(5):
283–290 412–423
86 L.K. Smith

14. Dunn PP, Day S, Williams S, Bendukidze N HLA Class I and II Loci, Version No: 6.0,
(2005) HLA-DQB1 sequencing-based typing Feb 2011
using newly identified conserved nucleotide 18. Applied Biosystems (2002) BigDye®
sequences in introns 1 and 2. Tissue Antigens Terminator v3.1 cycle sequencing kit protocol.
66(2):99–106 © Copyright 2002, Applied Biosystems
15. Van Dijk A, Melchers R, Hilkes Y, Rozemuller 19. Applied Biosystems (2002) 3730/3730xl DNA
E, Tilanus M (2007) HLA-DRB sequencing- analyzers user guide. © Copyright 2002.
based typing: an improved protocol which Applied Biosystems
shows complete DRB exon 2 sequences and 20. Rozemuller E (2000) SBT resolution and ambi-
includes exon 3 of HLA-DRB4/5. Tissue guities (Chapter 6). In: Tilanus MGJ, Hansen
Antigens 69(Suppl 1):61–63 JH, Hurley CK (eds) IHWG technical manual
16. Applied Biosystems (2002) 3730/3730xl DNA genomic analysis of the human MHC: DNA-
analyzers sequencing chemistry guide. © based typing for HLA alleles and linked
Copyright 2002, Applied Biosystems Polymorphisms. Publication of the 13th
17. Conexio Genomics Pty Ltd (2011) CG-HLA International Histocompatibility Working
sequence based typing kits™ Instructions for Group. Fred Hutchinson Cancer Research
Use, PCR amplification and sequencing of Centre, Seattle, USA
 Chapter 6

Data Analysis of HLA Sequencing Using Assign-SBT

v3.6+ from Conexio
Carla Wirtz and David Sayer

Abstract
DNA Sequencing is now a standard frontline high-throughput HLA typing procedure with some
unrelated bone marrow donor registry typing laboratories performing tens of thousands of tests per
year. The advantage of DNA sequencing is that, by definition, sequencing directly identifies all bases in
the DNA template. Alternative molecular-based assays such as the use of sequence-specific PCR primers
(PCR-SSP) and oligonucleotide probes (PCR-SSO) provide information only for those regions to which
the oligos are designed and no information is obtained for the regions between primers and probes.
The era of routine high-throughput sequencing-based typing (SBT) was made possible by the devel-
opment of locus-specific PCR-based assays and the development of the HLA sequencing-based typing
software, Assign-SBT v3.2.7 by Conexio Genomics. A single PCR per locus simplified the template prepa-
ration stage of the test and Assign-SBT simplified the sequence analysis and allele assignment stage.
Together these developments dramatically simplified the SBT procedure, making SBT cost effective.
This chapter provides a comprehensive description of Assign-SBT sequence analysis software for use
in a HLA typing laboratory.

Key words: High throughput DNA sequencing, Sequence analysis, Assign-SBT, HLA

1. Introduction

1.1. Background DNA Sequencing is now a standard frontline high-throughput

HLA typing procedure with some unrelated bone marrow donor
registry typing laboratories performing tens of thousands of tests
per year. The advantage of DNA sequencing is that, by definition,
sequencing directly identifies all bases in the DNA template.
Alternative molecular-based assays such as the use of sequence-
specific PCR primers (PCR-SSP) and oligonucleotide probes
(PCR-SSO) provide information only for those regions to which
the oligos are designed and no information is obtained for the
regions between primers and probes. As a result of the increase of

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_6, © Springer Science+Business Media New York 2012

87
88 C. Wirtz and D. Sayer

HLA typings that are now performed by sequencing, there has

been an explosion in the description of new HLA alleles (see
IMGT/HLA Database, http://www.ebi.ac.uk/imgt/hla/).
The era of routine high-throughput sequencing-based typing
(SBT) was made possible by the development of locus-specific
PCR-based assays by Malcolm McGinnis, Pete Krausa, and Jason
Stein at Applied Biosystems (Foster City, CA, USA) and then
Forensic Analytical (South San Francisco, CA, USA) and the devel-
opment of the HLA sequencing-based typing software, Assign-
SBT v3.2.7 by David Sayer and Damian Goodridge at Conexio
Genomics (Fremantle, WA, Australia). A single PCR per locus
simplified the template preparation stage of the test and Assign-
SBT simplified the sequence analysis and allele assignment stage.
Together these developments dramatically simplified the SBT pro-
cedure, making SBT cost effective.
Assign-SBT v3.2.7 was specifically designed for HLA typing in
a high-throughput routine laboratory. It contained a unique base
caller designed for accurate base calling of heterozygous sequence
and a unique quality scoring system that enabled poor quality sites
within a sequence to be identified and base call errors, if present to
be easily identified. Any editing resulted in an immediate change
to the results. Furthermore, the software was the first of its kind to
provide a quantitative quality control system, in the vein of
Shewhart analysis, which enabled sample-to-sample, run-to-run,
and assay-to-assay assessment of quality.
Assign-SBT has continued to evolve, not only to keep up with
the complexities of HLA typing but also to provide researchers
with the tool for continuing the research into HLA complexity and
its role in transplantation biology and disease susceptibility.
Assign-SBT v3.6+ is the latest version of HLA sequence anal-
ysis software from Conexio; it represents a new breakthrough in
HLA sequence analysis software.
Features of Assign-SBT v3.6+ include:
(a) The choice for the user to decide if they wish to genotype by
comparing against cDNA and genomic DNA reference
sequences from the IMGT/HLA database.
(b) The most significant improvement over Assign-SBT v3.6,
providing multiple report formats, includes:
– Indicators of which alleles are rare and which have been
reported in a quantifiable frequency, the so-called CWD
alleles.
– Grouping alleles according to their nucleotide sequence
identity in exons 2 and 3 of class I HLA genes and exon 2
of class II HLA genes (G groups).
– Grouping alleles according to their amino acid sequence
identity in exons 2 and 3 of class I HLA genes and exon 2
of class II HLA genes (P groups).
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 89

– Improved reporting of hemizygous sequencing primers

for the resolution of heterozygous ambiguities.
Maximum benefits of Assign-SBT v3.6+ are obtained when
using Conexio’s Resolver SBT assays.

1.2. About Assign SBT 1. Computer operator system’s

3.6+ Assign is a Windows-based program and will run on Windows
XP, Windows Vista, and Windows 7 operating systems. Microsoft
1.2.1. Compatibility
Excel 97 or above is required for the creation of reports.
2. Data files supported
Assign requires .ab1, .abd, or .scf files from automated DNA
sequencers. Sequence files from non-Applied Biosystems DNA
Sequencers can be analysed. All files must be processed with
Sequence Analysis software prior to importing into Assign.

1.2.2. Overview 1. Functions and features

(a) Sequences from multiple loci can be imported into the
same layout.
(b) The sample identifier, sequence electropherogram, and
allele assignment results are all visible on one screen.
(c) Sequence editing includes priority analysis of positions of
low quality and positions that are mismatched with closely
related alleles.
(d) Simultaneous analysis of sequences for the resolution of
heterozygous ambiguities is possible.
(e) Analysis of non-coding sequence for Class I alleles is
possible.
(f) Assign 3.6+ can report CWD alleles, G groups, and P
groups (see Note 1).
(g) Assign 3.6+ includes sample-to-sample and run-to-run QC
analysis.
2. Performance characteristics
(a) Throughput: Assign 3.6+ has successfully imported over
5,000 sequence electropherograms into a single project.
(b) Base Call Accuracy: Assign contains a unique base caller
developed to improve the accuracy of heterozygous base
calls. However, base call accuracy is influenced by sequence
data quality. Generally, samples with a base call quality
score (described in detail below) of >85 do not have incor-
rect base calls.
3. Limitations
(a) Sequence with background noise and poorly separated
peaks may result in incorrect base calls and the potential for
incorrect typing results. This is true for all sequence analysis
90 C. Wirtz and D. Sayer

software. However, this limitation is offset as poor quality

data are easily identified within the software, so that it can
be reviewed for incorrect base calls. Such data can be
excluded from analysis, so that an incorrect genotype
cannot be performed.
(b) Assign-SBT v3.6+ compares a sample sequence with a
library of sequences from known alleles. The report lists
those alleles/allele combinations in the library that are
identical to the sample sequence. It is possible that the same
sequence could be derived from alleles yet to be described
and whose sequence is not yet part of the library. Therefore
caution must be taken when interpreting the genotype
report as an HLA type.

2. Getting Started
and Using
the Software
Assign-SBT v3.6+ is a standalone computer software program and
2.1. Installation should be installed on the computer on which SBT analysis is
performed. It is recommended that Assign is installed by a user
with complete administrator access to the computer. The installer
package can be acquired by contacting Conexio via their website,
http://www.conexio-genomics.com. It is also helpful if the com-
puter has access to the Internet to facilitate the system updates
with new libraries and other files as needed.
1. To install: Double click on the installer file icon and follow the
instructions for installation. During installation, a 16-digit
computer-specific hardware identifier will be generated. Copy
and paste this ID into an e-mail and send to key@conexio-
genomics.com to obtain licence keys. The software will not be
functional until the key files are successfully installed.
2. Reference libraries and NMDP codes, if required, need to be
installed. Go to the Conexio Genomics website, http://www.
conexio-genomics.com, and click on the Downloads tab on
the left side of the screen. Click on the Libraries tab. Click
on the library version that you require. A zip file will be down-
loaded to your computer. Unzip the file and save the References
folder to the desktop or other convenient location. The
References folder will contain the gene-specific reference files,
P Group file, and G Group file.
(a) Next, launch the software. Use the default operator (admin)
and password (cg01) login. Click on Help | Update
System on the top menu bar.
(b) Click on the References button and navigate to the
unzipped References folder. Highlight the .xml reference
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 91

Fig. 1. The Update screen is used to update Keys, Reference files, and NMDP Codes.

files to update then click Open. The references will be

imported to the correct location within the software. A
message will appear indicating if the reference file update
is successful or not. If the update is unsuccessful, the pro-
cess should be repeated to verify that the correct file was
chosen for the update (Fig. 1).
3. Repeat the process with the NMDP codes by clicking on the
NMDP Codes tab on the Conexio website. The NMDP code
file is named “numer.txt” and must not be renamed. Save the
.txt file to your computer. Log into the software and click on
Help | Update System | NMDP codes. Navigate to the
numer.txt file and highlight it. Click Open. The code file will
be imported to the proper location within the software.
4. When the licence key(s) are received, use the Keys update
function within the software by clicking on Help | Update
System | Keys.

2.2. Login and Adding 1. Launch the software by double clicking on the Assign 3.6+
Users icon on the desktop. The default operator is “admin” and the
default current password is “cg01”. It is recommended that
the admin password is never changed.
2. Additional users can be added. Enter the admin login and
password then click “More”. Below will be a section to add
additional users. Type in the new operator’s name in the Edit
Operator section. Type in a password for that user. Re-type the
password. Select the Operator Level. Click Add/Update
directly next to the Retype Password box. Repeat for addi-
tional users (Fig. 2).
92 C. Wirtz and D. Sayer

Fig. 2. The Operator Login screen is used to add/edit users and select Settings files.

3. To launch the software under a particular user, double click on

the Assign icon. In the Operator drop down, select the user.
Type in the password then click Submit.

2.3. Settings The settings menu enables the user to configure the software for
their requirements. Settings can be saved as different settings files
to enable the software to be individualized for different users.

2.3.1. General General Settings allow modifications to the interface such as

changes in font size, electropherogram’s colours, and electro-
pherogram’s line thickness.
1. To open the default settings file, click on Edit | Settings on the
top menu bar in the software. The ‘General’ tab displays the
default settings and allows for creation of new settings files.
2. Customize the display
(a) Click on Display in the General tab.
(b) The Display options will appear.
(c) Adjust the base colours, background colours, text size, and
line width (EPG tracing width) then click Done (Fig. 3).
3. To create a new settings file if desired, type in a new settings
file name and proceed to the Advanced tab to create the
naming conventions and locus alias names.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 93

Fig. 3. The screen display is modified by selecting a colour scheme customized for
the user.

2.4. Naming In the Advanced tab, the user can enter the parameters that define
Conventions the sample name and the locus identifier in sample sequence
filename. The sequence filename must be unique for sequencing
reaction for a sample and contain the sample name and an identifier
that can be used as an alias for the locus being genotyped. If a stan-
dard system is used by the operator, analysis will occur automati-
cally and different data from different loci can be entered into the
same layout (Fig. 4).

2.4.1. Sample Delimiters Sample Delimiters have been used to separate the components
of the sequence-file name.
Example:
Sequence-file name A01[12345_DQ2F_A01
Delimiters have been used to separate the components of the
sequence-file name
[ has been used to separate the PCR number (A01) and
the sample name (12345).
_ has been used to separate the sample name and the locus
(HLA-DQB1) and primer (exon 2 Forward).
_ has also been used to separate the locus and well location
(A01).
Set the Naming convention by defining the Sample
Delimiters.
In the example above, the sample name begins with [ and
the sample name ends with _Enter [ in the Start string box, and
enter _ in the End string box.
94 C. Wirtz and D. Sayer

Fig. 4. Settings | Advanced tab is used to customize sample naming conventions and reference aliases, select Nomenclature
output naming, CWD Allele set, and to indicate if genomic references are to be used.

2.4.2. Alias Names 1. In the example above, DQ is used as a symbol for HLA-DQB1.
In Subheading 2.4.7, use the Ref: drop down menu to select
the locus. Next, in the Alias drop down menu select the alias
used (DQ). If the alias is not present, type it into the Alias box,
and then click Update directly to the right of the Ref box.
2. Repeat this for each locus alias you will be using. After all aliases
have been added, click on the Update button in the lower
right-hand corner of the Settings box.

2.4.3. Nomenclature The HLA nomenclature standards changed in April 2010 from v2
to v3.0. In Select the Output Naming Standard, either IMGT/
HLA3.0 or 2.0 can be selected for the naming convention of the
reported alleles.

2.4.4. Heterozygous Heterozygous Ambiguity Resolving Primers (HARPs) are sequenc-

Ambiguity Resolving ing primers designed to sequence only one of the alleles in a
Primers heterozygous sample. HARPs are used for the resolution of
heterozygous ambiguities, by producing hemizygous sequence for
one of the alleles. The sequence data from an HARP is combined
with the existing sequence data from the sample to produce a high-
resolution genotype report. Setting the Start Gap enables only
good quality sequence to be analysed by eliminating the poor
quality data at the start of a sequence. The Start Gap will vary
between laboratories according to the sequence reaction clean-up
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 95

method used, the DNA Sequencer make and model, and the polymer
in the sequencer capillaries.
1. Enter the number of base pairs between the end of the primer
and the first usable sequence generated in the Start Gap.

2.4.5. CWD Alleles The CWD alleles are stored in a text file. A default set of CWD
alleles based on those described by Cano et al. (1) are provided
with the software. However, laboratories may want to create their
own list.
1. Set the CWD allele set you will be utilizing, if any.

2.4.6. Extended Analysis By selecting Genomic References, the sequence analysis will
enable the comparison of the sample sequence with the Genomic
References provided by the IMGT/HLA database. This analysis
improves typing accuracy by enabling those alleles that are charac-
terized by polymorphisms to be genotyped. By leaving the
Genomic References box unchecked, analysis will be performed
against the standard cDNA reference.
1. If the laboratory will utilize analysis of the non-coding regions,
put a check in the Use Genomic References box.

2.4.7. Reference Aliases Reference aliases are used in the sequence filenames to indicate the
HARP being used for sequencing.
1. Click on the References tab of the Setting menu to establish
alias names for primers used.
2. Click on the Load Reference box at the top right of the screen.
Navigate to the References folder and click on the reference
needed. Click OK.
3. The reference information will populate the screen.
4. To establish a reference alias for the Codon 86 resolution
primer, highlight the DRB1.xml file and click OK.
5. In the References tab, in the lower left corner drop down
menu, select the Codon 86 primer. To the right of the primer
will be the alias drop down menu. Click the drop down menu
to determine if your naming alias is present.
6. To add an alias, type the alias into this box. For example, DR86
as an alias for the Codon 86 GTG primer. Next, click on Add
Alias directly below the alias drop down menu box. Then, click
on Update in the lower right corner. After all aliases have been
added, click on Done.

2.4.8. Variant Positions Variant Positions is a tool to draw attention to sequence artefacts
that may result in base call errors, or other positions within a
sequence where automated base calls may be frequently incorrect.
96 C. Wirtz and D. Sayer

Fig. 5. User Defined Variant positions are established using the References tab in the Update Settings menu.

Such positions can be included as those positions within the

sequence that must be validated by the user in order to generate a
report (Fig. 5).
1. Load the reference for which you want to create a variant posi-
tion following the instructions above.
2. Once the reference is loaded, click on Show then Variants.
This will open the variant position box.
3. In the lower left corner, in the blank box under Position, type
in the position for the variant.
4. In the Variant drop down, select the variant base type (usually
* so any call at that position is flagged).
5. Enter the length of the variant in bases.
6. In the Insert box, enter bases if this variant includes an inser-
tion. Leave blank if no insertion is expected.
7. Select the Class of variant (usually User Edited). Enter any
comments desired.
8. Click on Add/Update to add the variant position.
9. Repeat for additional variants.
10. Click on Update then Done when complete. A purple box will
be displayed at each variant position indicated.

2.5. Importing If the sequence file-naming convention is defined in the soft-

Sequences ware Settings, sequences can be imported by browsing to a
directory and importing the contents or by importing the
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 97

Fig. 6. The Import Files menu is accessed by selecting File | Import | Electropherograms
from the top menu bar.

sequence files individually. Importing from a directory also

allows filters to be applied, so that only specific samples are
imported, or only sequences from a particular locus or from a
sequencing primer can be imported.

2.5.1. Importing 1. To import sequences by directory (import all electrophero-

Sequences by Directory grams in a given folder), Select File | Import | Electro-
pherograms on the top menu bar.
2. In the pop up window, click on Browse and navigate to the
folder that contains the sequences. Highlight the folder and
click OK (Fig. 6).
3. The folder location will populate the Import Files menu. Check
the Import All Subdirectories box if all subdirectories are to be
imported. Click Go.

2.5.2. Importing 1. To import only selected electropherograms from a folder, click

Sequences Individually on Select Files Manually. Navigate to the folder containing
the necessary sequences. Highlight all the sequences to be
imported using the Ctrl or Shift key. Then click Open.
2. Use the Filters dialogue to import to search for an import
samples of a specific Name, or all samples for a Locus or
specific sequence files for a sequencing Primer.

2.6. The Screen Once the sequence data has been imported, the software screen
Layout, Editing, becomes populated with sample, sequence, and result information.
and Analysis There is a predominance of white to red shading used to demon-
strate sequence data quality.
98 C. Wirtz and D. Sayer

Fig. 7. Shading within the electropherogram coverage indicates quality of given sequences.

The use of shading to indicate sequence data quality

SBT errors may occur if a base call error is made, and the prob-
ability of a base call error is increased if the quality of the data is
poor. Assign-SBT contains a quality scoring algorithm that assesses
the quality of a sequence peak based on the peaks shape, whether
or not it is well separated from neighbouring peaks and whether or
not there is non-specific background.
A Base Call Score (BCS) from 0 to 50 is calculated from each
peak and is represented in a box under the base call as a shade of
red to white, where red is a BCS of 0 and white has a BCS of 50,
shading in between 0 and 50 is shaded accordingly. The consensus
sequence BCS is calculated from the BCS of sequences that con-
tribute to a consensus.
The BCS for positions within a sequence can be used to calcu-
late a quality score of a sample.
The use of colour coding enables a sample with poor quality
data and poor quality positions within the sequence to be readily
identified and checked for base call errors (Fig. 7).

2.6.1. Sample ID Pane The screen layout shows information for a particular sample.
It includes the sample ID, the electropherograms data, the aligned
sequences for a sample, and the best-matched allele combina-
tions (Fig. 8).
The samples imported are listed on the left side of the screen.
Sample names are colour coded to indicate data quality.
1. The sample pane also includes five columns of boxes.
6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 99

Fig. 8. The Sample ID Pane shows information about the samples in the project.

(a) A light blue box in the C column indicates if comments

have been made about a sample. Right clicking on the
sample allows the comments to be added and reviewed.
The comments are also included in the genotype report.
(b) A green box in the A column indicates that the sample has
been verified at all positions indicated in the Navigator.
This box changes to green once all the positions requiring
confirmation have been confirmed using the navigator (see
below for Navigator bar use).
(c) A green box in the 1 column indicates that the sample has
undergone first review. After the first reviewer has per-
formed the analysis, the yellow box in the 1 column is
clicked, changing it to green.
(d) A green box in the 2 column indicates the sample has
undergone secondary review. Checking this box will lock
the sample and prevent any further edits unless the box is
manually unchecked.
(e) A green box in the R column indicates the sample can be
reported using the Report Generator.
100 C. Wirtz and D. Sayer

Fig. 9. Right click on Sample ID to access Sample Options section.

2. Right click on the sample name in the Sample Pane to access

sample options (Fig. 9).
(a) Show Comments will display any quality warnings about a
sample.
(b) Edit Comments gives a text box to record any comments
about a sample. These comments will appear on the report.
A light blue box in the C column will indicate that a com-
ment is present.
(c) Reanalyze will remove any edits and trims that have been
made.
(d) Add New Samples will launch the import files menu.
(e) Remove Sample will remove the highlighted sample from
the project.
(f) Remove All will remove all samples from the project.

2.6.2. Sequence Importing the sequence electropherograms results in a display of

Electropherograms how the sequence files are orientated according to the gene
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 101

Fig. 10. Sequence Electropherogram information is contained in the Sample layout screen, including the structure of
the gene being sequenced, sample sequence alignments, consensus sequences, and sample sequence data.

Fig. 11. The structure of the gene being sequenced is represented at the top of the screen.

structure, the sequence electropherograms themselves, and the

Assign-SBT base calls and quality score information (Fig. 10).
The Assign layout contains important information to assist
with the analysis of DNA sequence data.
1. The Structure of the Gene being sequenced
The blue band shows the genetic structure of the reference
sequence and the yellow bars above indicate those positions
within the sequence that when the base call is changed will
result in change to an alternative allele combination in the
results pane (Fig. 11).
2. Sample Sequence Alignments
The bands shaded white to red show the sequence data align-
ments. They are shaded white to red according to quality. This
enables “at-a-glance” to be able to locate poor data quality
positions for manual review. The sequence filename and direc-
tion of sequencing (< or >) is included (Fig. 12).
3. The Library Consensus sequence
Beneath the sequence alignment map is consensus sequence
of alleles within the library (Fig. 13). The sequence is shaded
yellow to indicate exon sequence and white to indicate intron
102 C. Wirtz and D. Sayer

Fig. 12. The sample sequence alignment shows sequence quality shaded within each electropherogram. Lighter electro-
pherograms represent better quality sequence.

Fig. 13. Library Consensus sequence is colour coded depending on region. This sequence represents all alleles within the
reference library.

Fig. 14. Sample Consensus Sequence of the sample being sequenced appears just below the Library Consensus sequence,
just above the electropherogram tracings.

sequence. In addition, positions shaded light blue indicate

there are alleles in the library that contain deletions at this
position and dark blue regions indicate the position of inser-
tions in some alleles.
4. The Sample Consensus Sequence
The sequence data below the consensus sequence of alleles
within the library is the consensus sequence of the sample
being sequenced. The boxes underneath the base calls are
shaded white to red to indicate the quality of the consensus
base call. It is the consensus sequence that is compared to the
sequence library and it is only the consensus sequence where
base call edits can be made (Fig. 14).
5. The Sample Sequence Data
Beneath the sample consensus sequence is the sample sequence
electropherograms data, the software base calls, and the quality
indicator. The electropherogam panels contain the sequence
filename, the “Sensitivity” of heterozygous base calls. That is,
the percentage that one peak needs to be within another
before a heterozygous base call is made on data with no back-
ground. The signal intensities of the four bases are also indi-
cated (Fig. 15).
6. Right clicking on a given electropherogram gives access to
options for each electropherogram (Fig. 16).
(a) Set Start Base will trim off all data to the left of the cursor.
(b) Set End Base will trim off all data to the right of the cursor.
(c) Less Sensitivity will filter out background noise, raising
the threshold 5%.
(d) More Sensitivity will increase sensitivity, calling more
heterozygous bases.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 103

Fig. 15. Sample sequence data, software base calls, and quality indicators are represented within the electropherogram
tracing.

Fig. 16. Right click on the electropherogram to view Electropherogram options.

(e) Exclude Het InDels will filter out excess background in

the sequence due to a co-amplified stretch of DNA. When
the signal to noise is relatively low, the extra signal can be
interpreted as a heterozygous InDel. Choosing “Exclude
Het InDel” informs the software that the additional signal
104 C. Wirtz and D. Sayer

Fig. 17. Amino Acid sequence can be viewed instead of base pair sequence by pressing Ctrl + A in the electropherogram
pane.

should be interpreted as background and not as a real

sequence feature.
(f) Reanalyse EPG will remove any user edits and trims from
the electropherogram.
(g) Deactivate EPG will remove the electropherogram from
analysis, but does not remove it from the layout. Right
clicking on the deactivated electropherogram again will
result in the Reactivate EPG option.
(h) Show Warnings will display any quality warnings about
that particular electropherogram.
7. Amino Acid Sequence can be viewed instead of base pair
sequence (Fig. 17).
(a) Pressing Ctrl + A in the electropherogram pane shows the
amino acid sequence of the reference sequence and the
consensus sequence of the sample. “Z” is used for heterozy-
gous sequences. Clicking on the Z or scrolling to the
position within the sequence will show the amino acids at
these positions. This will assist with understanding the
consequence of novel alleles.
8. Coloured boxes above the reference sequence are used to
highlight specific sites within the sequence (Fig. 18).
(a) Yellow boxes indicate positions within the sequence that
differ from allele combinations in the results pane.
(b) Green boxes appear when the base call has been confirmed
using the Navigator (see below).
(c) Blue boxes appear when a position has been edited.
(d) Purple boxes indicate a user-defined variant position (see
Subheading 2.4.8).

2.6.3. Results Pane The Results pane includes the sample name, the start and stop
positions of the test sequence, and the allele combination with the
best-matched sequence to the test sequence. Bolded alleles are
CWD alleles whilst those unbolded are and rare alleles (Fig. 19).
1. Allele 1 and Allele 2 columns display the allele pairs best
matched to the test sequence.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 105

Fig. 18. Coloured boxes above the Sample Consensus sequence indicate points of
interest in the sequence.

Fig. 19. The Results pane includes sample name, start/stop positions, and best-matched allele combination.

2. The MM0 column is the number of mismatches between the

test sequence and the sequence of the alleles.
3. Additional columns, labelled MM1, MM2, etc., will be pres-
ent when heterozygous ambiguity resolving sequencing
primers are used. These columns indicate the number of mis-
matches between the test sequence and sequence of the allele.
4. The N-C column indicates the number of mismatches in the
non-coding region (Fig. 20).
5. The IND column contains mismatch information in the
heterozygous insertion/deletion (indel) data. The software is
sensitive to runs of mixed bases since many introns do have
heterozygous indels.
When there is background in the sequence due to a co-amplified
stretch of DNA or when the signal to noise is relatively low,
the extra signal can be interpreted as a heterozygous InDel.
Choosing “Exclude Het InDel” informs the software that the
106 C. Wirtz and D. Sayer

Fig. 20. Non-coding and Indel data are viewed in the N-C and IND columns, respectively.

Fig. 21. The Navigator enables editing, review, and movement from sample to sample
within the project.

additional signal should be interpreted as background and not

as a real sequence feature.
If either the N-C or IND column is shaded pink, the data from
these regions are NOT included in the analysis. To activate the
N-C and IND layers to include the data in the analysis, click on
the appropriate column. Clicking again will deactivate the
layers. When the N-C and IND layers are activated, the data
are included in the analysis and the data within these layers
must be edited (see Notes 2 and 3).
6. The Differences column indicates the regions within the
reference sequence that contain the sequence differences
between the ambiguities.

2.7. Navigator The navigator enables sequence editing, moving between samples,
and moving between positions within a sequence. Importantly, the
Navigator is used to Validate automated base calls (Fig. 21).
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 107

1. Positions for Validation can include positions with a low Base

Call Score (BCS), Edited positions, potential MisMatch
positions, and user-defined variant positions (box to the right
of the MM).
(a) Base Call Score (BCS) is a quality measurement that is
determined by peak spacing, presence of background noise,
and signal strength. The BCS for each base appears above
the BCS in the navigator.
(b) Bases with a BCS of lower than 70 (or lower than 35 for
single direction sequences) will be included in the priority
review using the red X or the double arrows.
2. Moving to positions for Validation is performed by selecting
the double arrow buttons or using the red X button. The pres-
ence of the red X means that the cursor is at a position that has
not been validated by the operator. Clicking the red X validates
the base call and changes the red X to a green tick.
3. Selecting either single arrow button moves the EPG one posi-
tion left or right.
4. Selecting the up arrow moves to the sample above and select-
ing the down arrow moves to the sample below in the Sample
Pane.
5. The Master drop down menu selects between master sequences,
HARP sequences, Master-intron, or Master-indel sequences.
6. The No Offset drop down menu allows the user to chose the
base numbering motif desired.
7. Underneath the No Offset drop down menu are the codon
and base location.
To navigate to a particular base, enter the base position in the
right drop down and hit enter.
8. Clicking on the red X will confirm the base call and move to
the next position that meets the “to be validated” criteria. As
the red X is clicked, a green box will appear above the base
indicating that it has been confirmed. Once ALL validation
positions have been verified, the yellow box under the Audit
column in the Sample Pane will turn green.
9. If a base call needs to be edited, the call can be changed manu-
ally using the base letters on the Navigator. The raw data are
not changed with edits. The consensus sequence is changed,
and these changes are recorded in the saved project. When the
project is opened, the changes are applied again to the raw data.

2.8. Other Sample 1. Resizing the EPG

and Sequence Editing The EPG can be resized by pressing the Shift key and the up/
Functions down or left/right arrows on the computer keyboard.
108 C. Wirtz and D. Sayer

Fig. 22. View options allows the data and reference library data to be viewed in different ways.

2. Hiding EPG traces

Pressing the computer keyboard Shift key and one of the
letters representing the four bases simultaneously will remove
the trace of this base from the EPG. Repeating the process will
return the trace.
This function is useful if heterozygyous peaks are perfectly
overlaid and the base call requires confirmation.

2.9. View Options The View options enable the samples sequence data and the
sequence data from the alleles in the library to be viewed in differ-
ent ways (Fig. 22).
1. Status Bar shows the status of the project at the bottom of the
screen.
2. Panes
(a) Electropherogram is the default and displays the electro-
pherogram tracings of the sample.
(b) Consensus displays the consensus sequence for all samples
in a project.
(c) Quality displays the consensus sequence shaded according
to the consensus sequence BCS for each base for every
sample in the project.
(d) Alignments displays the consensus sequence for each of the
possible allele combinations for a given sample. Mismatches
with the sample consensus appear highlighted in yellow.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 109

(e) Reference Alleles shows the sequence of alleles within the

library compared to the sequence of the selected sample.
Differences are highlighted in yellow. The alleles are shown
in the results pane.
The user can select specific allele sequences to align together
by typing the allele names in the Navigator.
3. Letters or Dots
(a) Selecting letters will show the bases for alignments and
reference as letters.
(b) Selecting Dots will show dots at each base where the align-
ment or reference matches with the consensus sequence.
Bases that differ will be shown as letters.
4. Nucleotides or Codons
(a) Selecting Nucleotides will show the base numbering.
(b) Selecting Codons will show the codon numbering.
5. When using the coding sequence reference, the View
Unaligned option will include or exclude the intronic overlap
of the sequences between exons.
6. View All EPG enables the HARP electropherograms to be
seen with the EPG of the F and R sequences. Note that the
highlighted sequence in the results pane (yellow column) is
highlighted in the EPG pane. By selecting the HARPs layer,
the HARPs EPGs become highlighted.
7. Filter Confirmed
As positions are confirmed using the red X, alleles within the
results pane are excluded and only those alleles with 0 mis-
matches with the test sequence remain. To restore the list of
possible allele combinations, unclick on the Filter Confirmed
option in the view pane.

2.10. Data Analysis Log into the Assign 3.6+ software and select the settings file desired
and Editing EPGs by clicking on Edit then Settings. Select the settings file and
click Done.
2.10.1. Logging on

2.10.2. Importing Data 1. Import the desired dataset using either the Browse option to
import an entire folder of data or the Select Files Manually to
selectively import files.
2. Imported samples will be displayed as a list on the left side of
the screen.
3. The Electropherogram data will be in the centre of the screen.
4. The Allele assignments for the active (highlighted in blue)
sample will be displayed at the right side of the screen.
110 C. Wirtz and D. Sayer

5. Each of these panes can be sized by dragging the frame to the

desired width to optimize the amount of electropherogram
data displayed.
6. Resize the electropherogram peaks if desired.

2.10.3. Navigation 1. The Navigator box is used to navigate through the data
checking the critical bases.
2. Set the desired bases to be audited in the Navigator box: BCS,
Edits, MM, and variant positions.
3. Highlight the sample to be reviewed by clicking on it in the left
sample pane.
4. Navigate to the first base in the sample by clicking on the left
arrow with the bar in the navigator.
5. Using the red X button, navigate through the sequence verify-
ing all desired bases. As each base is confirmed, the red X will
change to a green check and the cursor will move to the next
base to be verified. A green box will appear above each base
pair that has been verified using this method.
(a) There should be at least one allele pair in the Results Pane
on the right that indicates no mismatches in the MM0
column.
(b) Once the Master Layer has been reviewed, the Navigator
will take the user to any resolution primer layers that are
present. These layers must also be reviewed before the
sample analysis is complete.
(c) Once all priority review bases have been verified, the red X
will change to a green check mark indicating no additional
bases need confirming. In the Sample Pane, the box under
the A column (Audit) will turn green.
(d) Clicking on the box under the 1 column will indicate the
sample has been reviewed once.
(e) At this point, the project should be saved to prevent any
accidental loss of data review. Click on File then Save As.
Select a file name and location to save the project. The
saved project (.xml format) indicates which electrophero-
grams were used including their saved location, any edits
and confirmations that were done, and information about
the user.
(f) If Genomic References were selected, the N-C and IND
column may be present. Refer to the differences column to
determine if analysis of the non-coding region would be
beneficial. Click on the N-C column to highlight it in yellow.
Navigate to the mismatch positions and make any edits
necessary. Once all review and edits have been made, click
back on the MM0 column to view the allele pairs.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 111

Fig. 23. Example of a sample with an Insertion or Deletion (InDel) in exon 4.

(g) Repeat on the IND column if desired. The IND column

should be reviewed only when there is a clear insertion or
deletion in the sample. Poor quality or mobility shifts can
trigger the Heterozygous Insertion or Deletion warning.
Use the Exclude Het InDel option to eliminate any warn-
ings due to background noise and not true InDels. Below
is an example of a clear InDel (Fig. 23).
6. Once all the review has been completed on the first sample,
click on the second sample and repeat the process.
(a) Save your work often to prevent any loss of review.
7. If a secondary review is desired, the project should be opened
by the second reviewer to preserve the audit trail. The second
reviewer logs onto the software and selects Open in the File
menu. Navigate to the saved project (.xml file). The software
will locate the raw data, import it, and apply all changes and
verifications that have been performed by any previous
reviewers.
(a) Select the desired priority review positions in the Navigator
box. Use the left double arrow button to navigate to the
required positions. Using the big red X will cancel and
reapply the audited positions.
(b) Once all positions have been reviewed, the second reviewer
can click on the box in the 2 column in the sample pane.
(c) Once the second review box is checked, the sample is
“locked” and no more edits can be made unless the box
is manually unchecked.
(d) Save often to prevent the accidental loss of data review.
112 C. Wirtz and D. Sayer

Fig. 24. The MM2 column indicates the presence of a resolution primer to resolve heterozygous ambiguities.

Fig. 25. The Resolving layer is activated to allow review of resolution primer data.

2.10.4. Resolution Primer 1. HARPs are used to resolve heterozygous ambiguities within
Layers the regions sequenced. HARPs target one of the alleles present,
creating a hemizygous sequence. In the example below,
multiple ambiguities exist after analysing the master sequence
as indicated in the MM0 column highlighted in yellow.
Heterozygous sequence is present (Fig. 24).
2. Selection of an HARP that will detect only one of the alleles in
the pair will result in a hemizygous sequence in the MM1 layer.
This will eliminate some of the heterozygous ambiguities.
Allele combinations with no mismatches in the MM0 and
MM1 column will be included in the report. Alleles with mis-
matches in the MM1 column have been eliminated by the
HARP sequence (Fig. 25).

2.10.5. Auditing All user interaction is logged in the audit trail. The software logs
time and date any edit was made, when the project was saved and
the user performing the action. This information can be printed on
the allele report along with the sample allele assignments.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 113

Fig. 26. Genotyping Report options allows for customized reports.

2.11. Reporting The Assign-SBT reports enable a comprehensive assessment of the

sequence data. A genotyping report lists the best-matched alleles to
the sample sequence and also enables CWD alleles to be indicated.
The genotype report also enables alleles to be reported as functional
groups by reporting G groups and P groups and enables the user to
structure the report specific for their requirements. The software can
also report the HARPs required to resolve the heterozygous
ambiguities.
To access the report functions, click on Reports then Report
Generator on the top menu bar.

2.11.1. Genotyping Report The Genotyping report is used to report the allele combinations that
have identical sequence to the sequence of the sample (Fig. 26).
1. The Full Report Section enables laboratories to create their
report format. The Sample drop down menus enable the lab to
include or exclude specific items from the report.
(a) The Sample section contains the Auditing, Match Summary,
G Groups, and P Groups reports.
(b) Match Summary will list all the possible allele pair combi-
nations. If the CWD option is selected, the CWD alleles
will appear in bold type in the report (Table 1).
114 C. Wirtz and D. Sayer

Table 1
Match summary option listing all possible allele pair
combinations

Sample Q950050964
Reference IMGT/B 3.3.0 2011-01-14
The allele pairs listed below are compatible with the consensus sequence
B*18:01:01 B*44:02:01:01
B*18:01:01 B*44:02:01:02S Intron 4
B*18:01:05 B*44:02:01:01 Exon 2
B*18:01:05 B*44:02:01:02S Exon 2, Intron 4
B*18:01:05 B*44:27:01 Exon 2
B*18:09 B*44:09 Exon 2
B*18:12 B*44:12 Exon 2
B*18:20 B*44:51 Exon 3
B*18:43 B*44:55 Exon 2
CWD alleles appear in bold

(c) The G Group Report enables those Class I alleles with

identical nucleotide sequence in exons 2 and 3 to be
reported under the same code and class II alleles that are
identical in exon 2. If the CWD option is selected, those
alleles that have been reported at a significant allele fre-
quency groups will appear in bold type in the report. HLA
alleles that have identical nucleotide sequences across the
exons encoding the peptide binding domains (exons 2 and
3 for HLA class I and exon 2 only for HLA class II alleles)
will be designated by an upper case “G” which follows the
first three fields of the allele designation of the lowest num-
bered allele in the group (Table 2).
(d) The P Group Report will list the alleles grouped according
protein sequences as encoded by exons 2 and 3 for HLA class
I alleles, and for HLA class II alleles this is based on identical
protein sequences as encoded by exon 2. P groups containing
CWD alleles will be bolded on the report (Table 3).
(e) The Auditing option will include a comprehensive audit
report including date, time, and identification of the opera-
tor validating the results.
2. The Layers section contains the Edit List, Electropherogram List,
Sequences, Mismatch List, and Mismatch table. These can be
included or excluded according to the needs of the laboratory.
6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 115

Table 2
G group report

Sample Q950050964
Reference IMGT/B 3.3.0 2011-01-14
The allele pairs listed below are compatible with the consensus sequence
B*18:01:01G B*44:02:01G
B*18:09 B*44:09
B*18:12 B*44:12
B*18:20 B*44:51
B*18:43 B*44:55

Table 3
P group report

Sample Q950050964
Reference IMGT/B 3.3.0 2011-01-14
The allele pairs listed below are compatible with the consensus sequence
B*18:01P B*44:02P
B*18:09 B*44:09
B*18:12 B*44:12
B*18:20 B*44:51
B*18:43 B*44:55

3. The Additional Information section can be used to add com-

ments specific for a typing run. These comments appear at the
top of the report.
4. Sort by is used to sort the report by Sample Name or Locus.
5. Summary Options includes options to add to the report:
NMDP Codes, HARPs used, Full + Part indicates which alleles
are fully typed in the IMGT database, and Differences indi-
cates where each allele pair differs from the others.
6. Audit Options allows the operator to choose to record all the
Save events for the project by clicking the Save box. Confirm
records change and priority base confirmation.
116 C. Wirtz and D. Sayer

Fig. 27. The heterozygous ambiguity resolving primers (HARPs) report indicates which resolution primers are required
to resolve ambiguities.

2.11.2. HARPs Report 1. The HARPs report indicates to the operator which HARPs are
required to resolve heterozygous ambiguities and which exons
are required to be sequenced to resolve allele ambiguities
(Fig. 27).
(a) The Output Filters can be used to filter for a single sam-
ple/locus or all samples/loci in a project.
(b) The Start Gap under Parameters can be set depending on
the gap between the end of the sequencing primer and
usable data. This is generally around ten for POP6 users
and longer for POP7 users.
(c) The operator can choose between a Full Report and a G
Groups Report under the Type section.
(d) The Full Report will list the HARP resolution for each
allele pair (Table 4).
(e) The G Group Report will group the resolution based on
the broader G Grouping of the allele pairs (Table 5).
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 117

Table 4
Example of a full HARPs report listing resolution for each allele pair and HARPs
required to resolve the ambiguities

Sample Q950050964
Reference IMGT/B 3.3.0 2011-01-14
Use ONE primer from group 1 C1-AC559-R-520 C1-TG539-R-507 C1-CT559-R-494
Use ONE primer from group 2 C1-GAA309-R-953
Splits B*18:01:01 B*44:02:01:01
B*18:01:01 B*44:02:01:02S Intron 4
B*18:01:05 B*44:02:01:01
B*18:01:05 B*44:02:01:02S Intron 4
B*18:01:05 B*44:27:01 Exon 4
B*18:09 B*44:09
B*18:12 B*44:12
B*18:20 B*44:51
B*18:43 B*44:55
Unresolved ambiguities remain within exon 4, intron 4

Table 5
Example of a G group HARPs report

Sample Q950050964
Reference IMGT/B 3.3.0 2011-01-14
Use ONE primer from group 1 C1-AC559-R-520 C1-TG539-R-507 C1-CT559-R-494
Use ONE primer from group 2 C1-GAA309-R-953
Splits B*18:01:01G B*44:02:01G
B*18:01:05 B*44:02:01G
B*18:09 B*44:09
B*18:12 B*44:12
B*18:20 B*44:51
B*18:43 B*44:55
Unresolved ambiguities remain within exon 4, intron 4
Resolution is based on the broader G grouping of the allele pairs
118 C. Wirtz and D. Sayer

Fig. 28. The FASTA report enables sequences in FASTA text format to be produced.

(f) The Output Format can be in a Text file, Excel (default),

or XML format.
(g) The Options section will determine if the report is gener-
ated using Splits or not. Clicking Show Splits will list all
the allele combinations and the HARPs needed to resolve
them. Leaving it unclicked will result in a list of required
HARPs only.

2.11.3. FASTA The FASTA report enables sequences in FASTA text format to be
produced. Selecting the sample, locus, layer, group, and region
provides a detailed description of the FASTA file in the FASTA file
name (Fig. 28).

2.11.4. Quality Reports The quality reports utilizes the BCS at each position to create qual-
ity control information for a sample, which can then be compared
between different samples to create quality control information for
the assay. The principle is that if the mean and standard deviation of
BCS for a number of sequence positions can be calculated, this will
provide a quality score for the region of sequence from which the
mean and standard deviation were calculated. This information can
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 119

Fig. 29. The Quality Report utilizes the base caller score at each position to create a comparison between different sam-
ples, providing quality control information for the assay.

then be used to calculate quality information for specific sequencing

primers, for different assays and different samples. The data can be
used to monitor the performance of a test and set performance cri-
teria that can be used when assessing changes, such as reagent batch
changes, or DNA extraction procedures, for example (Fig. 29).
1. The Quality report dialogue enables the user to select parame-
ters for quality analysis. In addition to the sample information,
the user can select across regions (i.e., an exons) within the
sequence to analyse, or a specific range of bases within a region.
2. Leaving the Get Projects default at _Current_ will produce a
quality report in the active Assign-SBT layout.
3. By Clicking Get Projects and then browsing to directories in the
_Current_ drop down menu, a quality report can be generated
from saved layouts within the browsed directories (Fig. 30).
The Quality Report shows the mean BCS in blue and the stan-
dard deviation in pink for all samples HLA-B exon 2 for a SBT run.
The red dots show the number of edits made for each sample.
120 C. Wirtz and D. Sayer

Fig. 30. The Quality Report shows the mean Base Call Score (BCS) represented by squares, and the standard deviation in
diamonds for all samples. The dots show the number of edits made for each sample.

3. Notes

1. CWD alleles, or common and well-documented alleles, are

those alleles with a calculated allele frequency in one or more
populations, G groups are those groups of alleles that share the
same nucleotide sequence in exons 2 + 3 of class I and exon 2
of class II, and P groups are those alleles that share the same
amino acid sequence in exons 2 + 3 of class I and exon 2 of class
II alleles. These options enable laboratories to determine the
probability of a particular allele or allele combination within an
ambiguous report, or if an ambiguous report contains func-
tional differences between alleles).
2. Note: If the software does not detect heterozygous insertion/
deletion data, the IND column is not created.
3. Note: If the Use Genomic References box is not checked in
the Edit/Settings/Advanced tab, the N-C column is not
created.
 6 Data Analysis of HLA Sequencing Using Assign-SBT v3.6+ from Conexio 121

4. Conclusion

This chapter provides a comprehensive description of the use of

Assign-SBT sequence analysis software for use in HLA sequencing-
based typing.
The sequencing of HLA genes was one of the first applications
for DNA re-sequencing. Currently the applications of DNA
sequencing seem endless and sequencing is used for the identification
of sequence differences in a countless number of genetic systems.
In order to accommodate the increasing needs of DNA sequenc-
ing, Conexio has developed Assign-ATF. Assign-ATF includes all
the quality control, auditing, and editing features of Assign-SBT
but enables the user to create his own references and libraries and
use the software for whatever application is required. Assign-ATF
software operates in two modes: a genotyping mode, which com-
pares a sequence against a library of allele/or variant sequences
(similar to HLA typing) and is used for applications such as
Hepatitis C genotyping, and a variant detection mode, which com-
pares a sequence against a single reference sequence (usually a
“wildtype” sequence). In the variant detection mode, all sequence
differences between the sample and the reference sequence are
reported; the software also enables nucleotide changes to be
reported as predicted amino acid changes. Assign-ATF is ideal for
the detection of variants associated with genetic disorders.
DNA sequencing technology has advanced significantly in recent
years. Of the Next Generation (NG) technologies, the 454 Sequencing
Technology is the first to be used for HLA typing (2–4). As was the
case for Sanger Sequencing and the original development of SBT, it
is software that is making HLA typing using NG sequencing technol-
ogy possible. NG sequencing presents new challenges including,
shorter sequencing read lengths, the generation of up to hundreds of
thousand strings of sequence data and the presence of non-specific
sequence data. At Conexio, we have developed a version of Assign for
use with Next Generation sequencing technology, and, despite the
challenges, we have created a user friendly, high-throughput sequence
analysis software that enables high resolution HLA typing.

References

1. Cano P, Klitz W, Mack SJ et al (2007) Common 3. Lind C, Ferriola D, Mackiewicz K et al (2010)

and well-documented HLA alleles: report of Next-generation sequencing: the solution for
the Ad-Hoc committee of the American soci- high resolution, unambiguous human leuko-
ety for histocompatibility and immunogenet- cyte antigen typing. Hum Immunol
ics. Hum Immunol 68(5):392–417 71(10):1033–1042
2. Bentley G, Higuchi R, Hoglund B et al (2009) 4. Holcomb CL, Höglund B, Anderson MW et al
High-resolution, high-throughput HLA geno- (2011) A multi-site study using high-resolution
typing by next-generation sequencing. Tissue HLA genotyping by next generation sequencing.
Antigens 74(5):393–403 Tissue Antigens 77(3):206–217
 Chapter 7

Simple Methods for the Detection of HLA-G Variants

in Coding and Non-coding Regions
Holger Nückel, Erick C. Castelli, Philippe Moreau,
Crista Ochsenfarth, Peter A. Horn, and Vera Rebmann

Abstract
The non-classical human leukocyte antigen (HLA)-G plays a crucial role in the induction of tolerance at
the feto–maternal interface as well as in transplantation, cancer, inflammation, and autoimmune diseases.
To understand gene regulation and the impact of polymorphic sites on the function, simple and easy
feasible approaches are needed for the detection of HLA-G variants in coding and non-coding regions.
Here we summarize a set of methods for the identification of variants in the exon 2–4, in the 3¢ untrans-
lated region and in the gene promoter region of the HLA-G gene.

Key words: HLA-G, Polymorphism, 14 bp Fragment ins/del, 3¢ Untranslated region, Human

leukocyte antigen, Coding region, Non-coding region, Gene promoter region

1. Introduction

The non-classical human leukocyte antigen (HLA)-G is strongly

expressed at the feto–maternal interface but only marginally on
healthy tissues including thymus, cornea, and erythroid cells as well
as blood cells. In a non-physiological situation HLA-G is found to
be expressed in grafted tissue, in cancer, or in inflammation and
autoimmune diseases. The expression of this molecule is always
associated with the induction of tolerance by the inhibition of the
adaptive and innate immune response (for review see ref. (1)).
The gene structure of HLA-G is homologous to the classical
HLA class I genes: The HLA-G gene exhibits seven introns and
eight exons. Exon 1 encodes the peptide signal, exons 2, 3, and 4,
the extracellular a1, a2, and a3 domains, respectively, exon 5 the

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_7, © Springer Science+Business Media New York 2012

123
124 H. Nückel et al.

transmembrane region and exon 6 the cytoplasmic tail; exon 7 is

always absent from the mature mRNA; and exon 8 is not translated
because of a premature stop codon in exon 6 (2). HLA-G exhibits
only a limited variability within the coding region. But an extended
polymorphism has been reported in the 5¢ upstream regulatory
region (5¢URR) and in the 3¢-untranslated region (3¢UTR) of the
gene. According to the International Immunogenetics Database
(IMGT database 2.28.0, January 2010) (3), 44 different HLA-G
variants have been assigned; 33 single nucleotide polymorphisms
(SNPs) are found within the encoding region of HLA-G; and only
13 variations result in an amino acid exchange. Nevertheless it has
to be pointed out that at least 24 of the 44 alleles reveal a substitu-
tion detected only in one or in a few individuals worldwide (2).
The nucleotide variations that result in amino acid exchanges
at the protein level may lead to conformational changes and thereby
may affect the biological function. With regard to the consensus

Table 1
HLA-G polymorphisms in exon 1–4 and in 3¢UTR

Position Exon 2 Exon 3

Rel. to
ATG = 0 234 239 280 292 297 306 324 361 366 372 408 706 726 738 741

Codon 11 13 27 31 32 35 41 54 55 57 69 93 100 104 105

Amino acid A S/F Y/H T/S/M Q R A Q/R/- E P A H G/D G/V S/C

G*01:01:01 GCC TCC TAC ACG CAG CGG GCG CAG GAG CCG GCC CAC GGC GGG TCC

G*01:01:02 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:01:03 --- --- --- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:01:04 --- --- --- --- --- --- --- --- --- --A --T --- --- --- ---

G*01:01:05 --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---

G*01:01:06 --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---

G*01:01:07 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:01:08 --- --- --- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:01:09 --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---

G*01:01:11 --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---

G*01:01:12 --- --- --- --- --- --- --T --- --- --A --- --T --- --- ---

G*01:01:13 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:01:14 --- --- --- --- --- --A --- --- --- --A --- --T --- --- ---

G*01:01:15 --- --- --- --- --- --- --- --- --A --- --- --- --- --- ---
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 125

allele HLA-G*01:01 only 4 out of the 33 SNPs within the coding

region of HLA-G are related to an amino acid exchange, which do
present frequencies above 1% in population studies (4–9). One of
them is located at codon 31 in exon 2 exchanging a threonine for
serine, being typical for the G*01:03 allele (Table 1). Two are
found at codon 110 and codon 130 in exon 3. The first one results
in the amino acid exchange of leucine to isoleucine typical for the
G*01:04 group. A deletion of a cytosine at the first nucleotide of
codon 130 in exon 3 (Table 1) changes the reading frame and
leads to a stop signal (TGA) in codon 189 at exon 4. As this allele
misses substantial information of the full length transcript, it is des-
ignated as G*01:05N allele (N = null allele) (10, 11). The fourth
variation comprises a variation at codon 258, exchanging threo-
nine for methionine, being typical for the G*01:06 allele. The
other SNPs are synonymous substitutions.

Exon 4 3¢UTR (6)

748 755 778 814 871 902 394 1590 1659 1681 1734 1799 1827 3741

107 110 117 130 148 159 169 188 211 219 236 258 267

Frequencies in
G A L/I L/- E Y/H H H A R/W/Q A T/M p Brazilians (2)

GGA CTC GCC CTG GAG TAC CAC CAC GCG CGG GCA ACG CCG −14 bp 39.8

--- --- --- --- --- --- --- --- --- --- --- --- --- +14 bp 19.9

--T --- --- --- --- --- --- --- --- --- --- --- --- +14 bp 5.34

--- --- --- --- --- --- --- *** *** *** *** *** *** *** 0.49

--T --- --- --- --- --- --- --- --- --- --- --- --- −14 bp 0

--- --- --- --- --- --- --- --T --- --- --- --- --- *** 0.97

--T --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- −14 bp 4.37

--- --- --- --- --A --- --- --- --- --- --- --- --- *** 0

--- --- --G --- --- --- --- --- --- --- --- --- --- *** 0.49

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --T *** *** *** *** *** *** *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0
(continued)
126 H. Nückel et al.

Table 1
(continued)

Position Exon 2 Exon 3

Rel. to
ATG = 0 234 239 280 292 297 306 324 361 366 372 408 706 726 738 741

Codon 11 13 27 31 32 35 41 54 55 57 69 93 100 104 105

Amino acid A S/F Y/H T/S/M Q R A Q/R/- E P A H G/D G/V S/C

G*01:01:16 --G --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:01:17 --- --- --- --- --- --- --- --- --- --A --- --T --T --- ---

G*01:01:18 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:01:19 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:01:20 --- --- --- --- --- --- --- --- --- --- --- --T --- --- ---

G*01:02 --- --- --- --- --- --- --- -G- --- --- --- --- --- --- ---

G*01:03 --- --- --- T-- --- --- --- --- --- --- --- --- --- --- ---

G*01:04:01 --- --- --- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:04:02 --- --- --- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:04:03 --- --- --- --- --- --- --- --- --- --- --- --- --- --- ---

G*01:04:04 --- --- --- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:04:05 --- --- --- --- --A --- --- --- --- --A --- --- --- --- ---

G*01:05N --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:06 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:07 --- -T- --- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:08 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:09 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---

G*01:10 --- --- --- -T- --- --- --- --- --- --- --- --- --- --- ---

G*01:11 --- --- --- -T- --- --- --- --- --- --A --- --- --- --- ---

G*01:12 --- --- C-- --- --- --- --- --- --- --A --- --- --- --- ---

G*01:13N --- --- --- --- --- --- --- T-- --- --A --- --T --- --- ---

G*01:14 --- --- --- --- --- --- --- --- --- --A --- --- -A- --- ---

G*01:15 --- --- --- --- --- --- --- --- --- --A --- --- --- -T- ---

G*01:16 --- --- --- --- --- --- --- --- --- --A --- --T --- --- -G-

G*01:17 --- --- --- --- --- --- --- --- --- --A --- --T --- --- ---
Nucleotide sequence variations observed in exon 1–4 of the coding region of HLA-G according to the IMGT (3) A
alanine; S serin; F phenylalanine; Y tyrosine; T threonine; M methionine; Q glutamine; R arginine; E glutamine;
Table 1 indicates SNP identified by the primer mentioned in Table 2
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 127

Exon 4 3¢UTR (6)

748 755 778 814 871 902 394 1590 1659 1681 1734 1799 1827 3741

107 110 117 130 148 159 169 188 211 219 236 258 267

Frequencies in
G A L/I L/- E Y/H H H A R/W/Q A T/M p Brazilians (2)

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --A --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- --- --- --A --- *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- *** ***

--- --- --- --- --- --- --- --- --- --- --C --- --- *** 0

--- --- --- --- --- --- --- --T --- --- --- --- --- +14 bp 8.74

--- A-- --- --- --- --- --- --- --- --- --- --- --- −14 bp 8.25

--- A-- --- --- --- --- --- --T --- --- --- --- --- *** 0

--- A-- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- A-- --- --- --- --- --- --- --- --- --- --- --A *** 3.88

--- A-- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- ~TG --- --- --- --- --- --- --- --- --- +14 bp 0.97

--- --- --- --- --- --- --- --- --- --- --- -T- --- +14 bp 4.85

--- A-- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- T-- --- --- --- *** 0

--- --- --- --- --- C-- --- *** *** *** *** *** *** *** 0.49

--- --- --- --- --- --- --- *** *** *** *** *** *** *** 0

--- A-- --- --- --- --- --- *** *** *** *** *** *** *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- -A- --- -T- --- *** 0

--- A-- --- --- --- --- --- --- --- --- --- --- --- *** 0

--- --- --- --- --- --- --- --- --- --- --- -T- --- *** 0

--- --- --- --- --- --- -G- *** *** *** *** *** *** *** ***
+14 bp insertion (presence) of 14 bp; −14 bp deletion (absence) of 14 bp; *** not evaluated; ~ deleted; amino acid codes:
H histidine; P proline; G glycine; D asparagine acid; V valine; C cysteine; L leucine; W tryptophan; grey background in
128 H. Nückel et al.

As exon 7 is absent from the mature mRNA and exon 8 is not

translated due to the stop codon in exon 6, this gene segment is
considered to be the 3¢UTR of the mature RNA (2). Here the
HLA-G gene exhibits at 3¢UTR, a 14 bp fragment insertion/
deletion (ins/del) polymorphism. HLA-G alleles with a 14 bp
insertion are associated with alternative splicing products missing
92 bases from 3¢UTR. Although such truncated mRNAs are
reported be more stable, the 14 bp insertion itself is related to
low mRNA production (2).
Currently, there are more than 29 variation sites in the pro-
moter region (27 SNPs and 2 insertions/deletions) (8, 12, 13)
(Castelli et al. 2010, unpublished data). Many of the promoter
region polymorphisms either coincide with or are close to known
or putative regulatory elements, and thus may affect the binding of
HLA-G regulatory factors.
The realization that HLA-G strongly suppress the immune sys-
tem, and that certain HLA-G variations are associated with high
and low expression levels which might be of functional relevance,
has emphasized the need for simple, practical approaches for the
typing of the coding and non-coding regions of the HLA-G gene.
Here we describe a two-pronged approach to such typing. The first
uses PCR to generate six amplicons which are then used in either
(a) pyrosequencing (PSQ) to detect the most relevant SNPs (codon
31, 35, 57, 93, 100, 107, 100, 130, 188, or 258) or (b) a simple
electrophoretical separation of the specific PCR product for the
identification of the 14 bp fragment ins/del polymorphism in
3¢UTR. Using this strategy all HLA-G* alleles with frequencies
above 1% can be defined, whereas HLA-G* alleles with nucleotide
substitutions within introns are not taken into consideration.
Secondly, for the further evaluation of the HLA-G promoter region
variation, a PCR-amplified fragment of approximately 1,852 bp
encompassing the −1445 (5¢URR) and +407 (exon 2) nucleotides
is produced and sequenced.

2. Materials

2.1. Materials for 1. 2× Master Mix RED for PCR: 150 mM Tris–HCl, pH 8.5,
HLA-G* Allele Typing 40 mM (NH4)2SO4, 3 mM, MgCl2, 0.2% Tween 20®, 0.4 mM
and the Identification DNTPs, 0.05 units/mL Ampliqon Taq DNA polymerase, inert
of the 14 bp Fragment red dye, and a stabilizer (Ampliqon, Skovlunde, Denmark).
ins/del Polymorphism 2. Template DNA: A good quality DNA at a concentration of
in 3 ¢ UTR approximately 50 mg/mL.
2.1.1. Polymerase Chain 3. Oligonucleotides primers for PCR (Table 2): All primers are
Reactions diluted with sterile and nuclease-free water to a concentration
 Table 2
Primers to identify SNPs within the coding region and the 14 bp fragment insertion/deletion (ins/del) polymorphism at
3¢UTR of HLA-G

Primer number Name 5¢–3¢ Sequence Target Use

1 HLA-Gex2-SEBT TCATCGCCATGGGCTACG Exon 2 First amplification

2 HLA-Gex2-AS CTGGCCTCGCTCTGGTTGTAG
3 HLA-Gex2_31_35-seq CGAGTCGCTGTCGAA Exon 2; codons 31, 35 Pyrosequencing
4 HLA-Gex2_57-seq TGTCTCCTCTTCCCAATA Exon 2; codons 57 Pyrosequencing
5 HLA-Gex3-SEBT CCGGGTACTCCCGAGTCT Exon 3 First amplification
6 HLA-Gex3-AS TCGTTCAGGGCGAGGTAATC
7 HLA-G_93_100seq GGACCCCAGGTCGCA Exon 3, codons 93, 100 Pyrosequencing
8 HLA-G_107_110seq ATACTGTTCATACCCGC Exon 3; codons 107, 110 Pyrosequencing
9 HLA-G130-SE GGTGGGGCCAGGTTCTC Exon 3 codon 130 First amplification
10 HLA-G130-ASbt GGAGATCTGAGCCGCAGTGTC
11 HLA-G130seq GCCCTGAACGAGGAC Exon 3 codon 130 Pyrosequencing
12 codon188-SE TCAGACCCCCCCAAGACA Exon 4 First amplification
13 codon188-AS.bt GGGTGGCCTCATAGTCAAAGAC
14 codon188-seq GACCCCCCCAAGACA Exon 4, codon 188 Pyrosequencing
15 codon258SEBT TGCGGAGATCATACTGACCTG Exon 4 First amplification
16 codon258AS CGACCCTGTTAAAGGTCTTCAGAG
17 codon258-seq CATGCTGCACATGGCA Exon 4, codon 258 Pyrosequencing
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions

18 HLAG3utr-SE GTGATGGGCTGTTTAAAGTGTCACC Exon 8 First amplification

19 HLAG3utr-AS GGAAGGAATGCAGTTCAGCATGA
129

SE sense; AS anti-sense; BT 5¢biotinylation of the primer at the time of synthesis; seq sequencing primer
130 H. Nückel et al.

of 100 pmol for storage and diluted to a concentration of

10 pmol for usage.
4. PCR tubes or plates (depending on the amount of samples),
with corresponding lids or covers.
5. Thermal cycler for PCR.

2.1.2. Gel Electrophoresis 1. Biozym LE agarose for gel electrophoresis (Biozym Scientific
GmbH, Oldendorf, Germany).
2. 10× Tris buffered ethelene-diamine-tetra-acetic acid (EDTA):
108 g Tris base, 55 g boric acid, 7.44 g EDTA to 1 L with
double distilled water.
3. Sybr Safe DNA gel stain or ethidium bromide (Invitrogen,
Darmstadt, Germany).
4. Electrophoresis chambers and sufficient power supply.
5. UV-transilluminator of a Gel Documentation System with UV
light.
6. DNA marker—pBR322DNA/AluI Marker (Fermentas
GmBH, St. Leon-Rot, Germany).

2.1.3. Pyrosequencing 1. Pyrosequencing machine, for example, PSQ96MA (Biotage,

Uppsala, Sweden) to be used in conjunction with the com-
puter program PSQ96MA 2.1.1 which is a part of this system.
Software however may differ according to which pyrosequenc-
ing machine is used.
2. Vacuum Prep Tool (Biotage, Uppsala, Sweden).
3. Vacuum source with a minimum vacuum of 300 mmHg.
4. Liquid waste container (must withstand absolute vacuum).
5. Mixer/shaker for microtiter plates, one at room temperature
and one which can be heated to 80°C.
6. PSQ™ 96 Sample Prep Thermoplate (Biotage, Uppsala,
Sweden).
7. PyroMark™ Q96 Plate Low (Qiagen, Hilden, Germany).
8. PyroMark™ Gold Q96 reagents (Qiagen, Hilden, Germany).
9. PyroMark™ binding buffer, annealing buffer, and wash buffer
(Qiagen, Hilden, Germany).
10. 70% Ethanol to be prepared from 100% pure ethanol with
double distilled water.
11. 0.5 M NaOH to be prepared with double distilled water.
12. Biotinylated PCR product.
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 131

13. Streptavidin sepharose high performance beads (GE Healthcare,

Munich, Germany).
14. Sequencing primers (see Table 2).
15. High purity and nuclease free water.

2.2. Materials 1. DNA polymerase: this protocol was optimized and tested
for the HLA-G Gene using three different DNA polymerases (a) Platinum Taq
Promoter Region DNA polymerase (Invitrogen, Carlsbad, CA), (b) PCR Long
Variations PCR enzyme mix (Fermentas, Maryland, USA), and (c)
Platinum Taq DNA polymerase high-fidelity (Invitrogen,
Carlsbad, CA).
2. 0.2-mL PCR tubes or plates.
3. Ultra-pure deionized and autoclaved water, DNAse and
RNAse-free.
4. dNTPs: this method was optimized using a 100 mM dNTP set
(Invitrogen) but any set of dNTP should do.
5. 5 mM dNTP solution: 5 mL of each dNTP from the 100 mM
dNTP Set and 80 mL of ultra-pure deionized water, resulting
in a solution of 5 mM of each dNTP.
6. 50 mM Magnesium chloride or 50 mM magnesium sulphate:
typically this is provided with the polymerases described in step 1.
7. Primers: Primers used in this protocol are given in Table 3.
Each primer was diluted to a final concentration of
10 pmol/mL.
8. DNA samples diluted to 50 ng/mL.
9. This protocol was standardized using the Veriti 96-well
Thermal Cycler (Applied Biosystems), but any thermal cycler
should do.
10. Agarose, horizontal electrophoresis cube apparatus, and ethid-
ium bromide to assemble a 1% agarose gel.
11. Any non-denaturant load buffer.
12. DNA ladder that allows the identification of a 1,852-bp
fragment. This protocol was standardized using the 1 Kb
GeneRuler Express (Fermentas).
13. UV-transilluminator or Gel documentation system with UV
light.
14. Exonuclease I enzyme (EXO) and Shrimp alkaline phosphatase
(SAP).
15. Sequencer and reagents for sequencing according to the
sequencer available.
 132
H. Nückel et al.

Table 3
Primers to amplify and sequence the HLA-G promoter region

Primer name Type 5¢–3¢ Sequence Targeta Use References

GPromo.S Forward ACATTCTAGAAGCTTCACAAGAATG −1445 to −1421 First amplification (12, 22)

GPromo.R Reverse GCCTTGGTGTTCCGTGTCT +389 to +407 Castelli et al. 2010, unpublished data
G-908R Reverse TTCACCTCACAGTTGTAAGTGTTC −930 to −907 Sequencing (22)
G-830F Forward CACACGGAAACTTAGGGCTACG −829 to −808 Sequencing (22)
G-304R Reverse GCCAAGCGTTCTGTCTCAGTGT −302 to −282 Sequencing (22)
GPR-247 Forward CTCAAGCGTGGCTCTCAGGGTC −245 to −225 Sequencing (22)
HG01F Forward TAAAGTCCTCGCTCACCCAC −47 to −28 Sequencing (21)
GPROMO3R Reverse GTTGGTATATAAATGCATCTAAAAG −713 to −689 Sequencing Castelli et al. 2010, unpublished data
GIN1-98 Reverse GTTTCCCTCCTGACCCCGCACT +77 to +98 Sequencing (22)
a
According to the IMGT (3)
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 133

3. Methods

3.1. HLA-G* Allele See Table 2 for a detailed outline of individual PCR products and
Typing and the their representative mutations with codon names. A total of six
Identification of the PCR products will be produced: Exon 2 (including codons 31, 35,
14 bp Fragment ins/del and 57), exon 3 (including codons 93, 100, 107, 110), exon 3
Polymorphism in 3¢UTR codon 130, exon 4 codon 188, and exon 4 codon 258, and 3¢UTR.
It is recommended to make a master mix in order to simplify the
3.1.1. Polymerase Chain pipetting and reduce the chance of any single error (see Note 1).
Reaction
1. Exon 2 codons 31, 35, 57 PCR: Each individual PCR reaction
has a total volume of 50 mL; this includes 25 mL Ampliqon
Master Mix RED, 20 mL high purity water, 2 mL of each primer
numbers 1 and 2 (Table 2), and 1 mL DNA. Thermal cycler
was set as following: 5 min at 95°C (hot start), 30 s at 95°C
(denaturation), 40 s at 65°C (annealing), 30 s at 72°C (exten-
sion), and finally, 10 min at 72°C. Steps 2–5 are repeated 38
times before the final extension time. The amplified product is
208 bp long (see Note 2). This product will be used for pyros-
equencing—see Subheading 3.1.3, step 1.
2. Exon 3 codons 93, 100, 107, 110 PCR: Each individual reac-
tion has a total volume of 30 mL; this includes 15 mL Amplicon
Master Mix RED, 12 mL high purity water, 1 mL of each primer
numbers 5 and 6 (Table 2), and 1 mL DNA. Thermal cycler
conditions are exactly as for Exon 2 PCR. End product should
be 278 bp long. All of the 30 mL of PCR product will be for
pyrosequencing—see Subheading 3.1.3, step 2 (see Note 2).
3. Exon 3 codon 130 PCR: Use the same master mix recipe as for
exon 3 (30 mL) (step 2), using primer numbers 9 and 10
(Table 2). Thermal cycler conditions are the same as for Exon
2 PCR with the following exception: The annealing tempera-
ture is changed to 65.5°C and the extension is changed from
30 to 15 s. PCR product is 167 bp long. 15 mL of this product
will be used for pyrosequencing—see Subheading 3.1.3, step
3. The rest can be frozen at −20°C until all genotypes are
confirmed.
4. Exon 4 codon 188 PCR: Use the same master mix recipe as
for exon 3 (30 mL) (step 2), using primer numbers 12 and 13
(Table 2). Thermal cycler conditions are the same as for Exon
2 PCR with one exception: The annealing temperature is
changed to 60°C. PCR product is 58 bp long. 15 mL of this
PCR product will be used for pyrosequencing—see
Subheading 3.1.3, step 4. The rest can be frozen at −20°C
until all genotypes are confirmed.
5. Exon 4 codon 258 PCR: Use the same master mix recipe as for
exon 3 (30 mL) (step 2), using primer numbers 15 and 16
134 H. Nückel et al.

(Table 2). Thermal cycler conditions are the same as for Exon
2 PCR with one exception: The annealing temperature is
changed to 60°C. PCR product is 267 bp long. 15 mL of this
PCR product will be used for pyrosequencing—see
Subheading 3.1.3, step 5. The rest can be frozen at −20°C
until all genotypes are confirmed.
6. 3¢UTR PCR: Use the same master mix recipe as for exon 3
(30 mL), using primer numbers 18 and 19 (Table 2). Thermal
cycler conditions are exactly as for Exon 2 PCR. PCR product
should be 434 bp long. 10 mL of this PCR product will be used
for Gel Electrophoresis—see Subheading 3.1.2, steps 3–5. The
rest can be frozen at −20°C until all genotypes are confirmed.

3.1.2. Electrophoretical 1. To confirm that the PCR of the coding region worked properly,
Separation of the PCR the PCR products (i.e. products from Subheading 3.1.1, steps
Products of the Coding and 1–5) are applied to a 2% agarose gel. To prepare the gel add
the 3¢UTR Region 2 g agarose to 100 mL of 0.5× TBE and cook in the micro-
wave for approximately 3–5 min until all agarose is dissolved
(see Note 3). Thereafter Sybr Safe DNA gel stain or ethidium
bromide is added (0.5 mg/mL). Keep in mind that you need
enough wells in the gel for all the samples, plus negative and
positive PCR controls, as well as for the DNA Marker. A nega-
tive PCR control would consist of PCR master mix which was
run in the thermal cycler exactly like the patient samples, only
without template DNA. A positive PCR control would be PCR
master mix which was run in the thermal cycler exactly like the
test samples except that here a DNA sample with known
HLA-G allele (14) would be used being already proven to
function in PCR.
2. Place the agarose gel in the gel chamber and make sure it is
sufficiently covered by 0.5× TBE.
3. Load the samples completely in the gel and attach the power
supply. Run at 180 V until good separation is acquired.
4. Put the gel into a UV-Transilluminator of a Gel Documentation
System with UV light. The presence of a fragment with the
expected bp indicates that the amplification ran properly (see
Note 4).
5. For the SNP in the 3¢UTR, 10 mL of the product (i.e. from
Subheading 3.1.1, step 6) is applied to a 3% agarose. This is
prepared by mixing 3 g agarose with 100 mL 0.5× TBE and
cooking for approximately 5 min. Microwaving times may vary.
Thereafter Sybr Safe DNA gel stain or ethidium bromide is
added (0.5 mg/mL).
6. After the gel has cooled off, place in the gel chamber, make
sure that it is covered by 0.5× TBE, and load 10 mL of each
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 135

Fig. 1. Typing of 14 bp ins/del polymorphism in 3¢UTR. PCR product is run on a 3%

agarose gel until good separation is acquired.

PCR product onto the gel. Run at 180 V until good separation
is acquired. Put the gel into a UV-Transilluminator of a Gel
Documentation System with UV light. The two different PCR
products of 210 and 224 bp should be present.
7. A sample with only one band at 210 (deletion 14 bp) or 224
(insertion 14 bp) indicates a homozygosity for the given geno-
type, for example del/del or ins/ins. A sample with two bands
visible (one at 210 and the other at 224) indicates a heterozy-
gote del/ins (Fig. 1).

3.1.3. Pyrosequencing All steps here are to be performed at room temperature unless
otherwise stated. In addition, all buffers should be stored at 4°C,
but allowed to warm up to room temperature before starting. Also
before use, all sequences to be analysed should be entered into the
computer as indicated by the software instructions.
1. The exon 2 codons 31, 35, 57 PCR product will be used for
two individual sequencing reactions. The first, using primer
number 3, sequences SNPs in codon 31 and 35. The second,
using primer number 4, sequences one SNP in codon 57.
2. The exon 3 codons 93, 100, 107, 110 PCR product will also
be used for two individual sequencing reactions. The first,
using primer number 7, sequences SNPs in codons 93 and
100. The second, using primer number 8, sequences SNPs
in codons 107 and 110.
3. The exon 3 codon 130 PCR product will be used for only one
sequencing reaction using primer number 11 for the SNP in
codon 188.
4. The exon 4 codon 188 PCR product will be used for only one
sequencing reaction using primer number 14 for the SNP in
codon 188.
5. The exon 4 codon 258 PCR product will be used for only one
sequencing reaction using primer number 17 for the SNP in
codon 258.
136 H. Nückel et al.

6. Set up the vacuum prep tool according to manufacturer’s

instruction (this must only be performed once, at the time of
initial set-up).
7. Immobilize the various PCR products obtained from
Subheading 3.1.1 to the Streptavidin Sepharose beads. Do this
by preparing a master mix of 37 mL binding buffer, 15 mL high
purity water, and 3 mL of very well mixed Streptavidin Sepharose
beads per reaction. To each well in a Pyro Mark Low plate,
pipette 55 mL of binding buffer-beads mixture. To this “binding”
plate, add 15 mL of PCR product (biotinylated), see Note 5.
Using a shaker, incubate at room temperature with constant
agitation to keep the beads in motion.
8. While the binding plate is mixing, prepare the “annealing”
plate. Do this by preparing a master mix of 40 mL annealing
buffer and 2 mL of above mentioned (Subheading 3.1.3, steps
1–5) primer (10 pmol stock solution). To each well in a new
Pyro Mark Low plate, pipette 40 mL of annealing buffer mix-
ture. Each different SNP sequence must be separately analy-
sed. This means one cannot combine primers; you have to use
only one primer per reaction. Set this annealing plate aside.
9. Prepare the washing troughs for the vacuum prep tool. Fill
with approximately 180 mL of the solutions; 70% ethanol,
0.5 M NaOH, washing buffer, and double distilled water. Turn
the vacuum on.
10. After 5 min (see Note 6), remove binding plate from shaker
and quickly but gently lower the vacuum prep tool into the
binding plate. This will collect the beads, which are now bound
to the DNA from the PCR and filter out all other solutions.
Make sure that the liquid has been aspirated completely from
the binding plate and that little to no beads are remaining at
the bottom of the plate (see Note 7).
11. Place the vacuum prep tool in the 70% ethanol solution and let
the solution flush through the filters for approximately 5 s.
Move to the NaOH trough and once again, let the solution
flow through the filters for approximately 5 s. Finally, move the
prep tool to the washing buffer and flush through for 5 s.
Allow all liquid to drain completely from the probes (filter tips)
by turning the prep tool to a 90° angle, and then back to a
horizontal position a few times. Make sure it is very dry.
12. Holding the vacuum prep tool over the sequencing plate
(annealing), which has already been filled with annealing buf-
fer and specific primer, turn the vacuum off. Release the vac-
uum prep tool into the buffer and shake to remove all beads
from the filters into the annealing buffer.
13. Place the annealing plate, including primer and bead-DNA
complex, at 80°C and shake for 2 min using the PSQ 96 Sample
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 137

Prep thermoplate. During this time, start to wash the vacuum

prep tool in high purity water to prepare for the next run. Also
prepare the enzyme and substrate cartridge including the
necessary nucleotides.
14. Remove the annealing plate from the heat block and let cool to
room temperature. Continue with the sequencing reaction
according to the programme specifications for the PSQ96MA
2.1.1.
15. After the PSQ run has been completed, the results (genotypes)
are ready for proofing and adding to the text document. With
one simple click on “Analyze Results” the computer analyzes all
of the pyrograms created (results) and reports to you whether or
not the sample was successful (passed, check, or failed) as well
as the genotype of the sample. The results should also be man-
ually proofed in order to exclude mistakes possibly made by
the computer. First all other genotyping methods must be
completed. The quality of the PSQ results are checked by the
control samples (see Subheading 3.1.2, step 1).

3.2. Methods for the 1. Separate the reagents for PCR amplification (DNA polymerase,
HLA-G Gene Promoter dNTPs, ultra-pure deionized water, and the primers for the
Region Variations first amplification GPromo.S and GPromo.R—Table 3), thaw
and keep them on ice.
3.2.1. HLA-G Promoter
Amplification by PCR 2. On a separate bench, thaw the DNA samples and keep them
on ice.
3. Prepare a mix of reagents for the amplification. The following
protocol is for one reaction and the volumes are adjusted to
the reagents in proposed final concentration. Amplification is
performed in a final volume of 50 mL containing:
(a) 33.7 mL of ultra-pure water (or a quantity enough to a final
volume of 50 mL if any of the volumes given below changed);
(b) 1× PCR buffer (typically 5 mL of the buffer provided with
the polymerases described in Subheading 2);
(c) 3 mL of magnesium chloride (for Platinum Taq Polymerase
from Invitrogen and Long PCR enzyme Mix from
Fermentas) or 3 mL of magnesium sulphate (Platinum Taq
Polymerase High-Fidelity from Invitrogen), giving a final
concentration of 1.5 mM;
(d) 2 mL of the 5 mM dNTP, giving a final concentration of
0.2 mM of each dNTP;
(e) 2 mL of each primer for the first amplification, giving a final
amount of primer of 20 pmol each per reaction;
(f) 0.3 mL of DNA polymerase (1.5 units);
(g) Gently homogenize this solution by pipetting up and
down. Cap the tube.
138 H. Nückel et al.

4. On the other bench, add 2 mL of the diluted DNA sample into

the mix and gently homogenize the solution by pipetting up
and down. Cap the tube.
5. Put the reaction mix at the thermal cycler using the following
cycling protocol:
6. Initial denaturation cycle at 94°C for 2 min.
(a) 32 cycles at 94°C for 30 s, 59°C for 30 s, and 72°C (for
regular Platinum Polymerase) or 68°C (for High-Fidelity
Platinum and Long PCR enzyme Mix) for 135 s.
(b) Final extension step at 72°C (for regular Platinum
Polymerase) or 68°C (for High-Fidelity Platinum and
Long PCR enzyme Mix) for 5 min. Keep them on ice
before going to the next step.
7. Assemble a common 1% agarose gel (using TBE or TAE buffer)
stained with ethidium bromide and load 5 mL of the amplification
reaction together with a non-denaturant load buffer of your
preference. Keep the rest of the amplified product in the fridge
or on ice for DNA sequencing as described in Subheading 3.2.2.
Use a DNA ladder that allows the detection of a fragment of
approximately 1,852 bp. Then, run the electrophoresis for
about 2 h using 100 V or the necessary to separate the frag-
ments in the electrophoresis system available.
8. Put the gel into a UV-Transilluminator of a Gel Documentation
System with UV light. The presence of a fragment of approxi-
mately 1,852 bp (Fig. 2) indicates that the amplification ran
properly. No other fragments would be detected. You may dis-
pose the gel properly.

3.2.2. HLA-G Promoter In the previous section, the protocol was given to amplify the
Sequencing HLA-G promoter region by PCR. Once you have it amplified
with a verified clean product of approximately 1,852 bp (see
Subheading 3.2.1, step 7), you may sequence this PCR product by
using the appropriate primers described in Table 3. Alternatively,
you may use restriction endonucleases to screen for variability (not
within the scope of this chapter). Depending on the sequencer
available, the sequencing protocol must be adjusted for each
machine and sequencing reagents used. No sequencing protocol
is provided in this chapter since it depends on the infrastructure
available. Please refer to your sequencer’s documentation in order
to better adjust a sequencing reaction of the PCR product. In this
section, a strategy to evaluate the entire promoter by sequencing
using the primers listed in Table 3 is given.
1. To assure a purified product for sequencing, take 10 mL of the
PCR product in a separate 0.2 mL PCR tube and add 1 mL
of EXO and 1 mL of SAP. Mix the solution by pipetting up
and down. Cap the tube and put it on ice before using the
thermal cycler.
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 139

Fig. 2. HLA-G gene promoter amplification products. 1% Agarose gel stained by ethidium bromide. The DNA ladder used
was 1 Kb GeneRuler Express (Fermentas). From the heavier to the lighter fragment, the ladder indicates fragments of 5,000,
3,000, 2,000, 1,500, 1,000, 750, 500, 300, and 100 bp. The expected amplification is approximately 1,852 bp.

Fig. 3. Primer map used to amplify and sequence the HLA-G promoter region.

2. In the thermal cycler, use the following cycling protocol: 37°C

for 15 min, then remove from the thermal cycler 80°C for
15 min and keep in freezer or on ice until sequencing.
3. Use the sequencing primers in Table 3 to sequence the HLA-G
promoter. Use the sequencing protocol that best suits your
sequencer. The location of each primer is given in Table 3.
Figure 3 illustrates the position and direction of each primer.
Note that, by using all the primers in Table 3 and Fig. 3, it is
likely that all sequences of the promoter region will be covered
and any variation sites present in this segment will be detected.
In addition, these primers are optimized to detect all the 29
known variation sites in the promoter region using a sequencer
such as ABI3100 (Applied Biosystems) with 50-cm capillaries.

3.3. Analysis 1. One strategy to evaluate haplotypes is to use computational

of Genotypes inference to obtain haplotypes for each sample by using algo-
and Determination rithms such as the PHASE method (15, 16), the EM
of Haplotypes (Expectation-Maximization) (17, 18), and the ELB (Excoffier-
Laval-Balding) (19) algorithms. Please see the further refer-
ences for information about how to use these and the limitations
of the methods (20, 21).
140 H. Nückel et al.

4. Notes

1. For the PCR for the coding region you may want to run a
control on an electrophoresis gel to confirm that the PCR
worked properly. This will mean that you need more than
one PCR reaction for each patient because all of the 50 mL
(or 30 mL) of PCR product will be needed for further analysis.
We suggest this for optimization of DNA concentration as well.
2. We use DNA prepared from EDTA-blood. This is stable and of
relative good quality. When, for example, DNA has been
extracted from paraffin slices, it is possible that the amount of
DNA and/or primer concentration would not be enough or
possibly too much.
3. When making more than one gel, it is helpful to prepare them
individually. Cooking the gels for too long (which can be nec-
essary when making many gels at once) alters the concentra-
tion of the gel when too much TBE evaporates because of
the boiling.
4. After visualization of PCR products on the gel, it is possible
that no PCR product or an incorrect one has been amplified.
It is very likely that this is caused by problems with the tem-
plate DNA. Furthermore make sure that the correct annealing
temperature is used which otherwise might lead to false priming.
Possible causes of no PCR products could be either that simply
DNA was not added to the mixture or that the DNA was pre-
pared from paraffin slices. In this case, a long enough product
cannot be amplified. Suggestions for improvement include
simply trying the PCR again making sure to pipette DNA into
the PCR mix or measuring the concentration of the DNA to
make sure that enough template is present.
5. PSQ manufacturers suggest a maximum of 30 mL PCR prod-
uct per reaction. We have optimized, and for our purposes,
with our DNA templates, 15 mL of PCR product is usually
enough. Too much PCR product can also be detrimental for
results.
6. Leaving the binding plate to shake for longer than 5 min
(manufacturers recommended time) does not make a differ-
ence in the results. We have allowed the plate to shake for up
to 10 min with no difference in results.
7. The sepharose beads sediment quickly, and when they are at
the bottom of the plate, then they are harder to collect with the
vacuum prep tool. It is useful to act quickly and gently to collect
the maximum amount of beads bound to DNA. If too many
beads have been left at the bottom of the plate, it is possible to
add a bit (approximately 20 mL) of high purity water, shake for
7 Simple Methods for the Detection of HLA-G Variants in Coding and Non-coding Regions 141

a minute, and then aspirate one more time. The vacuum filter
tool should be left ON at all times in order to not to lose
any beads. Be cautious to always hold the filter at the same
point, and not to turn it 180° by accident thereby mixing
the samples!

References

1. Carosella ED, Favier B, Rouas-Freiss N, Moreau 10. Ober C, Aldrich CL (1997) HLA-G polymor-
P, Lemaoult J (2008) Beyond the increasing phisms: neutral evolution or novel function?
complexity of the immunomodulatory HLA-G J Reprod Immunol 36:1–21
molecule. Blood 111:4862–4870 11. Suarez MB, Morales P, Castro MJ, Fernandez
2. Donadi EA, Castelli EC, Arnaiz-Villena A, V, Varela P, Alvarez M, Martinez-Laso J, Arnaiz-
Roger M, Rey D, Moreau P (2011) Implications Villena A (1997) A new HLA-G allele (HLA-
of the polymorphism of HLA-G on its func- G*0105N) and its distribution in the Spanish
tion, regulation, evolution and disease associa- population. Immunogenetics 45:464–465
tion. Cell Mol Life Sci 68:369–395 12. Tan Z, Shon AM, Ober C (2005) Evidence of
3. Robinson J, Waller MJ, Fail SC, Marsh SG balancing selection at the HLA-G promoter
(2006) The IMGT/HLA and IPD databases. region. Hum Mol Genet 14:3619–3628
Hum Mutat 27:1192–1199 13. Berger DS, Hogge WA, Barmada MM, Ferrell
4. Castelli EC, Mendes-Junior CT, Deghaide NH, RE (2010) Comprehensive analysis of HLA-G:
de Albuquerque RS, Muniz YC, Simoes RT, implications for recurrent spontaneous abor-
Carosella ED, Moreau P, Donadi EA (2010) tion. Reprod Sci 17:331–338
The genetic structure of 3¢untranslated region 14. Rebmann V, van der Ven K, Passler M, Pfeiffer K,
of the HLA-G gene: polymorphisms and haplo- Krebs D, Grosse-Wilde H (2001) Association of
types. Genes Immun 11:134–141 soluble HLA-G plasma levels with HLA-G
5. Castelli EC, Mendes-Junior CT, Donadi EA alleles. Tissue Antigens 57:15–21
(2007) HLA-G alleles and HLA-G 14 bp poly- 15. Stephens M, Smith NJ, Donnelly P (2001) A
morphisms in a Brazilian population. Tissue new statistical method for haplotype recon-
Antigens 70:62–68 struction from population data. Am J Hum
6. Hviid TV (2006) HLA-G in human reproduc- Genet 68:978–989
tion: aspects of genetics, function and preg- 16. Stephens M, Donnelly P (2003) A comparison of
nancy complications. Hum Reprod Update bayesian methods for haplotype reconstruction
12:209–232 from population genotype data. Am J Hum
7. Moreau P, Contu L, Alba F, Lai S, Simoes R, Genet 73:1162–1169
Orru S, Carcassi C, Roger M, Rabreau M, 17. Excoffier L, Slatkin M (1995) Maximum-
Carosella ED (2008) HLA-G gene polymor- likelihood estimation of molecular haplotype
phism in human placentas: possible association frequencies in a diploid population. Mol Biol
of G*0106 allele with preeclampsia and miscar- Evol 12:921–927
riage. Biol Reprod 79:459–467
18. Excoffier L, Laval G, Schneider S (2005)
8. Rizzo R, Hviid TV, Govoni M, Padovan M,
Arlequin ver. 3.0: an integrated software pack-
Rubini M, Melchiorri L, Stignani M, Carturan
age for population genetics data analysis. Evol
S, Grappa MT, Fotinidi M, Ferretti S, Voss A,
Bioinform Online 1:47–50
Laustrup H, Junker P, Trotta F, Baricordi OR
(2008) HLA-G genotype and HLA-G expres- 19. Excoffier L, Laval G, Balding D (2003) Gametic
sion in systemic lupus erythematosus: HLA-G phase estimation over large genomic regions
as a putative susceptibility gene in systemic using an adaptive window approach. Hum
lupus erythematosus. Tissue Antigens 71: Genomics 1:7–19
520–529 20. Castelli EC, Mendes-Junior CT, Veiga-Castelli
9. Sipak-Szmigiel O, Cybulski C, Wokolorczyk D, LC, Pereira NF, Petzl-Erler ML, Donadi EA
Lubinski J, Kurzawa R, Baczkowski T, Radwan (2010) Evaluation of computational methods
M, Radwan P, Ronin-Walknowska E (2009) for the reconstruction of HLA haplotypes.
HLA-G polymorphism and in vitro fertilization Tissue Antigens 76:459–466
failure in a Polish population. Tissue Antigens 21. Castelli EC, Mendes-Junior CT, Deghaide
73:348–352 NH, de Albuquerque RS, Muniz YC, Simoes
142 H. Nückel et al.

RT, Carosella ED, Moreau P, Donadi EA 22. Ober C, Aldrich CL, Chervoneva I, Billstrand
(2010) The genetic structure of 3¢untrans- C, Rahimov F, Gray HL, Hyslop T (2003)
lated region of the HLA-G gene: polymor- Variation in the HLA-G promoter region
phisms and haplotypes. Genes Immun 11: influences miscarriage rates. Am J Hum Genet
134–141 72:1425–1435
 Chapter 8

Molecular Typing of HLA-E

Nina Lauterbach, Christina E.M. Voorter, and Marcel G.J. Tilanus

Abstract
Human leukocyte antigen-E (HLA-E) is a non-classical HLA class I gene that shows a limited degree of
polymorphism compared to the classical HLA genes. The HLA-E molecule can bind peptides derived from
the leader sequence of various HLA class I alleles and some viral homologues, including CMV. The HLA-E
peptide complex can act as a ligand for the CD94/NKG2 receptors expressed on the surface of natural
killer cells and T cell subsets. Differences in expression levels between the different HLA-E alleles have
been reported and a role for HLA-E polymorphism in stem cell transplantation has been postulated. This
chapter focuses on routine technologies for HLA-E typing: the sequence-specific primer-PCR method that
uses sequence-specific primers, the PCR sequence-specific oligonucleotides Luminex method, using
sequence-specific probes attached to beads and the sequencing-based typing method, where sequencing of
the alleles is performed.

Key words: Human leukocyte antigen-E, HLA typing method, Sequence-specific primer-PCR,
PCR-sequence-specific oligonucleotides, Sequencing-based typing

1. Introduction

The human leukocyte antigen-E (HLA-E) gene is located between

HLA-A and HLA-C on the short arm of chromosome 6. With
only nine alleles encoding three different proteins, HLA-E is the
least polymorphic of all MHC class I molecules. Two HLA-E
molecules, encoded by HLA-E*01:01 and HLA-E*01:03, exist
with about equal frequencies in the population (1, 2). The previ-
ously identified HLA-E*01:02 allele was removed from the
IMGT/HLA database because the sequence was found identical
to HLA-E*01:01:01:01. The allele HLA-E*01:04 has been origi-
nally reported in 1 out of 11 Japanese individuals (3), and has not
been identified thereafter, even since a population of 50 Japanese

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_8, © Springer Science+Business Media New York 2012

143
144 N. Lauterbach et al.

individuals has been investigated. It appears very likely that this

allele is the result of sequencing artefacts (2). In the remaining 8
alleles, only one non-synonymous substitution is present at posi-
tion 382, which results in one amino acid difference; an arginine at
position 107 in HLA-E*01:01 is replaced by a glycine in HLA-
E*01:03 (4). All other nucleotide differences are either located in
the non-coding region or are synonymous substitutions. The
typing techniques described here have mainly been focused on the
non-synonymous substitution, although with sequence-based
typing also the synonymous substitutions are identified.
It has been reported that some form of balancing selection is
acting on HLA-E to maintain the two alleles (4, 5), implicating that
there are functional differences. Indeed, a few studies show in vitro
a functional difference between HLA-E*01:01 and HLA-E*01:03
(6–8). Furthermore, studies focusing on the effect of HLA-E
polymorphism on stem cell transplantation outcome show a protec-
tive role for the HLA-E*01:03 genotype (9–12), albeit that the
various inconclusive and contradictive results remain to be resolved
and confirmed in larger cohort studies. In Behcet’s disease, HLA-
E*01:01 seems to be associated with reduced risk (13).
Different HLA-E typing techniques, like sequence-specific
primer-PCR (PCR-SSP), RFLP, Taqman and sequencing-based
typing (SBT), have been described (1, 2, 9, 13, 14). In this chap-
ter, we describe three different routine methods to identify HLA-E
polymorphism in genomic DNA. Method 1, a PCR-SSP method
uses two different 5¢ primers and a generic 3¢primer that distin-
guish the non-synonymous difference of HLA-E*01:01 and HLA-
E*01:03 at amino acid position 107. Method 2, an SSO method is
based on the Luminex technology, using different probes for HLA-
E*01:01, HLA-E*01:03, and HLA-E*01:04 attached to labelled
beads. Method 3 is a direct sequencing method based upon PCR
amplification and sequencing with specific primers enabling
sequencing of exons 1–5 and the intervening introns. The non-
coding differences between HLA-E*01:01:01:01, E*01:01:01:02,
and E*01:01:01:03 and between HLA-E*01:03:01:01 and
E*01:03:01:02 are located outside this region, whereas all synony-
mous and non-synonymous substitutions can be detected.

2. Materials

2.1. PCR-SSP Method 1. Optional Biohazard.

2.1.1. PCR Reaction 2. Thermocycler.
3. Vortex mixer with adjustable speed.
4. Centrifuge with Swing bucket rotor for 96-well microplate.
5. Micropipettes and tips.
 8 Molecular Typing of HLA-E 145

6. PCR tubes (Micronics) and caps or seal.

7. Mastermix (MM) for 1,000 reactions (store at −80°C):
– 1.0 mL 10× PCR buffer II (Perkin Elmer-buffer), (Applied
Biosystems).
– 0.6 mL MgCl2 25 mM, (Applied Biosystems).
– 20 mL Each dNTP 100 mM, (Amersham Pharmacia).
– Optional: 0.1 mL Cresol red 10 mg/mL, (Sigma).
– 0.5 mL Glycerol (99.5%), ICN.
– 2.3 mL Aqua Dest (AD) (or 2.4 mL if no cresol red is added).
8. AmpliTaq DNA polymerase (5 Units/mL), (Applied Biosystems).
Store at −20°C.
9. HLA-E amplification primers:
Name Direction Sequence 5¢-3¢ Location Position
MSSP08079 Forward CGAGCTGGGG Exon 3 740–756
CCCGACA
MSSP08080 Forward CGAGCTGGGG Exon 3 740–756
CCCGACG
MSSP08088 Reverse TTCCAGGTAGG Exon 3 902–920
CTCTCTGG

10. Internal control primers (located in the growth hormone gene):

Name Direction Sequence 5¢-3¢
IC1 Forward CAGTGCCTTCCCAACCATTCCCTTA
IC2 Reverse ATCCACTCACGGATTTCTGTTGTGTTTC

2.1.2. Agarose Gel 1. Agarose electrophoresis grade, (Invitrogen). Store at room

Electrophoresis temperature (RT).
2. 10× TBE buffer, (Gibco) (dilute with MilliQ to obtain 0.5×
TBE-dilution). Store at RT.
3. Ethidium bromide (50 mL, 1% in water), (Fluka Biochemika).
Store at 4°C.
4. 50 or 100 bp DNA ladder, (Invitrogen) dilute 50 mL DNA
marker with 950 mL aqua dest and 200 mL loading buffer
(dissolve in 10 mL AD: 100 mg Orange G, 10 mg SDS,
200 mL 0.5 M EDTA, 500 mL 1 M Tris/HCl pH 8.0, 1 g
Ficoll 400, add 6 mL glycerol and water to a final volume of
20 mL). Store at 4°C for maximum 1 year.
5. Geltray and combs.
6. Microgram scale.
7. 50°C Water bath.
146 N. Lauterbach et al.

8. Gel electrophoresis system, (Biorad).

9. Electrophoresis power supply.
10. UV transilluminator (Geldoc system), (Biorad).

2.2. PCR-SSO Method 1. HLA-E SSO Luminex kit (One Lambda, not commercially
available yet).
2.2.1. PCR, Hybridization,
and Luminex Bead – Denaturation buffer, store at RT.
Reactions – Neutralization buffer, store at RT.
– Hybridization buffer, store at RT.
– Wash buffer, store at RT.
– SAPE stock (100×), store at 4°C.
– SAPE buffer, store at 4°C.
– Primer Set D-Mix, store at −30°C (mix should be pink or
light purple, when necessary vortex).
– HLA-E specific amplification primers, store at −30°C.
– HLA-E SSO beads (including positive and negative control
beads), store at −30°C (after thawing store in the dark at
4°C, never re-freeze).
2. Optional: Biohazard.
3. AmpliTaq DNA polymerase (5 Units/mL), (Applied Biosystems).
Store at −20°C.
4. 1.5-mL microfuge tubes.
5. PCR tubes (micronics) and caps or seal.
6. Micropipettes and tips.
7. 96-Well, thin walled PCR plate and holder.
8. PCR Thermocycler.
9. Centrifuge
– Rotor for 1.5-mL microfuge tubes.
– Swing bucket rotor for 96-well microplate.
10. Vortex mixer with adjustable speed.
11. Luminex fluoranalyser (Luminex Corporation) or Labscan100
(One Lambda).
12. Luminex XMap Sheath fluid.

2.2.2. Optional: Agarose See Subheading 2.1.2.

Gel Electrophoresis

2.3. SBT Method 1. Expand High Fidelity PCR kit, containing 10× Expand High
Fidelity buffer and Expand High Fidelity enzyme mix, (Roche).
2.3.1. Amplification and
Sequencing Reaction 2. dNTP 10 mM each.
3. HLA-E amplification primers:
 8 Molecular Typing of HLA-E 147

Name Direction Sequence 5¢-3¢ Location Position

E08072 Forward CAGCGTCGCCA 5¢UT 75–54

CGACTCCCGAC
E08073 Reverse GGCTCGTGTGTGT Intron 5 2,219–
GGATGG 2,237
(Sequences from Paquay et al. (14), see Note 1)

4. HLA-E sequencing primers:

Name Direction Sequence 5¢-3¢ Location Position

E08074 Forward GAAGGACTC Intron 1 131–145

GGGGAG
E08075 Forward AGATTCACCCC Intron 2 549–567
AAGGCTG
E08076 Forward CTAAGTCCA Intron 3 1,465–1,481
GGCTGGTG
E08077 Reverse AGCCTTGGGG Intron 2 551–564
TGAATC
E08078 Reverse TCCCTGTTT- Intron 3 1,062–1,077
CTTCTAC
(Sequences from Paquay et al. (14), see Note 1).

5. ExoSAP-IT, (USB corporation), store at −20°C.

6. BigDye Terminator v1.1 Cycle Sequencing kit (Applied Bio-
systems) consisting of:
– 5× BigDye Terminator v1.1/3.1 sequencing buffer, store at
2–8°C.
– BigDye Terminator v1.1 cycle sequencing mix RR-2500
(BDT), store at temperatures between −15 and −25°C.
7. Montage SEQ96 Sequencing Reaction Cleanup kit consisting
of injection solution (Millipore) and 96-well SEQ filter plates,
(Montáge® life science).
8. 3700/3730 BigDye® Terminator v1.1 Sequencing standard,
(Applied Biosystems), store at temperatures between −15 and
−25°C.
9. 10× 3730 buffer with EDTA, prepare fresh by adding 16 mL
of buffer to 144 mL MilliQ water.
10. Performance optimized polymer (POP-7), (Applied Biosystems).
11. HiDi Formamide, aliquot, freeze-thaw only once, (Applied
Biosystems).
12. 96-well Optical Reaction plate with barcode (3730 plate),
(Applied Biosystems).
13. 3730 Trayholder, (Applied Biosystems).
148 N. Lauterbach et al.

14. Plate septa 96 well (Applied Biosystems).

15. Optional: Biohazard.
16. Vortex.
17. Centrifuge with Swing bucket rotor for 96-well microplate.
18. Micropipettes.
19. Pipette tips.
20. 1.5-mL microfuge tubes.
21. PCR tubes (micronics).
22. Thermocycler.
23. 3730 DNA analyzer, (Applied Biosystems).

2.3.2. Requirements See Subheading 2.1.2.

Agarose Gel Electrophoresis

3. Methods

3.1. PCR-SSP In the HLA-E PCR-SSP approach, two different PCR reactions are
used; one positive for HLA-E*01:01 and one positive for HLA-
E*01:03. Internal amplification control primers, located in the
growth hormone gene, are included to check for any inconsistencies
during the PCR. Analysis is performed by agarose gel electrophore-
sis and by the detection of ethidium bromide-stained fragments.

3.1.1. Amplification 1. Prepare the mastermix, the DNA sample and thaw primers
(see Note 2).
2. Provide a clean work environment, e.g. a Biohazard and keep
the reaction mixes on ice throughout the entire protocol.
3. For each sample, pipette the reaction mix for HLA-E*01:01
and HLA-E*01:03 amplification as described below in two
separate PCR tubes (It is most efficient to make reaction mixes
before starting, see Notes 3 and 4).
1× Reaction mix HLA-E*01:01:

4.6 mL MM
2 pmol IC1
2 pmol IC2
10 pmol Forward primer: MSSP08079
10 pmol Reverse primer: MSSP08088
Add aqua dest to a final volume of 8 mL
 8 Molecular Typing of HLA-E 149

1× Reaction mix HLA-E*01:03:

4.6 mL MM
2 pmol IC1
2 pmol IC2
10 pmol Forward primer: MSSP08080
10 pmol Reverse primer: MSSP08088
Add aqua dest to a final volume of 8 mL

4. Add 100 ng of DNA to each mix and centrifuge the PCR tubes
(pulse centrifugation until reaching 200 × g).
5. Dilute Taq polymerase with AD to obtain 0.33 Unit/mL and
add 1 mL per reaction/tube.
6. Centrifuge the PCR tubes again.
7. Cap or seal the PCR tubes and transfer them to a PCR
Thermocycler.
8. Run the PCR program:

2 min-96°C
10 cycles 10 s-94°C
1 min-65°C
20 cycles 10 s-94°C
50 s-61°C
30 s-72°C
Hold on 4°C

3.1.2. Electrophoresis 1. Transfer 200 mL of 0.5% TBE to an Erlenmeyer and add 3 g

agarose and mix.
2. Transfer the glass with TBE and agarose to a microwave and
boil until the agarose is fully solved (see Note 5).
3. Cool the solvent by holding the glass under cold running
water, swirling the glass to obtain an even temperature.
4. Put the glass covered with parafilm in a 50°C water bath for a
minimum of 10 min (see Note 6).
5. Add 10 mL ethidium bromide (final concentration 0.5 mg/
mL), swirl the glass, and pour the gel solvent in a geltray.
Discard the air bubbles and put combs in the gel.
6. Leave the gel for a minimum of 10 min at RT to solidify, fol-
lowed by approximately 30 min at 4°C.
7. Put sufficient 0.5× TBE buffer in the gel electrophoresis sys-
tem and add 10 mL ethidium bromide (the gel must be com-
pletely covered by buffer).
150 N. Lauterbach et al.

Fig. 1. Example of PCR-SSP reactions for detection of HLA-E*01:01 and HLA-E*01:03 in

three samples (see Note 7). The HLA-E-specific PCR product has a fragment length of
190 bp, the internal control is 429 bp. Sample 1 is typed as HLA-E*01:01 homozygous,
sample 2 as HLA-E*01:03 homozygous and sample 3 as HLA-E*01:01,01:03 heterozygous.

8. Remove the combs from the geltray and put the geltray in the
electrophoresis system.
9. Take the PCR tray out of the thermocycler and centrifuge the
tray (pulse centrifugation until reaching 200 × g).
10. Load 10 mL of each PCR product in each lane of the gel and
10 mL of 50 bp DNA marker in one lane.
11. Close the electrophoresis lid and run the gel for 36 min at
10 V/cm.
12. Following electrophoresis transfer the gel to a Geldoc system to
visualize the ethidium bromide-stained bands under UV-light.

3.1.3. Analysis 1. HLA-E*01:01 and HLA-E*01:03-specific PCR fragments

have a length of 190 bp and the internal control fragment
has a length of 429 bp. When a specific fragment is visible for
only the HLA-E*01:01 mix and negative for the HLA-E*01:03
mix, the sample is typed as HLA-E*01:01 homozygous.
A sample is HLA-E*01:03 homozygous, when a specific band
is visible for only the HLA-E*01:03 mix and negative for the
HLA-E*01:01 mix. When for both mixes the specific frag-
ments are present, the sample is heterozygous, HLA-
E*01:01,01:03. An example is shown in Fig. 1.

3.2. PCR-SSO For the SSO method, different probes specific for HLA-E*01:01,
*01:03, and *01:04 are bound to fluorescently coded beads, devel-
oped in collaboration with Dr J. Lee (One Lambda). HLA-E is
amplified using two amplification primers each of which is labelled
with biotin. After annealing of the biotinylated DNA amplicon to
the beads, the DNA is labelled by addition of streptavidin coated
phyco-erythrin. The fluorescence intensity is measured by luminex
 8 Molecular Typing of HLA-E 151

equipment and analysed manually to determine if the specific

probes have bound the complementary DNA and thereby identify
if a sample is positive for a certain allele. This method is suitable for
large-scale testing and easy to implement in routine luminex typing
approaches.

3.2.1. Amplification 1. Adjust the concentration of DNA to 20 ng/mL using sterile

water.
2. Prepare mastermix (see Note 8) for all reactions:

Per reaction 13.8 mL D-mix*

4.0 mL HLA-E-specific amplification primers
0.2 mL Taq polymerase
18.0 mL
Add the Taq polymerase immediately before use
*
see methods part 2.2.1

3. Pipette 2 mL of DNA onto the bottom of a PCR tube.

4. Mix and centrifuge the MM and aliquot 18 mL into each well
containing DNA.
5. Cap or seal and transfer tubes to a PCR Thermocycler.
6. Run PCR program:

3 min-96°C
5 cycles 20 s-96°C
20 s-60°C
20 s-72°C
30 cycles 10 s-96°C
15 s-60°C
20 s-72°C
10 min-72°C
Hold on 4°C

7. Amplified DNA is now ready to be tested by hybridization.

Optionally, 2–5 mL of the PCR product can be used for analysis
by gel electrophoresis. For method, see electrophoresis in
Subheading 3.1.2 (see Note 9).

3.2.2. Denaturation/ 1. Prepare a crushed ice bath and place a clean 96-well plate in a
Neutralization Procedure tray holder.
2. Transfer 5 mL of PCR product into a well of the plate.
3. Add 2.5 mL denaturation buffer per reaction, mix thoroughly, seal
or cap the tray and incubate for 10 min at room temperature.
4. Vortex the neutralization buffer, add 5 mL to each reaction,
mix thoroughly (notice the colour change to clear or pale
yellow) and place the plate on the ice bath.
152 N. Lauterbach et al.

3.2.3. Hybridization 1. Make sure that the thermocycler has been turned on and is
Procedure pre-warmed at 60°C.
2. Prepare hybridization mixture by combining 34 mL of hybrid-
ization buffer with 4 mL of bead mixture per reaction, see Note
10. Protect the mixture with the fluorescently labelled beads
from the light as much as possible.
3. Keep the tray on ice to prevent early annealing. Vortex the
hybridization mixture and add 38 mL to each well, mix well by
pipetting up and down.
4. Cover tray with tray seal and place PCR plate into the pre-
warmed thermocycler.
5. Incubate for 15 min at 60°C.
6. Following incubation, place tray in tray holder, remove seal,
and quickly add 100 mL wash buffer to each well.
7. Cover tray with tray seal and centrifuge for 5 min at
1,000–1,300 × g.
8. Remove wash buffer by flicking (see Note 11).
9. Repeat the washing step twice.
10. Vortex SAPE stock and prepare 1× SAPE solution during third
centrifugation (see Note 12):
0.5 mL SAPE stock + 49.5 mL SAPE buffer, keep solution in
the dark until use.

3.2.4. Labelling 1. Add 50 mL of 1× SAPE solution to each well, seal tray, and
vortex thoroughly at low speed.
2. Place PCR plate in the pre-warmed thermocycler and incubate
for 5 min at 60°C.
3. Following incubation, place tray in holder, remove seal, and
quickly add 100 mL wash buffer to each well.
4. Cover tray and centrifuge for 5 min at 1,000–1,300 × g.
5. Remove supernatant by flicking.
6. Add 70 mL wash buffer to each well, mix by pipetting, and
transfer 80 mL to reading plate.
7. Read the plate in the Luminex apparatus.
8. If tray is not immediately read, keep it in the dark and at 4°C
(see Note 13).

3.2.5. Analysis 1. The luminex output of one reaction will show the fluorescent
intensity (FI) value measured for each bead in the reaction.
There are 6 beads present, bead 35 is an internal negative con-
trol, no DNA must be bound to this bead. Bead 57 is a positive
control, each PCR product must be bound to this bead. Bead 58
contains a probe for HLA-E*01:01, bead 59 for HLA-E*01:03
 8 Molecular Typing of HLA-E 153

Table 1
FI values of the various beads and the calculated normalized
values

Sample 1 2 3

Bead 35 14 12 9
Bead 57 3,137 3,302 3,304
Bead 58 2,942 100 2,414
Normalized value bead 58 94 3 73
Bead 59 41 2,552 2,150
Normalized value bead 59 1 77 65
Bead 60 5,184 4,965 5,314
Normalized value bead 60 166 151 161
Bead 61 436 415 623
Normalized value bead 61 14 12 19
HLA-E typing result *01:01 *01:03 *01:01,01:03
Sample 1 is typed as HLA-E*01:01 homozygous, sample 2 HLA-E*01:03 homozy-
gous, and sample 3 heterozygous HLA-E*01:01,01:03

and HLA-E*01:04, bead 60 for HLA-E*01:01 and HLA-

E*01:03, and bead 61 for HLA-E*01:04. The measured
fluorescence intensity for each bead is an indication of the
amount of DNA annealed to the probe. A normalized value
can be calculated by [FI (bead)—FI (neg bead)]/[FI (pos
bead)–FI (neg bead)] *100. If the normalized value exceeds
30%, the bead is positive. Values below 30% are regarded as
negative (see Note 14). Based on the positive and negative
reactions, HLA-E typing can be obtained (see Table 1). The FI
of the positive control bead must be >1,000 and the FI of the
negative control bead must be <100 (see Note 15). Bead
counts need to be 100 or greater.

3.3. Sequencing- For sequence-based typing, the HLA-E gene is amplified using
Based Typing primers in the 5¢ UT region and in intron 5, enabling sequencing
of exons 1–5 and the intervening introns. The obtained sequences
are analysed using the Lasergene software program, or equivalents,
by aligning the sequence of a sample with the known HLA-E
alleles. Sequence comparison will enable typing of the alleles HLA-
E*01:01:01, 01:03:01, 01:03:02, 01:03:03, 01:03:04, 01:04.

3.3.1. Amplification 1. Provide a clean work environment, e.g. a Biohazard. Keep the
reaction mixes on ice throughout the entire protocol.
154 N. Lauterbach et al.

2. Transfer PCR tubes to a tray and put on ice, calculate the

number of reactions needed, and prepare one large batch of
mastermix (see Note 16):

Per reaction 5 mL Expand high fidelity buffer, 10× conc.

with 15 mL MgCl2
1 mL Each dNTP (10 mM)
1 mL Forward primer E08072 (10 pmol/mL)
1 mL Reverse primer E08073 (10 pmol/mL)
0.75 mL Expand high fidelity enzyme mix

Add aqua dest to a final volume of 48 mL

3. Pipette 48 mL MM/per PCR tube and add 100 ng of DNA.

4. Centrifuge (pulse centrifugation until reaching 200 × g).
5. Move the PCR tubes directly to the PCR Thermocycler.
6. Run PCR program:

2 min-94°C
10 cycles 15 s-94°C
30 s-65°C
2 min-72°C
10 cycles 15 s-94°C
30 s-61°C
2.5 min-72°C
10 cycles 15 s-94°C
30 s-61°C
3.5 min-72°C
7 min-72°C
Hold on 4°C

7. After the PCR reaction, 5–10 mL of the PCR products can be

checked by agarose gel electrophoresis according to the proto-
col in Subheading 3.1.2, running the gel for 60 min and using a
100 bp marker. A band with a length of 2.1 kb must be visible.

3.3.2. Purification of the 1. Pipette ExoSAP-IT in a PCR tube (all steps need to be per-
PCR Product formed on ice) followed by addition of the PCR product (4 mL
ExoSAP-IT/10 mL of PCR product), mix well and centrifuge
(impulse centrifugation until reaching 200 × g).
2. Transfer the PCR tubes to the PCR Thermocycler.
3. Run PCR program:

15 min 37°C
15 min 80°C
Hold on 4°C (see Note 17)
 8 Molecular Typing of HLA-E 155

3.3.3. Sequencing 1. In order to perform the sequencing reaction the sequencing

Reaction (primer) mix has to be made (see Note 18):

Per reaction 1.0 mL BDT sequencing mix

1.5 mL BDT sequencing buffer (5×)
0.5 mL Sequencing primer (5 pmol)
6.0 mL Aqua Dest
9.0 mL

2. Pipette 9 mL sequencing mix in a new PCR tube on ice and

add 1 mL purified PCR product, mix well.
3. Transfer the tubes to a PCR Thermocycler.
4. Run PCR program:

1 min-96°C
25 cycles 10 s-96°C
5 s-50°C
4 min-60°C
Hold on 4°C

3.3.4. Purification 1. Following the sequencing reaction, purify the products by

Following Sequence filtration using a vacuum system.
Reaction 2. Mix each sequencing reaction product with 20 mL of Millipore
buffer (from the Montage SEQ96 Sequencing Reaction Cleanup
kit) and pipette in a filter plate.
3. Put the filter plate on the vacuum system and extract for
5 min.
4. Add another 40 mL of millipore buffer to the filter plate
and extract again until all solution has been extracted (see
Note 19).
5. Add 30 mL of millipore buffer, put the filter plate on a vortex,
and shake for 10 min at max. 600/min.
6. Transfer 25 mL of the sample to a 3730 plate, add 10 mL of
sequencing standard to one well on the same plate (see Note 20)
and cover with a septa.
7. Put the plate in the 3730 DNA analyser (or store for maximum
1 day at 4°C) and run the sequencing program (see Note 21).

3.3.5. Analysis 1. The sequence output is analysed by the software Lasergene or

equivalents (see Note 22). Based upon the A → G nucleotide
substitution at position 382 in exon 3 (position refers to cDNA
reference sequence IMGT/HLA), HLA-E*01:01 can be dis-
tinguished from HLA-E*01:03 and HLA-E*01:04. HLA-
E*01:03 can be distinguished from HLA-E*01:04 by the
difference at position 532, *01:03 has an A at this position,
156 N. Lauterbach et al.

whereas *01:04 has a G. HLA-E*01:03:01, *01:03:02,

*01:03:03, and *01:03:04 can be identified based upon the
nucleotide substitutions at positions 294, 513, and 696.

4. Notes

1. The method described by Paquay et al. (14) is the sole

proprietary knowledge of Genome Diagnostics BV Utrecht,
the Netherlands.
PCR-SSP Method
2. It is most efficient to prepare the master mixes in large amounts
for multiple samples and aliquot.
3. Vortex the primers before adding to the mixes.
4. It is most efficient to prepare reaction mixes in larger amounts
for multiple samples. Prepared reaction mixes can be stored
at −80°C.
5. It is important to boil long enough until all agarose particles
are fully dissolved in the TBE buffer and when cooling down
to swirl the glass to prevent unequal solidification of the gel.
6. The temperature of the gel mixture must not exceed 60°C
when ethidium bromide is added because this high tempera-
ture will destroy the ethidium bromide.
7. Very weak or invisible DNA bands (including internal control)
can be caused by insufficient mixing of Taq polymerase before
adding it to the primer/DNA mixes.
PCR-SSO Method
8. Vortex D-mix and amplification primers before adding to the mix.
9. PCR products can be stored at −80°C to −20°C.
10. Vortex beads well before adding. Once beads are thawed, store
beads at 2–8°C and use within 3 months. Do not refreeze bead
mixture after thawing.
11. Try to remove (flick) as much wash buffer as possible by invert-
ing the tray and pouring into the waste in one movement. To
perform this accurately, training is needed. If you do not suc-
ceed in removing all wash buffer at once, you have to repeat
the flicking; however, you will have a higher risk of loosing
beads. Leaving too much wash buffer can affect your results
(see Note 15).
12. Remove the SAPE bottle from storage only when needed, and
return immediately to 2–8°C.
13. Prolonged storage of samples (more than 4 h) may result in
loss of signal.
 8 Molecular Typing of HLA-E 157

14. Since HLA-E*01:04 has never been detected after its initial
identification, it has not been possible to check the specificity
of the HLA-E*01:04 bead.
15. In case MFI values of the control beads are out of range (too
low positive or too high negative control values), we consider it
necessary to repeat the procedure. When the MFI value of the
positive beads is too low, this could be caused by inadequate
PCR amplification, that can be checked by agarose gel electro-
phoresis. If a correct strong PCR band is observed, the too low
positive control value might be caused by insufficient flicking,
resulting in a lower concentration of SAPE. A too high negative
control value might be due to some contamination. Repeating
the reactions with new reagents might be necessary.
SBT Method
16. Vortex the primers before adding to the mixes.
17. The purified PCR product can be stored for a maximum of
1 week at 4°C.
18. The amplification primers can also be used as sequencing prim-
ers in addition to the described sequencing primers.
19. It is important that all fluid has been extracted; otherwise, it
can result in a higher background because the remaining
labelled dNTPs could interfere with the sequence analysis.
20. The sequencing standard can be run to check for any inconsisten-
cies during the capillary electrophoresis and analysis procedure.
21. The sequencing reaction products can be analysed by any auto-
mated DNA sequencer that is able to detect the Big Dye
Terminator-labelled nucleotides.
22. After DNA sequence analysis, the sequences are stored in .AB1
files. DNASTAR software Lasergene contains EDITSEQ, ALIGN,
and SEQMAN modules. Individual sequences are selected, these
are the sample AB1 files, and HLA-E sequences obtained from the
IMGT/HLA database/EBI (www.ebi.ac.uk/imgt/hla) are used
as reference. The uploaded reference and sample sequence files are
assembled by SEQMAN. From the resulting contig, the nucle-
otide peaks can be viewed, checked and if necessary adjusted.
Extended sequences obtained as displayed in the contig can be
used as reference sequence additionally. Lasergene software is a
product from DNASTAR, Madison, USA.

Acknowledgements

The authors thank Jarhow Lee (One Lambda, Los Angeles, USA)
for the development of the HLA-E SSO kit and reagents and Els
Bielen and Timo Olieslagers for their practical assistance.
158 N. Lauterbach et al.

References
1. Antoun A et al (2009) Ethnic variability in diversity for the regulation of NK cells. Hum
human leukocyte antigen-E haplotypes. Tissue Immunol 61(11):1059–1065
Antigens 73(1):39–45 9. Tamouza R et al (2005) Association of HLA-E
2. Grimsley C et al (2002) Definitive high resolu- polymorphism with severe bacterial infection
tion typing of HLA-E allelic polymorphisms: and early transplant-related mortality in
identifying potential errors in existing allele matched unrelated bone marrow transplanta-
data. Tissue Antigens 60(3):206–212 tion. Transplantation 80(1):140–144
3. Ohya K, Kondo K, Mizuno S (1990) 10. Tamouza R et al (2006) Homozygous status
Polymorphism in the human class I MHC locus for HLA-E*0103 confers protection from acute
HLA-E in Japanese. Immunogenetics graft-versus-host disease and transplant-related
32(3):205–209 mortality in HLA-matched sibling hematopoi-
4. Grimsley C, Ober C (1997) Population genetic etic stem cell transplantation. Transplantation
studies of HLA-E: evidence for selection. Hum 82(11):1436–1440
Immunol 52(1):33–40 11. Danzer M et al (2009) Clinical significance of
5. Geraghty DE et al (1992) Polymorphism at HLA-E*0103 homozygosity on survival after
the HLA-E locus predates most HLA-A and -B allogeneic hematopoietic stem-cell transplanta-
polymorphism. Hum Immunol 33(3): tion. Transplantation 88(4):528–532
174–184 12. Ludajic K et al (2009) Association of HLA-E
6. Ulbrecht M et al (1999) Cell surface expression polymorphism with the outcome of hematopoi-
of HLA-E: interaction with human beta2- etic stem-cell transplantation with unrelated
microglobulin and allelic differences. Eur J donors. Transplantation 88(10):1227–1228
Immunol 29(2):537–547 13. Park KS et al (2007) HLA-E*0101 and HLA-
7. Strong RK et al (2003) HLA-E allelic variants. G*010101 reduce the risk of Behcet’s disease.
Correlating differential expression, peptide Tissue Antigens 69(2):139–144
affinities, crystal structures, and thermal stabili- 14. Paquay MM, Schellekens J, Tilanus MG (2009)
ties. J Biol Chem 278(7):5082–5090 A high-throughput Taqman(R) approach for
8. Maier S et al (2000) Implications of HLA-E the discrimination of HLA-E alleles. Tissue
allele expression and different HLA-E ligand Antigens 74(6):514–519
 Chapter 9

Molecular Analysis of Complement Component

C4 Gene Copy Number
Alison S.L. Castley and O. Patricia Martinez

Abstract
Classical, alternative, or lectin pathways may activate the complement system cascade. The classical pathway
includes the C4 protein and functions in the prevention of immune complex precipitation and in clearance
of immune complexes.
Two isotypes of C4—C4A and C4B—are coded by genes located at two loci within the major histo-
compatibility complex (MHC) on chromosome 6. While these isotypes share over 99% amino acid sequence
homology, five nucleotide differences located in exon 26 are responsible for major structural and functional
differences between the C4 isotypes.
C4A and C4B are highly polymorphic with over 40 alleles, gene duplications, and “null alleles”. C4
genes may be short (14.6 kb) or long (21 kb), due to the absence or presence of an endogenous retroviral
sequence—HERV-K(C4)—in intron 9, respectively. The C4 gene copy number (GCN) can vary from 1–3
per haplotype or 2–6 per diploid genome. The variation in GCN leads to a range of C4 plasma protein
concentrations among healthy subjects. In subjects with equal numbers of C4 genes, subjects with short
genes have C4 plasma levels relatively higher than subjects with long genes.
Variation of the C4 GCN, the gene size (long or short) and the C4 isotypes (C4A and C4B) may also
lead to susceptibility to autoimmune disease. Therefore, in subjects with autoimmune disease, a low serum
C4 level may be due to ongoing disease activity associated with complement activation and consumption
or it may be due to genetic factors. Distinguishing between these will have clinical implications.
Exact determination of GCN can be difficult, at least in part due to the high degree of homology
between C4A and C4B and a variety of techniques has been described. This chapter describes a quantitative
TaqMan real-time PCR (qPCR) copy number assay, based on our laboratory experience using this assay.

Key words: Human complement genes, C4A, C4B, Gene copy number, Real-time PCR

1. Introduction

The complement system consists of at least 26 proteins that are

expressed constitutively and non-specifically. These proteins partici-
pate in a cascade of reactions that need to be activated to function.
There are three major pathways for activation of the complement

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_9, © Springer Science+Business Media New York 2012

159
160 A.S.L. Castley and O.P. Martinez

cascade—classical, alternative, and lectin pathways. Different mech-

anisms or substances activate each of these.
The classical pathway is involved in the prevention of immune
complex precipitation and in enhancing binding to CR1 and clearance
of immune complexes (1). Homozygous deficiency of components of
the classical pathway—C4, C2 and C1—predispose to immune com-
plex-mediated autoimmune disease (2).
The C4 protein is coded by genes located at two loci within
the major histocompatibility complex (MHC) on chromosome 6.
The length of each of these genes varies depending on the presence
or absence of a 6.5 kb endogenous retrovirus insertion—HERV-
K(C4)—in intron 9, producing “long” (22 kb) and “short”
(14.6 kb) C4 genes (3–7).
The products of these loci—C4A and C4B (8) exhibit an
extremely high degree of genomic sequence homology through-
out the C4 molecule (9) and share over 99% amino acid sequence
(7). Despite this amino acid homology, C4A and C4B are func-
tionally different. C4A binds preferentially to amino groups, such
as those found in immune complexes (10), while C4B binds pref-
erentially to acceptor hydroxyl groups found predominantly on the
carbohydrates of red blood cell surfaces (11).
C4A and C4B are very polymorphic, with over 40 alleles,
including “null alleles” and gene duplications (4, 12). C4 null alleles
are defined by the absence of C4 protein in plasma (8, 13, 14).
They may be due to gene deletion (4, 15), homo-duplication (16),
gene conversion (12), or mutations leading to non-expression.
The variation in gene copy number (GCN) leads to a range of
C4 plasma protein concentrations. Uko et al. defined reference
ranges for serum C4 concentration in subjects with 0, 1, and 2 C4
null alleles (17) and Rebmann et al. used the C4 plasma concentra-
tion to define C4 null alleles and C4A duplications (18). The size
of C4 genes seems to determine C4 protein level and activity, with
higher C4 protein levels and higher C4 haemolytic activities being
detected in subjects with short C4 genes than those with long
genes only. A good correlations between the number of long genes
with serum levels of C4A, and the number of short genes with
serum levels of C4B has also been reported (19–21).
The GCN of C4 and surrounding genes can vary from 1–3 per
haplotype (4) or 2–6 per diploid genome, with 60.8% of people
with four copies of C4 genes, 27.2% with less than four copies, and
12% with more than four copies (22). Different ancestral haplo-
types (AHs) of the MHC carry specific numbers of several genes,
including the C4 genes (23).
It has been suggested that variation of the C4 gene copy number,
the polymorphism of gene size (long or short) and the C4 isotypes—
C4A and C4B—may lead to susceptibility to autoimmune disease.
An increased frequency of C4A null alleles in SLE was first reported
in 1983 (24). More recently, a variation of 2–6 total C4 GCN has
9 Molecular Analysis of Complement Component C4 Gene Copy Number 161

been reported in 1,241 European Americans, with 0–5 C4A and 0–4
C4B isotypes. The risk of SLE significantly increased in subjects with
only two copies of C4 and patients with SLE had lower total C4 and
C4A genes (OR = 6.514; P = 0.00002) (21). Another study has shown
susceptibility to SLE decreased in subjects with five or more total
copies of C4 (OR = 0.466; P = 0.016). Both zero copies (OR = 5.267;
P = 0.001) and one copy (OR = 1.613; P = 0.022) of C4A were risk
factors for SLE, whereas three or more copies of C4A appeared to be
protective (OR = 0.574; P = 0.012) (25).
Polymorphism of C4 can be defined by a number of methods,
including electrophoretic mobility of the intact protein or its sub-
units, serologically using the Rodgers (Rg) and Chido (Ch) deter-
minants (26) and by DNA-based methods (27). Exact determination
of GCN can be difficult, at least in part due to the high degree of
homology between C4A and C4B. Furthermore, the polymor-
phisms are located in the C4d fragment of the a-chain of the C4
molecule, and only five nucleotide polymorphisms located in exon
26 are responsible for the major structural and functional differ-
ences between the C4 isotypes (28).
A number of approaches have been tried to determine the C4
GCN, including counting of fragment numbers following restriction
fragment length polymorphism (RFLP) analysis (25), densitometry
of products obtained by pulsed field gel electrophoresis (PFGE) and
Southern blotting (22), and a multiplex ligation-dependent probe
amplification (MLPA) assay combined with isotype-specific ELISAs
to determine the level of C4A and C4B (20).
Wu et al. developed quantitative TaqMan real-time PCR (qPCR)
assays based on nucleotide sequences specific for the C4A and C4B
genes, structural characteristics corresponding to long and short C4
genes, and the breakpoint region of RP-C4-CYP21-TNX (RCCX)
modular duplication (29). Using cloned C4 genomic DNA (gDNA)
covering six logs of DNA concentrations for calibration and a relative
standard curve they described eight HLA haplotypes with single C4
genes in mono-modular RCCX that are associated with multiple auto-
immune and infectious diseases and 32 bi-modular, 4 tri-modular, and
one quadri-modular RCCX (29).
In this chapter, a qPCR copy number assay based on the
method of Wu et al. (29) is described.
In a qPCR, a fluorescence dye is incorporated into the amplified
DNA as the PCR proceeds. The dye emits light proportional to the
number of amplified copies. The TaqMan dye chemistry is specific
and precise for monitoring the kinetics of the amplification process.
Quantification of the initial amount of DNA present in the reaction
is based on the number of cycles required to reach a threshold (CT),
at which the fluorescence is increased significantly and passes an
arbitrarily defined value. The greater the initial amount of DNA
template, the fewer cycles that are required for the reaction to reach
the fluorescence threshold and a smaller CT observed.
162 A.S.L. Castley and O.P. Martinez

The TaqMan® Copy Number Assay is designed to determine copy

number variation within the genome. A duplex PCR of the target
gene and a reference sequence known to be present as two copies in a
diploid genome is performed. The copy number of the target sequence
in a gDNA sample is determined relative to the reference sequence.
The comparative method measures the comparative differ-
ence between the target and the reference sequences, and then
compares the comparative difference values of test samples to a
calibrator sample(s) known to have two copies of the target
sequence. The copy number of the target is calculated to be two
times the relative quantity.

2. Materials

1. Purified patient gDNA at 5 ng/mL (see Notes 1 and 2).

2. C4A TaqMan® Copy Number Assay (100× or 20×—Supplied
by Applied Biosystems) (see Notes 3–5).
(a) Unlabelled, target-specific forward and reverse primers to
amplify the C4A gene(s).
(b) FAM™ dye-labelled TaqMan® MGB probe to detect the
C4A sequence.
3. C4B TaqMan® Copy Number Assay (100× or 20×—Supplied
by Applied Biosystems) (see Notes 3–5).
(a) Unlabelled, target-specific forward and reverse primers to
amplify the C4B gene(s).
(b) FAM™ dye-labelled TaqMan® MGB probe to detect the
C4B sequence.
4. TaqMan® Copy Number Reference Assay RNase P (Supplied
by Applied Biosystems) (see Notes 4–6).
(a) Two primers to amplify the genomic reference sequence.
(b) VIC® dye-labelled TAMRA™ probe to detect the genomic
reference sequence.
5. TaqMan® Genotyping Master Mix containing AmpliTaq Gold®
DNA Polymerase, UP (Ultra-Pure) and dNTPs (Supplied by
Applied Biosystems).
6. Applied Biosystems Real-Time PCR Genetic Analyser (i.e. 7900,
7900HT) (purchased from Applied Biosystems).
7. Applied Biosystems CopyCaller™ Software—for post-PCR
data analysis of copy number quantitation (Supplied by Applied
Biosystems). This is free to download from AB Web site:
http://www.appliedbiosystems.com.au (see Note 7).
8. Sterile deionised water (CSL).
 9 Molecular Analysis of Complement Component C4 Gene Copy Number 163

9. Adjustable pipettes with aerosol barrier tips (manual or electronic,

P10, P20, P100, P200).
10. MicroAmp® Optical 96-Well Reaction Plates (Applied
Biosystems).
11. MicroAmp® Optical Adhesive Film or MicroAmp® Optical
Caps (Applied Biosystems).
12. Sterile 0.5-mL Eppendorf microcentrifuge tubes (Crown
Scientific).
13. Sterile biological safety cabinet.
14. Vortex mixer (Ratek).
15. Plate centrifuge.
16. Separate pre-PCR (an area to set up the C4 reactions) and
post-PCR areas (see Note 8).

3. Methods

The qPCR reaction occurs in a single tube or well, using universal

PCR cycling conditions on an Applied Biosystems Real-Time PCR
System. Following amplification, data files containing the sample
replicate comparative values for each reporter dye can be exported
from the qPCR instrument software and imported into a software
analysis tool.

3.1. Pre-PCR Set Up 1. Ensure blood has been collected in either ACD or EDTA
collection tubes.
3.1.1. Prepare Genomic DNA
2. Extract, purify, and quantify DNA using your usual laboratory
methods.
3. Dilute the gDNA. Use sterile deionised water to obtain a con-
centration of 5 ng/mL (see Notes 1 and 2).
4. Use the diluted gDNA immediately or keep at 4°C for long-
term storage.
5. Vortex the diluted gDNA for 5 s before use.

3.1.2. Samples to Be 1. Test samples for target copy number assessment.

Included 2. Control samples with known C4 copy number for each C4
assay (see Note 9).
3. A “no template” control (sterile deionised water) used to
detect background fluorescence and exclude contamination.
4. For accurate copy number assignment, four replicates of each
gDNA test sample, each gDNA control sample and the “no
template” control must be set up.
164 A.S.L. Castley and O.P. Martinez

Table 1
Reaction mixture composition – Volume (mL) of reaction mixture components
required in a C4 assay

Volume of reagents (mL)

1 Sample (mL) Number of reactionsa

C4A or C4B TaqMan® Copy 1 1 mL × number of reactions
number assay (20×)
TaqMan® Copy number reference 1 1 mL × number of reactions
assay RNase P (20×)
TaqMan® Genotyping master mix (2×) 10 10 mL × number of reactions
Sterile deionised H2O 4 4 mL × number of reactions
Total 16 16 mL × number of reactions
a b
Number of reactions = 4 × (number of samples + controls ) + 2
b
Samples known to contain 0–4 copies of C4A or C4B (see Note 9). One of these samples, usually the sample containing
two copies of C4A or C4B is used as the calibrator sample (see Note 10)

5. A calibrator sample is also required. This is a gDNA sample

with known copy number for target of interest—C4A or C4B,
depending on the assay being performed (see Note 10).

3.2. Prepare PCRs 1. Calculate the volume of components needed, based on the
reaction volume and the number of reactions (see Table 1).
Include excess volume to provide for losses that occur during
reagent transfers (equivalent to two extra samples). Thus,
number of reactions = 4 × [number of samples + controls] + 2.
2. Thaw the TaqMan® Copy Number and TaqMan® Copy Number
Reference reagents at room temperature, vortex gently for 5 s
and then centrifuge tubes briefly (500 rpm for 10 s).
3. Thoroughly mix the TaqMan® Genotyping Master Mix by
flicking and briefly centrifuge (500 rpm for 10 s).
4. Combine the required volumes of reaction components in a
microcentrifuge tube (see Table 1). Use separate microcentri-
fuge tubes for the C4A and for the C4B assays and ensure the
tube caps are firmly in place to avoid leakage.
5. Mix the contents thoroughly by flicking or inverting the micro-
centrifuge tubes, then centrifuge briefly at 500 rpm for 10 s.
6. Pipette 16 mL of reaction mixture into each of the appropriate
wells of a MicroAmp® Optical 96-well reaction plate.
7. Briefly vortex the diluted gDNA samples for 5 s (from
Subheading 3.1.1, step 5).
8. Using a different pipette tip for each sample, add 4 mL of
gDNA (5 ng/mL) and the control samples to the appropriate
 9 Molecular Analysis of Complement Component C4 Gene Copy Number 165

replicate wells of the MicroAmp® Optical 96-well reaction

plate containing the reaction mixture (see Note 11).
9. Mix the reaction mixture and gDNA by pipetting up and down
several times (see Note 12).
10. Seal the reaction plate with optical adhesive film or optical caps,
then centrifuge the reaction plate briefly (500 rpm for 10 s).
11. Inspect all the wells to ensure a uniform volume.
12. Run the reactions.

3.3. Running the PCRs 1. Place the reaction plate on the 7900 qPCR instrument to be
on the Real-Time used.
Instrument 2. Under “Instrument” ensure the following parameters are
correctly identified on the template.
Parameter Setting

Run Standard
Reaction plate 96-Well standard
Ramp speed/model 9600 Emulation
Sample volume 20 mL

3. Ensure all samples within a C4 copy number assay are uniquely

identified and ensure all replicates have the same identifier (see
Note 13).
4. Clearly identify the copy number assay for the appropriate wells
and omit any unused wells.
5. Run the plate under these conditions:
(a) Hold at 95°C for 10 min.
(b) 40 Cycles each of:
(i) 95°C for 15 s.
(ii) Followed by 60°C for 60 s.

3.4. Post Real-Time 1. Open the Analysis Settings window of the RT-PCR Instrument
PCR Analysis software and set:
(a) Manual CT Threshold of 0.2.
(b) Autobaseline—On.
2. Apply settings and close window.
3. Select “Analyse” to analyse experiment and review data.
4. Select the VIC and FAM tabs in the analysis programme indi-
vidually and make sure that the amplification curves have a dis-
tinct exponential amplification phase for (see Figs. 1 and 2):
(a) Reference Assay (VIC® dye signal) in all samples.
(b) Copy Number Assay (FAM® dye signal) in most wells.
166 A.S.L. Castley and O.P. Martinez

Fig. 1. Principles of the real-time PCR. A fluorescent dye is incorporated into the amplified DNA and emits light proportional
to the number of amplified copies. Quantification of the initial amount of target DNA sequence present in the reaction is
based on the number of cycles required for the measured fluorescence to reach a given threshold. The number of cycles
required to reach that threshold is the CT. The CT value is highly proportional to the copy number of the target sequence in
the template. The greater the copy number present, the fewer cycles of PCR required for the accumulated product to be
detected in the PCR process, and the lower the CT value.

Fig. 2. An amplification profile from a C4B copy number assay from the 7900 (FAM representation). Samples with 1, 2, and
3–4 copy numbers are identified. Zero copy numbers are seen as background fluorescence. The CT values for samples with
one and two C4B copies are 26 and 25 respectively. The CT value for the sample with three C4B copies is just over 24 and
for the sample with four C4B copies is just under 24 (see Figs. 1 and 3).
 9 Molecular Analysis of Complement Component C4 Gene Copy Number 167

5. Review the amplification curves for all samples and controls.

If there has been a problem with amplification (either no, weak
or different shaped amplification curves) review the method
(see Notes 14–17).
6. Review the RT-PCR data of replicate samples. Ensure three or
more replicates have amplified adequately (see Figs. 1 and 2).
Adequate amplification is required for Copy Caller™ analysis.
7. Export results. Export each experiment result or results table
to one or more exported RT-PCR files (.txt file).

3.5. Utilising 1. Open the CopyCaller™ software on the computer.

CopyCaller™ Software 2. Import the exported (.txt) data into the CopyCaller™ software.
for Final Analysis
3. Go to “Assay Selection” and highlight one assay at a time.
Identify the calibrator sample for use in the analysis go to
“Tools” → “Analysis setting” and ensure the 2-copy number
calibrator is selected, then “Apply” the parameters.
4. Place a “tick” in the C4 assay you analysed (under “Assay
Selection”). The results for the assay should appear.
5. The CopyCaller™ software will assign a copy number to each
sample for each assay. The results can have the 0–8 copy num-
bers (see Fig. 3 and Note 18).

4. Notes

1. Patient gDNA must have a UV absorbance A260/A280 ratio

greater than 1.7. The ratio of absorption at 260 vs. 280 nm is
commonly used to assess protein contamination in solutions.
2. All samples must have the same amplification rate. This is
achieved by diluting all gDNA samples to 5 ng/mL. Use the
same amount of gDNA for each sample and its replicates to be
run in the same assay.
3. If your Taqman® Copy number assay is supplied at 100× con-
centration, dilute it to 20× with sterile deionised water.
4. To minimise the freeze–thaw cycles of the Taqman® copy num-
ber kits and reference kits, sub-aliquot small volumes (15 mL)
into small Eppendorf tubes.
5. The Taqman® copy number kits and reference kit should be
stored protected from light at −15 to −25°C.
6. The TaqMan® Copy Number Reference RNase P has been
optimised for copy number assays by Applied Biosystems. This
kit detects the ribonuclease P RNA component H1 (H1RNA)
gene (RPPH1) on chromosome 14 and is known to be present
as two copies in a diploid genome.
168 A.S.L. Castley and O.P. Martinez

Fig. 3. Results generated by the CopyCaller™ software from the C4B amplification profiles shown in Fig. 2. (a) A bar graph
of the C4 copy number calculated by the software for each sample from the results shown in (b). A copy number range bar
(asterisk) indicates the minimum and maximum copy number calculated for the replicates of each sample. (b) The sample
location in the 96-well reaction plate (well), the sample identification (sample), the CT for FAM (C4) and for VIC (reference
sequence) and the difference between the CT values for C4 and the reference (DCT). For example—wells 57–60 contain
replicates of the control that has two copies of C4B (2 CN CTL), with FAM CT values of 24.9362; 24.8713; 24.8222, and
24.9778. The same samples have VIC CT values of 25.5299; 25.3803; 25.4582, and 25.4402, respectively. (c) The results
of the copy number analysis of individual samples. Each row shows the combined values of the replicates used to generate
the results for a single sample, the expected and the observed copy number for each sample and the mean value of the
FAM and VIC CT for C4 and the reference assay, respectively.

7. To download the Applied Biosystems CopyCaller™ Software,

Microsoft Windows® XP Service pack 2 or a later version is
required.
8. To avoid contamination of the gDNA, the Taqman® copy
number kits and the reference kit are important to have sepa-
rate areas for PCR reaction set-up (Pre-PCR) and for DNA
amplification (Post-PCR).
9. It is important to include gDNA control samples for each C4
copy number assay. Samples with 0–4 copies of C4A and C4B
can be purchased from the Coriell Institute for Molecular
Research (www.coriell.org). Once you have determined the
copy number of C4A and C4B in your in-house gDNA stocks,
9 Molecular Analysis of Complement Component C4 Gene Copy Number 169

these can be used as alternative control samples. The known

copy number should always be correctly assigned. The 1–4
copy number controls should always amplify both the C4 assay
(FAM) and the reference assay (VIC) control. The 0 copy
number controls should amplify only the reference assay
(VIC).
10. The calibrator sample is required by the CopyCaller™ software
for correct assignment of C4 copy numbers to the samples
tested. It is advised to use one of the control samples as the
calibrator sample. For the detection of C4A copy numbers, use
a gDNA sample with two copies of C4A and for the detection
of C4B copy numbers use a gDNA sample with two copies of
C4B.
11. To avoid sample mix-up and for ease of identifying samples for
analysis, place all replicates of a sample in adjacent wells.
12. To avoid cross-contamination, use a fresh aerosol pipette tip to
add and mix each set of replicates. Discard each tip after each
set of replicate samples are mixed.
13. Unique identification of all samples for a copy number assay is
an important requirement for analysis using the CopyCaller™
software. Ensure all replicates of a sample have the same
identifier.
14. If problems with DNA amplification have occurred, ensure
that:
(a) Samples are not contaminated—check the 260/280 ratio
and re-extract gDNA if required.
(b) The gDNA used was at the correct concentration of
5 ng/mL.
(c) The correct thermal cycling conditions were utilised.
(d) Reagents have not expired—correct handling, mixing, and
storage of reagents are essential.
(e) There is no inhibitor to the reaction. A serial dilution of
the sample can be evaluated to test for this.
(f) Pipetting techniques are accurate. Use an aerosol-resistant
pipette tip and follow appropriate dispensing techniques
to prevent bubbles.
(g) The workspace if free from contaminants. Wear clean
gloves and a clean lab coat. Open and close samples and
reagents carefully to avoid splashing and clean down work
space appropriately.
(h) For each well the appropriate detector and dye informa-
tion are selected (FAM for C4 assays and VIC for refer-
ence assay).
170 A.S.L. Castley and O.P. Martinez

If all of these have been fulfilled and amplification of the samples

continues to fail, the samples may have a polymorphism in the
primer- or probe-binding site.
15. The “no template control” should not amplify with either the
C4 assays or the reference assay.
16. New batches of C4 TaqMan® Copy Number Assay and
TaqMan® Copy Number Reference Assay RNase P should be
tested using the control gDNA before routine use.
17. New batches of TaqMan® Genotyping Master Mix should be
tested before routine use by running the new batch and the
current batch in parallel.
18. The controls utilised in this assay have known C4A and C4B
copy numbers (i.e. 0–4 C4 copies). Therefore, the observed
copy number for these samples should be the same as the
expected copy number for the assay to be valid. All of these
should amplify with the reference assay (VIC). All samples,
except those with 0 copies of C4, should also amplify with the
C4 assay (FAM).

References
1. Schifferli JA, Ng YC, Peters DK (1986) The component C4 gene is due to an ancient retro-
role of complement and its receptor in the viral integration. Exp Clin Immunogenet 12:74
elimination of immune complexes. N Engl J 8. Awdeh ZL, Alper CA (1980) Inherited struc-
Med 21:488 tural polymorphism of the fourth component
2. Lachmann PJ (1987) Heberden Oration 1986. of human complement. Proc Natl Acad Sci
Complement—friend or foe. Br J Rheumatol USA 77:3576
26:409–415 9. Ulgiati D, Abraham LJ (1996) Comparative anal-
3. Prentice HL, Schneider PM, Strominger JL ysis of the C4B1 gene from the HLA-A1, B8,
(1986) C4B gene polymorphism detected in a DR3 haplotype. Exp Clin Immunogenet 13:43
human cosmid clone. Immunogenetics 23: 10. Law SA, Doods AW, Porter RR (1984) A com-
274 parison of the properties of two classes, C4A
4. Schneider PM, Carroll MC, Alper CA et al and C4B, of the human complement compo-
(1986) Polymorphism of the human comple- nent C4. EMBO J 3:1819
ment C4 and steroid 21-hydroxylase genes: 11. Isenman DE, Young JR (1984) The molecular
restriction fragment length polymorphisms basis for the difference in immune hemolysis
revealing structural deletions, homoduplica- activity of Chido and Rodgers isotypes of human
tions, and size variants. J Clin Invest 78:650 complement component C4. J Immunol 132:
5. Palsdottir A, Fossdal R, Arnason A et al (1987) 3019
Heterogeneity of human C4 gene size: a large 12. Braun L, Schneider PM, Giles CM et al (1990)
intron (6.5 kb) is present in all C4A genes and Null alleles of human complement C4. Evidence
some C4B genes. Immunogenetics 25:299 for pseudogenes at the C4A locus and for gene
6. Dangel AW, Mendoza AR, Baker BJ et al conversion at the C4B locus. J Exp Med
(1994) The dichotomous size variation of the 171:129
human complement C4 genes is mediated by a 13. O’Neill GJ, Yang SY, Dupont B (1978) Two
novel family of endogenous retroviruses, which HLA-linked loci controlling the fourth compo-
also establish species-specific genomic patterns nent of human complement. Proc Natl Acad
among Old World primates. Immunogenetics Sci USA 75:5165
40:425 14. Mauff G, Brenden M, Braun-Stilwell M et al
7. Chu X, Rittner C, Schneider PM (1995) Length (1990) C4 reference typing report. Complement
polymorphism of the human complement Inflamm 7:193
 9 Molecular Analysis of Complement Component C4 Gene Copy Number 171

15. Carroll MC, Palsdottir A, Belt KT, Porter RR temic lupus erythematosus. Cytogenet
(1985) Deletion of complement C4 and steroid Genome Res 123(1–4):131
21-hydroxylase genes in the HLA class III 23. Zhang WJ, Degli-Esposti MA, Cobain TJ et al
region. EMBO J 4(10):2547 (1990) Differences in gene copy number car-
16. Yu CY, Campbell RD (1987) Definitive RFLPs ried by different MHC ancestral haplotypes.
to distinguish between the human complement Quantitation after physical separation of haplo-
C4A/C4B isotypes and the major Rodgers/ types by pulsed field gel electrophoresis. J Exp
Chido determinants: application to the study of Med 171:2101
C4 null alleles. Immunogenetics 25:383 24. Christiansen FT, McCluskey J, Dawkins RL
17. Uko G, Christiansen FT, Dawkins RL, McCann et al (1983) Complement allotyping in SLE:
VJ (1986) Reference ranges for serum C4 association with C4A null. Aust N Z J Med
concentrations in subjects with and without C4 13:483
null alleles. J Clin Pathol 39:573 25. Bohme JM, Andersson G, Andersson E et al
18. Rebmann V, Doxiadis I, Kubens BS, Grosse- (1985) HLA-DR beta genes vary in number
Wilder H (1992) Quantitation of the human between different DR specificities whereas the
component C4: definition of C4 Q0 alleles and number of DQ beta genes is constant.
C4A duplications. Vox Sang 62:117 J Immunol 135:2149
19. Yang Y, Chung EK, Zhou B et al (2003) 26. Giles CM, Uring-Lambert B, Goetz J et al
Diversity in intrinsic strengths of the human (1988) Antigenic determinants expressed by
complement system: serum C4 protein concen- human C4 allotypes; a study of 325 families
trations correlate with C4 gene size and poly- provides evidence for the structural antigenic
genic variations, haemolytic activities, and model. Immunogenetics 27:442
body mass index. J Immunol 171(5): 27. Schneider PM, Stradmann-Bellinghausen B,
2734–2745 Rittner C (1996) Genetic polymorphism of the
20. Wouters D, van Schouwenburg P, van der Horst fourth component of human complement:
A et al (2009) High-throughput analysis of the population study and proposal for a revised
C4 polymorphism by a combination of MLPA nomenclature based on genomic PCR typing of
and isotype-specific ELISA’s. Mol Immunol Rodgers and Chido determinants. Eur J
46(4):592 Immunogenet 23:335
21. Yang Y, Chung EK, Wu YL et al (2007) Gene 28. Yu CY, Belt KT, Giles CM et al (1986)
copy-number variation and associated polymor- Structural basis of the polymorphism of
phisms of complement component C4 in human human complement components C4A and
systemic lupus erythematosus (SLE): low copy C4B: gene size, reactivity and antigenicity.
number is a risk factor for and high copy number EMBO J 5:2873
is a protective factor against SLE susceptibility in 29. Wu YL, Savelli SL, Yang Y et al (2007) Sensitive
European Americans. Am J Hum Genet 80: and specific real-time polymerase chain reac-
1037 tion assays to accurately determine copy num-
22. Wu YL, Yang Y, Chung EK et al (2008) ber variations (CNVs) of human complement
Phenotypes, genotypes and disease suscepti- C4A, C4B, C4-long, C4-short, and RCCX
bility associated with gene copy number vari- modules: elucidation of C4 CNVs in 50 con-
ations: complement C4 CNVs in European sanguineous subjects with defined HLA geno-
American healthy subjects and those with sys- types. J Immunol 179(5):3012
 Chapter 10

Genotyping of Single Nucleotide Polymorphisms

by 5¢ Nuclease Allelic Discrimination
Mari Malkki and Effie W. Petersdorf

Abstract
Real-time quantitative PCR is an efficient method for high-throughput genotyping of single nucleotide
polymorphisms (SNPs). In this chapter, we describe the 5¢ nuclease allelic discrimination assay for geno-
typing biallelic SNPs.

Key words: Single nucleotide polymorphism, Real-time polymerase chain reaction, TaqMan®
chemistry, Reporter dye, Fluorescent label

1. Introduction

Single nucleotide polymorphisms (SNPs) are the most common

form of human genetic variation (1–8). SNPs are biallelic and occur
approximately every 1,000 base pairs (bp) throughout the human
genome (1–8). SNPs are powerful markers for mapping genes that
cause disease (9–21). Association of SNPs residing within the major
histocompatibility complex (MHC) have recently been described
for diseases with well-known associations to HLA phenotypes and
haplotypes, including rheumatoid arthritis, diabetes, and multiple
sclerosis (12–21).
Because of their biallelic nature, SNPs can be readily geno-
typed using techniques that discriminate any two-way combination
of adenine, guanine, cytosine, and thymine nucleotide bases.
Among the most robust assays for SNP genotyping, real-time
quantitative assays rely on sensitive and specific quantification of
amplified DNA or cDNA using polymerase chain reaction (PCR)
(22–24). Real-time PCR methods offer a variety of different labels

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_10, © Springer Science+Business Media New York 2012

173
174 M. Malkki and E.W. Petersdorf

for the detection of genetic variation, including DNA-intercalating

agents, such as SYBR® Green (25), fluorogenic probes (26–28),
and the widely used TaqMan® chemistry (29, 30).
In this chapter, we describe the real-time 5¢-nuclease genotyping
assay known as TaqMan® for discriminating between two alleles of
a specific SNP. The method may be used for genotyping individual
SNPs that reside throughout the human genome. In brief, a wild-
type SNP Allele “A” is amplified separately from the alternative
Allele “B” using region specific forward and reverse primers and
two allele-specific TaqMan® probes designed to target the poly-
morphism. TaqMan® probes have a fluorescent reporter dye (VIC®
specific for allele “A” and 6-carboxyfluorescein [FAM™] specific
for allele “B”) attached to its 5¢ end and a quencher dye (minor
groove binder [MGB] or 6-carboxy-tetramethyl-rhodamine
[TAMRA]) at its 3¢ end. The amplification is performed using a
thermal cycler or a real-time PCR system. During amplification
each uniquely labeled probe binds preferentially to one of the two
alleles of the SNP of interest with different affinity. As amplification
proceeds, the Taq polymerase enzyme cleaves the bound probe,
and a fluorescent signal is generated. Fluorescent signals are inter-
preted automatically using sequence detection software dedicated
to real-time PCR instrumentation. A fluorescent signal from only
the VIC dye indicates homozygosity for Allele “A”; the presence of
only FAM dye fluorescence indicates homozygosity for Allele “B,”
and the presence of both fluorescent signals indicates Allele “A”/
Allele “B” heterozygosity.

2. Materials

2.1. TaqMan Assay Supplies needed for performing the TaqMan® assay are as follows:
1. 96-Well optical plates and adhesive optical covers (Applied
Biosystems, [ABI]).
2. Splash-free support base for 96-well plates.
3. Plate centrifuge.
4. 10, 200 and 1,000-mL sterile tips.
5. 1–10, 20–200 and 100–1,000-mL single or multiple tip electronic
pipets or 10, 20, 200 and 1,000-mL manual pipets.
6. Eppendorf tubes.
7. TaqMan® Universal Master Mix.

2.2. Real-Time PCR ABI 7500 Real-Time PCR system (see Note 1).
Instrumentation
10 Genotyping of Single Nucleotide Polymorphisms by 5¢ Nuclease Allelic Discrimination 175

2.3. Design of Probes Currently, over 160,000 inventoried, ready-to-use SNP assays and
and Primers over 4.5 million predesigned, made-to-order SNPs assays are avail-
able from ABI (https://products.appliedbiosystems.com). The par-
ticular SNP of interest can be searched by its rs number (http://
www.ncbi.nlm.nih.gov/SNP/) from the ABI Web site. ABI 40× or
80× TaqMan® SNP genotyping assays can be ordered for different
sample sizes: small scale for 1,500 reactions, medium for 5,000 reac-
tions, and large for 12,000 reactions calculated from 5 mL reaction/
sample. The ready-to-use assay mixture is preloaded with (1) the
forward and reverse primers for amplification of the polymorphic
sequence, and (2) the two allele-specific TaqMan® probes descriptive
of the SNP of interest, one labeled with VIC® dye and the other with
6FAM™ dye. Every assay comes with an information file containing
details on the chromosomal location of the SNP, allele frequency
(only for validated assays) and the nucleotide sequence to which the
VIC/FAM reporter dye is linked. Upon delivery, all assays should be
protected from light by wrapping the tube in foil. The 40× and 80×
assays should be diluted with 1× TE buffer (10 mM Tris–HCl, 1 mM
EDTA, pH 8.0) to 20× working mixture for best stability. This assay
mixture should be aliquoted in batches to avoid freeze–thaw cycling.
Assay should be stored at −15 to −25°C.
For SNPs which do not have ready-made assays available, there
are several approaches for the user to design probe and primer sets:
(1) use ABI Primer Express Software (see below); (2) use ABI
Custom TaqMan® design service; and (3) search of the literature or
available databases, such as SNP500Cancer (http://variantgps.nci.
nih.gov/cgfseq/pages/snp500.do) for TaqMan® genotyping reagents
for the SNP of interest.
Should the user require probes and primers to be designed, the
ABI Primer Express software v2.0 provides assay design guidelines
developed specifically for quantification assays. The general guide-
lines are as follows:
1. Import the DNA sequence containing the SNP of interest
directly from NCBI dbSNP database (http://www.ncbi.nlm.
nih.gov/SNP/).
2. Design forward and reverse amplification primers each with
optimal lengths of 20 nucleotides and with matching Tm
(59°C optimal) within 1°C of each other. Strive for an ampli-
con length of 75–150 bp. Avoid a guanine residue at the 5¢
end of the probe because a G residue adjacent to reporter dye
will quench the reporter fluorescent even after cleavage. Target
a Tm of 65–67°C for each probe and a minimum length of 13
nucleotides. Avoid runs of an identical nucleotide in the probe
sequence (especially runs of four or more consecutive Gs
176 M. Malkki and E.W. Petersdorf

should be avoided). Position the SNP site in the central third

of the probe; if this is not possible, then shift the position of
SNP site towards the 3¢ end of the probe but do not place it in
the last two 3¢ nucleotides.

3. Methods

3.1. Preparation Prepare 5–25 ng/mL of DNA in TE buffer (10 mM Tris pH 8.0,
of Genomic DNA 0.1 mM EDTA) or in nuclease-free water. All samples should
preferably have the same DNA concentration for genotyping to
avoid a wide range of fluorescence activity within a genotype group
(see Notes 2 and 3).

3.2. Preparation Dissolve self-designed MGB and TAMRA quenched probes in TE

of Probes and SNP (10 mM Tris–HCl, 1 mM EDTA, pH8.0) or nuclease-free water at
Assay Mixture 100 mM concentration and the nonlabeled reverse and forward
primers in TE (10 mM Tris–HCl, 1 mM EDTA, pH 8.0) at
1,000 mM concentration. To make 20× SNP genotyping assay
cocktail containing MGB quencher, use 4 mM for each of the VIC
and FAM probe and 18 mM for each of the nonlabeled primer. To
make 20× SNP genotyping assay containing TAMRA quencher,
use 2 mM for each of the VIC and FAM probe and 14 mM for each
of the nonlabeled primer. For example, to make 500 mL of 20×
SNP assay mix with MGB quencher use 20 mL each of the VIC and
FAM probes, 9 mL each of the nonlabeled primers and 442 mL
nuclease-free water.
SNP assays purchased from ABI are ready to use.

3.3. TaqMan® This can be purchased from ABI. To avoid PCR contamination
Universal PCR from previously amplified material, “carry-over,” inclusion of
Master Mix AmpErase® uracil-N-glycosylase (UNG) in the buffer is recom-
mended (see Note 4). A single lot of master mix should be used in
the same experiment, since lot-to-lot variability can contribute to
variation in fluorescence values. The PCR master mix includes a
passive reference dye ROX™, which is used to normalize fluorescent
signals in individuals wells caused by small differences in PCR reac-
tion mixture volume (see Note 5).

3.4. PCR Mixture For one well, a basic cocktail volume is 3.75 mL containing
2.5 mL 2× Taqman universal PCR master mix, 0.25 mL 20× SNP
assay mix, and 1 mL of nuclease-free water. For full 96-well plate
this cocktail is made as a 105× mixture to allow for dead volume
and pipetting errors as follows: 262.5 mL 2× TaqMan® universal
PCR master mix, 26.25 mL 20× SNP assay mix, and 105 mL of
10 Genotyping of Single Nucleotide Polymorphisms by 5¢ Nuclease Allelic Discrimination 177

nuclease-free water. This cocktail can be scaled up for as many

plates as needed.

3.5. Amplification 1. Clean up the working surface with ethanol and water.
of Target Sequence 2. Label the 96-well optical plate with assay name and plate
identifier and put the plate into a PCR splash-free support
base (see Note 6).
3. Prepare the PCR mixture for as many plates as needed.
4. Pipet 1.3 mL of individual DNAs into the bottom of wells,
changing the tip between every DNA sample. Leave two wells
in each plate devoid of template DNA as negative controls; for
these wells, add 1.3 mL of water in place of DNA.
5. Add 3.75 mL of PCR master mix onto wall of each well. Change
tips if necessary to prevent carry over from well to well.
6. Cover the plate with adhesive optical cover and centrifuge the
plates at 1,500 rpm for 1 min to concentrate reagents at the
bottom of the wells, and to remove any air bubbles.
7. Place the plate/s in ABI 9700 PCR thermal cycler or 7500
Real-time PCR system (see Note 1). If a real-time machine is
used for the amplification step, please follow the manufacturer’s
instructions for plate setups and use of the machine. For MGB
quencher probes, program the following cycling conditions:
2 min 50°C (this step only if UNG containing buffer is used),
10 min 92°C followed by 45 cycles of 30 s 92°C and 1 min 30 s
60°C. For TAMRA quencher probes, recommended conditions
are: 2 min 50°C (perform this step only if UNG containing
buffer is used), 10 min 95°C followed by 45 cycles of 30 s 95°C
and 1 min 30 s 60°C. After cycling, keep plates at 4°C.
8. Read the plates in 7500 Real-Time PCR system using allelic
discrimination with automated allele calling settings for the
SDS 2.1 software (please refer to manufacturer’s instructions)
(see Note 7).

4. Notes

1. When the Taqman assay for a particular SNP is performed for

the first time, it is recommended that a real-time PCR system
(such as the ABI 7500 Real-time PCR system) is used to opti-
mize the assay. After optimization, a regular PCR thermocy-
cler, such as the ABI 9700 PCR system, can be used for the
amplification step and the real-time PCR machine for the
post-PCR data analysis.
178 M. Malkki and E.W. Petersdorf

Fig. 1. Example of “trailing” genotype clusters AA (circle), AB (triangle), and BB (diamond ). Trailing describes the variation
of fluorescence signals among a given genotype that has the appearance of being spread across an imaginary line that
extends from the no template control (square). This phenomenon is often caused by variability in concentration of individual
genomic DNAs.

2. If specific controls for unique SNP genotypes are desired, the

user has several resources for verifying whether the SNP of interest
has been characterized for a reference cell from the International
HapMap Project (http://hapmap.ncbi.nlm.nih.gov/) or the
1000 Human Genomes Project (www.1000genomes.org/). The
user may determine whether the cells of interest can be purchased
from Coriell Institute for Medical Research (http://ccr.coriell.
org/sections/Collections/NHGRI/?SsId=11).
3. The most common cause for trailing of genotype clusters, as
illustrated in Fig. 1, is variation in the DNA concentration of
the test samples. Trailing may be minimized by accurate mea-
surement of DNA concentration using the picogreen method
(following the manufacturer’s instructions or by using a stan-
dard spectrometer) and use of concentrations of 5–25 ng/mL.
Trailing of genotype clustering can also be caused by pipetting
error/s, use of expired reagents, the presence of PCR inhibitor
in DNA or degraded DNA.
4. TaqMan® Universal PCR Master Mix can be purchased with or
without AmpErase® UNG. AmpErase® UNG containing
Taqman PCR master mix is recommended for high-throughput
PCR applications. AmpErase® UNG can prevent the
reamplification of carryover PCR products by removing any
uracil incorporated to single- of double-stranded DNA.
5. The fluorescence signal from the passive reference dye ROX™
should be consistent across all wells and, therefore, it serves as
an internal fluorescence signal reference to which the reporter
dye (FAM and VIC) signal is normalized during data analysis.
10 Genotyping of Single Nucleotide Polymorphisms by 5¢ Nuclease Allelic Discrimination 179

Fig. 2. Example of a good separation of genotype clusters AA (circle), AB (triangle), and BB (diamond ). The no template
controls (square) are distant from any clusters. Samples within a given cluster are close to the center of the cluster and the
separation of the three clusters is very clear. Contrast this figure with that shown in Fig. 1.

Signal normalization is necessary to correct for well-to-well

non-PCR-related fluorescence signal fluctuations caused by
minor differences in concentrations or volume from well-to-well.
In the data analysis, the normalized reporter (Rn) is the ratio
of the fluorescence intensity of the reporter dye (FAM or
VIC) signal to the fluorescence intensity of the passive refer-
ence dye signal.
6. Keeping the plate off the work surface prevents dirt getting
onto outside of wells. Dirt can get into the real-time machine
plate block and cause carryover fluorescence for all plates.
Routine background fluorescence check of real-time machine
plate block is recommended (please refer to manufacturer’s
instructions).
7. The vast majority of TaqMan® assays perform well using stan-
dard assay and PCR cycling conditions (Fig. 2). However,
should modifications be required, it is recommended that for
every new assay, the first assay is run on a real-time PCR
machine to observe PCR amplification plots (please refer to
manufacturer’s instructions). An Rn (fluorescence of the
reporter dye divided by the fluorescence of a passive reference
dye ROX™) vs. PCR amplification cycle number plot and delta
Rn (Rn minus baseline) vs. PCR amplification cycle number
plot are useful to identify and examine irregularities in
amplification; a Ct (threshold cycle) vs. individual well plot is
useful to locate outliers in detector data sets and to identify
dirty wells in the instrument block.
180 M. Malkki and E.W. Petersdorf

Fig. 3. Example of SNPs assay requiring modification of PCR annealing temperature from 60 to 62°C. (a) At 60°C
annealing, AB (triangle) and BB (diamond ) clusters are not well separated, (b) the same assay with 62°C annealing
temperature showing better separation of AB (triangle) and BB (diamond ) clusters.

If assay clustering for AA, AB, and BB genotypes can be

discriminated as shown in Fig. 2, the rest of the amplifications for
that SNP can be performed using a standard PCR machine fol-
lowed by real-time PCR machine analysis for post-PCR end-point
allelic discrimination plate reads. After amplification, the assay
plates can be covered in foil and left at 4°C for few days before
doing the end-point reads.
When the genotype clusters are not well separated (Figs. 3 and 4),
adjustment of the annealing PCR temperature and/or reducing
the number of PCR cycles may help. Start by increasing annealing
temperature 1°C at a time from 60 to 61°C and then to 62º as
needed (Fig. 3a, b) or decrease it by 1°C from 60 to 59°C (Fig. 4a, b).
Although 5 mL reaction volumes are economical, some assays will
require 10–25 mL for optimal results.
Failure to achieve clean separation of genotype clusters may
indicate the presence of novel polymorphism in the probe sequence
and requires sequencing of the target region.
10 Genotyping of Single Nucleotide Polymorphisms by 5¢ Nuclease Allelic Discrimination 181

Fig. 4. Example of SNPs assay requiring modification of PCR annealing temperature from 60 to 59ºC. (a) At 60°C annealing,
AB (triangle) and AA (circle) clusters are not well separated, (b) the same assay with 59°C annealing temperature showing
better separation of AB (triangle) and AA (circle) clusters.

References

1. Kruglyak L (1997) The use of a genetic map of 7. International HapMap 3 Consortium et al

biallelic markers in linkage studies. Nat Genet (2010) Integrating common and rare genetic
17:21–24 variation in diverse human populations. Nature
2. Wang DG et al (1998) Large-scale identification, 467:52–58
mapping, and genotyping of single-nucleotide 8. 1000 Genomes Project Consortium et al
polymorphisms in the human genome. Science (2010) A map of human genome variation from
280:1077–1082 population-scale sequencing. Nature 467:
3. Kruglyak L (1999) Prospects for whole-genome 1061–1073
linkage disequilibrium mapping of common 9. Johnson GC et al (2001) Haplotype tagging
disease genes. Nat Genet 22:139–144 for the identification of common disease genes.
4. The International SNP Map Working Group Nat Genet 29:233–237
(2001) A map of human genome sequence 10. Carlson CS et al (2004) Selecting a maxi-
variation containing 1.42 million single nucle- mally informative set of single-nucleotide
otide polymorphisms. Nature 409:928–933 polymorphisms for association analyses using
5. Kruglyak L, Nickerson DA (2001) Variation is linkage disequilibrium. Am J Hum Genet 74:
the spice of life. Nat Genet 27:234–236 106–120
6. Sherry ST et al (2001) dbSNP: the NCBI data- 11. Johnson AD, O’Donnell CJ (2009) An open
base of genetic variation. Nucleic Acids Res access database of genome-wide association
29:308–311 studies. BMC Med Genet 10:6–23
182 M. Malkki and E.W. Petersdorf

12. The MHC Sequencing Consortium (1999) and disease associations: 2004. Tissue Antigens
Complete sequence and gene map of a human 64:631–649
major histocompatibility complex. Nature 22. Saiki RK et al (1985) Enzymatic amplification
401:921–923 of beta-globin genomic sequences and restric-
13. Walsh EC et al (2001) An integrated haplotype tion site analysis for diagnosis of sickle cell
map of the human major histocompatibility anemia. Science 230:1350–1354
complex. Am J Hum Genet 73:580–590 23. Mullis KB, Faloona FA (1987) Specific syn-
14. Horton R et al (2004) Gene map of the extended thesis of DNA in vitro via a polymerase-cata-
human MHC. Nat Rev Genet 5:889–899 lyzed chain reaction. Methods Enzymol
15. Miretti MM et al (2005) A high-resolution 155:335–350
linkage-disequilibrium map of the human major 24. Higuchi R et al (1993) Kinetic PCR analysis:
histocompatibility complex and first generation real-time monitoring of DNA amplification
of tag single-nucleotide polymorphisms. Am J reactions. Biotechnology (NY) 11:
Hum Genet 76:634–646 1026–1030
16. de Bakker PI et al (2006) A high-resolution 25. Morrison TB, Weis JJ, Wittwer CT (1998)
HLA and SNP haplotype map for disease asso- Quantification of low-copy transcripts by con-
ciation studies in the extended human MHC. tinuous SYBR Green I monitoring during
Nat Genet 38:1166–1172 amplification. Biotechniques 24:954–958;
17. Cree BA et al (2010) A major histocompatibility 960; 962
Class I locus contributes to multiple sclerosis 26. Piatek AS et al (1998) Molecular beacon
susceptibility independently from HLA-DRB1* sequence analysis for detecting drug resistance
15:01. PLoS One 5:e11296 in Mycobacterium tuberculosis. Nat Biotechnol
18. Baschal EE et al (2009) Defining multiple 16:359–363
common “completely” conserved major histo- 27. Marras SA, Kramer FR, Tyagi S (1999)
compatibility complex SNP haplotypes. Clin Multiplex detection of single-nucleotide varia-
Immunol 132:203–214 tions using molecular beacons. Genet Anal
19. International MHC and Autoimmunity 14:151–156
Genetics Network et al (2009) Mapping of 28. Mhlanga MM, Malmberg L (2001) Using
multiple susceptibility variants within the MHC molecular beacons to detect single-nucleotide
region for 7 immune-mediated diseases. Proc polymorphisms with real-time PCR. Methods
Natl Acad Sci USA 106(44):18680–18685 25:463–471
20. Valdes AM, Thomson G, Type 1 Diabetes 29. Holland PM et al (1991) Detection of specific
Genetics Consortium (2009) Several loci in the polymerase chain reaction product by utilizing
HLA class III region are associated with T1D the 5¢—3¢ exonuclease activity of Thermus
risk after adjusting for DRB1-DQB1. Diabetes aquaticus DNA polymerase. Proc Natl Acad Sci
Obes Metab 11(suppl 1):46–52 USA 88:7276–7280
21. Shiina T, Inoko H, Kulski JK (2004) An update 30. Heid CA et al (1996) Real time quantitative
of the HLA genomic region, locus information PCR. Genome Res 6:986–994
 Chapter 11

High Resolution MICA Genotyping by Sequence-Based

Typing (SBT)
Yizhou Zou and Peter Stastny

Abstract
We have developed a MICA typing method based on polymerase chain reaction (PCR) sequence-based
typing and a computer program that determines the polymorphisms and distinguishes the GCT repeats in
exon 5. One PCR amplification was performed to obtain templates of 2.2 kb, including exons 2, 3, 4, and
5 of MICA to be sequenced with two forward and two reverse primers. Overlay of nucleotide sequencing
signals resulting from presence of different GCT repeats in exon 5 from two different MICA alleles can be
identified by a computer program that analyses the combined signal string containing the 35 bases.

Key words: MICA typing, SBT, External domains, Transmembrane region, Computer analysis

1. Introduction

The major histocompatibility complex class I chain-related genes

A (MICA) are located 46.4 kb centromeric to HLA-B and have been
shown to be quite polymorphic, with 65 alleles described so far. It
appears that the expression of the MICA gene products is limited
mainly to certain epithelial cells, endothelial cells, fibroblasts, and
tumor cells (1). MICA genes, although located in the MHC class
I region, are structurally quite different from the HLA class I genes
and have been reported to be associated with some diseases (2) and
with the immune response to transplants (3). Recent studies have
shown that the MICA gene products may play a role in the activa-
tion of natural killer (NK) cells and other cells that express the
NKG2D receptor (4). Because the gut is also one of the major sites
of graft-vs.-host disease following bone marrow transplantation,
it is of interest that the MIC genes play a role in this process (5).
We recently found that organ transplant recipients make antibodies

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_11, © Springer Science+Business Media New York 2012

183
184 Y. Zou and P. Stastny

against specific MICA alleles (3) and antibodies against MICA

antigens were associated with kidney transplant rejection (6). Some
DNA-based typing techniques have previously been described for
MICA genes, including sequence-based typing (SBT) (7) and anal-
yses of single strand conformational polymorphisms (SSCP) (8),
sequence-specific oligonucleotide probes (SSOP), and sequence-
specific primers (SSP) (9). The transmembrane (TM) region of
MICA encodes the repeat polymorphism (GCT/Ala), and seven
types of repeats have been described as A4, A5, A5.1, A6, A7, A9,
and A10. Analysis of the number of repeats has usually been based
on the size of the TM fragments determined by a fragment analysis
software program following fluorescent PCR amplification (10).
In view of the large number of MICA alleles, the typing of the TM
region has become more important for resolving ambiguities in
MICA allele typing. We have developed a method for analyzing
MICA polymorphisms in exons 2–4 and the TM types in exon 5
from the PCR products obtained after generic MICA amplification
of exons 2–5 (11). This typing approach uses computerized assign-
ment of polymorphic sites and resolves also the short tandem
repeats within the transmembrane region. This method can resolve
all of the MICA alleles and performs MICA allele typing without
any additional steps.

2. Materials

2.1. Specimen 1. 5-mL genomic DNA (20 ng/mL) in molecular biology-grade

water (see Note 1).

2.2. Reagents 1. MICA locus-specific PCR primers (Table 1).

2. High Fidelity Taq (4 U/mL) and 10× PCR buffer with
MgCl2.
3. dNTP mix (25 pMol/mL).
4. BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied
Biosystems (ABI), Forest, CA, USA).
5. Four sequencing primers: 1F, 2R, 3F, and 4R (Table 1).
6. NaOAc/EDTA buffer.
7. Molecular biology-grade water.
8. Ethanol 99% anhydrous, molecular biology grade.
9. POP-6 (Performance Optimized Polymer) from Applied
Biosystems.
10. Hi-Di Formamide from Applied Biosystems.
 11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) 185

Table 1
Sequences of primers

Use for Name Sequences (5¢–3¢)

PCR MICA6823_F CGTTCTTGTCCCTTTGCCCGTGTGC

MICA9023_R GATGCTGCCCCATTCCCTTCCCAA
Sequencing 1F ATTTCCTGCCCCAGGAAGGTTGG
2R CAACTCTAGCAGAATTGGAG
3F AAGAGAAACAGCCCTGTTCCTCTCC
4R GATGCTGCCCCATTCCCTTCCCAA

2.3. Supplies 1. Thin-wall 96-well PCR plates (ISC BioExpress, Kaysville, UT,
Cat. T3031-21).
2. Thermal seal sealing films (ISC BioExpress, Cat. T2417-5).
3. Quanti-Marker 100 bp (GeneMate Inc., Kaysville, UT, USA.
Cat. QM4-104K).
4. MicroAmp Optical Plate (Applied Biosystems, Part No. N801-
0560).
5. Running Buffer, 10× (Applied Biosystems, Part No. 402824).
6. Metal Plate (Applied Biosystems, Cat. MACL09645).
7. MultiScreen®-HV (Millipore, Billerica, MA, USA. Cat.
MAHVN4550).
8. Pipette tips for 1–20, 21–200, and 100–1,000 mL.
9. PCR tubes (0.2 thin walled) Applied Biosystems, #N801-
0838.
10. Genetic analyzer plate Septa 96-wells Applied Biosystems,
#4315933.
11. Genetic analyzer plate Retainer Applied Biosystems,
#4317241.
12. Capillary array 50 cm long, Applied Biosystems, #4315930.

2.4. Equipment 1. Vertical laminar flow hood.

2. Thermocycler with heated lid and 96-well format.
3. Pipettes, volumes 1–20, 20–200, 100–1,000 mL.
4. Agarose gel apparatus.
5. Gel documentation camera.
186 Y. Zou and P. Stastny

6. Electrophoresis power supply.

7. Variable speed vortex.
8. Table top centrifuge (96-well tray holder).
9. ABI 3130 XL Analyzer (16 capillary).
10. Computer and Printer.

3. Methods

One set of amplifications was performed for each sample with

generic primers (MICA6823-F and MICA9023-R) (see Table 1) to
amplify a MICA gene fragment of approximately 2,200 bp includ-
ing exons 2, 3, 4, and 5. Four sequencing primers (see Table 1,
Fig. 1) generate four sequencing fragments for allele-level typing
analysis.

3.1. DNA Preparation 1. DNA can be isolated from acid-citrate-dextrose (ACD) anti-
coagulated whole blood, culture cell lines, or buccal epithelial
cells. DNA isolation methods vary considerably depending on
the starting material. The goal is to prepare high-molecular-
weight DNA of sufficient purity to allow HLA and MICA
allele-level typing (see Note 1). Preparations of high-quality
DNA are critical for amplification of MICA alleles since the
length of the amplified fragment is in the range of 2,200 bp.
Therefore, a protocol that produces a high-quality DNA prep-
aration should be selected. DNA samples should be of high
quality as determined by 260/280 nm absorbance ratios (1.7–
2.0). Prepare DNA samples with concentration at 20 ng/mL in
water. For example, to make a 100 mL DNA sample at 20 ng/mL
from a 100 ng/mL DNA sample, add 20 mL of DNA
(100 ng/mL) to 80 mL of water.

Fig. 1. Strategy for MICA allele-level typing by sequence-based typing (SBT). DNA frag-
ments (2.2 kb) containing the polymorphic sites in exons 2, 3, and 4 and the STR in exon
5 are amplified with the generic MICA-specific primers MICA6823F and MICA9023R. The
PCR primer and sequencing locations and their orientation are shown by the arrows.
 11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) 187

3.2. Amplification of 1. Primers for MICA allele amplification:

MICA Alleles by PCR Forward primer MICA6823-F and reverse primer MICA9023-R
are used to amplify a 2,200 bp fragment including exon 2, 3, 4,
and 5 of the MICA gene.
2. Prepare the samples for PCR.
Determine how many DNA samples are to be tested and add
positive and negative controls. Calculate as follows:

Water 8.8 mL (N* + 2)

10× Buffer 2.0 mL (N + 2)
2.5 mM dNTP 2.0 mL (N + 2)
10 pmol/mL primer 6823_F 1.0 mL (N + 2)
10 pmol/mL primer 9023_R 1.0 mL (N + 2)
4 U/mL Taq Polymerase 0.2 mL (N + 2)
20 ng/mL DNA 5.0 mL/each
Total 20.0 mL
N* = number of samples + 1 positive control + 1 negative control

3. First add 5 mL of DNA samples (20 ng/mL) to PCR tubes

separately, then vortex master mix well, spin down briefly, and
pipette 15 mL of the PCR master mix into each well.
4. Cap the tube strips and place in pre-warmed thermocycler.
5. Run the following profile (see Note 2).
MICA locus specific SBT-PCR profile:

1 Cycle 96˚C 2 min

10 Cycles 96˚C 30 s
69˚C 50 s
72˚C 90 s
20 Cycles 96˚C 30 s
67˚C 50 s
72˚C 90 s
1 Cycle 72˚C 10 min
4˚C Infinite

6. Check the PCR products.

(a) Remove samples from the thermocycler.
(b) Run the electrophoresis gel (2%) with 4 mL of the PCR
product mixed with 4 mL of 2× loading buffer. The gel
should be run at 130 V for approximately 20 min.
(c) To estimate the concentration of the PCR products, com-
pare the bands to the Quanti-Marker 100-bp map. The
188 Y. Zou and P. Stastny

Fig. 2. PCR products of MICA gene amplification. 4 mL of PCR products from DNA samples
were loaded to 2% gel for electrophoreses analysis. Bands with molecular weight at
2.2 kb are shown.

map has markers shown at various base pair lengths and

various concentrations in ng/band (see Fig. 2, Note 3).

3.3. Cleaning the MICA The PCR products have to be purified before they are used as
DNA Templates sequencing templates, because residual PCR primers and nucle-
otide triphosphates (dNTPs) can interfere with the SBT chemistry
resulting in lower data quality. An easy way to perform PCR
purification is by using exonuclease I and shrimp alkaline phos-
phatase (ExoSAP-IT) that will remove unincorporated primers and
dNTPs.
1. Take the vial of ExoSAP-IT from the freezer (see Note 4).
2. Add 3.6 mL of ExoSAP-IT to the well containing the PCR
products.
3. Seal the purification tray with micro amp full plate covers or
caps.
4. Centrifuge at 1,000 × g for 30 s.
5. Place in the thermal cycler and run the following profile.

37°C 30 min
80°C 15 min
4°C Infinite

6. Remove the treated PCR products from the thermocycler.

3.4. Preparation There are four sequencing primers (Table 1) for SBT.
for MICA Allele
1. Prepare the Sequencing Mix. The following table is for each
Sequencing
reaction. To the sequencing plate add 7.25 mL of water, 2 mL
of purified PCR product, and 2 mL of primers. Then mix with
BigDye and buffer.
 11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) 189

Sequencing reaction (15 mL):

Water 7.25 mL (adjustable if more PCR product is
needed, see Note 5)
PCR product (treated) 2.0 mL (Adjustable, see Note 6)
Sequencing Primer 2.0 mL (see Note 7)
5× buffer for BigDye 2.25 mL
BigDye 1.5 mL
Total 15.0 mL

2. Centrifuge the samples briefly for about 1 min at 800 ¥ g to

remove any bubbles.
3. Put the samples on the thermocycler with the following program.
This takes about 1½ h.
96°C 1 min
25 cycles 96°C 10 s
50°C 5 s
60°C 2 min

Keep at 4°C.
4. Following cycle sequencing, the sequencing reaction products
have to be purified to remove non-incorporated dye termina-
tors which would otherwise cause sequencing artifacts.

3.5. EDTA/NaOAc/ After Cycle sequencing is complete, remove plate from thermo-
Ethanol Precipitation cycler and place at 4 or −20°C until ready to precipitate. Do not let
of Cycle-Sequenced plate sit in thermocycler overnight (low sample volume leads to
Products easy evaporation). This protocol assumes a 15-mL cycle-sequencing
reaction.
1. To each well add 2.5 mL of NaOAc/EDTA (4 part of 3M
sodium acetate pH 5.5 plus 1 part of 0.5M EDTA pH 8.0).
Mix with pipette tips.
2. Add 25 mL 100% EtOH (from freezer) to each. Mix with
pipette tips. Replace sealing tape, seal very well and vortex for
1 min, and incubate at room temperature for 15 min.
3. Centrifuge at 2,500 × g for 30 min.
4. Invert plate onto paper towel and spin at 185 × g for 1 min to
dry.
5. Add 50 mL 80% EtOH to each well and mix with pipette tips.
6. Do not vortex and make EtOH fresh to ensure accurate con-
centration (see Note 8).
7. Centrifuge at 2,500 × g for 15 min.
190 Y. Zou and P. Stastny

8. Flick plate contents into sink, invert plate onto paper towel,
and spin at 185 × g for 1 min to dry. Start timer immediately.
9. Resuspend samples in 10 mL Hi-Diformamide. Cover plate
with sealing tape, vortex thoroughly (15 s), and denature in
thermocycler (96°C for 3 min).
10. Centrifuge for 30 s at 1,000 rpm and place it on the support
base, specific for each analyzer.
11. Place a full plate over the tray and follow with the 96-well plate
retainer.
12. Link tray map to tray ID and click on the green arrow to start
the run.
13. The run module for this procedure takes about 90 min for
each injection (see Note 9) on the 3130XL ABI Sequencer,
when using a 50 cm long capillary array (see Note 10).

3.6. Analysis The raw data results are analyzed by the ABI Sequence Analysis
of Allele-Level Typing program and saved in four electronic files (SampleID_MICA_1F.
by Computer Program ab1, SampleID_MICA_2R.ab1, SampleID_MICA_3F.ab1, and
SampleID_MICA_4R.ab1). The segments of MICA containing
exons 2, 3, 4, and 5 are assembled and the aligned sequences (see
Note 11), with MICA*001 as consensus, are analyzed with a pro-
gram for assignment of MICA alleles (see Note 12, Fig. 3). This
software was designed to analyze nucleotide substitutions at poly-
morphic sites using an updated MICA allele database (see Note 13).

Fig. 3. Alignment of MICA sequences by SBT. This aligned sequence is generated from four sequencing files (see Note
15). Sequences in exon 2, 3, 4, and 5 are assembled as in this alignment window by a computer program. Their intron
sequences were removed.
 11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) 191

Table 2
Alignment of the STR in the transmembrane region
of the MICA genes (5¢, 5¢–3¢)

A4 GCTGCTGCTGCTATTTTTGTTATTATTATTTTCTATGT
A5 GCTGCTGCTGCTGCTATTTTTGTTATTATTATTTTCTA
A5.1 GCTGCTGGCTGCTGCTATTTTTGTTATTATTATTTTCT
A6 GCTGCTGCTGCTGCTGCTATTTTTGTTATTATTATTTT
A7 GCTGCTGCTGCTGCTGCTGCTATTTTTGTTATTATTAT
A9 GCTGCTGCTGCTGCTGCTGCTGCTGCTATTTTTGTTAT
A10 GCTGCTGCTGCTGCTGCTGCTGCTGCTGCTATTTTTGT
MICA STR is located in the exon 5 with the different number of “GCT” repeat. MICA
allele named by WHO HLA nomenclature is based on the sequences of MICA gene
including STR type

Table 2 shows the alignment of the 35 nucleotides containing

different STR from (GCT)4 to (GCT)10 in the TM region. Overlay
of the sequencing signals may occur due to potentially different
number of (GCT) repeats in exon 5 from two different MICA
alleles in one sample, and they can be identified using the computer
program or manually (see Note 14, Fig. 4).

4. Notes

1. Measurement of DNA concentration must be accurate. The

integrity of the genomic DNA used is also important to obtain
a good yield in the PCR amplification. Always check the elec-
tropherogram of each sequencing report to make sure the
assignment of nucleotides is correct and there are two positive
signals for each assigned heterozygous position. ACD antico-
agulant is preferred. Heparin has been shown to inhibit some
PCR reactions. Blood samples older than 1 week may produce
poor yields and/or poor quality DNA unless they have been
stored frozen. One milliliter whole blood has about ten million
white cells and yields ~100 mg DNA suitable for ~200
amplification reactions. DNA isolation and PCR set up should
be restricted to the pre-PCR area to prevent DNA contamina-
tion. The pre-PCR manipulations should be handled in a lami-
nar flow hood. All the reagents should be freshly prepared
and/or autoclaved, if appropriate.
2. All Pre-PCR reagents are thawed to room temperature, vortexed,
spun, and kept cold until ready to be used. It is better to
192 Y. Zou and P. Stastny

Fig. 4. The combined signals from two different STR types in one sample. Five different sequencing results of the TM
polymorphisms are shown with the overlay signals matching the combined TM types from A4/A9, A5/A5.1, A5.1/A6, A5/
A6, or A6/A9. Computer typing for MICA TM is based on this algorithm. It is available at this web address: http://www4.
utsouthwestern.edu/stastny-lab/download_page.htm.
11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) 193

prepare fresh PCR mixture before testing. Preparation of the

positive and negative control of PCR is to monitor the quality
of MICA gene application.
3. If some samples failed in PCR, it is likely due to specimen
problems. Often adjusting the PCR components, for example,
adding more DNA, more Taq polymerase, and/or adding
more Mg2+, the PCR will overcome this difficulty. Allele drop-
out can be caused by mismatch between the primers and target
DNA that may present a novel allele differing at the binding
sites. Allele drop-out may also occur when a DNA sample with
poor quality is used, if this is the reason, the typing will be
repeated using a re-purified DNA isolate with optimal quality.
Check DNA on a 0.8% agarose gel for DNA sample quality.
4. ExoSAP-IT treated PCR products can stand a “small dilution”
sometimes necessary to repeat a given Cycle-Seq reaction
(usually 6–8 mL of molecular grade water).
5. Add water, cover the tray, vortex briefly, and spin for 1 min
before using it for a new cycle sequencing reaction.
6. Noisy baseline, inappropriate PCR product purification, or
poor quality reaction precipitation. Re-purify the sequencing
reaction products or purify the PCR products and repeat the
sequencing reactions.
7. MICA sequencing primers include four separate primers (1F,
2R, 3F, 4R). The fragments of the sequencing products should
be more than 600 bp.
8. Incorrect ethanol precipitation may result in excess salt remain-
ing in the sequence reaction. These salt ions will preferentially
inject onto the capillary over the sequence DNA molecules
resulting in weak sequence. Be sure to perform the ethanol
precipitation correctly and do a well-measured/correct con-
centration wash step afterwards.
9. Weak signal strength may occur in inappropriate injection time
or injection voltage because of variations between instruments.
Adjustments of the injection time and/or the injection voltage
may be needed to get a signal range from 10 to 2,000 relative
fluorescent units.
10. Plates waiting to be run may be stored at −20°C, but probably
should be re-denatured before running on the genetic analyzer.
Stored plates should be wrapped in foil to protect dyes from
light.
11. Novel sequence outside of exon 2–5 will not be detected in
this method. Some MICA alleles have the polymorphism in
exon 6, which is not part of the segment analyzed in this
approach (see Table 3). There are more than 20 groups of
194 Y. Zou and P. Stastny

Table 3
Alleles with identical extracellular domains (exons 2, 3, and 4)
and their associated transmembrane types (exon 5)

Groupsa Alleles TM types

1 MICA002:01 A9
MICA020 A10
MICA023 A5.1
MICA050 A7
2 MICA00801 A5.1
MICA027 A5
MICA048b A5
3 MICA009:01 A6
MICA049b A6
4 MICA007:01 A4
MICA026 A6
a
Groups of alleles with identical sequences in exons 2, 3, and 4. The differences are
either in TM type (group 1, 4) or
b
Different only in exon 6 (group3) or both (group 2)

known possible ambiguities with random combination of the

65 alleles. Additional testing may be needed to resolve such
ambiguities if required. Ambiguous allele assignments can
occur when two alleles are present and the composite sequence
is identical for more than one combination (cis/trans
ambiguities).
12. With dye terminator sequencing, the peak incorporation pat-
terns are not completely uniform. For homozygous positions,
this is not a problem, but for certain heterozygote positions,
one peak may be present at a much lower level than the other.
Consequently, the software may not correctly identify these
positions as heterozygous sequencing in both orientations.
They have to be carefully reviewed and corrected before sending
for computer program analysis.
13. Confirm that the allele database utilized is up to date. Web
address for the IMGT/HLA database: http://www.ebi.uk/
imgt/hla/align.html.
14. In most cases, 3F and 4R sequencing files will be found to have
multiple heterozygote positions either due to variation in
intron 5 containing 7 or 8T repeats or because the STR types
 11 High Resolution MICA Genotyping by Sequence-Based Typing (SBT) 195

are different in the two alleles. Those overlaid signals are

important and required for successful TM typing by this
method.
15. The following file format is suggested when the sequencing
files are setup.
DNASampleName_LocuName_SequencingPrimerName,
for example:
N0123456789_MICA_1F.ab1
N0123456789_MICA_2R.ab1
N0123456789_MICA_3F.ab1
N0123456789_MICA_4R.ab1
The file names are useful to organize the data by sample
and sequencing primers used.

References

1. Zwirner NW, Fernandez-Vina MA, Stastny P 7. Petersdorf EW, Shuler KB, Longton GM et al
et al (1998) MICA, a new polymorphic HLA- (1999) Population study of allelic diversity in
related antigen, is expressed mainly by kerati- the human MHC class I-related MIC-A gene.
nocytes, endothelial cells, and monocytes. Immunogenetics 49:605–612
Immunogenetics 47:139–148 8. Komatsu-Wakui M, Tokunaga K, Ishikawa Y
2. Stephens HA (2001) MICA and MICB genes: et al (1999) MIC-A polymorphism in Japanese
can the enigma of their polymorphism be and a MIC-A-MIC-B null haplotype.
resolved? Trends Immunol 22:378–385 Immunogenetics 49:620–628
3. Zwirner NW, Marcos CY, Mirbaha F et al 9. Ahmad T, Marshall SE, Mulcahy-Hawes Y et al
(2000) Identification of MICA as a new poly- (2002) High resolution MIC genotyping:
morphic alloantigen recognized by antibodies design and application to the investigation of
in sera of organ transplant recipients. Hum inflammatory bowel disease susceptibility.
Immunol 61:917–924 Tissue Antigens 60:164–179
4. Cerwenka A, Lanier LL (2003) NKG2D 10. Ota M, Katsuyama Y, Mizuki N et al (1997)
ligands: unconventional MHC class I-like mol- Trinucleotide repeat polymorphism within exon
ecules exploited by viruses and cancer. Tissue 5 of the MICA gene (MHC class I chain-related
Antigens 61:335–343 gene A): allele frequency data in the nine popu-
5. Parmar S, Del Lima M, Zou Y et al (2009) lation groups Japanese, Northern Han, Hui,
Donor-recipient mismatches in MHC class I Uygur, Kazakhstan, Iranian, Saudi Arabian,
chain-related gene A in unrelated donor trans- Greek and Italian. Tissue Antigens 49:448–454
plantation lead to increased incidence of acute 11. Zou Y, Han M, Wang Z et al (2006) MICA
graft-versus-host disease. Blood 114: allele-level typing by sequence-based typing
2884–2887 with computerized assignment of polymorphic
6. Zou Y, Stastny P, Süsal C et al (2007) Antibodies sites and short tandem repeats within the trans-
against MICA antigens and kidney-transplant membrane region. Hum Immunol 67:
rejection. N Engl J Med 357:1293–1300 145–151
 Chapter 12

Standard Methods for the Management

of Immunogenetic Data
Pierre-Antoine Gourraud, Jill A. Hollenbach, Thomas Barnetche,
Richard M. Single, and Steven J. Mack

Abstract
In this chapter, we outline some basic principles for the consistent management of immunogenetic data.
These include the preparation of a single master data file that can serve as the basis for all subsequent analyses,
a focus on the quality and homogeneity of the data to be analyzed, the documentation of the coding
systems used to represent the data, and the application of nomenclature standards specific for each immu-
nogenetic system being evaluated. The data management principles discussed here are intended to provide
a foundation for the data analysis methods detailed in Chaps. 13 and 14. The relationship between the data
management and analysis methods covered in these three chapters is illustrated in Fig. 3.
The application of these data management principles is a first step toward consistent and reproducible
data analyses. While it may take extra time and effort to apply them, we feel that it is better to take this
approach than to assume that low data quality can be compensated for by large sample sizes.
In addition to their relevance for analytical reproducibility, it is important to consider these data man-
agement principles from an ethical perspective. The reliability of the data collected and generated as part
of a research study should be as important a component of the ethical review of a research application as
the security of those data. Finally, in addition to ensuring the integrity of the data from collection to pub-
lication, the application of these data management principles will provide a means to foster research integ-
rity and to improve the potential for collaborative data sharing.

Key words: Data management, Data standards, High polymorphism, HLA, Immunogenetics, KIR

1. Introduction

1.1. Avoiding the In recent years, large amounts of genetic data have become avail-
“Garbage-In Garbage able to the research community, increasing the potential for new
Out” Predicament findings in ways that have not previously been possible. The need
for standardized data management and statistical analysis increases
as more data become available in order to mitigate the variability

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_12, © Springer Science+Business Media New York 2012

197
198 P.-A. Gourraud et al.

between individual studies. The difficulty of dealing with heterogeneity

between studies is not really new, but it becomes more significant
as the amount of accessible data increases.
Sir Arthur Conan Doyle summarized the problem in 1890 in his
second novel, “The Sign of Four”:
while the individual man is an insoluble puzzle, in the aggregate he
becomes a mathematical certainty. You can, for example, never fore-
tell what any one man will do, but you can say with precision what
an average number will be up to. Individuals vary, but percentages
remain constant. So says the statistician.
Throughout its history, the field of immunogenetics has pro-
vided a continuously changing perspective of individual immu-
nological characteristics. Different levels of information were
involved: the source materials level (sera, DNA, RNA, etc.), the
experimental level (with the switch from pure serological tech-
niques to a mix of serological and molecular techniques, and the
advent of parallel data acquisition instruments), and the analytical
level (with development of computer-based data storage and sta-
tistical modeling). While immunogenetic data production capacity
has clearly advanced significantly, it is not clear that the biostatisti-
cal capacity to analyze such quantities of data has developed in a
similar manner.
In addition, immunogenetics is distinctive in that it is relevant
to both very basic and applied fields of research. Transplantation is
the seminal example of gene-based personalized medicine; basic
knowledge in molecular anthropology, the history and genetic
structure of populations, the genetic components of disease, foren-
sic medicine, and genome evolution all benefit from the progress in
immunogenetics. With rivers of data now flowing from high-
throughput typing systems, a rigorous statistical approach is needed
more than ever to navigate this new immunogenetic sea. In this
chapter, we propose some foundational guidelines for the analysis
of immunogenetic datasets, focusing on ensuring quality and
homogeneity of data in order to escape from the “Garbage-in
Garbage-out” predicament. Supplementary materials for this chap-
ter can be found online at http://www.immunogenomics.org/
methods.html.

2. Data Organization
and Storage
2.1. Organization The “master data file” is the foundation of a quality data analysis.
of a Master Data File Ideally, a master data file will contain all the data needed for the
analysis in a single electronic document, which can be validated,
transferred, and exported. While individual software tools may
require data to be organized in different ways (i.e., different data
 12 Standard Methods for the Management of Immunogenetic Data 199

Fig. 1. The benefits of a single master data file. A master data file integrates all data pertinent to a given study in a standard
fashion. Genotype and phenotype data that must be compiled for a research project are shown on the right. These are all
incorporated into a single master data file, which is stored in a database, or as an MS Excel table.

input formats), the master data file should be the source of all data
analyzed regardless of the organization of those data. This first step
crystallizes all the thought that has been put in the study design
and the development of the hypothesis(ses) under investigation.
Study design is beyond the focus of this chapter but sample selec-
tion, phenotype assessment, and required sample size computations
clearly warrant close attention. As shown in Fig. 1, the master data
file gathers demographic, phenotypic, and genotypic information;
see supplementary Table S1 for a sample master data file.

2.2. Characteristics A typical master data file consists of (1) a series of rows dedicated
of a Master Data File to each analytical unit of the study and (2) a series of columns each
consisting of either a “primary” or “derived” variable.

2.2.1. Units of Analysis The master data file is generally structured using a single row dedi-
cated to each analytical unit of the study. The analytical unit is the
entity for which the required items of data have been collected and
compiled. In general, analytical units are individual subjects,
although in some cases the major entity being analyzed in a study
is not an individual; for example, a transplantation study may focus
on donor/recipient pairs, or an epidemiological study may focus on
cases and controls. When nuclear families are analyzed, it may also
be more appropriate to consider the nuclear family itself as the
analytical unit of the study rather than its constituent members.
In these cases, data on more than one individual may be included
on a single row of the master data file. Conversely, when the study
design involves repeated measures (over time or under different
200 P.-A. Gourraud et al.

circumstances), observations of a single subject are usually split

over several rows. This may be the case for example, when investi-
gating the consistency of a measure between different techniques
or between different operators.

2.2.2. Variables Each column (or field) in the master data file should be dedicated to
a variable. A variable is an observed or measured characteristic that
can have more than one value and to which a numerical measure
or a category from a classification can be assigned. It describes a
single item in the data set. Where possible, each column should be
numerically coded, and missing values must be given a conven-
tional code as well. It is often useful to give each variable a short
(but informative) name, which is stored in the first row of each
column (also known as the column header). These variables are
sometime called “raw” or “primary” variables, because they directly
encode information; “analysis” or “derived” variables encode
information that can be derived from the raw variables, and can
be included in the master data file as well. For example, body mass
index (BMI) can be calculated from weight and height variables,
but it may be useful to include a calculated BMI in the master data
file as well. The master data file may include other variable fields
that will contain variable data generated as part of the analyses.
In general, the master data file should be designed to exclude
redundancies between fields. At this early stage of the analysis, it is
also important to identify the type of the variables being included
in the master data file; the two most important variable types are
“quantitative” and “qualitative” variables, and the statistical
method that will be applied will depend on the type of variable
being analyzed.
1. Quantitative variables
Quantitative variables provide numerical measures of a
given characteristic being observed, and can be either “continu-
ous” or “discrete.” Continuous quantitative variables have an
infinite number of possible values that are not countable,
whereas discrete quantitative variables have either a finite
number of possible values or a countable number of possible
values. For example, in hematopoietic stem cell allogeneic
transplantation, the dose of CD34+ cells (e.g., 4.61 × 106 kg−1
of body weight) is a continuous quantitative variable. When
these values are rounded (e.g., to 5 × 106 kg−1 of body weight),
the variable may be considered as a discrete quantitative
variable. In general, mathematical and statistical operations
can be carried out on quantitative variables.
2. Qualitative variables
Qualitative (or categorical) variables allow for classification
of individuals based on a characteristic, and can be either
“nominal” or “ordinal.” Nominal qualitative variables define
unordered or non-hierarchical categories, whereas ordinal
 12 Standard Methods for the Management of Immunogenetic Data 201

qualitative variables can be used to order or rank the observations.

For example, leukemia subtypes can be defined with a nominal
qualitative variable; 1 = acute lymphoblastic leukemia (ALL);
2 = acute myelogenous leukemia (AML), 3 = chronic lympho-
cytic leukemia (CLL); 4 = chronic myelogenous leukemia
(CML), 5 = unclassified leukemia, 999 = data missing, whereas
stages of progression to type-1-diabetes (T1D) can be defined
with an ordinal qualitative variable; 0 = normal glucose regula-
tion, 1 = impaired fasting glucose, 2 = impaired glucose tolerance,
3 = diabetes mellitus, 999 = data missing. In general, qualitative
variables are descriptors and not numbers, and mathematical
operations are not appropriate. Instead, qualitative variables
are useful for stratifying and dissecting patterns derived from
the analysis of quantitative variables.

2.2.3. Data Dictionary In addition to the numerically encoded variables, a master data file
should also contain (or be associated with) a dictionary defining
each variable and its associated coded values. The minimal items
for this data dictionary are the short name of the variable, a detailed
description of its content, a specification of the meaning of each
associated numerical code, and the code used for missing values.
Any additional information or comment that can be useful for a
better characterization of the variable has to be mentioned. This
effort will ensure sustainability of the data.
Figure 2 shows an example of the data structure in a master
data file: numerically encoded data and data dictionary; see

Fig. 2. Characteristics of a Master Data File and associated Data Dictionary. A data dictionary provides context for the
numerically encoded observations recorded in the master data file. Each field (column) in the master data file corresponds
to a row in the data dictionary, where it is described in a standardized fashion.
202 P.-A. Gourraud et al.

supplementary Table S2 for a sample data dictionary for Table S1.

Figure 3 presents a “roadmap” to the landscape of statistical tech-
niques; it summarizes the analytical approach: it goes from raw data
storage, through analysis of missingness of the data, the use of
classical descriptive statistical techniques (both numerical and graph-
ical) to final statistical modeling fueling the interpretation of the
data. Figure 3 is also available online as supplementary Figure S1.

2.3. Storage of Data The best way to appropriately organize the data usually requires
building a database. Because many research groups do not have the
information technology (IT) resources to do so, a spreadsheet
application (e.g., Microsoft Excel or its equivalent) is often used
instead. Although very easy to use, accidental sorting, deletions,
and unintended changes to the data can occur with such spread-
sheet applications. For example, when HLA allele names are stored
in Excel, it is common for allele names that should be treated as
text (e.g., “0101” or “01:01”) to be treated as numbers or times
(e.g., 101 or 1:01). Such data rearrangements may be difficult to
detect and will certainly affect the results of the study. If spread-
sheet applications are used to manage allele name data, individual
datasets should be stored and transmitted as text-formatted “flat
files,” and measures should be taken to ensure the validity of allele
names prior to analysis.

3. Nomenclature
Management
The high degree of polymorphism observed for immunogenetic
genes and loci, coupled with the structural variation of immune
genes, has resulted in regularly updated allele name nomenclatures
that keep pace with the growing complexity of these systems, but
each nomenclature revision brings with it the potential for confu-
sion when data generated using different nomenclature versions
are analyzed together (the so-called immunogenetic tower of
Babel) (1). This is a common problem for HLA and KIR genes,
alleles, and haplotypes; MHC microsatellites, single nucleotide
polymorphism (SNP), copy number variants (CNVs), and small
insertion-deletion changes (indels). It is essential that the data in
an analysis share a common nomenclature, and that the specific
nomenclature version under which a dataset was generated (or
validated if the dataset is compiled from multiple sources) is
identified in the master data file.

3.1. HLA Nomenclature The HLA nomenclature for allele names has changed considerably
Variation over the last several decades. Each allele name contains a string of
numbers arranged in “domains” that correspond to polymorphisms
at the serological, protein, synonymous nucleotide, noncoding
 12
Standard Methods for the Management of Immunogenetic Data

Fig. 3. A roadmap to the landscape of statistical techniques. This roadmap is a guide summarizing potential approaches to data analysis. Starting with raw data in the center, analyses
203

proceed first through an assessment of missing data, and then the application of classical descriptive statistical techniques (both numerical and graphical), before a set of final
statistical tests, specific to the type of data and the hypothesis being tested, and which inform the interpretation of the data, are selected.
204 P.-A. Gourraud et al.

nucleotide, and expression levels. The nomenclature has been

expanded as new alleles, and new forms of polymorphism have
been identified, and the features of three major versions of this
nomenclature are outlined in Table 1.
With each improvement to the nomenclature, some allele
names have been deleted (2) or changed (e.g., B*1522 changed to
B*3543 in 2002 after DNA sequences outside of exons 2 and 3
became available). With the adoption of the current nomenclature
version, many allele names have changed in nonobvious ways (e.g.,
the DPB1*0502 allele changed to *104:01 in order to maintain
the sequential order of DPB1 protein sequences), and the locus
identifier for HLA-C locus alleles changed from Cw* to C* (3).
If the nomenclature version pertinent to a given dataset has not
been explicitly identified, it is important to be able to make an
educated guess about the likely version under which the data were
generated.

3.1.1. Antigen Recognition Rather than treating each allele as distinct, it is often useful to ana-
Sequence Nomenclature lyze alleles that share the peptide sequences that constitute the
antigen recognition sequence (ARS) or that share the nucleotide
sequences that encode the ARS as a combined category. The ARS is
alternatively referred to as the peptide-binding domain or peptide-
binding region (respectively abbreviated as PBD or PBR). The ARS
of class II HLA molecules is encoded by exon 2, and the class I
ARS is encoded by exons 2 and 3. While the version 2 HLA nomen-
clature was in effect, allele-codes identifying alleles that share the
same ARS encoding exon sequences were developed. For example,
the “A*010101g” and “A*02 G1” codes have been used to repre-
sent all HLA-A alleles that shared exon 2 and 3 nucleotide sequences
with the A*010101 allele (4, 5).
Version 3 of the HLA nomenclature includes official codes for
identifying alleles that share an ARS or ARS-encoding exon
sequence. Alleles that share an identical ARS are included in a P
group (e.g., A*01:01P includes all HLA-A alleles that share the
ARS with the A*01:01:01:01 alleles), and alleles that share identi-
cal exon 2 or exons 2 and 3 nucleotide sequences are included in a
G group (e.g., A*01:01:01G includes all HLA-A alleles that share
their exon 2 and exon 3 sequences with A*01:01:01:01) (3).

3.1.2. Nomenclature Several resources are available for the conversion of allele names
Conversion Resources between nomenclature versions. The ImMunoGeneTics (IMGT)/
HLA web site and database provides tables for the conversion of
version 2 allele names to version 3 names, and houses a web-based
conversion tool for individual allele names at http://www.ebi.
ac.uk/imgt/hla/convert_name.html (6).
Entire datasets can be converted from nomenclature version 2
to version 3 using either the Allele Name Translation Tool (ANTT)
or Update NomenCLature (UNCL), tools available from the
 12

Table 1
Features of the three nomenclature versions for classical HLA allele names

Nomenclature Domain size (number of characters)

Version Epoch Serologica Protein Synonymous Noncoding Expression variants Examples

1 1987–2002 2b 2b 1c 2d Nd, Le A*01011
2 2002–2010 2 2 2f 2 L, N, Ag, Cg, Sg A*01010101 A*01010102N
3 2010+ 2 2+g 2+g 2+g A, C, N, L, S, Qh A*01:01:01:01 A*01:01:01:02N
a
The concept of a serologic specificity does not apply to DPB1 alleles. The first two digits of allele names for unique DPB1 protein sequences have been assigned in numerically (14)
b
Implemented in 1987 (15)
c
Implemented in 1990 (16)
d
Implemented in 1995 (17)
e
Implemented in 1996 (18)
f
This domain was expanded to two digits in 2002 (2)
g
Implemented in 2002 (2)
h
These domains were explicitly defined as colon-delimited fields in 2010, and can accommodate any number of variants (3)
i
Implemented in 2010 (3)
Standard Methods for the Management of Immunogenetic Data
205
206 P.-A. Gourraud et al.

immunogenomics data analysis working group (IDAWG) at

http://immunogenomics.org/software.html (7). UNCL is a web-
based tool, and the ANTT is a locally run tool that can be custom-
ized to convert between other nomenclatures and name
conventions. Both tools accept data in the form of tab-delimited
text files, and generate tab-delimited text files of translated data.
The current version of the Helmberg Sequence COmpilation
and Rearrangement Evaluation (SCORE) virtual-DNA analysis
database software (8, 9) will convert between version 2 and 3
nomenclatures for allele name data in its database. SCORE will
also convert allele names to user-defined nomenclatures.

3.1.3. Allele Name Because the polymorphic domains that make up an HLA allele
Truncation name are arranged in a hierarchical fashion, it is possible to delete
domains from the right end of an allele name while retaining
important information for analysis. This process of deleting
domains is known as truncation (aka right-truncation), and it is
important to ensure that truncation has been carried out consis-
tently within a dataset. For example, the full-length A*01:01:01:01
allele name can be truncated to A*01:01:01, A*01:01, and A*01.
In some cases, truncation of allele names may be equivalent to
applying a P or G group code, but this is not always the case.
Depending on the research question being investigated, different
levels of truncation may be appropriate (e.g., it may be appropriate
to truncate allele names to the peptide level for a study of peptide
presentation, but not for a study of allelic diversity and evolution),
but all versions of a given allele name must be truncated to the
lowest common level for analysis to avoid spurious results.

3.2. Microsatellite In general, microsatellite alleles should be identified using both the
Nomenclature number of repeats and the fragment size (10). In order to compare
(14th Workshop) the results of several studies, the correspondence between the
repeat number and the various fragment lengths (depending on
primer pairs used) should be established. This correspondence list
must include specific details (e.g., in the form of UniSTS numbers)
of the primer pairs used to genotype each microsatellite (http://
www.ncbi.nlm.nih.gov/sites/entrez?db=unists). When multiple
synonymous names are in use for a given microsatellite allele, the
lowest numbered DS6 number should be used as a reference. For
example D6S273 should be used to refer to the following synony-
mous names—142XH6, AFM142xh6, GC378-D6S273. More
details can be found on the NCBI’s MHC database, “dbMHC”
(http://www.ncbi.nlm.nih.gov/projects/gv/mhc/).

3.3. Single Nucleotide As recently discussed (1), multiple distinct identifiers may exist
Polymorphism and for a given simple genetic marker (e.g., a SNP or indel), some-
Insertion/Deletion times leading to confusion in the comparison of markers across
Nomenclature studies. This is partly due to the fact that the unique reference
sequences applied to structurally variable and highly polymorphic
 12 Standard Methods for the Management of Immunogenetic Data 207

3.4. SNPs: RS# and regions like the KIR cluster and the MHC region are not appro-
HGVS Nomenclature priate for these markers. One commonly accepted solution to
this problem is to use the NCBI’s dbSNP reference SNP
(RefSNP) accession ID (rs number or rs#) (11). However, because
rs numbers are not necessarily stable identifiers across successive
dbSNP releases/builds, all references to an rs# should be accom-
panied by dbSNP build number. For example, rs16375 is not in
use anymore in build 130 of dbSNP; rs1704 should be used
instead.
Many other possibilities can be used as complementary indica-
tions which directly refer to the sequence changes (e.g.,
rs2306220:A>G). The other acceptable option is to use the Human
Genome Variation Society (HGVS) nomenclature (12, 13). For
rs16375/rs1704, an accessioned sequence from EBI, GenBank, or
DDBJ can be used as the reference sequence NT 007592.14:
20656832insATTTGTTCATGCCT, but many different acces-
sioned genomic or mRNA sequences can be used.

3.5. KIR Nomenclature Ideally, the goal in storing and analyzing KIR data is to have
some data representation for each chromosome at every locus.
The limitations of the typing technology and the variation in
KIR haplotype structure dictate that in many cases the second
chromosome will be typed as “unknown” (?). Nevertheless, it is
important to have data representation for the full genotype in
order to facilitate downstream analyses. Each KIR locus typed
can have at least one full genotypic record (more in the case of
genotypic ambiguity, which is likely for allelic typing results).
When a locus is absent from a given haplotype, this absence must
be coded in the data (i.e., as an “absent” allele), and should be
denoted as “0.”
Where allelic typing is available, a genotype with two alleles
represents the simplest case scenario, a heterozygote with the locus
present on both haplotypes. However, when we have only one
allele detected, e.g., KIR2DL2*002, there are two possible geno-
types: “002,002” or “002,0.” At present, most typing systems can-
not distinguish these. In these cases, the genotype should be treated
as ambiguous, i.e., 002,0/002.
For typing data that is strictly presence/absence, the data can
be treated as for a biallelic locus: allele “1” = “present” and allele
“0” = “absent.” If the locus is absent, we have the full genotype
0,0; however for locus present, we have an ambiguous genotype,
which may be either “1,0” or “1,1.” This can also be represented
as an ambiguous genotype : 1,0/1. Because this notation does not
consider order, a genotype coding of “0,1” is not used, but would
be treated the same as “1,0.”
Some KIR loci have particular additional details to consider
prior to analyzing data:
208 P.-A. Gourraud et al.

1. KIR2DL2/L3:
KIR2DL2 and KIR2DL3 were formerly treated as separate
loci. The alleles were not named in a series; therefore, it is
important to distinguish between KIR2DL2 and KIR2DL3
when the data is recorded. In the case of this locus, since all
typing systems are able to detect both KIR2DL2 and KIR2DL3
an investigator should always have a full genotype, i.e., no
missing data or ambiguity with presence/absence typing.
2. KIR2DL5:
KIR2DL5 may be either centromeric or telomeric in the
KIR cluster, and possibly on both ends. As such, an individual
may have four copies of KIR2DL5; at this time there is no way
to distinguish these definitively. Many typing systems will simply
type for the presence of KIR2DL5; if typed as “absent,” there is
confirmation of genotypes of 0,0 both centromerically and
telomerically. However, a typing of ‘present’ gives the genotype
1,0/1 on either or both sides of the KIR complex. Some typing
systems currently distinguish “KIR2DL5A” and “KIR2DL5B”;
all current data suggest that KIR2DL5A corresponds to the
telomeric position and that KIR2DL5B corresponds to the
centromeric position.
3. KIR2DS3/S5:
As with KIR2DL5, KIR2DS3/S5 may be either centro-
meric or telomeric in the KIR complex, and possibly on both
ends. As such, an individual may have four copies of the gene.
Previously, KIR2DS3 and KIR2DS5 were thought to be dif-
ferent loci, and all current typing systems can distinguish
between them. However, an individual who is typed, for exam-
ple, as positive for both KIR2DS3 and KIR2DS5 in a pres-
ence/absence typing system may have these two alleles on
either or both sides of the KIR cluster. Some data is emerging
that suggests that particular alleles of each of these may be
either centromeric or telomeric, but this is not yet confirmed.
Presently, the only way to ascertain definitively that an indi-
vidual has both centromeric and telomeric copies of either
KIR2DS3 or KIR2DS5 is the case where we have allelic typing
with more than two allele calls for the locus.
4. KIR3DL1/S1:
As above for KIR2DL2/L3, KIR3DL1 and KIR3DS1
were formerly treated as separate loci. However, it has been
recognized for some time that they are allotypes of the same
locus, and in this case alleles have been named in a series. As with
KIR2DL2/L3, all typing systems minimally detect KIR3DL1
and KIR3DS1, so there should be no missing data/ambiguity
in presence/absence typing systems at this locus. If a typing is
 12 Standard Methods for the Management of Immunogenetic Data 209

submitted as, e.g., “3DL1,” we know that the genotype is

“3DL1,3DL1.”

Acknowledgments

This work was supported by National Institutes of Health (NIH)

grants U01AI067068 (JAH, SJM) and U19 AI067152 (PAG)
awarded by the National Institute of Allergy and Infectious Diseases
(NIAID) and by NIH/NIAID contract AI40076 (RMS). The con-
tent is solely the responsibility of the authors and does not necessar-
ily represent the official views of the National Institute of Allergy
and Infectious Diseases or the National Institutes of Health.

Glossary

Genetic data 1. Allele: Any of the alternative forms (sets of forms) of DNA
sequence at locus. These variants may occur for genes and/or
genetic markers.
Example: B, HLA-DRB1*01:01:01, HLA-A*01, D6S1666
(184).
2. Diplotype: The pair of haplotypes within a given genotype.
The chromosomal phase between alleles is always known.
Example: HLA (A*01 B*08 DR*17, A*26 B*27 DR*17).
HLA (A*01 B*27 DR*17, A*26 B*08 DR*17).
When analyzing data for more than one locus, diplotypic data
must be distinguished from genotypic data. Because the alleles
at different loci in a given genotype can be combined to make
many possible haplotype pairs, a genotype must be considered
to correspond to multiple diplotypes. Unfortunately, the term
“genotype” is sometimes used to refer to a given pair of haplo-
types, especially when familial segregation has been studied.
3. Gene: The functional and physical unit of heredity. A gene
consists of a DNA segment with a specific sequence. It includes
information for the synthesis of mRNA molecules that direct
the synthesis of proteins.
Example: ABO Glycosyltransferase gene, HLA-DRB1 gene,
KIR-2DS3/S5 gene.
4. Genotype: The genetic makeup at one or more loci of an indi-
vidual. It refers to a set of alleles carried by an individual
(regardless of the expression of those alleles). The chromo-
somal phase (chromosomal identity of alleles at different loci)
between alleles may not be known.
210 P.-A. Gourraud et al.

Example: HLA-A (1, 2); HLA-B (8, 44); HLA-DRB1 (03, 04).
KIR: KIR (A, A).
Microsatellites: D6S273 (134, 136) D6S273
(*(GT)19, *(GT)20).
SNP: RS345336443 (G/G).
5. Haplotype: Set of alleles of contiguous loci. They are usually
co-transmitted on a parental chromosome.
Example: HLA-A*01-B*08-DRB1*03.
6. Locus: Literally, “place” in Latin, it is the specific usual physi-
cal location of genes, an individual genetic marker, or set of
genetic markers in a genome.
Example: ABO locus, HLA-DRB locus, KIR-2DS3/S5cen
locus, D6S1666 microsatellite locus.
7. Phenotype: The observable expression of alleles as a physical or
biochemical trait resulting from the interaction of the genome,
the environment, and the experimental settings. In disease
studies it may refer to the presence or a manifestation of the
disease under study. Disease phenotypes may be reflected in a
variety of ways as quantitative or qualitative variables.
This term may also refer to a set of alleles (expressed or not)
detected by a technique. In codominant or heterozygous situ-
ations, phenotypes are noted as pairs of data; each pair is specific
to a particular gene and locus.
Example: ABO system: [A]. HLA system: [HLA-A (1, 2);
HLA-B (8,44); HLA-DR (3, 4)].

Phenotypic 1. Admixture: The outcome of interbreeding between members

and demographic of different populations. An admixed population is generally
data derived from populations in different geographic regions.
2. Collection site: The location where the sample was collected.
This can be identified using latitude and longitude coordinates,
or by specifying the country or nation, and city/town/village,
or other locale where the collection took place.
3. Complexity: An ordinal variable that represents an estimate of
the degree of admixture and population sub-structure in each
population sample.
Example:
Complexity 1: a population sample collected from a single set-
tlement or group of closely related settlements.
Complexity 2: a population sample collected from a group of
separate but discrete settlements.
Complexity 3: a population sample collected in a metropolitan
area or across an entire nation.
Complexity 4: an admixed population.
4. Data management methods: The approaches used in storing
and processing the data in preparation for analysis. This can
12 Standard Methods for the Management of Immunogenetic Data 211

include the formats and programs used to store and edit the
data (e.g., a specific spreadsheet program or database system),
as well as any modifications that were made to the data between
the generation of the data resulting from the typing assay and
the inclusion of the data in the master data file. For example, if
ambiguities were resolved, the approach used to resolve them
should be documented in the data dictionary; if HLA allele
data were truncated to a common level, or “binned” into a
common sequence category (e.g., treating all alleles that
encode the same peptide-binding region as the same allele)
this should documented.
5. Ethnicity: A group of individuals (or populations) sharing a
common language, culture, or religion, and who are assumed
to share a common ancestry. Ethnicity should be distinguished
from geography (e.g., “North American” is not an ethnicity),
and though ethnicity is often associated with indigenous
nationality (e.g., “Irish,” “Chinese”) qualifiers are often neces-
sary to distinguish ethnicity from nationality (e.g., “Han
Chinese”).
6. Family: If individuals in the study belong to discrete familial
groups, a family ID is qualitative variable identifying member-
ship in a particular pedigree, as well as the relationship to the
index case (proband).
7. Geographic region: A specific continental or subcontinental
area comprised by multiple nations in which the population is
located, or from which the population was derived, if the popu-
lation is a migrant population. For example, European Americans
or European Australians would be assigned to the European
region, or to a specific subregion of Europe. Conversely North
America would only pertain to Native American/Amerindian/
Aleut/Eskimo populations. Populations derived from more
than one region (admixed populations) can be assigned to a
specific class for the type of admixture (depending on the
regions of origin) or included in a single class for all admixed
populations. The definitions of each region and admixed class
should be defined in the data dictionary.
8. Latitude and longitude: Geographic coordinates that specify
specific locations on the surface of the Earth. Latitude and
longitude values should be recorded in a decimal format,
with minutes and seconds indicated as factions of each
degree value. North latitudes and east longitudes should be
recorded with positive values, and south latitudes and west
longitudes should be recorded with negative values. For
example, 35° 20 min south latitude would be recorded as
−35.333, and 2° 30 min east longitude should be recorded
as 2.5 or +2.5.
212 P.-A. Gourraud et al.

10. Population: A group of individuals living in a specific geo-

graphic area. More specifically, a population is defined such
that all pairs of individual members have the opportunity to
mate, and are more likely to mate with each other than with
members of other populations. A population should be docu-
mented in the data dictionary in terms of the pertinent geo-
graphic area and the approximate number of included
individuals. Population sample: A unique descriptor for the
individuals from a given population that were included in the
study. If the study involves multiple sets of individuals (sam-
ples) from the same population, each set of individuals should
be given a unique name; usually it is sufficient to append the
number of individuals to the end of the population name (e.g.,
antarctica_87, antarctica_207, antarctica_597).
11. Population substructure: A barrier to the opportunity of mating
between all pairs of individuals in a population.
12. Proband: The individual under study, primarily used in family-
based disease association studies.
13. Status: The status of an individual as affected or unaffected
with respect to a disease phenotype, or belonging to a case or
a control group.
14. Typing assay: The laboratory method(s) and associated proto-
cols used to generate the data included in the analysis. Many of
them are described in this volume. Commonly used molecular
methods for HLA and KIR genotyping include sequence-
specific priming (SSP), sequence-specific oligo probe (SSO or
SSOP), sequence/sequencing-based typing (SBT), matrix-
assisted laser desorption/ionization time-of-flight (MALDI-
TOF), and reference strand conformation analysis (RSCA).
Serology has been used historically for HLA phenotype data
generation. When possible, a description of the assay identify-
ing the assay manufacturer and reagent version/lot employed
should be included in the data dictionary. Literature citations
or references to specific protocols should also be associated
with the methods used, especially if multiple distinct methods
have been employed in generating the data.

References
1. Gourraud PA, Feolo M (2010) The Babel Nomenclature for factors of the HLA system,
Tower revisited: SNPs—Indels—CNVs. 2002. Tissue Antigens 60:407–464
Confusion in naming sequence variant always 3. Marsh SG, Albert ED, Bodmer WF, Bontrop
rises from ashes Tissue Antigens 75:199–200 RE, Dupont B, Erlich HA, Fernández-Viña M,
2. Marsh SG, Albert ED, Bodmer WF, Bontrop Geraghty DE, Holdsworth R, Hurley CK, Lau
RE, Dupont B, Erlich HA, Geraghty DE, M, Lee KW, Mach B, Maiers M, Mayr WR,
Hansen JA, Mach B, Mayr WR, Parham P, Müller CR, Parham P, Petersdorf EW, Sasazuki
Petersdorf EW, Sasazuki T, Schreuder GM, T, Strominger JL, Svejgaard A, Terasaki PI,
Strominger JL, Svejgaard A, Terasaki PI (2002) Tiercy JM, Trowsdale J (2010) Nomenclature
 12 Standard Methods for the Management of Immunogenetic Data 213

for factors of the HLA system, 2010. Tissue 11. Sherry ST, Ward MH, Kholodov M, Baker J,
Antigens 75:291–455 Phan L, Smigielski EM, Sirotkin K (2001)
4. Cano P, Klitz W, Mack SJ, Maiers M, Marsh dbSNP: the NCBI database of genetic varia-
SG, Noreen H, Reed EF, Senitzer D, Setterholm tion. Nucleic Acids Res 29:308–311
M, Smith A, Fernández-Viña M (2007) 12. den Dunnen JT, Antonarakis SE (2000)
Common and well-documented HLA alleles: Mutation nomenclature extensions and sug-
report of the Ad-Hoc committee of the gestions to describe complex mutations: a dis-
American society for histocompatibility and cussion. Hum Mut 15:7–12
immunogenetics. Hum Immunol 68: 13. den Dunnen J (2010) Nomenclature for the
392–417 description of sequence variants. Human
5. Robinson J, Mistry K, Marsh SGE (2010) Exon Genome Variation Society. http://www.hgvs.
identity and ambiguous typing combinations. org/mutnomen/
Anthony Nolan Research Institute. http:// 14. Bodmer JG, Marsh SGE, Parham P, Erlich HA,
www.ebi.ac.uk/imgt/hla/pdf/ambiguity_ Albert E, Bodmer WF, Dupont B, Mach B,
v2280.pdf Mayr WR, Sasasuki T, Schreuder GMT,
6. Robinson J, Mistry K, McWilliam H, Lopez R, Strominger JL, Svejgaard A, Terasaki PI (1990)
Parham P, Marsh SG (2011) The IMGT/HLA Nomenclature for factors of the HLA system,
database. Nucleic Acids Res 39(Database 1989. Tissue Antigens 35(1):1990
Issue):D1171–D1176 15. Who Nomenclature Committee (1988)
7. Mack SJ, Hollenbach JA (2010) Allele Name Nomenclature for factors of the HLA system,
Translation Tool and Update NomenCLature: 1987. Tissue Antigens 32:177–187
software tools for the automated translation of 16. Bodmer JG, Marsh SG, Albert ED, Bodmer
HLA allele names between successive nomen- WF, Dupont B, Erlich HA, Mach B, Mayr WR,
clatures. Tissue Antigens 75:457–461 Parham P, Sasazuki T et al (1991) Nomenclature
8. Helmberg W, Lanzer G, Zahn R, Weinmayr B, for factors of the HLA system, 1990. Hum
Wagner T, Albert E (1998) Virtual DNA Immunol 31(3):186–194
analysis—a new tool for combination and stan- 17. Bodmer JG, Marsh SG, Albert ED, Bodmer
dardised evaluation of SSO, SSP and sequencing- WF, Bontrop RE, Charron D, Dupont B,
based typing results. Tissue Antigens 51: Erlich HA, Mach B, Mayr WR (1995)
587–592 Nomenclature for factors of the HLA system,
9. Helmberg W (2000) Storage and utilization of 1995. Tissue Antigens 46:1–18
HLA genomic data—new approaches to HLA 18. Bodmer JG, Marsh SG, Albert ED, Bodmer
typing. Rev Immunogenet 2:468–476 WF, Bontrop RE, Charron D, Dupont B,
10. Gourraud PA, Cambon-Thomsen A, Dauber Erlich HA, Fauchet R, Mach B, Mayr WR,
EM, Feolo M, Hansen J, Mickelson E, Single Parham P, Sasazuki T, Schreuder GM,
RM, Thomsen M, Mayr WR (2007) Strominger JL, Svejgaard A, Terasaki PI (1997)
Nomenclature for HLA microsatellites. Tissue Nomenclature for factors of the HLA system,
Antigens 69(Suppl 1):210–213 1996. Tissue Antigens 49:297–321
 Chapter 13

Analytical Methods for Immunogenetic Population Data

Steven J. Mack, Pierre-Antoine Gourraud, Richard M. Single,
Glenys Thomson, and Jill A. Hollenbach

Abstract
In this chapter, we describe analyses commonly applied to immunogenetic population data, along with
software tools that are currently available to perform those analyses. Where possible, we focus on tools that
have been developed specifically for the analysis of highly polymorphic immunogenetic data. These analytical
methods serve both as a means to examine the appropriateness of a dataset for testing a specific hypothesis,
as well as a means of testing hypotheses. Rather than treat this chapter as a protocol for analyzing any
population dataset, each researcher and analyst should first consider their data, the possible analyses,
and any available tools in light of the hypothesis being tested. The extent to which the data and analyses
are appropriate to each other should be determined before any analyses are performed.

Key words: Data analysis, Highly polymorphic, HLA, Immunogenetics, KIR, Population study

1. Introduction

The Analytical Engine has no pretensions whatever to originate anything.

It can do whatever we know how to order it to perform. It can follow
analysis, but it has no power of anticipating any analytical revelations or
truths. Its province is to assist us in making available what we are already
acquainted with. For, in so distributing and combining the truths and the
formulae of analysis, that they may become most easily and rapidly ame-
nable to the mechanical combinations of the engine, the relations and
the nature of many subjects in that science are necessarily thrown into
new lights, and more profoundly investigated (Ada Augusta, Countess of
Lovelace).

While data analysis is a central component of modern genetic and

genomic research approaches, most analytical methods have not

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_13, © Springer Science+Business Media New York 2012

215
216 S.J. Mack et al.

been developed specifically for immunogenetic data; the high level

of polymorphism at the HLA and KIR loci necessitates analytical
methods and software tools with the capacity to process 20 or
more alleles per locus for many loci. In addition, the extensive link-
age disequilibrium (LD) between immunogenetic loci (e.g., cover-
ing 3 MB in the MHC) requires the simultaneous computation of
association measures between as many as 14 (e.g., in the KIR com-
plex) highly polymorphic loci. Because there is no single tool that
can carry out all analyses, and because available tools are not per-
fect for immunogenetic data, specific concessions must sometimes
be made to enable analysis. Researchers and analysts should always
remain aware that even when an analytical application is appropri-
ate to the data, analysis can still be confounded by variation in
nomenclature and data-resolution (see Chap. 12).
As Ada Lovelace recognized in 1843, analytical tools do not
provide new insights or reveal truths; they may allow the imple-
mentation of complex methods, but these methods and their
assumptions should always be known and understood to the
researcher. Therefore, in addition to the discussions in this chapter,
researchers should be familiar with the literature describing each
method and the manual describing the use of for each analytical
tool. No analytical result should be accepted without being pre-
sented in the appropriate context.
In general, while some of the analyses described below can be
performed on paper or in a spreadsheet application, all are best car-
ried out by a dedicated software tool that has been specifically
developed with that analysis in mind. This provides a facile means
to describe what was done, minimizes error on the part of the ana-
lyst, and maximizes reproducibility both by the researcher and by
other researchers. Software tools generally run in either a Microsoft
Windows, Apple, or Linux/Unix environment (or in a web-
browser), but because many tools do not run in all three of these
operating systems, we do not recommend one over the others.
Researchers should maintain access to all three operating systems
in order to take advantage of all available tools.
Many of these calculations are “computationally expensive” in
that they are CPU- and memory-intensive. In general, the faster
the CPU and the more RAM on a system, the faster an analysis will
progress, but contemporary computer systems should be sufficient
to the task of running the applications described here. However,
care should be taken to ensure that the most recently released ver-
sion of any application is used for analysis; this will ensure that any
known software issues have been addressed. All of the tools that we
describe here are free, so there is no reason to rely on an outdated
version. Finally, many of the analyses described here have been
implemented as functions and packages written in R, the language
and environment for statistical computing (1). While we recom-
mend specific R packages for analysis, and generally recommend
 13 Analytical Methods for Immunogenetic Population Data 217

using the R language, a tutorial on R programming is beyond the

scope of this chapter.
In the sections that follow, we may discuss one particular soft-
ware tool or application in the context of a given analysis.
Relationships between analytical methods and some available tools
are summarized in Table 1. However, this table is not intended to
be comprehensive; there are many more possible analyses than
those described here, and there are many more tools available than
are discussed. For example, the compilation of genetic analysis
software at http://www.nslij-genetics.org/soft/ (2) includes 520
applications. Finally, supplementary population datasets (derived
from the master data set included in the supplementary materials
for Chap. 12) and associated files are available online at http://
www.immunogenomics.org/methods.html; these datasets dem-
onstrate input formats and were used to generate the example
figures and tables associated with each analysis.

2. Data Reporting

The verification of published findings through independent repli-

cation is a core element of scientific research. Whereas replication
is often interpreted to pertain primarily to the generation of experi-
mental data, it is equally important that analyses of those data be
replicated. To facilitate replication of analytical outcomes, the data
analyzed must be reported in an accurate and thorough manner.
Data described in the body of an immunogenetic paper should be
presented as both raw allele counts and the allele frequencies calcu-
lated from them; this will allow other investigators to perform
additional analyses (using counts) and will permit easy identification
of the extent of differences in frequencies. In addition, all alleles,
genotypes, and haplotypes (including rare variants) should be made
available, either within the body of the paper or as supplementary
material.

3. Analyses

3.1. Calculation 1. Direct Counting

of Gene (Allele) With the advent of molecular typing techniques, the need
Frequencies to estimate gene (allele) frequencies (GF) from phenotype data
has diminished. In most cases, gene frequencies for HLA data
can be obtained via direct counting, where the number of
observations for a given allele is divided by the number
of chromosomes (2n, where n = sample size) under study.
 Table 1
Matrix of population genetic analyses and available software tools
218

General
statistical
software
Analysis (e.g., SAS) MS Excel GenAlExa PyPopb Arlequinc EstiHaplod PHYLIPe Structuref CLUTOg PLINKh R packagei GenePopj

Carrier frequency + + ++ − − − − −
S.J. Mack et al.

estimation
Hardy–Weinberg − − + +++ ++ − − − + ++
Haplotype estimation − − + +++ ++ +++ − − +++ + haplo.stats
++
Linkage disequilibrium − − +++ ++ − − − +++ + ++
Measures of selection − − ++ +++ ++ − − − − −
Measurement of genetic − − ++ − ++ − ++ ++ +++ − ++ ++
differentiation
Principal component − − ++ − − − − − − ++
analysis
Phylogenetic analysis − − − − − ++ + +++ − ++
Population structure − − − ++ − − ++ +++ −
analysis
− = analysis is not possible with that tool; + = analysis is possible, but the tool is not recommended for this analysis; ++ = analysis is possible with this tool; +++ = this tool has been
optimized for the analysis of immunogenetic data
a
http://www.anu.edu.au/BoZo/GenAlEx/
b
http://www.pypop.org
c
http://cmpg.unibe.ch/software/arlequin3/
d
http://www.methodomics.com/estihaplo
e
http://evolution.genetics.washington.edu/phylip.html
f
http://pritch.bsd.uchicago.edu/structure.html
g
http://glaros.dtc.umn.edu/gkhome/views/cluto/
h
http://pngu.mgh.harvard.edu/~purcell/plink/
i
http://www.r-project.org/
j
http://genepop.curtin.edu.au/
 13 Analytical Methods for Immunogenetic Population Data 219

The PyPop (python for population genomics) application

(3) can be used to calculate allele frequencies in this manner.
Supplementary Tables S1, S2, S3, and S4 are PyPop-formatted
synthetic genotype data files for four populations. Supplementary
Table S5 is a PyPop configuration file specific for these data.
Further examples of PyPop-formatted HLA allele-frequency
data and configuration files, developed by Solberg et al. (4),
can be found at http://www.pypop.org/popdata. These cal-
culations are also easily accomplished directly from a Microsoft
Excel spreadsheet via the Genetic Analysis in Excel (GenAlEx)
Excel add-on (5).
2. Estimation from carrier frequencies
Direct counting cannot be used for all immunogenetic
data. A notable exception remains for the KIR loci, where
much of the available data have been generated on a presence/
absence basis for many KIR loci, yielding phenotypes for which
only carrier frequencies (CF, the presence of 1 or two of the
considered allele) can be obtained by direct counting. In these
cases, it is necessary to estimate gene frequencies. This can be
done with the assumption that the population under study is in
Hardy–Weinberg equilibrium (HWE) (see the final equation
in step 3). The most simple equation is given by:

GF = 1 − 1 − CF.
However, Lynch and Milligan (6) have shown that this
provides a downwardly biased estimate and suggest that a better
estimate is obtained by:

GF = 1 − (x 1/ 2 (1 − (var(x ) / 8x 2 ))−1 ).
where x = 1−CF, and var(x) = x(1−x)/N, where N is the num-
ber of sampled individuals.
Table 2 presents gene frequencies estimated using each of
these methods (formula a and formula b) in comparison to
frequencies calculated by direct counting.
3. Confidence intervals
Any estimated GF^ should be accompanied by a confidence
interval (CI), a range of values that reveals the precision of the
estimated frequency. The likelihood that the CI includes the
actual population GF is given by the confidence level (usually,
a 95% chance). For GF estimated from a sample of size n, the
CI has a lower bound of:

GF^ − ε (GF(1 − GF) / n).

And an upper bound of:

GF^ + ε (GF(1 − GF) / n),

220 S.J. Mack et al.

Table 2
Comparison of methods for determining gene frequencies for present/absent data

Gene frequency estimate from

carrier frequency Direct countinga

Locus Count Carrier frequency Formula a Formula b Count Gene frequency

Present 109 0.4225 0.2401 0.2398 62 0.2403
Absent 243 0.9419 0.7589 0.7570 196 0.7597
b
Total 352 1.3643 0.9989 0.9968 258 1
a
Assuming a molecular method that can distinguish homozygotes from heterozygotes
b
2n = 258

where ε invokes the Normal distribution to determine the

probability of the CI associated with the estimated GF. For a
95% CI, ε is 1.96, and for a 99% CI, ε is 2.576. (GF(1 − GF) / n)
is an approximation of the standard error of the estimated GF.

3.2. Hardy–Weinberg The Hardy–Weinberg (HW) principle provides a useful model for
Testing primary quality control (QC) verification of the integrity of geno-
type data, as genotyping errors may result in both individual
genotype deviations and overall deviations from HW equilibrium
(HWE) (see Note 1). In addition, HW testing is also useful for
detecting sampling errors (see below) in population samples.
Confidence in the accuracy of Hardy–Weinberg testing is therefore
crucial for confidence in subsequent analyses, as many analytical
methods (e.g., LD and haplotype estimation, Ewens–Watterson
analyses of selection) are predicated on an assumption of HWE in
the data set. In a Hardy–Weinberg test, observed genotype counts
are compared to those expected under Hardy–Weinberg equilib-
rium proportions (HWEP), as calculated by generating a table
of all possible genotypes, using an appropriate statistical method.
The relationship between the allele and genotype frequencies under
HWEP is given as:

f (Ai Ai ) = pi 2

and

( )
f Ai A j = 2 pi p j ,

where pi is the allele frequency of Ai. and pj is the allele frequency

of Aj. When a population is in HWE, there will not be a significant
departure from these allele and genotype frequencies and there will
be no change in allele frequencies between generations.
 13 Analytical Methods for Immunogenetic Population Data 221

Tests of overall locus-level HWEP compute a p-value to esti-

mate the significance of observed deviations across all genotypes.
Significant deviation of observed genotype counts from expected
HWEP can result from factors that include sampling errors (the
sampling of admixed, stratified, or some other form of blended
populations), inbreeding or other nonrandom mating, natural
selection, and genotyping errors. Tests for deviation from HWEP
have low power, and significant deviation from HWEP is not com-
mon. Genotyping errors (e.g., failure to detect a specific allele,
resulting in an excess of homozygotes) are the first consideration
when significant deviations from HWEP are detected (especially
when such deviations are detected only at a single locus in a
multilocus analysis), rather than the operation of selection, admixture,
or nonrandom mating, unless the sample is suspected to be from
an unusual population.
1. Chi-square testing
Hardy–Weinberg testing can be particularly challenging
for highly polymorphic datasets. Historically, the chi-square
test has been the standard approach for testing fit to HWEP
at the overall locus-level (regardless of the level of polymor-
phism at that locus). However, this test can lead to false accep-
tance or rejection of the null hypothesis when individual
expected genotype counts in the table of all possible genotypes
are small or close to zero (also known as, “sparse cells” in the
table of all genotypes, as represented by the shaded cells with
low numbers of expected genotypes in the upper half of Fig. 1).
Sparse cells can be problematic because the minimum number of
observed genotypes must be 1, while the minimum number
of expected genotypes will be the square of the frequency of
the rarest allele. It is not unusual for 30–40 or more alleles to
be observed at the highly polymorphic HLA loci, with a wide
range of frequencies, resulting in many such sparse cells. As a
result, the ratio of observed to expected (O/E) genotypes can
be as large as 100 or 1000 for rare genotypes in large popula-
tions with no actual HWEP deviation, while the O/E ratio for
common genotypes will usually be much smaller (usually
between 0.2 and 5), even in cases of actual deviation from
HWEP. Three approaches can be taken to increase the accu-
racy of Hardy–Weinberg tests of immunogenetic data; (1) rare
alleles can be “lumped” together in a combined class (as in
the lower half of Fig. 1) for the chi-square test to be effective;
(2) a complete enumeration of all possible tables of all possible
genotypes (an exact test) can be undertaken, or (3) approxi-
mations to such complete enumerations can be made via
resampling.
222 S.J. Mack et al.

No Allele "Lumping"
Number of Each Genotype Expected
Allele Count Frequency 1 2 3 4 5 6 7 8
1 12 0.29 1 3.43
2 10 0.24 2 5.71 2.38
3 9 0.21 3 5.14 4.29 1.93
4 5 0.12 4 2.86 2.38 2.14 0.60
5 2 0.05 5 1.14 0.95 0.86 0.48 0.10
6 2 0.05 6 1.14 0.95 0.86 0.48 0.19 0.10
7 1 0.02 7 0.57 0.48 0.43 0.24 0.10 0.10 0.02
8 1 0.02 8 0.57 0.48 0.43 0.24 0.10 0.10 0.05 0.02
Total 42 1

Alleles 5-8 "Lumped"

Number of Each Genotype Expected
Allele Count Frequency 1 2 3 4 5-8
1 12 0.29 1 3.43
2 10 0.24 2 5.71 2.38
3 9 0.21 3 5.14 4.29 1.93
4 5 0.12 4 2.86 2.38 2.14 0.60
5-8 6 0.14 5-8 3.43 2.86 2.57 1.43 0.86
Total 42 1

Fig. 1. Sparse cells and the “lumping” of alleles into combined classes. The effect of creating a combined, “lumped” allele
class on expected genotype counts is illustrated in the upper and lower halves. In the upper half, expected genotype counts
are calculated for all eight alleles; the expected counts for the ten genotypes comprising alleles 5, 6, 7, and 8 (shaded) are
much less than 1. In the lower half, alleles 5–8 have been lumped into the “5–8” allele class, and no genotype in the table
has an expected count less than 0.6.

2. Exact tests
An exact test for HWEP was developed by Louis and
Dempster (7). This test generated all possible tables of geno-
types (based on observed allele frequencies) when the sample
size and allele frequencies are held constant in accordance with
the exact distribution. The p-value was given by the cumulative
conditional probability of obtaining a table of genotypes (with
sample size and allele frequencies equal to the observed sam-
ple) with a conditional probability less than or equal to that of
the genotypes in the observed sample (8, 9). This test provides
the exact p-value for every sample and it does not require input
parameters that may affect the result. However, the number of
possible tables of genotypes grows exponentially as either the
sample size (n) or the number of distinct alleles (k) increases,
reducing the feasibility of this test when n and k are large.
3. Resampling approximations
Resampling approximations to complete enumeration of all
possible tables were developed for data sets with larger num-
bers of alleles, where the asymptotic chi-square test may be par-
ticularly problematic and exact tests with complete enumeration
were not possible (10–13). While these approximation or resa-
mpling tests are often erroneously referred to as “exact” tests,
they do not perform a true exhaustive search for all possible
tables of genotypes. These methods generally use the Monte
Carlo (MC) simulation method to approximate to the exact
 13 Analytical Methods for Immunogenetic Population Data 223

p-value and, therefore, represent an acceptable alternative to the

exact test.
Guo and Thompson (10) developed the first conventional
MC test of HWEP based on Levene’s conditional sampling
distribution and also proposed a MC test that uses a finite and
irreducible Markov Chain (MCMC) (14) to randomly gener-
ate tables of all possible genotypes. In these MC-based tests,
the p-value is given by the fraction of randomly generated
tables with a conditional probability less than or equal to the
conditional probability of the observed genotypes. Resampling
MC and MCMC tests perform very favorably when compared
to the exact test and always outperform the chi-square test.
However, the MCMC method may fail to approximate to the
exact p-value in a few cases, and the MC test is preferred in
cases where the exact test cannot be performed.
Chi-square tests, and MC and MCMC resampling approx-
imations of the exact test are performed by PyPop, which has
been designed specifically for the analysis of highly polymor-
phic immunogenetic data. For the chi-square test, PyPop auto-
matically creates combined categories of rare alleles based on a
user-defined “lumping” threshold, with a default value of 5.
4. Hardy–Weinberg testing of individual genotypes
Chen et al. (15) measured the goodness of fit of individual
genotypes to expected HWEP in MC approximations of the
exact test by comparing disequilibrium coefficients (16, 17).
PyPop calculates two measures (Chen and Diff) of the good-
ness of fit of individual genotypes when the MC or MCMC
test is implemented. In cases when locus-level deviations from
HWEP are detected, these individual genotype tests may help
identify the specific genotypes contributing to the deviation,
but should be considered only when the number of expected
genotypes is at least 1. As noted above, large p-values may
result when a genotype with an expected count much less than
1 is observed once or twice; researchers should avoid making
analytical inferences on the basis of these p-values.

3.3. Haplotype Estimated haplotypes and haplotype frequencies play a central role
Estimation in most genetic studies. Haplotype-level analyses are important to
studies of the etiology of human disease, selective forces acting on
populations, and optimal sizes for bone-marrow donor registries
(BMDRs). Associations between markers and disease loci that are
not evident with a single-marker locus may be identified in mul-
tilocus marker analyses using estimated haplotype frequencies
(HFs). The design of studies and the recruitment of the samples
are dependent on the possibility of identifying haplotypes by segre-
gation analysis in families or estimating haplotypes from population
samples of phase-unknown unrelated individuals (18). Haplotypes
224 S.J. Mack et al.

are used for disease association mapping, QTL mapping, and

even imputing underlying genetic markers (19).
The term “haplotype” now includes any set of genetic polymor-
phisms (i.e., all DNA sequence variation including deletion/inser-
tions) at contiguous loci. Except when recombination occurs, these
neighboring genetic polymorphisms are cotransmitted by a single
parental chromosome. Haplotypes may be represented as blocks of
DNA sequence variants (e.g., SNP haplotype blocks), or groups
of sequence variants can be abstracted into an allelic nomenclature
at the level of a functional locus, as in the HLA and KIR systems.

3.3.1. Expectation- Early work on the estimation of haplotype frequencies from unre-
Maximization Algorithm lated genotype data was based on the expectation-maximization
(EM) algorithm with the assumption of HWEP at the locus-level
(20–24). Later work refined, explored, and extended aspects of the
algorithm (25–30). Application to haplotypes of SNPs (31–33)
and Bayesian methods (34, 35) are commonly used. It remains
unclear whether the Bayesian algorithms perform better than max-
imum likelihood implemented in EM algorithm (35).
Haplotypes can be estimated using a number of software tools,
for example, in a standard implementation of the Expectation-
Maximization (EM) algorithm in the haplo.stats package for R, the
open source language, and environment for statistical computing
(1). Although there is a great desire within the immunogenetic
community for applications capable of analyzing very large (>mil-
lion individuals) data sets, available HF and LD estimation soft-
ware are generally limited in their capacity to a few thousands of
individuals. For example, precompiled versions of PyPop are cur-
rently limited to 7 loci and 5,000 individuals when it comes to
estimating haplotypes and calculating LD values. In contrast,
haplo.stats can accommodate very large datasets, depending on the
number of alleles at each locus; for example, haplo.stats estimates
haplotypes for 240,000 individuals over four loci with an average
of 25.5 alleles per locus, or 60,000 individuals over 50 loci with
the same mean number of alleles. Supplementary Table S6 is a
haplo.stats-formatted version of the synthetic genotype data in
Supplementary Tables S1–S4. The “master data file” described in
Chap. 12 of this volume (Chap. 12 Supplementary Table S1) can
also serve as a haplo.stats input file.
Population-level haplotype frequencies are estimated via EM
using simultaneous maximum-likelihood estimation of n-locus
haplotype frequencies. The expectation step determines the
expected number of copies for each haplotype contributing to a
given genotype. For a three locus haplotype, this is calculated as:

E [nabc | Pi ] = 2 f abc Sf abc / Pr (Pi ),

 13 Analytical Methods for Immunogenetic Population Data 225

where S is the number of ambiguous haplotypes in Pi, E [nabc|Pi] is

the expected number of copies of haplotype Habc within Pi, and fabc
is the frequency of each other possible haplotype Habc to form the
genotype of frequency Pi. The maximization step determines new
estimates for fabc for the next iteration of the algorithm. At each
iteration, the estimations globally improve.

3.3.2. Challenges 1. Rare estimated haplotypes

to the Use of Estimated The performance of haplotype frequency estimation algo-
Haplotypes rithms is sensitive to various aspects of the population under
study (35). Estimated frequencies for rare haplotypes (n = 1 or
2 in a dataset), which incorporate low-frequency alleles, are
often incorrect, even when the EM algorithm finds the global
maximum likelihood (27, 36, 37). The accuracy of haplotype
estimates is critical for association and candidate gene studies,
fine-mapping of disease genes, and for microsatellite, SNP, and
protein level variation, and the presence or absence of specific
low-frequency alleles and haplotypes must inform the robust-
ness of associations. Analytical inferences should not be made
on the basis of these rare haplotypes.
2. Haplotype estimation for immunogenetic data
The diversity and complexity of Immunogenetic data poses
additional challenges for haplotype estimation. Over the past
30 years, the Immunogenetic community has seen an expo-
nential increase of the number of HLA alleles leading to regular
nomenclature revisions, and this phenomenon now extends to
the KIR genes (38). In both the MHC and KIR regions we
have: heterogeneity of typing resolution, heterogeneity of
typing techniques, heterogeneity of allele nomenclatures, con-
tinual discovery of new alleles, large numbers of allele per loci
(roughly >50), and high haplotype diversity (roughly >1,000).
In addition, KIR and HLA data are very sensitive to ethnic
background diversity. The potential for population substruc-
ture is particularly relevant for immunogenetic data due to the
fact that MHC and KIR genes can reflect both the selective
and demographic histories of populations. These issues are
exacerbated in BMDRs where sample sizes for specific research
questions are often very large (>100,000).
Little is known about the behavior of estimated haplotypes
in the extreme situations described above for the HLA and
KIR regions and little attention has been paid to the biases
affecting haplotype frequency estimation. The frequency of
the alleles, the sample size of the dataset, the various levels of
missing information, and the various levels of linkage disequi-
librium surely influence the accuracy of the estimation.
Haplotype frequency estimations are primarily affected by sam-
pling fluctuation. In HLA and KIR, it is highly likely that the
226 S.J. Mack et al.

sample sizes are usually too small to cover the extent of the
haplotype diversity. As a result, the haplotype frequencies and
linkage disequilibrium between alleles are overestimated; such
a bias would occur even if the chromosome phase was known.
3. HF Estimation for KIR
Because some KIR genes are present only on certain haplo-
types, the space of possible KIR haplotypes excludes some locus
combinations that could be generated from the observed geno-
typic data. The EM algorithm for estimating KIR HFs must
be modified to account for this reduced combinatorial space,
e.g., using an a priori list of known/possible haplotypes to
constrain the EM algorithm (39, 40). The user-designated a
priori haplotype list is said to span a set of observed genotypes
if each observed genotype can be generated from at least one
pair of haplotypes in the list. If the list does not span the
observed genotypes, the resulting estimates must be carefully
interpreted.
Several recent KIR HF estimation studies have noted short-
comings in the use of such constraints, imposed by the need to
specify predefined haplotype patterns (Fig. 2). Yoo et al. (40)
found that accuracy measures related to haplotype identification
were particularly low for fewer than 200 individuals and sug-

Generate
HLA All Possible
KIR Reference
Haplotypes
Pairs

Determine
Observed HLA Observed KIR
All Possible
Genotypes Haplotypes KIR Genotypes

2 Classes of
Observed Genotypes:

a) Unambiguous Haplotypes
b) Ambiguous Haplotypes

E-M
Algorithm

Haplotype
Frequency
Estimates

Fig. 2. Overview of HLA haplotype estimation and KIR haplotype estimation strategies. The EM algorithm for estimating KIR
haplotype frequencies (HFs) can be modified from the standard approach applied to HLA genotypes (upper-left box) to
account for a reduced combinatorial space using a set of reference haplotypes as an a priori list of known/possible haplo-
types to constrain the algorithm (upper-right box).
 13 Analytical Methods for Immunogenetic Population Data 227

gested that more than 500 individuals would provide accept-

able estimation accuracy. In describing their HAPLO-IHP soft-
ware, Yoo et al. noted that unusual haplotypes incompatible
with constraints may be incorrectly rejected. When the a priori
list of user-defined haplotypes does not span the observed gen-
otypes, haplotypes that may not exist are “constructed” in an
attempt to satisfy user-defined haplotype patterns.

3.4. Measures Linkage disequilibrium is defined as the nonrandom association of

of Linkage alleles at two loci. High levels of LD combined with high levels
Disequilibrium of polymorphism are the defining characteristic of immunogenetic
loci. Measurement of LD provides a means to assess the degree to
which pairs of alleles are likely to be observed on the same haplo-
type and has important implications in analyzing immunogenetic
data for population and disease association studies.

3.4.1. Haplotype-Level LD 1. Dij and Dij′

statistics Pairwise disequilibrium statistics can be calculated for each
haplotype for polymorphic loci:
Dij = xij − pi × q j ,

where xij is the estimated haplotype frequency (see previous

section) and pi and qj are the ith and jth allele frequencies at the
two loci. In order to account for differing allele frequencies at
the loci, a normalized disequilibrium value can be used (41).
This is given by:
Dij′ = Dij / D max ,

where Dmax is the lesser of piqj and (1−pi)(1−qj), when Dij is <0
and pi(1−pi) and qj(1−qj), when Dij is >0.
2. r2
The r2 measure is another means of normalizing Dij to
account for differing allele frequencies. This is the square of the
correlation coefficient (r) between the alleles at the p and q loci
(42). Because r is given as:

r = (Dij / pi × q j (1 − pi )(1 − q j ))1/ 2,

r2 is therefore given as:

r 2 = Dij2 / (pi × q j (1 − pi )(1 − q j )),

3.4.2. Global LD Statistics For loci with more than two alleles, global LD statistics extend the
haplotype-level statistics to account for all possible combinations
of alleles at each locus (43).
1. Wn
Wn is a multiallelic extension of the correlation measure r.
The chi-square value for testing the significance of LD can be
written as W/(2N) where:
228 S.J. Mack et al.

W = (∑∑ Dij / pi × q j )1/ 2 ,

and pi and qj are the observed allele frequencies at each of
the two loci having k and l alleles, respectively. Wn, or Cramer’s
V statistic, is a normalized value that addresses differing num-
bers of alleles at the two loci (44, 45).

Wn = W / (min(k, l ) − 1).
The values of Wn fall between 0 and 1, and the significance
of the overall disequilibrium is assessed using the abovemen-
tioned chi-square test. It should be noted that the Wn measure
is always symmetric with respect to two loci, whereas the num-
ber of alleles reported at each locus can differ considerably.
It is therefore important not to overinterpret values of Wn
for locus pairs with highly asymmetric numbers of alleles.
Finally, for biallelic loci, Wn is equivalent to r.
2. D¢
D¢ is a second global disequilibrium statistic, which sums
the absolute value of normalized Dij values over all haplo-
types, weighted by the frequencies of the alleles in each
haplotype (46). As with Wn, D¢ values fall between 0 (equilib-
rium) and 1 (linkage). This is given as:

D ′ = ΣΣpi × q j | Dij′ |,

PyPop calculates Dij , Dij′ , D¢ and Wn values.

3.4.3. Graphical The interpretation of LD values between many markers can be facili-
Representation of LD tated through the graphical representation of LD patterns. Compared
Patterns to a tabular presentation of LD values, such visual representations
facilitate the identification of patterns and interesting subsets of the
data. So-called “heat maps” are a common means of representing
pairwise LD values across markers, as a half-matrix in which the
strength of the LD (e.g., the log of the p-value) is represented by a
color scale. However, most of the software tools developed for
graphical LD presentations represent biallelic markers and can there-
fore only represent average LD between multiallelic loci. Popular
software tools for this purpose include graphical overview of linkage
disequilibrium (GOLD) (47), Haploview (48), MIDAS (49), and
various packages in R (e.g., LDHeatmap (50)). While PyPop does
not generate graphical LD representations, PyPop-generated LD
data can be imported directly to R. To our knowledge, only MIDAS
will simultaneously represent the interallelic component of LD.

3.5. Measurement 1. Ewens–Watterson homozygosity statistic

of Selection The expected proportion of homozygotes under Hardy–
Weinberg, for an observed value of k and a given sample size
(n), is used as a measure of the allele-frequency distribution
 13 Analytical Methods for Immunogenetic Population Data 229

and compared to the distribution expected under the neutral

model for the same values of k and n (51). Allele-frequency
distributions are used to calculate Watterson’s homozygosity F
statistic (52). This is given by:
F = ∑ pi2 ,
where pi is the frequency of the ith allele at a locus. The
homozygosity test can be accomplished using the exact test
described by Slatkin (53, 54). For given values of n and k, all
possible configurations of alleles are listed (each configuration
is a distinct way of distributing the n sampled genes into k
allelic categories). The probability of obtaining a particular
configuration can be computed under the null hypothesis of
neutrality using the Ewens sampling formula (51). The
homozygosity value of each configuration along with its prob-
ability gives the sampling distribution for F under neutrality.
This distribution is used to find the probability of obtaining
homozygosity values equal to or larger than that observed, for
a test of positive selection, by examining how many
configurations result in homozygosities greater than this
observed value (53). Similarly, a test of balancing selection is
based on the probability of obtaining a homozygosity value as
small as or smaller than the observed value. Significant p-values
reject the null hypothesis that the sample came from a popula-
tion that is undergoing neutral evolution.
2. Normalized deviate of homozygosity
Homozygosity values calculated for different values of n
and k can be directly compared by calculating the normalized
deviate of homozygosity (Fnd) (55). This is given by:

Fnd = (Fobs − Fexp ) / var(Fexp ),

where Fobs is the homozygosity value calculated for an observed

frequency distribution, Fexp is the mean homozygosity expected
under the neutral model. While Fnd is a normalized deviate
(similar to a z-score), the sampling distribution for Fnd is not
normally distributed, so that p-values cannot be inferred from
a given Fnd value using traditional parametric methods.
Statistical significance for an Fnd value is given by the significance
of the corresponding Fobs value.
The normalized deviate of homozygosity can also be used
to characterize homozygosity values that deviate significantly
from the null hypothesis in terms of modes of evolution. Fnd
values significantly lower than 0 result from allele-frequency
distributions that are more “even” than expected and are
consistent with the action of balancing selection. Fnd values
signi fi cantly higher than 0 result from allele-frequency
230 S.J. Mack et al.

distributions that are more skewed than expected toward

specific alleles and are consistent with either directional
selection or an extreme demographic effect.
In addition, because Fnd is equal to 0 under the null hypoth-
esis, a paired sign test (56) can be used to compare multiple Fnd
values against the expectation of neutrality.
PyPop calculates F and Fnd values.

3.6. Measures of 1. Heterozygosity

Genetic Diversity The level of genetic diversity at a given locus is dependent
upon the allele frequencies of the marker and the number of
alleles observed in the sample. Within a given population,
variation may be described by the heterozygosity (H ), which
ranges between 0 and 1:
H = 1 − Σpi2 ,
where pi is the frequency of the ith allele. For a population in
HWE, this is the probability that a random individual in the
population is a heterozygote. Heterozygosity will be maxi-
mized when all alleles are at an equal frequency.
2. Polymorphism information content
The polymorphism information content (PIC) value is an
additional statistic based on allele frequencies at a locus that
describes the ability of a marker to differentiate individuals
within a population (57):
PIC = 2ΣΣpi × p j (1 − pi × p j ),
where pi is the frequency of the ith allele, and pj is the fre-
quency of the (i + 1)th allele. This is the probability that one of
two individuals in a randomly mating population is a heterozy-
gote and that the other is a different genotype. As with
heterozygosity (H), PIC is maximized when all alleles are at
equal frequency. The values of H and PIC are very similar at
high heterozygosities, but PIC will never exceed H and its val-
ues are less than H when heterozygosity is low.

3.7. Measures of The measures below are used to quantify genetic variation within
Genetic Differentiation and between populations and to determine subdivisions (subpopu-
lations) of a single source (total population).
1. FST
FST values quantify levels of population differentiation by
assessing the proportion of genetic variance in subpopulations
relative to the total genetic variance (58). FST can be calculated
based on a partitioning of heterozygosity:
FST = (H t − H S ) / H t ,
 13 Analytical Methods for Immunogenetic Population Data 231

where Ht is the heterozygosity of the total population, and Hs

is the average heterozygosity of the subpopulations. Nei (59)
has shown that this can be expressed in terms of allele
frequencies:
FST = Σvar(pi ) / Σvar(1 − pi ),
where pi is the allele frequency of the ith allele in the total
population and var(pi) is the variance of the ith allele over
subpopulations. FST is a qualitative measure, and Wright (60)
has suggested guidelines for interpretation of these values:
0–0.05 indicates little genetic differentiation between
subpopulations
0.05–0.15 indicates moderate genetic differentiation between
subpopulations
0.15–0.25 indicates great genetic differentiation between
subpopulations
0.25 and above indicates very great genetic differentiation
between subpopulations
It has been shown that there is good concordance between
FST and average divergence times within and between subpop-
ulations, given neutral loci and assuming the infinite alleles
mutation model (61). FST has been traditionally applied to data
such as those obtained for allozyme variation. This measure
may lose some power when applied to loci with a relatively
high mutation rate, as has been suggested for microsatellite
loci. Several different methods have been applied to estimate
mutation rates for microsatellites (62, 63) and rates ranging
from 10−3 to 10−5 have been reported.
2. RST
Additional measures related to FST have been described for
application to microsatellite data. Slatkin (64) introduced RST,
which has similar properties to FST, but assumes a stepwise
mutation process, as well as a relatively high mutation rate. It is
calculated as:
RST = (S − S w ) / S ,
where Sw is the sum over all loci of twice the weighted mean of
the within population variances, V(A) and V(B), and S is the
sum over all loci of twice the variance of the combined popula-
tions, V(A + B). In computer simulations, it was demonstrated
that RST may provide a relatively more unbiased estimate of
coalescence times compared to FST.
3. Population-pairwise FST
The degree of differentiation between pairs of popula-
tions (population-pairwise FST) can be used to investigate
232 S.J. Mack et al.

the existence of population structure (64–66). In accounting

for small differences in subpopulation sample sizes, the
population-pairwise FST calculation may result in small negative
pairwise FST values; it is common practice to treat these nega-
tive values as being equivalent to zero. Pairwise standardized
FST values ( FST ′ values) are generated using Hedrick’s method
of dividing each value by the maximum population-pairwise
FST value (67), allowing comparison of genetic differentiation
between loci with different mutation rates and between popu-
lations with different effective sizes. For n subpopulations, this
population-pairwise approach results in a matrix of n(n−1) FST
or FST ′ values.
Arlequin will calculate FST and population-pairwise FST val-
ues. Supplementary Table S7 includes Arlequin-formatted
genotype data for the synthetic datasets discussed in this
chapter.

3.8. Graphical There are a variety of methods for representing genetic difference
Representations data between subpopulations in a graphical (as opposed to tabular)
of Genetic format. In many cases, the graphical representation can be applied
Difference Data independently of the measure of differentiation, so that multiple
different genetic differentiation measures can be compared using
the same graphical representation and multiple graphical represen-
tations can be applied to the same genetic differentiation measure.
Because the graphical representation usually depends on an addi-
tional analysis, we describe some of the commonly used represen-
tations here as individual analyses.
In general, these representations should not necessarily be
thought of as providing the definitive answer to a question so much
as they serve as aids for the interpretation of genetic differentiation
data that may be too complex to present in a tabular format (as
with LD values). As Ada Lovelace noted, “the Analytical Engine
has no pretensions whatever to originate anything.” The results of
these methods should always be interpreted critically, and the
researcher who uses these methods should develop a set of criteria
for accepting or rejecting the results of a method before using that
method. Overinterpretation of any of these representations should
be avoided when there is no obvious historical, functional, or bio-
logical basis for them.
1. Principal component analysis
Principal component analysis (PCA) is used for dimension-
ality reduction in a data set, identifying those elements that
contribute most to its variance, and is particularly useful as an
exploratory tool in a complex data set (68). For the representa-
tion of genetic differentiation analyses, it is common to present
the results of a PCA via multidimensional scaling (MDS),
where each range for a given component is presented along a
corresponding MDS axis (69, 70). Each data-element (a popu-
lation or an individual) is represented by its position relative to
 13 Analytical Methods for Immunogenetic Population Data 233

each axis. This MDS approach allows the comparison of simi-

larities and differences between populations, or the individuals
that they comprise. For a 2D PCA MDS plot, distribution of
points along the primary (x) axis will correspond to the great-
est amount of variation in genetic distances in the data set (the
first principal component); distribution along the y-axis corre-
sponds to the next highest degree of the remaining variation
that is not correlated with the x-axis (the second principal
component). Because there can be many more than two prin-
cipal components (as long as there remains variation that is not
correlated with higher order components), comparisons can be
related with multiple 2D PCA plots, representing the intersec-
tion of different components, or with multiple 3D visualiza-
tions. However, increasingly smaller percentages of the variance
are represented by the higher numbered components, and
these are usually not presented. In some cases, it may be neces-
sary to present multiple plots for the same components. For
example, when some populations display extensive genetic dif-
ferentiation relative to others, it is often difficult to illustrate
the differences between populations with relatively low degrees
of differentiation; the PCA for these latter populations can be
presented in a MDS plot with a smaller scale.
As noted above, PCA can be used to investigate differentia-
tion between sampled individuals or between populations. The
PCA-mediated comparison of individuals in multiple popula-
tions is in essence a population structure analysis, which is
described below. For population-level analyses, genetic distances
are first calculated in a pairwise fashion (e.g., as population-
pairwise FST values) between populations, and PCA is performed
to assess the variation in distance between populations. PCA
MDS analysis is available in a wide variety of statistical software
applications (e.g., GenAlEx, and R packages). A population-
level PCA would proceed via the following steps:
(i) Calculate allele or haplotype frequencies in a set of
subpopulations.
(ii) Use a differentiation measure to generate a genetic
distance matrix for the subpopulations.
(iii) Calculate the principal components from the distance matrix.
(iv) Generate a (series of) MDS figure(s) representing two or
three of the principal components.
A 2D population-level PCA MDS plot generated in
GenAlEx for the synthetic datasets discussed in this chapter is
presented in Fig. 3. GenAlEx formatted data are included in
Supplementary Table S8. (see Note 2).
2. Population structure analysis
Population substructure and population admixture can be
directly investigated by estimating the likelihood that a given
genotype belongs to a specific population. Likelihood values
234 S.J. Mack et al.

Victoria

Second Principal Component

Wilkes

Marie Byrd

Queen
Maud

First Principal Component

Fig. 3. Population-level principal component analysis multidimensional scaling plot

generated using the supplementary data. Population-level principal component (PC)
analysis of the synthetic data in Supplemental Table S8. The internal axes represent
values of 0.0 for each component. The first PC in this plot represents 70% of the variance
in these data, and the second PC represents 26%. Higher order PCs describe only 4% of
the variance and do not need to be presented.

can be calculated based on HWEP, allele frequencies, and LD

(if available) and are used to assign individual genotypes to
specific groups or clusters, which can range from individual
populations to geographic regions. As with phylogenetic
trees, these clustering results can be displayed using multiple
graphical representations. A 2D individual-level PCA plot
generated in GenAlEx for the synthetic datasets discussed in
this chapter is presented in Fig. 4.
While there are many clustering tools available, the most
widely used is Structure (71, 72), which Rosenberg et al. (73)
used to cluster the populations of the CEPH Human Genetic
Diversity Panel, largely by geographic origin, on the basis of
genotype data for a genome-wide set of 377 microsatellites.
Structure iteratively resamples individuals into a number of
user-defined clusters (K) and calculates the likelihood for each
organization via Bayesian inference from expected HWEPs. In
addition, other types of clustering analyses can also be carried
out with Arlequin, CLUTO (74), and various R packages.
Supplementary Table S9 contains Structure-formatted
genotype data for the synthetic datasets discussed in this chap-
ter. Figure 5 includes structure plots generated with these syn-
thetic datasets for K = 2 and 4 (see Note 3).
3. Phylogenetic analysis
A phylogenetic tree (aka dendrogram) is a branching represen-
tation of the evolutionary history between populations, individuals,
or gene/protein sequences (taxa) based upon similarities and
differences in some characteristic (for our purposes, immuno-
genetic allele and haplotype frequencies) shared by all taxa.
 13 Analytical Methods for Immunogenetic Population Data 235

Marie Byrd
Queen Maud
Victoria

Second Principal Component

Wilkes

First Principal Component

Fig. 4. Individual-level principal component analysis multidimensional scaling plot gener-

ated using the supplementary data. Individual-level principal component (PC) analysis of
the synthetic data in Supplemental Table S8. The internal axes represent values of 0.0 for
each component. The first PC in this plot represents 21% of the variance in these data,
and the second PC represents 19%. Because higher order PCs describe 60% of the vari-
ance (evident from the extensive clustering of individuals in the second PC dimension),
additional PC dimensions should be presented.

Fig. 5. Structure bar plot generated using the supplementary data. Structure analysis of the
synthetic data in Supplementary Table S9 with the number of clusters (K) set to 2 or 4.
Vertical bars represent each individual included in the analysis, and each tone (or color in
the electronic version) indicates the extent to which that individual’s genotype is derived
from one of the K clusters, with each tone (or color) corresponding to a cluster. Because of
the low number of loci and the extensive sharing of alleles between populations, very few
individuals are assigned to a single cluster (tone/color). However, the relative relatedness
of each population can be inferred from the tone/color compositions of their constituents.

Trees generated using gene or protein sequences (sequence

trees) often allow inferences regarding the relative “age” of
sequence variants and the inference of ancestral sequences. We
do not discuss sequence trees here. Trees generated using pop-
ulation-level allele-frequency data (population trees) can repre-
sent relative degrees of shared ancestry between populations.
Where sequence trees can be interpreted as gene genealogies,
population trees should be considered as graphs of the general
trends in relationships between modern populations, which can
236 S.J. Mack et al.

change in ways (e.g., admixture, splitting, fusion, bottlenecks,

etc.) that nucleotide and protein sequences cannot. In particu-
lar, the relationships represented by population trees are
generally representative of the first (and sometimes second)
principal components of the frequency data used to generate them.
Population trees are generated via the following general steps.
(i) Calculate allele or haplotype frequencies in a set of
subpopulations.
(ii) Use a differentiation measure to generate an estimated
genetic distance matrix for the subpopulations.
(iii) Calculate the tree topology from the distance matrix.
(iv) Generate a tree figure representing that tree topology.
Allele-frequency-based genetic distance calculations do not
take the sequence relationships between individual alleles into
account, so that all alleles are considered to be equidistant from
each other; for many immunogenetic alleles (e.g., those that
diverged prior to the radiation of human populations from
Africa), this should not be an issue, but when populations dis-
play large differences in frequency for alleles or haplotypes that
have been relatively recently generated (e.g., DRB1*08:02:01
and DRB1*08:07), genetic distances may be overestimated,
resulting in very large branch lengths. Similarly, alleles that may
be reported differently depending on the typing method used
(e.g., DRB1*14:01:01 and DRB1*14:54) may result in similar
overestimates. In these cases, the datasets should be reviewed for
consistency in the level of resolution of typing, and alternate
names for what is potentially the same allele should be binned
into a common category (e.g., DRB1*14:01:01G). Finally, trees
including isolated populations with low values of k, in which a
few alleles are subject to genetic drift, may suffer similar prob-
lems. It is oftentimes useful to establish threshold for sample size
(e.g., 2n > 49) and k (e.g., >5) to exclude populations that are
too small or display too few alleles.
In general, phylogenies can be drawn as either “rooted” or
“unrooted” trees, as illustrated in Fig. 6; rooted trees identify
a common ancestor for all of the taxa, giving the tree direc-
tionality from root to twigs. We recommend presenting popu-
lation trees constructed using immunogenetic allele and
haplotype frequencies as unrooted, as it is difficult to know
exactly how or where to place the root. For example, the
influence of low values of k on branch lengths (discussed above)
raises questions about the effectiveness and appropriateness of
midpoint rooting. Similarly, because these trees are based on
frequency distributions, the lack of a nonhuman population
sharing allelic and haplotypic diversity with human populations
makes outgroup rooting difficult.
 13 Analytical Methods for Immunogenetic Population Data 237

Fig. 6. Examples of phylogenetic trees. Three representations of the same phylogeny for
four taxa (A–D). Black dots indicate nodes in each tree. The branches between each taxon
and the nearest node are known as “twigs” or “leaves”. Grey dots in the rooted trees
indicate the position of the root node. Taxa A and C are more similar to each other than
either is to taxon B or D, and taxa B and C are more similar to each other than either is to
taxon A or D. In the unrooted tree and the midpoint rooted tree, A and C are in one clade,
and B and D are in a second clade. In the outgroup rooted tree, A, B, and C are in a single
clade, to the exclusion of D.

PHYLIP
The PHYLogeny Inference Package (PHYLIP) (75, 76) is a
software suite of applications for building phylogenetic trees
using a variety of methods. Supplementary Table S10 is a PHYLIP
GENDIST-formatted allele-frequency data file for the synthetic
population datasets analyzed in this chapter. Figure 7 includes a
pair of unrooted Neighbor-Joining (NJ) (77) trees generated
using the data included in Supplementary Tables S7 and S10.
The tree in Fig. 7a is based on Nei’s standard genetic distances
(SGD) (78) calculated in PHYLIP, whereas the tree in Fig. 7b
is based on population-pairwise FST values calculated in Arlequin.
A genetic distance scale should be included with every tree.
Steps (ii)–(iv) outlined above were carried out with PHYLIP
to generate Fig. 7a using the GENDIST (for step ii.), NEIGHBOR
(for step iii.), and DRAWTREE (for step iv) programs to draw
unrooted NJ trees based on Nei’s Standard Genetic Distances.
Figure 7b was generated using the same procedure for steps (iii)
and (iv), but steps (i) and (ii) were carried out using Arlequin to
generate population-pairwise FST values.
PHYLIP’s GENDIST estimates genetic distance with three
different measures—Nei’s SGD, Cavalli-Sforza’s chord dis-
tance (79), and Reynold’s genetic distance (65)—and each
measure is based on implicit assumptions that may not always
apply to immunogenetic data. For example, all three measures
assume that population differentiation derives from genetic
drift, yet the HLA loci have been shown to be under balancing
selection in numerous studies (4, 80, 81).
238 S.J. Mack et al.

Fig. 7. Phylogenetic trees generated using the supplementary data. (a) Unrooted
neighbor-joining tree generated in PHYLIP using Nei’s standard genetic distances (included
in Supplemental Table S10) generated for the synthetic data in Supplementary Tables
S1–S4. Inset bar shows a genetic distance of 0.082. (b). Unrooted neighbor-joining tree
generated in PHYLIP using population-pairwise FST distances generated in Arlequin for
the synthetic data in Supplementary Table S7. Inset bar shows a population-pairwise FST
distance of 0.007.

Nei’s SGD assumes that new alleles arise by neutral muta-

tion, and that the mutation rate is equal across all loci; again,
the latter assumption clearly does not hold for HLA loci, where
there are many more class I alleles than class II alleles, and
where many HLA-B alleles are observed to be restricted to
specific regions of the world, whereas most HLA-DQA1 and
DQB1 alleles are observed in all populations (82).
However, the Cavalli-Sforza and Reynolds distance mod-
els assume no mutation; frequency differences between popu-
lations are assumed to be the result of genetic drift alone. This
assumption of no mutation seems even a further departure
from observed immunogenetic biology than the assumption of
locus-identical neutral mutation, as unique HLA alleles are
observed on a regular basis (83), and natural selection appears
to have favored novel HLA allele variants over older variants in
North and South American populations (84, 85).
Clearly, none of these models applies perfectly to immuno-
genetic data. Our empirical experience has been that HLA
gene-trees conform best to expectations when generated using
Nei’s SGD; this distance estimate includes a mutational com-
ponent, which clearly applies to HLA data.
PHYLIP’s NEIGHBOR (see Note 4) builds trees with
either of two clustering methods—NJ and Unweighted Pair
Group Method with Arithmetic Mean (UPGMA) (86, 87).
For the purposes of this discussion, the primary difference
between these methods is the assumption of the rate of evolu-
tion. The NJ method does not assume a regular stochastic
 13 Analytical Methods for Immunogenetic Population Data 239

evolutionary rate (i.e., a molecular clock), while the UPGMA

method does. This difference means that NJ trees are inher-
ently unrooted (and should be drawn as such), while UPGMA
trees are inherently rooted. The assumption of a molecular
clock in the UPGMA method makes these trees particularly
ill-suited for building immunogenetic gene-frequency trees, as
exonic immunogenetic allelic differentiation does not conform
to a molecular clock model; for the HLA loci, different modes
of evolution are in effect for introns, non-ARS-encoding exon
sequences, and ARS-encoding exon sequences (85, 88, 89).
For the generation of immunogenetic gene-frequency trees in
PHYLIP, we recommend building NJ trees with Nei’s SGD.
4. Criteria for rejection or acceptance of trees
As discussed above, it is important to develop criteria for
the purpose of evaluating trees prior to analysis. Sequence trees
can be evaluated via bootstrapping (90, 91), in which a popu-
lation of trees is generated by resampling subsets of the pri-
mary data. The topologies of the resulting trees are compared
and each branch in the consensus tree is evaluated based on
degree of sharing of topological features across the population
of trees.
For sequence trees, resampled datasets are generated by
randomly sampling and duplicating subsets of nucleotide or
peptide positions, but resampled gene-tree datasets are gener-
ated by randomly sampling and duplicating entire loci; there-
fore bootstrapping cannot be used to evaluate gene-trees
constructed for single loci, and bootstrapping performs poorly
for gene-trees constructed with a small number of loci. For
example, gene-trees constructed using two loci will yield boot-
strap values of 0.0, 0.5, or 1.0.
Studies of population relationships at non-MHC loci have
generally shown a close correlation between genetics and geog-
raphy, so that populations tend to share similar allele frequen-
cies with their neighbors on a local level (92, 93). Therefore, it
is not unreasonable to expect that most populations will be rep-
resented as being more closely related to their neighbors (pop-
ulations in the same global region) than to nonneighboring
populations in gene-trees, and that most relationships in a tree
will corroborate geographic, historical, anthropologic, and lin-
guistic evidence. Mack and Erlich (94) proposed that HLA
gene-trees be rejected as invalid if more than 6% of the intrare-
gional population relationships in that tree do not meet this
expectation. This criterion is necessarily conservative; if a tree
meets no expectations, it is difficult to say which relationships
are genuine, and which might be spurious. Trees such as these
may reveal more about the diversification of the loci inves-
tigated than than they do about population relationships.
Finally, population relationships, however unexpected, that are
240 S.J. Mack et al.

repeatedly observed in trees derived from different markers,

clearly merit serious consideration as reflecting actual relation-
ships rather than as a spurious artifact of the tree-building
process.

4. Notes

1. For the most part, publications that use HW testing in this

manner, including this chapter, share a common approach:
regardless of the method, all attempt to detect a statistically
significant difference between the observed genotype frequencies
and those expected under HWEP; these expected frequencies
derive from a Null hypothesis (H0) that assumes the HW model
is true. While much effort has gone into the development of
these statistical methods, the approach itself is rarely ques-
tioned. Perhaps, rather than evaluating data on the basis of a
failure to detect statistically significant HW deviations, it should
be demonstrated that any detected departure from HWE is
below a critical threshold, allowing one to assume that the HW
model still applies.
The field of bioequivalence clinical trials has shown that
classical difference testing may not be optimal for assessing
absence of association when dealing with the modest effects
that characterize departure from the HW model in genetic epi-
demiology studies. Whereas difference testing returns the
probability of observing a difference by chance, equivalency
testing returns the probability of observing a lack of departure
(or a modest departure) by chance. Equivalence testing is often
implemented as two one-sided tests: one returns the probabil-
ity of observing a lack of difference if the actual departure is
positive (e.g., homozygote excess), and the other returns the
probability of a lack of difference if the actual departure is
respectively negative (heterozygote excess). In the future, a
more natural approach to the HW testing may better quantify
the extent to which data do not depart from HWEP.
2. To generate a population-level PCA plot using GenAlEx, pop-
ulation allele frequencies are calculated and a genetic distance
matrix is derived from the frequency distributions for each
locus. In this case Nei’s unbiased genetic distances (95)
were used, but Nei’s standard genetic distances (78) can also
be used. The PCA is based on the genetic distance between
population groups. For a PCA plot at the individual level (see
Fig. 4), a matrix of genetic distances is computed from the
raw genotype data, and the PCA plot is generated from the
between-individual distance matrix.
 13 Analytical Methods for Immunogenetic Population Data 241

3. When creating a Structure project with this file, indicate the

number of individuals (694), the ploidy of the data (2, dip-
loid), the number of loci (2), the value provided for missing
data (−1), the presence of a header row of marker names, the
special format for the file including all data for each individual
on a single line, the inclusion of a column for the sample ID of
each sampled individual, and the inclusion of a column identifying
the population origin for each sampled individual. Then, create
a parameter set that defines the ancestry model (admixture),
the allele-frequency model (independent), and the run-length
in terms of the length of Burnin Period (50,000), and the
number of Monte Carlo Markov Chain (MCMC) reps after
Burnin (50,000). For other datasets, these last two values will
need to be determined empirically, by observing the number of
reps necessary for the alpha-value to converge. To start a
Structure run, specify the number of clusters (K) assumed for
the data (2 or 4). Group the resulting Bar Plot by population
ID to generate results similar to those in Fig. 5. Because
Structure clustering is accomplished via Bayesian inference,
Structure should be run multiple times for datasets that include
large numbers of individuals at many markers, and the results
compared for overall trends.
4. When using NEIGHBOR, researchers should always use the
Jumble (J) option to randomize the input order of taxa.

Acknowledgments

This work was supported by National Institutes of Health (NIH)

grants U01AI067068 (JAH, SJM) and U19 AI067152 (PAG)
awarded by the National Institute of Allergy and Infectious
Diseases (NIAID) and by NIH/NIAID contract AI40076 (RMS,
GT). The content is solely the responsibility of the authors and
does not necessarily represent the official views of the National
Institute of Allergy and Infectious Diseases or the National
Institutes of Health.

References
1. R Core Development Team (2008) R: A lan- 3. Lancaster AK, Single RM, Solberg OD, Nelson
guage and environment for statistical comput- MP, Thomson G (2007) PyPop update—a soft-
ing. R Foundation for Statistical Computing, ware pipeline for large-scale multilocus popula-
Vienna, Austria tion genomics. Tissue Antigens 69(s1):192–197
2. Li W, Hu Z, Jiang W (2010) An alphabetic list 4. Solberg OD, Mack SJ, Lancaster AK, Single
of genetic analysis software. North Shore LIJ RM, Tsai Y, Sanchez-Mazas A, Thomson G
Research Institute. http://www.nslij-genetics. (2008) Balancing selection and heterogeneity
org/soft/. Accessed 7 Oct 2010 across the classical human leukocyte antigen
242 S.J. Mack et al.

loci: a meta-analytic review of 497 population 20. Piazza A (1975) Haplotypes and linkage disequi-
studies. Hum Immunol 69(7):443–464 libria from three-locus phenotypes. In: Kissmeyer-
5. Peakall R, Smouse PE (2006) GENALEX 6: Nielsen F (ed) Histocompatibility testing.
genetic analysis in Excel. Population genetic Munskgaard, Copenhagen, pp 923–927
software for teaching and research. Mol Ecol 21. Dempster AP, Laird NM, Rubin DB (1977)
Notes 6:288–295 Maximum likelihood from incomplete data via
6. Lynch M, Milligan BG (1994) Analysis of pop- the EM algorithm. J R Stat Soc 39:1–38
ulation genetic structure with RAPD markers. 22. Ott J (1977) Counting methods (EM algorithm)
Mol Ecol 3:91–99 in human pedigree analysis: linkage and segrega-
7. Louis EJ, Dempster ER (1987) An exact test tion analysis. Ann Hum Genet 40:443–454
for Hardy-Weinberg and multiple alleles. 23. Yasuda N (1978) Estimation of haplotype fre-
Biometrics 43(4):805–811 quency and linkage disequilibrium parameter
8. Levene H (1949) On a matching problem aris- in the HLA system. Tissue Antigens
ing in genetics. Ann Math Stat 20(1):91–94 12:315–322
9. Emigh TH (1980) A comparison of tests for 24. Morton NE, Simpson SP, Lew R, Yee S (1983)
Hardy-Weinberg equilibrium. Biometrics Estimation of haplotype frequencies. Tissue
36(4):627–642 Antigens 22(4):257–262
10. Guo SW, Thompson EA (1992) Performing 25. Hawley ME, Kidd KK (1995) HAPLO: a pro-
the exact test of Hardy-Weinberg proportion gram using the EM algorithm to estimate the
for multiple alleles. Biometrics 48(2): frequencies of multisite haplotypes.
361–372 J Hered 86:409–411
11. Huber M, Chen Y, Dinwoodie I, Dobra A, 26. Long JC, Williams RC, Urbanek M (1995) An
Nicholas M (2006) Monte Carlo algorithms E-M algorithm and testing strategy for
for Hardy-Weinberg proportions. Biometrics multiple-locus haplotypes. Am J Hum Genet
62:49–53 56:799–810
12. Yuan A, Bonney GE (2003) Exact test of 27. Fallin D, Schork NJ (2000) Accuracy of haplo-
Hardy-Weinberg equilibrium by Markov chain type frequency estimation for biallelic loci, via
Monte Carlo. Math Med Biol 20:327–340 the expectation-maximization algorithm for
13. Ebrahimi N, Bilgili D (2007) A new method of unphased diploid genotype data. Am J Hum
testing for Hardy-Weinberg equilibrium and Genet 67:947–959
ordering populations. J Genet 86:1–7 28. Tishkoff SA, Pakstis AJ, Ruano G, Kidd KK
14. Metropolis N, Rosenbluth AW, Rosenbluth (2000) The accuracy of statistical methods for
MN, Teller AH, Teller E (1953) Equations of estimation of haplotype frequencies: an example
state calculations by fast computing machines. from the CD4 locus. Am J Hum Genet
J Chem Phys 21:1087–1092 67:518–522
15. Chen JJ, Hollenbach JA, Trachtenberg EA, 29. Kirk KM, Cardon LR (2002) The impact of
Just JJ, Carrington M, Rønningen KS, Begovich genotyping error on haplotype reconstruction
A, King MC, McWeeney SK, Mack SJ, Erlich and frequency estimation. Eur J Hum Genet
HA, Thomson G (1999) Hardy-Weinberg test- 10:616–622
ing for HLA class II (DRB1, DQA1, DQB1 30. Single RM, Meyer D, Hollenbeck J, Nelson M,
and DPB1) loci in 26 human ethnic groups. Noble JA, Erlich HA, Thomson G (2002)
Tissue Antigens 54:533–542 Haplotype frequency estimation in patient
16. Hernández JL, Weir BS (1989) A disequilib- populations: the effect of departures from
rium coefficient approach to Hardy-Weinberg Hardy-Weinberg proportions and collapsing
testing. Biometrics 45(1):53–70 over a locus in the HLA region. Gen Epidemiol
22:186–195
17. Chen JJ, Thomson G (1999) The variance for
the disequilibrium coefficient in the individual 31. Stephens M, Smith NJ, Donnelly P (2001) A
Hardy-Weinberg test. Biometrics new statistical method for haplotype recon-
55:1269–1272 struction from population data. Am J Hum
Genet 68:978–989
18. Barnetche T, Gourraud PA, Cambon-Thomsen
A (2005) Strategies in analysis of the genetic 32. Niu T, Qin ZS, Xu X, Liu JS (2002) Bayesian
component of multifactorial diseases; biostatis- haplotype inference for multiple linked single-
tical aspects. Transpl Immunol 14(3–4): nucleotide polymorphisms. Am J Hum Genet
255–266 70:157–169
19. Guan Y, Stephens M (2008) Practical issues in 33. Qin ZS, Niu T, Liu JS (2002) Partition-
imputation-based association mapping. PLoS ligation-expectationmaximization algorithm
Genet 4(12):e1000279 for haplotype inference with singlenucleotide
 13 Analytical Methods for Immunogenetic Population Data 243

polymorphisms. Am J Hum Genet 71: alisation of interallelic disequilibrium between

1242–1247 multiallelic markers. BMC Bioinformatics
34. Stephens M, Donnelly P (2003) A comparison 7:227–237
of Bayesian methods for haplotype reconstruc- 50. Shin J-H, Blay S, McNeney B, Graham
tion from population genotype data. Am J J (2006) LDheatmap: an R function for graph-
Hum Genet 73:1162–1169 ical display of pairwise linkage disequilibria
35. Niu T (2004) Algorithms for inferring haplo- between single nucleotide polymorphisms.
types. Genet Epidemiol 27(4):334–347 J Stat Soft 16 Code Snippet 3
36. Slatkin M, Excoffier L (1996) Testing for 51. Ewens WJ (1972) The sampling theory of
linkage disequilibrium in genotypic data using selectively neutral alleles. Theor Popul Biol
the Expectation-Maximization algorithm. 3(1):87–112
Heredity 76:377–383 52. Watterson G (1978) The homozygosity test of
37. Tishkoff SA, Kidd KK (2004) Implications of neutrality. Genetics 88:405–417
biogeography of human populations for ‘race’ 53. Slatkin M (1994) An exact test for neutrality
and medicine. Nat Genet 36(Suppl 11):S21–S27 based on the Ewens sampling distribution.
38. Robinson J, Waller MJ, Fail SC, Marsh SG Genet Res 64:71–74
(2006) The IMGT/HLA and IPD databases. 54. Slatkin M (1996) A correction to the exact test
Hum Mutat 27:1192–1199 based on the Ewens sampling distribution.
39. Gourraud PA, Gagne K, Bignon JD, Cambon- Genet Res 68:259–260
Thomsen A, Middleton D (2007) Preliminary 55. Salamon H, Klitz W, Easteal S, Gao X, Erlich
analysis of a KIR haplotype estimation algo- HA, Fernandez-Vina M, Trachtenberg EA
rithm: a simulation study. Tissue Antigens (1999) Evolution of HLA class II molecules:
69(Suppl 1):96–100 allelic and amino acid site variability across
40. Yoo YJ, Tang J, Kaslow RA, Zhang K (2007) populations. Genetics 152:393–400
Haplotype inference for present-absent geno- 56. Conover W (1980) Practical nonparametric
type data using previously identified haplotypes statistics. Wiley, New York
and haplotype patterns. Bioinformatics 57. Chakravarti A (1991) Information content of
23(18):2399–2406 the Cen tre d’Etude du Polymorphisme
41. Lewontin RC (1964) The interaction of selec- Humain (CEPH) family structures for linkage
tion and linkage. I. General considerations; studies. Hum Genet 87:721–724
heterotic models. Genetics 49:49–67 58. Wright S (1951) The genetic structure of pop-
42. Hill WG, Robertson A (1968) Linkage dis- ulations. Ann Eugen 15:323–354
equilibrium in finite populations. Theor Appl 59. Nei M (1977) F-statistics and analysis of gene
Genet (Der Ziichter) 38:226–231 diversity in subdivided populations. Ann Hum
43. Klitz W, Stephen JC, Grote M, Carrington M Genet 41:225–233
(1995) Discordant patterns of linkage disequi- 60. Wright S (1978) Evolution and the genetics of
librium of the peptide transporter loci within populations, vol 4. The University of Chicago
the HLA class II region. Am J Hum Genetics Press, Chicago, Variability Within and Among
57:1436–1444 Natural Populations
44. Cramer H (1946) Mathematical methods 61. Slatkin M (1991) Inbreeding coefficients and
of statistics. Princeton University Press, coalescence times. Genet Res 58:167–175
Princeton, NJ 62. Weber JL, Wong C (1993) Mutation of human
45. Cohen J (1988) Statistical power analysis for short tandem repeats. Hum Mol Genet
the behavioral sciences. Erlbaum, Hillsdale, NJ 2(8):1123–1128
46. Hedrick PW (1987) Gametic disequilibrium 63. Gyapay G, Morissette J, Vignal A, Dib C,
measures: proceed with caution. Genetics Fizames C, Millasseau P, Marc S, Bernardi G,
117(2):331–341 Lathrop M, Weissenbach J (1994) The 1993–
47. Abecasis GR, Cookson WO (2000) GOLD- 1994 Genethon human genetic linkage map.
graphical overview of linkage disequilibrium. Nat Genet 7:246–339
Bioinformatics 16:182–183 64. Slatkin M (1995) A measure of population
48. Barrett JC, Fry B, Maller J, Daly MJ (2005) subdivision based on microsatellite allele fre-
Haploview: analysis and visualization of LD quencies. Genetics 139:457–462
and haplotype maps. Bioinformatics 65. Reynolds J, Weir BS, Cockerham CC (1983)
21(2):263–265 Estimation for the coancestry coefficient: basis
49. Gaunt TR, Rodriguez S, Zapata C, Day IN for a short-term genetic distance. Genetics
(2006) MIDAS: software for analysis and visu- 105:767–779
244 S.J. Mack et al.

66. Weir BS, Cockerham CC (1984) Estimating 83. Middleton D, Gonzalez F, Fernandez-Vina M,
F-statistics for the analysis of population struc- Tiercy JM, Marsh SG, Aubrey M, Bicalho MG,
ture. Evolution 38:1358–1370 Canossi A, Carter V, Cate S, Guerini FR,
67. Hedrick PW (2005) A standardized genetic Loiseau P, Martinetti M, Moraes ME, Morales
differentiation measure. Evolution V, Perasaari J, Setterholm M, Sprague M,
59:1633–1638 Tavoularis S, Torres M, Vidal S, Witt C,
68. Pearson K (1901) “On lines and planes of Wohlwend G, Yang KL (2009) A bioinformat-
closest fit to systems of points in space” (PDF). ics approach to ascertaining the rarity of HLA
Phil Mag 2(6):559–572 alleles. Tissue Antigens 74:480–485
69. Cox TF, Cox MAA (2001) Multidimensional 84. Cadavid LF, Watkins DI (1997) Heirs of the
Scaling, 2nd edn. Chapman and Hall, Boca jaguar and the anaconda: HLA, conquest and
Raton, FL disease in the indigenous populations of the
Americas. Tissue Antigens 6:702–711
70. Borg I, Groenen P (2005) Modern multidi-
mensional scaling: theory and applications, 2nd 85. Erlich HA, Mack SJ, Bergström T, Gyllensten
edn. Springer, New York UB (1997) HLA class II alleles in Amerindian
populations: implications for the evolution of
71. Pritchard JK, Stephens M, Donnelly P (2000) HLA polymorphism and the colonization of
Inference of population structure using multilo- the Americas. Hereditas 127(1–2):19–24
cus genotype data. Genetics 155(2):945–959
86. Sokal R, Michener C (1958) A statistical
72. Hubisz MJ, Falush D, Stephens M, Pritchard method for evaluating systematic relationships.
JK (2009) Inferring weak population structure Univ Kansas Sci Bull 38:1409–1438
with the assistance of sample group informa-
tion. Mol Ecol Resour 9(5):1322–1332 87. Murtagh F (1984) Complexities of hierarchic
clustering algorithms: the state of the art.
73. Rosenberg NA, Pritchard JK, Weber JL, Cann Comput Stat Quart 1:101–113
HM, Kidd KK, Zhivotovsky LA, Feldman MW
(2002) Genetic structure of human populations. 88. Bergström TF, Josefsson A, Erlich HA,
Science 298(5602):2381–2385 Gyllensten UB (1997) Analysis of intron
sequences at the class II HLA-DRB1 locus:
74. Karypis G. (2002) CLUTO: a clustering toolkit. implications for the age of allelic diversity.
Technical Report 02-017. University of Hereditas 127(1–2):1–5
Minnesota, Minneapolis, MN
89. Bergström TF, Josefsson A, Erlich HA,
75. Felsenstein J (1989) PHYLIP—Phylogeny Gyllensten U (1998) Recent origin of HLA-
Inference Package (Version 3.2). Cladistics DRB1 alleles and implications for human evo-
5:164–166 lution. Nat Genet 18(3):237–242
76. Felsenstein, J. (2005) PHYLIP (Phylogeny 90. Felsenstein J (1985) Confidence limits on phy-
Inference Package) version 3.6 (Distributed by logenies: an approach using the bootstrap.
the author). Department of Genome Sciences, Evolution 39:783–791
University of Washington, Seattle
91. Penny D, Hendry MD (1985) Testing meth-
77. Saitou N, Nei M (1987) The neighbor-joining ods of evolutionary tree construction. Cladistics
method: a new method for reconstructing phy- 1:266–278
logenetic trees. Mol Biol Evol 4(4):406–425
92. Cavalli-Sforza LL, Menozzi P, Piazza A (1994)
78. Nei M (1972) Genetic distance between popu- The history and geography of human genes.
lations. Am Nat 106:283–292 Princeton University Press, Pinceton, NJ
79. Cavalli-Sforza LL, Edwards AFW (1967) 93. Cavalli-Sforza LL, Feldman MW (2003) The
Phylogenetic analysis: models and estimation application of molecular genetic approaches to
procedures. Am J Hum Genet 19:233–257 the study of human evolution. Nat Genet
80. Hughes AL, Nei M (1988) Pattern of nucle- 33:266–275
otide substitution at MHC class I loci reveals 94. Mack SJ, Erlich HA (2007) Population relation-
overdominant selection. Nature 335:167–170 ships as inferred from classical HLA genes. 13th
81. Hughes AL, Yeager M (1998) Natural selection International histocompatibility workshop
at major histocompatibility complex loci of ver- anthropology/human genetic diversity joint
tebrates. Ann Rev Genet 32:415–435 report. In: Hansen JA (ed) Immunobiology of
82. Sanchez-Mazas A, Fernandez-Viña M, the human MHC: Proceedings of the 13th
Middleton D, Hollenbach JA, Buhler S, Di D, international histocompatibility workshop and
Rajalingam R, Dugoujon JM, Mack SJ, conference, vol I. IHWG, Seattle, pp. 747–757
Thorsby E (2011) Immunogenetics as a tool in 95. Nei M (1978) Estimation of average heterozy-
anthropological studies. Immunology gosity and genetic distance from a small num-
133(2):143–164 ber of individuals. Genetics 89:583–590
 Chapter 14

Analytical Methods for Disease Association Studies

with Immunogenetic Data
Jill A. Hollenbach, Steven J. Mack, Glenys Thomson,
and Pierre-Antoine Gourraud

Abstract
Disease association studies involving highly polymorphic immunogenetic data may involve analyses at one
or many units of analysis, including amino acid, allele, genotype and haplotype levels, as well as consideration
of gene–gene or gene–environment interactions. The selection of the appropriate statistical tests is critical
and will be dependent on the nature of the dataset (e.g., case-control vs. family data) as well as the specific
research hypotheses being tested. This paper describes the various study and analysis categories used for
such analyses, including the advantages and limitations of such techniques.

Key words: HLA, KIR, Immunogenetic, Data analysis, Disease association, Case-control, Family

1. Introduction

Statistics may be defined as “a body of methods for making wise

decisions in the face of uncertainty.”
W.A. Wallis

There is no shortage of statistical methods and tools available for

the analysis of genetic data; in disease association studies these pro-
vide the results upon which we draw inferences about human
health, and often influence the direction of future research, clinical
protocols, and public health policy. However, just as Wallis sug-
gested statistics assist us in making wise decisions in an uncertain
world, it is also imperative that we first make wise decisions about

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_14, © Springer Science+Business Media New York 2012

245
246 J.A. Hollenbach et al.

how we apply statistical methods. Unless utilized properly, statistics

can provide misleading results in disease association studies, con-
suming precious resources and often leading investigators down
blind alleys.
The human leukocyte antigen (HLA) loci serve as a model for
the study of genetic variation in human health and disease (1).
Their critical role in disease predisposition in a variety of autoim-
mune, infectious diseases and cancers (2–25), as well as transplant
outcomes, has long been recognized. Likewise, a role for the killer
cell immunoglobulin-like receptors (KIR) loci and their HLA
ligands in autoimmune diseases (26–29) and infectious diseases
(HIV and Hepatitis C (30, 31)) as well as solid organ and
hematopoietic stem cell transplantation (32–34) and pregnancy
(35, 36) is now established. However, the high levels of polymor-
phism associated with these loci mean that particular care must be
taken when analyzing data for disease association studies, in order
to avoid confounding results. In this chapter, we discuss the appro-
priate application of common statistical methods in the analysis of
highly polymorphic immunogenetic data. While most of these
analyses can be undertaken with either customized or commercial
computer software, no existing programs, either individually, or in
aggregate, can handle the breadth and complexity of the analyses
needed to detect primary and secondary disease genes in polymor-
phic immunogenetic regions. The complex linkage disequilibrium
(LD) patterns common to many immunogenetic data are a major
complicating factor in deciphering specific causal genetic variants.
Further, many analysis programs do not work with the high degree
of polymorphism found in some regions, for example HLA, while
others only allow a limited number of highly polymorphic loci to
be evaluated.
The effects of environmental factors, age of onset and gender-
specific effects, undetected heterogeneity of disease, gene–gene
and gene–environment interactions further complicate our attempts
to detect predisposing (PRE) genes for complex diseases. We thus
strongly recommend a multistrategy approach using a variety of
complementary methods. The data should be studied carefully to
understand differences in significance obtained from different
methods. There is no “best practice” for the types of analyses
needed for these projects. The power of different methods will
vary depending on the specific genetic and environmental features
of the disease under consideration; for most complex diseases this
involves many unknown factors. Although we must rely heavily on
computers, in the final analysis of multiple effects in a genetic
region and/or interaction or independent effects between unlinked
genes, manipulation of the data by the individual investigator must
play a crucial role.
 14 Analytical Methods for Disease Association Studies with Immunogenetic Data 247

2. Study
and Analysis
Categories
Case-control studies are a critical, population-based tool for epide-
2.1. Case-Control miological studies, including genetic association studies. These
Studies studies generally involve samples of unrelated individuals with the
(disease) phenotype of interest and a corresponding control sample
of unrelated, unaffected (or randomly ascertained) individuals
drawn from the same (ethnic and preferably geographic) popula-
tion. While a discussion of study design, including power calcula-
tions and sample size considerations is out of the scope of this
chapter, a general recommendation in immunogenetic studies is to
have a minimum of 200 chromosomes (100 individuals) in each of
the case and control cohorts, although much larger sample sizes
are preferable. Among the advantages of case-control studies is
elimination of the necessity to collect data from family members,
which is often logistically difficult and costly. However, case-con-
trol studies can be very sensitive to population stratification within
the sample cohorts, a particularly important issue in immunoge-
netic data, where allele frequency distributions can vary consider-
ably between human ethnic groups. If the samples are not collected
with scrupulous attention to homogeneity of ancestry background,
investigators run the risk of misinterpreting genetic difference
between cases and controls. In these cases, heterogeneity between
cases and controls due to allele frequency differences related to
population stratification may be mistaken for association with a
particular locus.
Population analyses, including calculations of gene (allele)
frequencies, tests for fit to expectations under Hardy–Weinberg
Equilibrium (HWE), and estimations of haplotype frequencies,
are the first step in any study involving immunogenetic data,
and will provide additional means to assess any underlying pop-
ulation substructure, as well as additional units of analysis on
which association analyses will be performed. These calculations
are detailed in Chap. 13. Many computer software programs are
available to perform such calculations on genetic data, but as
mentioned in Chap. 13, care must be taken to utilize tools
either specifically designed for immunogenetic data (e.g., the
PYPOP analytical package (http://www.pypop.org)) or those
that are able to handle the high levels of polymorphism characteristic
of these loci.

2.1.1. Contingency Tables 1. Basic construction and analysis of contingency tables, with par-
ticular attention to sparse cells.
A contingency table, or cross-tabulation, is used to test the
difference, or independence, of frequency distributions for cat-
egorical variables. Tests for heterogeneity between specific
248 J.A. Hollenbach et al.

groups (e.g., cases and controls) for immunogenetic data can

be performed using contingency table testing and a standard
Chi-square (χ2) measure. Software implementations of contin-
gency table analyses are available through a variety of commercial
packages (e.g., SAS (37), SPSS (38), GraphPad Prism (http://
www.grahpad.com)); through open source statistical software
packages (e.g., the “chisq.test” function in the R language for
statistical computing (39)); and various online calculators (e.g.,
http://www.physics.csbsju.edu/stats/contingency_NROW_
NCOLUMN_forqm.html).
A contingency table is always constructed utilizing raw
counts, rather than frequency data. The preferred approach in
this type of analysis for highly polymorphic immunogenetic
data is to first examine heterogeneity in overall allele frequency
distributions in cases vs. controls for a particular locus, using a
2 × k row by column (r × c) contingency table. Analyses of these
data historically relied on series of 2 × 2 tests for each allele
observed due to the presence of untyped (the so-called blank)
alleles; due to the often high frequencies of blanks in initial
studies, allele counts and frequencies could not easily be calcu-
lated. As described in Chap. 13, the widespread use of DNA
typing in most immunogenetic studies now allows for direct
ascertainment of allele frequencies.
Analysis based on a series of 2 × 2 tests requires corrections
for multiple comparisons (usually utilizing the Bonferroni
method), with a correction factor minimally equivalent to the
number of alleles tested. However, when significance of the
association of all individual alleles is assessed with the a priori
knowledge of overall heterogeneity at the locus, it is not neces-
sary to correct for multiple comparisons in subsequent 2 × 2
tests intended to identify the allele(s) with significant contribu-
tions to the overall deviation at the locus. An example of a 2 × k
contingency table is given in Table 1 and a related 2 × 2 table
for one allele is given in Table 2.
Rare alleles pose a particular challenge in traditional asso-
ciation tests in disease studies that use the χ2 test for goodness-
of-fit. Although the χ2 test has been the standard approach for
testing fit at the overall locus-level, this test can lead to false
acceptance or rejection of the null hypothesis when the
expected genotype counts in a contingency table are small (aka,
“sparse cells”), as discussed in Chap. 13 with regard to Hardy–
Weinberg (HW) testing; the χ2 test is inappropriate if any
expected count is less than one or if the expected count is less
than five in more than 20% of all cells in a contingency table.
It is not unusual for 30 or more alleles to be observed at the HLA
loci for example, with a wide range of frequencies, resulting in
many sparse cells. Although Yates’ correction for continuity is
sometimes applied to account for numerous sparse cells, this
14 Analytical Methods for Disease Association Studies with Immunogenetic Data 249

Table 1
A 2 ¥ k contingency table for case-control data

Allele Case (count) Control (count) Row total

0101 75 29 104
0102 5 5 10
0301 59 56 115
0401 6 15 21
0701 9 25 34
0701 6 11 17
0801 73 39 112
1101 39 35 74
1103 9 4 13
1104 40 24 64
Binned 33 25 57
Total 354 268 622

Table 2
A 2 ¥ 2 contingency table for allele 0101

Allele Case (count) Control (count) Row total

0101 75 29 104
Others (not 0101) 279 239 517
Total 354 268 622

may be overly conservative, and the preferred method is to

combine low frequency alleles into a single category, often
referred to as the “binned” category (Table 1). A conservative
(and recommended) approach is to combine alleles with an
expected value of less than five in cases or controls into the
binned category prior to calculation of the χ2 statistic. The χ2
statistic for a contingency table analysis of case-control data for
a genetic association is calculated as:

(Oi - Ei )2
c2 = å ,
Ei
where
Oi = the observed count of allele i
250 J.A. Hollenbach et al.

Ei = the expected count of allele i

And the derived values are summed over all cells in the tables.
The expected count for each cell in the r × c table is calculated as

(row total allele i) ´ (column total) ,

2n
where
Column total = sum of the counts in the column
Row total = number observations of allele i in all subjects
n = number individuals (cases + controls)
2n = number chromosomes (cases + controls)
The degrees of freedom (df) for the goodness of fit χ2 anal-
ysis are calculated from the number of alleles with expected
values in cases and controls of five or greater, plus the combined
category, −1 (i.e., k − 1). A p-value is obtained by comparing the
test statistic to the χ2 distribution for the appropriate degrees of
freedom.
In the example given in Table 1, we obtain a value of
χ = 39.71, with df = 10. This indicated significant differences
2

between cases and controls at the overall locus level, with

p < 0.001. In the example in Table 2 for a 2 × 2 table for an
individual allele, χ2 = 11.88, df = 1, p < 0.001.
These analyses can be performed at the allele level, as
shown in the example, as well as for genotypes and haplotypes.
In a haplotype-level analysis, population-level haplotype fre-
quencies are first estimated (see Chap. 13) and the frequency
estimates are converted to counts based on the sample size.
Particular care must be taken when working with estimated
haplotypes to combine the many rare types that are typically
estimated, as well as to not over-interpret the data.
Summary: Basic steps for contingency table analysis in
case-control studies with immunogenetic data
(a) Construct a 2 × k table of allele (or genotype or haplotype)
counts for cases and controls (see Table 1 example).
(b) Combine all alleles with expected values of four or less in
cases or controls into a common “binned” category.
(c) Calculate the χ2 test statistic for the table and assess
significance.
(d) If results are significant at the overall locus level, perform
additional testing using 2 × 2 contingency tables for each
allele (expected counts of five or more only) against all
other alleles.
14 Analytical Methods for Disease Association Studies with Immunogenetic Data 251

2. Relative Predispositional Effects (RPE).

While contingency table testing will reveal whether the
overall allele frequency distributions at a locus differ significantly
between cases and controls, identifying all allele(s) that con-
tribute significantly to these differences can be difficult. The
RPE method can be used to identify all heterogeneity in dis-
ease risk at the primary disease gene; alleles, haplotypes, or
genotypes with the strongest predisposing (PRE) or protective
(PRO) effects are sequentially removed from the analysis until
no further heterogeneity in risk effects is seen. Determining
the order in which haplotypes, genotypes or alleles are sequen-
tially removed requires interplay between the contribution to
the χ2 heterogeneity test, the ORs (see below) or Patient/
Control (P/C) ratio (40) and the control frequencies of the
allele, haplotype, or genotype. In the simplest application of
this method, the allele with the greatest contribution to the
overall χ2 values is removed, and the test statistic is recom-
puted, with degrees of freedom correspondingly reduced by
one. If the new χ2 is found to be statistically significant at the
test level, the process is repeated, with each subsequent analysis
involving the removal of the greatest contributor to the overall
χ2 from the previous analysis, until the test statistic is no longer
found to be significant. In some cases more than one allele can
be removed at a time, for example, where there are two PRE
alleles with strong effects or a PRE and PRO allele when prior
evidence has validated both as significant and strong effects.
The application of the RPE method to HLA data for juve-
nile idiopathic arthritis (JIA) (41, 42) is summarized in Table 3.
The effect of individual alleles is indicated as either predispos-
ing, neutral, or intermediate (INT) or protective. Rare alleles
which cannot be placed into one of these categories either
by the RPE method or all pairwise comparisons below have
no effect given. Significant p-values for these comparisons
are shaded gray.
The overall analysis of the complete dataset shows consid-
erable heterogeneity in risk (p < 1.1E − 27). DRB1*0801
(PRE), DRB1*1501, and DRB1*0701 (PRO) are the stron-
gest effects (the set of alleles labeled 1 in column 1). Repeat
analysis after the removal of these three alleles still gives a
highly significant result (p < 4.1E − 10), with DRB1*1104
(PRE) and DRB1*0401 (PRO) as the strongest effects (set
labeled 2 in column 1). Note that a strong argument could be
made for deleting these alleles also at the previous round, but
this does not alter the outcome. With the removal of these
strong effects, there is only minimal evidence of remaining risk
heterogeneity (p < 0.02), with DRB1*1103 (p < 0.01) (PRE)
and DRB1*0103 (p < 0.02) (PRO) the strongest effects
252 J.A. Hollenbach et al.

Table 3
JIA-OP HLA-DRB1 allele data ranked by odds ratio (OR)

RPEa Ab Bc DRB1 Patient Control Chi-square (c2) p-Valued Effect OR CIe CIe

Ix *1103 12 1 6.80 0.009 PRE 9.40 1.22 72.49

1 I Ix *0801 102 13 48.61 3.1E − 12 PRE 6.90 3.83 12.43
2 I Ix *1104 57 11 20.71 5.3E − 06 PRE 4.26 2.21 8.20
*0403 9 3 1.68 0.20 2.33 0.63 8.65
3 II IIx *1301 90 38 9.99 0.002 INT 1.95 1.31 2.90
IIx *0102 9 5 0.35 0.55 INT 1.39 0.46 4.18
II IIx *1101 60 36 1.42 0.23 INT 1.31 0.85 2.02
IIx *0901 9 6 0.08 0.78 INT 1.16 0.41 3.28
II IIx *0101 74 50 0.52 0.47 INT 1.16 0.79 1.69
II IIx *0301 89 61 0.50 0.48 INT 1.14 0.81 1.62
IIx *1201 10 8 0.006 0.94 INT 0.96 0.38 2.46
II IIx *1302 28 23 0.05 0.82 0.94 0.53 1.64
*1303 10 9 0.11 0.74 0.86 0.34 2.12
f
Binned 27 27 0.92 0.34 0.76 0.44 1.31
*1601 6 8 1.05 0.30 0.58 0.20 1.67
*1401 11 18 4.05 0.04 0.46 0.22 0.99
*1502 5 10 3.26 0.07 0.38 0.13 1.12
III IIIx *0404 7 16 6.34 0.01 PRO 0.33 0.14 0.81
1 III IIIx *1501 38 80 28.24 1.1E − 07 PRO 0.33 0.22 0.49
1 III IIIx *0701 30 65 23.92 1.0E − 06 PRO 0.33 0.21 0.51
2 III IIIx *0401 21 47 18.10 2.1E − 05 PRO 0.33 0.19 0.55
3 IIIx *0103 4 11 5.42 0.02 PRO 0.28 0.09 0.87
Total 708 546 182.1 1.1E − 27 PRO
a
Numbers denote the order of removal due to largest effect(s) in the relative predispositional effect (RPE) analysis
b
Set A: Based on pairwise allele comparisons, the common alleles are divided into mutually exclusive, and significantly
different, predisposing (PRE) (I), neutral (intermediate (INT)) (II), and protective (PRO) (III) categories for use later
in amino acid comparisons (described below)
c
Set B: The sets I, II, and III above are expanded (indicated by Ix, IIx, and IIIx) to include rare alleles, while excluding
those alleles which do not clearly fall into one of the three risk categories
d
The individual p-values are biased as the assumption of a 1 df χ2 is incorrect, and conservative; the p-values can be used
however for a relative ranking of the allelic effects
e
The upper and lower 95% confidence intervals (CIs) for the odds ratio (OR) are given (see below)
f
The binned category consists of all alleles with an expected value <5 under the χ2 test of heterogeneity of patient and
control allele counts
14 Analytical Methods for Disease Association Studies with Immunogenetic Data 253

(set labeled 3 in column 1). Note that these latter p-values

would not be significant with corrections for multiple compari-
sons of either the overall tests, or those for individual alleles.
Notwithstanding, we mention these results, since our principal
aim is to detect heterogeneity that may be relevant to detecting
additional disease genes in a genetic region or identifying the
causal amino acids in the primary disease gene.
3. Conditional Haplotype Method (CHM).
With immunogenetic disease associations, the high level of
LD between many of the genes means that multiple disease
associations are often observed; some of these may indicate a
causative genetic factor, and others may be due to LD with this
causative factor. In order to account for LD within a genetic
region, stratification by the disease-associated alleles using the
CHM can be employed (43). We stress that it is important to
identify all heterogeneity in disease risk at the primary disease
PRE gene, including relatively weak effects, before proceeding
with conditional analyses to detect secondary genetic effects.
Otherwise, spurious results may be seen resulting from hetero-
geneity in relative risk effects at the primary disease gene at the
allele, haplotype, or genotype levels. If all genes directly
involved in disease susceptibility have been identified, then the
relative frequencies of alleles at the other loci on high-risk hap-
lotypes should be the same in cases and controls; similarly for
INT and PRO haplotypes. While fit to these expectations does
not exclude the possibility that other genes in the genetic com-
plex are involved in disease, lack of fit unequivocally shows that
all disease-PRE genes in the region have not been identified.
For the CHM, we denote the primary disease locus by A
(or the primary plus secondary loci that have been identified),
that is, all putative disease PRE loci in the region, and alleles or
haplotypes thereof by Ai, i = 1,2,…,iA, and a linked marker
locus by B, with alleles Bk, k = 1,2,…,kB, then, under the null
that the A locus defines all disease predisposition in the
region:
f pat (Ai - Bk) / f pat (Ai - Bl ) = f con (Ai - Bk) / f con (Ai - Bl ),

where fpat(·) and fcon(·) represent patient and control frequen-

cies. That is, although the frequencies of the haplotypes Ai − Bk
and Ai − Bl will differ between patients and controls, the rela-
tive frequency of their ratios will be the same in patients and
controls, for each Ai.
Inequality of these relative frequencies in patients and con-
trols is expected if the allele or SNP or haplotypes thereof
under study does not include all genes involved in the disease
process and in LD with the marker loci. While fit to these
expectations does not exclude the possibility that other genes
254 J.A. Hollenbach et al.

in the region are involved in disease, lack of fit unequivocally

shows that all disease-PRE genes in the region have not been
identified (provided that stratification effects have not pro-
duced spurious results).
4. Amino acid-level analyses.
In some cases, the analytical challenges posed by rare alleles
can be avoided by analyzing the data at the amino-acid level.
Immunogenomic studies have generally focused on allele- and
locus-level polymorphisms, considering individual amino-acid
or nucleotide polymorphisms only as they pertain to specific
hypotheses. However, HW, LD, association, and selective neu-
trality analyses can be performed entirely at the amino-acid
level, considering each amino-acid polymorphism indepen-
dently of the others that constitute an allele sequence. The
issues surrounding rare alleles are attenuated in these analyses,
as the level of polymorphism at any individual amino-acid posi-
tion is not as high as at the allele or locus level. However, varia-
tion in the number of variants at each amino-position poses
challenges for LD estimates and attempts to correlate statistics
between positions.
5. Sequence feature variant type analysis.
Recently, a novel approach to genetic association analyses
has been developed by subdividing genes/proteins into bio-
logically relevant structural (e.g., beta-strand 1) and functional
(e.g., peptide binding site) sequence features (SFs), and their
“alleles,” called variant types (VTs) (44). In SFVT analyses,
association tests are performed on the VTs of these biologically
relevant SFs.

2.1.2. Odds Ratios Odds ratios (OR) are used in disease association studies as means
and Relative Risk to describe the strength of association between two variables, and
reflect the ratio of odds of an outcome (e.g., disease) occurring in
one group (e.g., individuals with allele a) to the odds of it occur-
ring in another group (individuals without allele a). Relative risk
(RR) is a related but different measure that describes the risk of
having the outcome of interest relative to exposure (e.g., positive
or negative for allele a). In case-control studies, the OR is the pre-
ferred measure, due in part to its use in logistic regression (see next
section), allowing a more ready comparison of results derived from
different methods. Another important advantage to the OR is that
it is by its mathematical nature invertible, i.e., the OR for predis-
position to disease is the direct inverse of the OR for protection;
this is extremely important in terms of immunogenetic studies,
where alleles may be associated in either a PRO or PRE role. It is
not unusual to find both PRO and PRE alleles in a given disease
study for a given locus; the ability to invert ORs allows more direct
14 Analytical Methods for Disease Association Studies with Immunogenetic Data 255

Table 4
Generic 2 × 2 table

Allele Case (count) Control (count)

Y a b
Not Y c d

comparison of the relative strengths of predisposition vs. protec-

tion. In contrast, RR is not invertible. RR is more applicable to
prospective studies, but for retrospective studies OR is preferred.
1. Calculation of OR and RR.
Both OR and RR are calculated for each allele of interest
in a 2 × 2 table. The OR is a simple cross-product of a 2 × 2
table. A generic example is given in Table 4.
The OR for Table 4 is given by:
ad .
OR =
bc
The RR is given by
a / (a + b) .
RR =
c / (c + d)

For the example given in Table 2, OR = 2.22 and RR = 1.34.

2. Confidence intervals.
Values for OR and RR are only meaningful when presented
in terms of a corresponding confidence interval (CI). A
confidence interval provides a likely range of values for an esti-
mate of a particular parameter within a particular level of
significance. The standard probability level is generally set at
95%, meaning that there will be a 95% chance that the value of
interest (here, the OR or RR) will fall within that range. For
OR and RR, a CI that includes one is not considered statisti-
cally significant, as the null hypothesis in this test is OR or
RR = 1.
Derivation of the CI is approached in a variety of ways,
most of which require the use of computer software, and may
include resampling (bootstrap), likelihood or other statistical
methods. A full discussion of these methods is not possible
here, and their application will ultimately depend on the type
of data and analytical methods and software used by the inves-
tigator. Most statistical software packages compute basic CIs
256 J.A. Hollenbach et al.

Table 5
Example of input variables for logistic regression analysis

Covariable
Dependent
Independent variable Risk factor: Risk factor: Risk factor: Genetic:
variable (HLA (disease smoking (yes/ smoking (years)- smoking snp-
or KIR)a,b status)c no)-Binaryd Quantitativee (duration)-Ordinalf Categoricalg
0101,0401 0 1 15.8 1 1
0401,0501 0 1 10 2 2
0301,0401 1 1 2.5 3 3
0101,0501 1 0 1 2 4
a
Data can be analyzed at the allele, genotype, haplotype, or other user-defined level
b
User will need to specify the genetic model, e.g., recessive, codominant, etc.
c
By convention, 0 = control; 1 = case
d
Here, risk factor (smoking) is analyzed as present or absent. 0 = nonsmoker, 1 = smoker
e
Here, risk factor (smoking) is analyzed as a continuous quantitative variable, number of years smoked
f
Here, risk factor (smoking) is analyzed as an ordered discrete variable, number of years smoked categorized 1 = 0–5
years; 2 = 6–10 years; 3 = 11–20 years; 4 = more than 20 years
g
Here, SNP data categorized by genotype, e.g., 1 = AG; 2 = AT; 3 = GG; 4 = GT

according to Clopper and Pearson (45), often known as the

“exact” method.

2.1.3. Logistic Regression Analysis using logistic regression can also be used in case-control
studies. Logistic regression is a form of generalized linear modeling
for data with a dichotomous, or binary, outcome variable, such as
case-control data where the outcome is either “affected” or “unaf-
fected.” Among its advantages, logistic regression provides a means
to develop association models that include the contribution of
both quantitative and qualitative covariables, which can involve
cumbersome stratification procedures in contingency table testing;
logistic regression is a critical tool in the analysis of complex multi-
variate datasets. Example of potential types of input variables for a
logistic regression analysis with immunogenetic data are given in
Table 5. Attention must be paid to how both quantitative and
qualitative variables are coded, because interpretation of the ORs
can greatly vary. For example in Table 5, including number of years
smoking as a covariate will provide a single OR associated with
each increment of 1 year of smoking and assumes that the risk
increment is fully proportional to years smoking. This model may
not make sense, so in some cases it is more appropriate to recode
quantitative variables as binary (e.g., smoking yes/no in Table 5)
or discrete, ordinal (e.g., smoking duration, Table 5), rather than
continuous, variables.
A disadvantage of logistic regression analysis is a tendency to
overestimate Ors (46) in sample sizes <500, possibly leading to
 14 Analytical Methods for Disease Association Studies with Immunogenetic Data 257

erroneous conclusions. Therefore, as with any statistical method,

care must be taken in both the application and interpretation of
results in logistic regression analyses. In addition, it is important
that the data be reported in a manner that allows readers to reanalyze
the data (see below, data reporting).
Logistic regression analysis is always performed using com-
puter software. While many commercial (e.g., SAS, SPSS, STATA)
and free (e.g., PLINK (47)) software packages can be used .for
logistic regression, many are not suitable for polymorphic immu-
nogenetic data. For example, PLINK performs logistic regression
for SNP or copy number variant(CNV) data and will not handle
the level of polymorphism typical of HLA or KIR data. HLA or
KIR data may be recorded for conditional analysis of SNP or CNV
data in PLINK, but this will depend on the specific research ques-
tion. There are also several packages and standard functions for the
R language (e.g., the “glm” function in the base package) that can
perform logistic regression, but extreme care must be taken in the
coding of immunogenetic data, as many of these functions are not
designed to handle high levels of polymorphism.
A regression analysis produces a model that includes all of the
variables that are useful in predicting the (dichotomous) outcome
variable. Logistic regression provides an OR for each variable
involved in predicting outcome, and can be particularly useful in
immunogenetic studies that must account for a number of cofac-
tors in assessing associations. Several options are available in build-
ing the most parsimonious model, i.e., the one explaining the
maximum of the variance with the minimal number of variables.
A backward stepwise approach can be particularly productive in an
exploratory analysis; independent variables are removed from the
full model and the fit of the new model is tested at each step. This
procedure allows the investigator to eliminate noncontributing
variables and build a more parsimonious model. A forward step-
wise approach can be more robust when large numbers of variables
are studied. In this procedure, the most significant variables are
successively added to an empty model, until the addition of vari-
ables does not contribute to significantly increase the variance
explained by the model.

2.2. Family-Based Family studies have long been the bread and butter of genetic epi-
Studies demiology. Family studies provide a means to assess genetic linkage
and/or association by testing within a nuclear family.

2.2.1. Pedigree Types The simplest type of family study involves (1) “trios,” sets of
in Family Studies “simplex” families consisting of a proband (an individual with the
disease phenotype of interest) and both of the proband’s parents
(Fig. 1a). Additional family collections may include (2) multiplex
sib pairs (MSP) pedigrees with two or more affected siblings
(Fig. 1c), (3) multiplex parent child (MPC), with one affected par-
ent and child (Fig. 1b) and (4) multiplex parent sibs (MPS) with at
258 J.A. Hollenbach et al.

Fig. 1. The four family-based ascertainment schemes are selected in the presence of at
least: (a) one affected child (S), (b) one affected child and one affected parent (multiplex
parent child (MPC)), (c) two affected sibs (multiplex sib pairs (MSP)), and (d) two affected
sibs and one affected parent (multiplex parent sibs (MPS)). As noted in the text, additional
family members (parent or sib) may be affected within each of these schemes; they occur
with frequencies expected under the disease model and are retained in analyses. For the
MSP and MPS ascertainment schemes, using the ordered notation of Thomson (55) the
first affected sib ascertained is always denoted by genotype AC, with the four possible
genotypes for the second affected sib listed.

least one affected parent and two or more affected siblings (Fig. 1d).
The type of analyses available will depend on the nature of the
pedigrees collected.

2.2.2. Affected Sib Pair When considering related individuals, the loci linked to disease
Method susceptibility are expected to be found in the region of the genome
inherited from the same ancestors. Affected sib pair methods can
be used in the study of HLA-associated diseases. Deviations from
the Mendelian random expectations of 25, 50, and 25% that two
affected sibs will on average share 2, 1, and 0 parental chromo-
 14 Analytical Methods for Disease Association Studies with Immunogenetic Data 259

somes in common identical by descent (IBD) suggest a disease

PRE gene in the region. Several statistics can be used, a goodness-
of-fit χ2 is commonly used and tests the null hypothesis of the
absence of linkage with the disease locus.

2.2.3. Transmission Originally developed as a test for genetic linkage, the Transmission
Disequilibrium Test and Disequilibrium Test (TDT) is applicable as an association test as
Affected Family-Based well, since it always tests a compound hypothesis of linkage and
Controls association: only linked markers that are associated with disease can
give a significant result. The fundamental basis for the TDT involves
the likelihood that if a marker is linked to disease, there will be a
bias in transmission from parents to the affected individual. In the
TDT test, only parents who are heterozygous for the marker of
interest are considered. The TDT test has the advantage of being
robust with respect to population structure, thereby sidestepping
the possible pitfall of population stratification often encountered in
case-control studies. The TDT test looks at deviations from the
expected 50:50 ratio of transmitted and nontransmitted alleles
from parents heterozygous for the marker alleles.
The use of nuclear family data (two parents and their children)
to estimate control marker allele frequencies (and similarly haplo-
type frequencies) was introduced, with application to HLA data,
by Rubinstein et al. (48), and Falk and Rubinstein (49). In families
ascertained for the presence of at least one affected child, termed S
(simplex) pedigrees (also referred to as trio families), the two
parental marker alleles not transmitted to the affected child (the
proband) are used as population (control) alleles (see Fig. 1a).
These are referred to as affected family-based controls (AFBACs).
This matched design for patient (parental transmitted) and “con-
trol” (parental nontransmitted) marker alleles avoids ethnic con-
founding in the case of a stratified population (50–52).
Field (53) and Thomson (54)—extended this AFBAC approach
to nuclear pedigrees ascertained for the presence of at least two
affected sibs, termed MSP pedigrees (although referred to then as
MS pedigrees, for multiplex sibs); using the alleles from both par-
ents that are never transmitted to either sib in the affected sib pair
as the “control” population (Fig. 1c). Additionally, MPC and MPS
ascertainment schemes were considered: these extend the S and
MSP ascertainment schemes to also include at least one affected
parent. The non- or never-transmitted alleles, respectively, from
the unaffected parent form the respective AFBAC populations
(Fig. 1b, d). Similar extensions apply to other ascertainment
schemes, e.g., with ascertainment based on the presence of at least
three affected sibs in a nuclear family, the AFBACs are the parental
alleles never transmitted to any of the three affected sibs.
AFBAC is similar to the TDT in that it examines the transmis-
sion of marker alleles in family trios. The AFBAC method has often
mistakenly been equated to a heterogeneity test of association of
260 J.A. Hollenbach et al.

patient and AFBAC frequencies. The AFBAC method per se is for

obtaining an unbiased estimate of marker allele frequencies, and
does not refer to any specific test of association/linkage. The TDT
is the appropriate statistical test to apply. However, note that in the
case of a homogenous population sample, the results from a TDT
and contingency table analysis will be very similar. Here, all non-
transmitted parental alleles are considered, regardless of whether
the parents are homozygous or heterozygous.
Statistical tests to apply to the AFBAC data (alleles or haplo-
types) are as follows (55):
(a) Equality of AFBAC maternal and paternal frequencies (a significant
difference may imply a maternal/fetal interaction effect, in which
case only the paternal AFBACs should be used).
(b) Test for equality of AFBAC frequencies with control marker
frequencies available from any other sources, e.g., population
data, other disease studies, and AFBACs ascertained under dif-
ferent criteria. With case/control data, one must be careful to
understand the ascertainment scheme for the controls, i.e.,
whether it is from a random or nondiseased population. For
rare diseases, there will be little difference, but for common
diseases this is not the case.
(c) Test for multiplicative structure of the paternal transmitted
and nontransmitted alleles for S ascertainment, with appropri-
ate modifications for other ascertainment schemes, and simi-
larly for the maternal alleles. A significant difference in the
latter case may reflect a maternal/fetal interaction effect.
(d) Test for a multiplicative structure of the paternal nontransmit-
ted alleles and the child’s genotype for S ascertainment, with
appropriate modifications for other ascertainment schemes,
and similarly for the maternal alleles. A significant difference in
the latter case may reflect a maternal/fetal interaction effect.
(e) Test for fit to HWP of the AFBAC genotypes created, with S
ascertainment, from the maternal and paternal nontransmitted
alleles in each family. For MS ascertainment, only the families
where both sibs share the same two parental alleles (AC)
(Fig. 1c) are used in this test. A significant deviation from
HWP may reflect a maternal/fetal interaction effect.
 14 Analytical Methods for Disease Association Studies with Immunogenetic Data 261

Table 6
Potential units of analysis for killer cell immunoglobulin-like receptors (KIR) data
in disease association

Analytical unit Examples

KIR ligands HLA-C1 carrier; HLA-C1 homozygous; HLA-C2 carrier; HLA-C2 homozy-
gous; HLA-Bw4 carrier; HLA-Bw4 homozygous; HLA-Bw6 carrier;
HLA-Bw6 homozygous; HLA-Bw4_Thr80 carrier; HLA-Bw4_Ile80 carrier
KIR genes KIR2DL1; KIR2DP1; KIR2DL2; KIR2DL3; KIR2DL5; KIR3DL1; KIR2DS1;
KIR2DS2; KIR2DS3; KIR2DS4; KIR2DS5; KIR3DS1
KIR diplotypes AA; AA (with no activating); AB; BB; centromeric AA; telomeric AA; centro-
meric AB; telomeric AB; centromeric BB; telomeric BB
KIR haplotypes A; B; B variants
KIR and ligands KIR2DL1 with HLA-C2; KIR2DL1 without HLA-C2; KIR2DL2 with
HLA-C1; KIR2DL2 without HLA-C1; KIR2DL3 with HLA-C1; KIR2DL3
without HLA-C1; KIR3DL1 with HLA-Bw4; KIR3DL1 with HLA-Bw4_
Ile80; KIR3DL1 with HLA-Bw4 homozygous; KIR2DL3 with HLA-C1;
KIR3DS1 with HLA-Bw4; KIR3DS1 with HLA-Bw4_Ile80; KIR3DS1 with
HLA-Bw4 homozygous

3. KIR Data:
Special
Considerations
While KIR data are in many ways typical of immunogenetic data,
with high levels of polymorphism and linkage disequilibrium
through the region, as well as a clear history of extensive recombi-
nation and gene conversion events, the limitations of most typing
techniques mean that in many cases investigators will have to deal
with large amounts of missing data. Likewise, most typing tech-
niques cannot detect or distinguish between loci that may be dupli-
cated on a KIR haplotype. In addition, the relationship between
the KIR and their HLA ligands dictates consideration of interac-
tion effects between these two important polymorphic regions
when analyzing for disease associations. Table 6 illustrates poten-
tial units of analysis in disease association studies with KIR data.
This list is by no means exhaustive, and the ultimate choice of
analyses performed should be determined by the specific research
hypotheses being tested in the study.
Issues specific to the analysis of KIR data for disease association
studies include:
1. Treatment of presence/absence or gene-content typing data. As
allele level KIR genotyping improves, datasets will contain a
mixture of gene-content and allele level typing data. Analysis
of KIR data at the gene-content level may at times be more
robust than at the allele level, or vice-versa. In gene-content
262 J.A. Hollenbach et al.

data analyses, allele frequencies must be estimated using

expected frequencies under Hardy–Weinberg Equilibrium
Proportions (HWEP) or phenotype frequencies. Allele- and
locus-level experimental data can be compared to published
allele-level KIR typing datasets.
2. Incorporation of HLA interaction effects. It is not clear how
best to model the KIR interaction with its HLA ligand; analy-
ses could define HLA at a broad functional level or use molecular
(HLA allele) designations. Current research depends almost
exclusively on broad designations, but recent work suggests
some crossover in HLA-C-group KIR ligand specificity. We
will assess the impact of HLA-ligand definition using both
broad designations and allelic designations for HLA ligands
using published data and model.
3. Controlling for other HLA associations. Many disease associa-
tions with KIR also show strong HLA associations (e.g., auto-
immune disorders). Accounting for primary HLA associations
while accounting for HLA’s function as a KIR ligand involves
consideration of the extensive linkage disequilibrium through-
out the HLA region.
4. The use of estimated haplotypes vs. the standard A/A and B/x
designations.
5. The treatment of haplotypic data. It is unclear how chromo-
somal location should be considered for otherwise identical
loci (e.g., KIR2DS3), and there are no standards for (locus
level) CNV.

4. Data Reporting

A basic premise in scientific research involves the necessity of repli-

cation of studies. Although this is usually taken to mean replication
of the laboratory findings, it is important that data analyses be rep-
licated as well. In order to facilitate replication of analytical out-
comes, it is imperative that there be a careful and thorough
reporting of data. For immunogenetic studies, the data should be
listed as both raw counts and frequencies; this will allow other
investigators to perform additional analyses (using counts) as well
as meta-analysis. We recommend the authors provide either within
the body of the paper or as supplemental material, a list of all alleles,
haplotypes and genotypes, including rare variants.
 14 Analytical Methods for Disease Association Studies with Immunogenetic Data 263

Acknowledgments

This work was supported by National Institutes of Health (NIH)

grants U01AI067068 (JAH, SJM) and U19 AI067152 (PAG) awarded
by the National Institute of Allergy and Infectious Diseases (NIAID)
and 1R01DK061722 (JAH) awarded by the National Institute of
Diabetes and Digestive and Kidney Diseases (NIDDK). The content is
solely the responsibility of the authors and does not necessarily repre-
sent the official views of the National Institutes of Health.

Glossary

Bonferroni correction The most commonly used method of

correcting for multiple comparisons. Generally, the test significance
level is divided by the number of comparisons made in the study,
thereby increasing the overall stringency of the significance testing.
Confidence interval A likely range of values for an estimate of a
particular parameter within a particular level of significance.
Contingency table Also known as a cross-tabulation. A 2 × n table
used to analyze heterogeneity between two sets of observations of
two or more categorical variables.
Genetic association The occurrence within a population, greater
than that expected by chance, of a genetic trait with a particular
phenotype.
Logistic regression A statistical method to determine which in
a set of independent variables has a predictive relationship to a
binary-dependent outcome variable.
Multiple comparisons Also know as multiple testing. Performing
a statistical test multiple times in the same analysis, thereby increas-
ing the number of chances that the null hypothesis will be incor-
rectly rejected, leading to false positive associations.
Odds ratio The ratio of odds of an outcome occurring in one
group to the odds of it occurring in another group.
Population stratification Also referred to as population substruc-
ture. Allele frequency differences between subpopulations within a
study population due to ancestry differences or selection biases.
Relative risk A measure that describes the risk of having the out-
come of interest relative to exposure.
Yates correction for continuity An adjustment to the χ2 test
statistic performed by subtracting 0.5 from the (O-E) value for
each cell in a contingency table. The purpose of this correction is
to account for sparse cells in the table which may introduce discon-
tinuity with regard to the χ2 distribution.
264 J.A. Hollenbach et al.

References

1. Vandiedonck C, Knight JC (2009) The human Rickards O, Klitz W (2001) Genetic variability
major histocompatibility complex as a para- and linkage disequilibrium within the HLA-DP
digm in genomics research. Brief Funct region: analysis of 15 different populations.
Genomic Proteomic 8(5):379–394 Tissue Antigens 57:424–439
2. Mather KA, President M, Eastel S, Hollenbach 8. Bergstrom TF, Engkvist H, Erlandsson R,
J, Klitz W, Huttley G, Thomson G (2007) Josefsson A, Mack SJ, Erlich HA, Gyllensten U
Long distance linkage disequilibrium between (1999) Tracing the origin of HLA-DRB1
HLA microsatellites in two human popula- alleles by microsatellite polymorphism. Am J
tions. In: Hansen JA (ed) Immunobiology of Hum Genet 64:1709–1718
the human MHC: proceedings of the 13th 9. Bergstrom TF, Mack SJ, Gyllensten U, Ehrlich
international histocompatibility workshop and HA (2000) Evolution of HLA-DRB loci,
congress. IHWG Press, Seattle DRB1 lineages, and alleles: analyses of intron-1
3. Ann B, Begovich WK, Steiner LL, Grams S, Suraj- and -2 sequences. In: Kasahara M (ed) Major
Baker V, Hollenbach J, Trachtenberg E, Louie L, histocompatibility complex: evolution, struc-
Zimmerman PA, Hill AVS, Stoneking M, Sasazuki ture and function. Springer, Tokyo
T, Rickards O, Titanji VPK, Konenkov VI, 10. Crawford MH, Reddy BM, Martinez-Laso J,
Sartakova ML (1999) HLA DQ hapotypes in 15 Mack SJ, Erlich HA (2001) Genetic variation
different populations. In: Kasahara M (ed) Major among the Golla pastoral caste subdivisions of
histocompatibility complex: evolution, structure Andhra Pradesh, India, according to the HLA
and function. Springer, Tokyo system. Hum Immunol 62:1031–1041
4. Petzel-Erler ML, Gorodezky C, Layrisse Z, 11. Ellis JM, Mack SJ, Leke RF, Quakyi I, Johnson
Klitz W, Fainboim L, Vullo C, Bodmer J, Egea AH, Hurley CK (2000) Diversity is demonstrated
E, Navarrete C, Infante E, Alaez C, Olivo A, in class I HLA-A and HLA-B alleles in
Debaz H, Bautista N, de la Rosa G, Vazquez Cameroon, Africa: description of HLA-
MN, Navarro JL, Pujol MJ, Duran C, A*03012, *2612, *3006 and HLA-B*1403,
Schafhauser C, Faucz FR, Janzen M, Maciag P, *4016, *4703. Tissue Antigens 56:291–302
Boldt ABW, Souza PSA, Probst CM, da Silva 12. Erlich HA, Mack SJ, Bergstrom T, Gyllensten
GF, Makhatadze N, Dominguez E, Montagnani UB (1997) HLA class II alleles in Amerindian
S, Matos M, Martinez A, Herrera F, Hollenbach populations: implications for the evolution of
J, Thomson G, Pando M, Satz L, Larriba J, HLA polymorphism and the colonization
Fernandez G, Pesoa SA, Borosky A, Garavito of the Americas. Hereditas 127:19–24
G, Angel L, Brown J, Llop E (1997)
Anthropology component report for region 13. Hildesheim A, Apple RJ, Chen CJ, Wang SS,
Latin-America: Amerindian and admixed pop- Cheng YJ, Klitz W, Mack SJ, Chen IH, Hsu
ulations. In: Charron D (ed) Genetic diversity MM, Yang CS, Brinton LA, Levine PH, Erlich
of HLA: functional and medical implications. HA (2002) Association of HLA class I and II
EDK Press, Paris alleles and extended haplotypes with nasopha-
ryngeal carcinoma in Taiwan. J Natl Cancer
5. Trachtenberg EA, Hayes E, Hollenbach JA,
Inst 94:1780–1789
Keyeaux G, Bernal J, Klitz W (1997) HLA
class II variation and linkage disequilibrium in 14. Johansson A, Ingman M, Mack SJ, Erlich H,
nine Amerindian and three African American Gyllensten U (2008) Genetic origin of the
tribes from Columbia. Results of expedicion Swedish Sami inferred from HLA class I and
humana. In: Charron D (ed) Genetic diversity class II allele frequencies. Eur J Hum Genet
of HLA: functional and medical implications. 16:1341–1349
EDK Press, Paris 15. Leffell MS, Fallin MD, Erlich HA, Fernandez-
6. Trachtenberg E, Vinson M, Hayes E, Hsu YM, Vijna M, Hildebrand WH, Mack SJ, Zachary
Houtchens K, Erlich H, Klitz W, Hsia Y, AA (2002) HLA antigens, alleles and haplo-
Hollenbach J (2007) HLA class I (A, B, C) and types among the Yup’ik Alaska natives: report
class II (DRB1, DQA1, DQB1, DPB1) alleles of the ASHI Minority Workshops, Part II.
and haplotypes in the Han from southern Hum Immunol 63:614–625
China. Tissue Antigens 70:455–463 16. Mack SJ, Bugawan TL, Moonsamy PV, Erlich
7. Begovich AB, Moonsamy PV, Mack SJ, JA, Trachtenberg EA, Paik YK, Begovich AB,
Barcellos LF, Steiner LL, Grams S, Suraj-Baker Saha N, Beck HP, Stoneking M, Erlich HA
V, Hollenbach J, Trachtenberg E, Louie L, (2000) Evolution of Pacific/Asian populations
Zimmerman P, Hill AV, Stoneking M, Sasazuki inferred from HLA class II allele frequency
T, Konenkov VI, Sartakova ML, Titanji VP, distributions. Tissue Antigens 55:383–400
 14 Analytical Methods for Disease Association Studies with Immunogenetic Data 265

17. Mack SJ, Erlich HA (1998) HLA class II 27. Williams AP, Bateman AR, Khakoo SI (2005)
polymorphism in the Ticuna of Brazil: evolu- Hanging in the balance. KIR and their role in
tionary implications of the DRB1*0807 allele. disease. Mol Interv 5:226–240
Tissue Antigens 51:41–50 28. Li Y, Zhang T, Ho C, Orange JS, Douglas SD,
18. Mack SJ, Tu B, Lazaro A, Yang R, Lancaster Ho WZ (2004) Natural killer cells inhibit hep-
AK, Cao K, Ng J, Hurley CK (2009) HLA-A, atitis C virus expression. J Leukoc Biol
-B, -C, and -DRB1 allele and haplotype fre- 76:1171–1179
quencies distinguish Eastern European 29. Hollenbach JA, Ladner MB, Saeteurn K, Taylor
Americans from the general European American KD, Mei L, Haritunians T, McGovern DPB,
population. Tissue Antigens 73:17–32 Erlich HA, Rotter JI, Trachtenberg EA (2009)
19. MacKenzie J, Perry J, Ford AM, Jarrett RF, Susceptibility to Crohn’s disease is mediated by
Greaves M (1999) JC and BK virus sequences KIR2DL2/KIR2DL3 heterozygosity and the
are not detectable in leukaemic samples from HLA-C ligand. Immunogenetics 61(10):
children with common acute lymphoblastic 663–671
leukaemia. Br J Cancer 81:898–899 30. Gaudieri S, Desantis D, McKinnon E, Moore
20. Meyer D, Single RM, Mack SJ, Erlich HA, C, Nolan D, Witt CS, Mallal SA, Christiansen
Thomson G (2006) Signatures of demographic FT (2005) Killer immunoglobulin-like recep-
history and natural selection in the human tors and HLA act both independently and syn-
major histocompatibility complex loci. Genetics ergistically to modify HIV disease progression.
173:2121–2142 Genes Immun 6:683–690
21. Solberg OD, Mack SJ, Lancaster AK, Single 31. Khakoo SI, Carrington M (2006) KIR and dis-
RM, Tsai Y, Sanchez-Mazas A, Thomson G ease: a model system or system of models?
(2008) Balancing selection and heterogeneity Immunol Rev 214:186–201
across the classical human leukocyte antigen 32. Kunert K, Seiler M, Mashreghi MF, Klippert
loci: a meta-analytic review of 497 population K, Schonemann C, Neumann K, Pratschke J,
studies. Hum Immunol 69:443–464 Reinke P, Volk HD, Kotsch K (2007) KIR/
22. Tang TF, Hou L, Chen M, Belle I, Mack S, HLA ligand incompatibility in kidney trans-
Lancaster A, Ho GY, Hwang WY, Alsagoff F, plantation. Transplantation 84:1527–1533
Ng J, Hurley CK (2007) HLA haplotypes in 33. Gedil MA, Steiner NK, Hurley CK (2007)
Singapore: a study of mothers and their cord KIR3DL2: diversity in a hematopoietic stem
blood units. Hum Immunol 68:430–438 cell transplant population. Tissue Antigens 70:
23. Tu B, Mack SJ, Lazaro A, Lancaster A, 228–232
Thomson G, Cao K, Chen M, Ling G, 34. Sun JY, Gaidulis L, Dagis A, Palmer J, Rodriguez
Hartzman R, Ng J, Hurley CK (2007) HLA- R, Miller MM, Forman SJ, Senitzer D (2005)
A, -B, -C, -DRB1 allele and haplotype frequen- Killer Ig-like receptor (KIR) compatibility plays
cies in an African American population. Tissue a role in the prevalence of acute GVHD in unre-
Antigens 69:73–85 lated hematopoietic cell transplants for AML.
24. Cao K, Hollenbach J, Shi X, Shi W, Chopek M, Bone Marrow Transplant 36:525–530
Fernandez-Vina MA (2001) Analysis of the 35. Lanier LL (1999) Natural killer cells fertile
frequencies of HLA-A, B, and C alleles and with receptors for HLA-G? Proc Natl Acad Sci
haplotypes in the five major ethnic groups of USA 96:5343–5345
the United States reveals high levels of diversity 36. Moffett A, Hiby SE (2007) How does the
in these loci and contrasting distribution pat- maternal immune system contribute to the
terns in these populations. Hum Immunol development of pre-eclampsia? Placenta 28
62:1009–1030 (suppl A):S51–S56
25. Hollenbach JA, Thomson G, Cao K, Fernandez- 37. Hu LP, Tao LX (2009) (Statistical analysis
Vina M, Erlich HA, Bugawan TL, Winkler C, using SAS software package for data of RxC
Winter M, Klitz W (2001) HLA diversity, dif- contingency table (part two)]. Zhong Xi Yi Jie
ferentiation, and haplotype evolution in He Xue Bao 7:878–882
Mesoamerican Natives. Hum Immunol 62:
38. Horstmann GA (1988) [Product of the month:
378–390
SPSS statistical package for the sciences].
26. Khakoo SI, Chloe LT, Martin MP, Brooks CR, Radiologe 28:438–439
Gao X, Astemborski J, Cheng J, Goedert JJ,
39. R Development Core Team (2008) R: a lan-
Vlahov D, Hilgartner M, Cox S, Little AM,
guage and environment for statistical comput-
Alexander GJ, Cramp ME, O’Brien S,
ing. R Foundation for Statistical Computing,
Rosenberg WMC, Thomas DL, Carrington M
Vienna
(2004) HLA and NK cell inhibitory receptor
genes in resolving hepatitis C virus infection. 40. Thomson G, Valdes AM, Noble JA, Kockum I,
Science 305:872–874 Grote MN, Najman J, Erlich HA et al (2007)
266 J.A. Hollenbach et al.

Relative predispositional effects of HLA class II 47. Purcell S, Neale B, Todd-Brown K, Thomas L,
DRB1-DQB1 haplotypes and genotypes on Ferreira MA, Bender D, Maller J, Sklar P, de
type 1 diabetes: a meta-analysis. Tissue Antigens Bakker PI, Daly MJ, Sham PC (2007) PLINK:
70:110–127 a tool set for whole-genome association and
41. Hollenbach JA, Thompson SD, Bugawan TL, population-based linkage analyses. Am J Hum
Ryan M, Sudman M, Marion M, Langefeld Genet 81:559–575
CD, Thomson G, Erlich HA, Glass DN (2010) 48. Rubinstein P, Walker M, Carpenter C, Carrier
Juvenile idiopathic arthritis and HLA class I C, Krassner J, Falk C, Ginsberg F (1981)
and class II interactions and age-at-onset Genetics of HLA disease associations: the use
effects. Arthritis Rheum 62:1781–1791 of the haplotype relative risk (HRR) and the
42. Thomson G, Marthandan N, Hollenbach JA, “haplo-delta” (Dh) estimates in juvenile diabetes
Mack SJ, Erlich HA, Single RM, Waller MJ, from three racial groups. Hum Immunol 3:
Marsh SG, Guidry PA, Karp DR, Scheuermann 384
RH, Thompson SD, Glass DN, Helmberg W 49. Falk CT, Rubinstein P (1987) Haplotype rela-
(2010) Sequence feature variant type (SFVT) tive risk: an easy reliable way to construct a
analysis of the HLA genetic association in proper control sample for risk calculations. Ann
juvenile idiopathic arthritis. Pac Symp Hum Genet 51:227–233
Biocomput:359–370 50. Khoury MJ, Beaty TH (1994) Applications of
43. Thomson G, Valdes AM (2007) Conditional the case-control method in genetic epidemiol-
genotype analysis: detecting secondary disease ogy. Epidemiol Rev 16:134–150
loci in linkage disequilibrium with a primary 51. Schaid DJ, Sommer SS (1994) Comparison of
disease locus. BMC Proc 1(suppl 1):S163 statistics for candidate-gene association studies
44. Thomson G, Marthandan N, Hollenbach JA, using cases and parents. Am J Hum Genet
Mack SJ, Erlich HA, Single RM, Waller MJ, 55:402–409
Marsh SGE, Guidry PA, Karp DR, Scheuermann 52. Thomson G (1995) Mapping disease genes:
RH, Thompson SD, Glass DB, Helmberg W family-based association studies. Am J Hum
(2009) Sequence feature variant type (SFVT) Genet 57:487–498
analysis of the HLA genetic association in juve-
nile idiopathic arthritis. Proceedings of the 53. Field LL (1991) Non-HLA region genes in
2010 pacific society of biocomputing confer- insulin dependent diabetes mellitus. Baillieres
ence. Pac Symp Biocomput:359–370 Clin Endocrinol Metab 5:413–438
45. Clopper CJ, Pearson ES (1934) The use of 54. Thomson G (1993) AGFAP method: applica-
confidence or fiducial limits illustrated in bility under different ascertainment schemes
the case of the binomial. Biometrika 26: and a parental contributions test. Genet
404–413 Epidemiol 10:289–310
46. Nemes S, Jonasson JM, Genell A, Steineck G 55. Thomson G (1995) Analysis of complex human
(2009) Bias in odds ratios by logistic regression genetic traits: an ordered-notation method and
modelling and sample size. BMC Med Res new tests for mode of inheritance. Am J Hum
Methodol 9:56 Genet 57:474–486
 Chapter 15

Impact of HLA Matching and HLA Antibodies in Organ

Transplantation: A Collaborative Transplant Study View
Caner Süsal and Gerhard Opelz

Abstract
The Collaborative Transplant Study (CTS) was initiated in 1982 in Heidelberg, Germany, and originated
from the need to gain further insight into the complex problems and risks involved in human organ trans-
plantation. Currently, more than 400 transplant centers in 45 countries are contributing to this voluntary
international effort, and from the beginning of the study, the impact on graft outcome of immunological
factors, such as matching for HLA antigens and allosensitization to HLA and non-HLA antigens, have
been areas of interest. Herein, we summarize the recent findings from the CTS on these two topics.

Key words: HLA matching, HLA antibodies, Crossmatch, Kidney transplantation, Liver transplantation,
Heart transplantation

1. Introduction

With graft survival rates steadily improving during recent years,

there is debate whether organs should be allocated according to
compatibility for HLA, or whether HLA matching should be
ignored and kidneys from deceased donors should be transplanted
locally with the shortest possible cold ischemia time. For example,
in a survey of transplantations in which one kidney was shipped
and the other was transplanted locally, Mange et al. found that the
shipment of deceased-donor kidneys was associated with an
increased risk of graft failure during the first posttransplant year
(1). This result, however, might not necessarily be a result of lon-
ger ischemia but could be due to a better physiological quality of
donor organs retrieved and transplanted locally. Results from the
Collaborative Transplant Study (CTS) (2) indicate that short cold
ischemia does not eliminate the effect of HLA matching and argue
for continued kidney exchange efforts for achieving better HLA

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_15, © Springer Science+Business Media New York 2012

267
268 C. Süsal and G. Opelz

compatibility (3, 4). Recent CTS work has focused on factors other
than graft survival and showed that better HLA matching in renal
transplantation is associated with the administration of lower dos-
ages of immunosuppressive agents (5) and a lower incidence of
side effects of immunosuppression, such as non-Hodgkin lym-
phoma and hip fractures (5, 6). Others have shown that better
HLA matching is associated with lower grade sensitization if a
patient lost a kidney graft and is relisted for a retransplant (7). A
recent CTS analysis indicated that HLA matching might also
improve the currently poor long-term outcome in lung transplan-
tation (8).

2. HLA Matching
in Kidney
Transplantation
In a comparative analysis of more than 130,000 deceased-donor
2.1. HLA Matching, kidney transplantations performed during two successive decades
Cold Ischemia Time (1985–1994 and 1995–2004), we were able to demonstrate that—in
and Graft Outcome in spite of improvements in immunosuppression and a significant
Kidney Transplantation decline in the length of ischemic preservation—the effect of HLA
matching on the survival rate of kidney transplants from deceased
donors continues to be influential (4). While numerically the
percent improvement in graft survival through HLA matching was
smaller in recent years, the relative contribution of HLA mis-
matches to the fraction of graft failures remained nearly unchanged
(Fig. 1). The main difference between the results obtained in the
comparative analysis of two decades was an improvement in the
overall graft survival rate. However, when 3 HLA-A + B + DR mis-
matches were taken as reference, transplants with no HLA mis-
matches showed a relative risk (RR) of failure of 0.76 (95%
confidence interval [CI] 0.71–0.80) among transplants performed
1985–1994 and 0.83 (CI 0.77–0.89) in the decade 1995–2004,
and likewise the increased risk of failure associated with six mis-
matches changed only little, from 1.23 (CI 1.14–1.32) in 1985–
1994 to 1.16 (CI 1.05–1.27) in 1995–2004. Analysis of transplants
performed from 2000 to 2004, during which most recipients pre-
sumably received “modern” immunosuppressive treatment, con-
tinued to show a strong and statistically significant HLA matching
effect.
In a separate analysis, cold ischemia up to 18 h was found not to
be detrimental for graft outcome, and short cold ischemia did not
eliminate the effect of HLA matching (3). These findings altogether
indicated that an abandonment of HLA matching in exchange for
transplanting with short ischemia times would be detrimental to
clinical kidney transplantation. For practical/logistical reasons, we
recommend emphasis on the avoidance of poorly matched grafts
(4, 5 or 6 HLA-A + B + DR mismatches) which is feasible even in
programs with relatively short transplant waiting lists.
 15 Impact of HLA Matching and HLA Antibodies in Organ Transplantation… 269

Fig. 1. Graft survival rates in relation to the number of HLA-A, -B, -DR mismatches for transplants performed between 1985
and 1994 (left) and between 1995 and 2004 (right). Transplants from deceased donors were analyzed. Numbers of mis-
matched HLA antigens (MM) and numbers of patients studied are indicated together with graft survival rates in percent
observed at 5 years posttransplant. The association of HLA matching with graft survival was statistically significant in both
decades (weighted regression P < 0.001). Reproduced from ref. (4) with permission from Wolters Kluwer Health.

2.2. Impact of HLA Posttransplant non-Hodgkin lymphoma is a rare but serious compli-
Mismatching cation after organ transplantation which is associated with high mor-
on Incidence tality. In a recent analysis of more than 150,000 deceased-donor
of Posttransplant kidney transplant recipients reported to the CTS study, we found
Non-Hodgkin that one and two HLA-DR mismatches were a significant risk factor
Lymphoma After for posttransplant non-Hodgkin lymphoma, particularly for the loca-
Kidney Transplantation tion kidney and central nervous system, whereas two HLA-B mis-
matches increased the risk for the development of lymphoma of the
kidney (5). An independent analysis of some 9,000 pediatric kidney
transplant recipients showed that two HLA-DR mismatches were
associated with a twofold higher risk of developing non-Hodgkin
lymphoma, which led us to conclude that two HLA-DR mismatches
should be avoided altogether in pediatric renal transplantation (9).

2.3. Association Bone fractures are a frequent complication after kidney transplanta-
of Mismatches tion for which various predisposing factors, such as older age, female
for HLA-DR with gender, diabetes (10–12), extended duration of ESRD disease, (11,
Incidence of 13) and prolonged pretransplant dialysis (10, 11), have been
Posttransplant Hip identified. Analysis of some 20,000 kidney transplant recipients
Fracture in Kidney reported to CTS showed that there was a significant association
Transplant Recipients between the number of HLA-DR mismatches and the occurrence
of osteoporosis within 5 years after transplantation (6). Importantly,
the risk of hip fracture increased for grafts with one HLA-DR mis-
match to a hazard ratio (HR) of 1.85 (95% CI 1.18–2.89, P = 0.007)
270 C. Süsal and G. Opelz

and with two HLA-DR mismatches to HR 2.24 (CI 1.25–4.02,

P = 0.007). These findings were in agreement with previous findings
of Nikkel et al. who, in a retrospective analysis of United States
Renal Data System data from 68,814 patients undergoing kidney
transplantation during 1988–1998, observed that HLA mismatches
were associated with a 9% increase in risk of overall bone fracture as
compared with patients with no mismatch (10).
Further examination revealed additional independent risk factors
for hip fracture: rejection treatment during the first year posttrans-
plant but not during subsequent years, serum creatinine more than
260 mmol/L at 1 year posttransplant, female gender, and age (³40
years for females and ³60 years for males). There were no significant
differences between patients with or without hip fracture with
respect to their body mass index or the maintenance dose of cal-
cineurin inhibitors, antimetabolites, or corticosteroids at 1 year (6).

3. Impact of HLA
Compatibility on
Lung Transplant
Survival The impact of HLA compatibility on graft survival was examined
in some 8,000 deceased-donor lung transplants performed during
1989–2009 and reported to CTS (8). While graft survival rates
showed a stepwise decrease as the combined number of
HLA-A + B + DR mismatches increased from one to six, it was strik-
ing that lung transplants with “perfect” matches, that is zero HLA
mismatches, showed an extremely poor outcome with half of all
grafts failing during the first year. It thus appears that compatibility
at all three HLA loci represents a particular constellation that can
override the beneficial effect of HLA matching in terms of amelio-
rating immunologic rejection, probably by increasing the risk of
severe lung transplant damage due to HLA-restricted immune
phenomena. Donor lung preservation for up to 12 h was not asso-
ciated with inferior graft survival vs. shorter preservation times,
thereby providing a window for efforts at HLA matching. These
results indicate that it may be appropriate to consider HLA match-
ing of donor lungs within restricted geographical areas to achieve
a moderate level of matching in order to help to improve the current
poor long-term outcomes after lung transplantation.

4. HLA Matching
and Presensitization
in Organ
Transplantation Meier-Kriesche et al., who examined some 16,000 renal transplant
candidates in the USA that were relisted after loss of a primary
4.1. Poor HLA Match in kidney transplant, found that with increasing numbers of HLA
Previous Transplant as a mismatches of the first transplant there was a significant increase in
panel reactive antibody (PRA) at relisting (7). Only 10% of the
Cause of Presensitization
 15 Impact of HLA Matching and HLA Antibodies in Organ Transplantation… 271

patients became sensitized after failure of a 0 HLA-A, -B-mismatched

transplant, whereas the proportion rose up to 37% with increasing
HLA mismatches. Because high PRA is associated with prolonged
waiting times, the authors concluded that many patients are nega-
tively impacted from poor HLA matching of their first kidney
transplant when needing a second transplant.

4.2. Pronounced In the CTS study, we continue to observe a pronounced influence

Impact of HLA of HLA matching on graft outcome in patients with pretransplant
Matching in Patients HLA antibodies or high serum levels of the T-cell activation marker
with High sCD30 (14–16), indicating that presensitization at the B- and
Alloreactivity T-cell level influences the impact of HLA compatibility. In patients
without HLA antibodies and low sCD30, HLA mismatches have
only a modest influence on graft outcome, suggesting that these
patients can tolerate relatively poorly matched organs. It follows
that particular attention to HLA matching should be paid in sensi-
tized patients with increased B- and/or T-cell alloreactivity.
Altogether, data from the CTS as well as other transplant reg-
istries indicate that kidney exchange programs for the purpose of
achieving better HLA matches continue to be meaningful. We
believe that at a time when the shortage of donor organs is the
predominant issue in discussions on how to provide patients with
end-stage renal failure with optimal access to renal transplantation,
the transplant community would be ill advised to eliminate HLA
matching, a factor that significantly reduces the rate of allograft
failure, severe side effects of immunosuppressive agents, and
sensitization.

4.3. Presensitization Although graft survival rates improved greatly during the last 25
as a Major Problem in years, sensitization of the recipient at the B-cell level still represents
Organ Transplantation a major problem in kidney transplantation. Dependent on the anti-
body detection method, approximately 20–40% of patients on the
kidney transplant waiting list possess serum alloantibodies prior to
transplantation, a phenomenon called presensitization. Mainly
three events can lead to presensitization: blood transfusion, preg-
nancy, and previous transplantation. Although there are great indi-
vidual variations, generally the probability that a patient will
develop antibodies increases with the frequency of exposure to for-
eign antigen. According to older literature, 15–25% of women
develop antibodies after the first pregnancy and this rate increases
to 50–60% after the second pregnancy. Lymphocytotoxic B-cell
reactivity was detectable in 20% of waiting list patients who were
treated with 1 blood transfusion, 40% after 10 transfusions, and
70% after 20 transfusions (17). Using an ELISA screening assay,
we found HLA antibodies of the IgG isotype in 17% of patients
waiting for a first transplant, 64% for a second, 84% for a third,
and 92% for a fourth graft. Using the more sensitive Luminex
technology, El-Awar et al. reported that almost all patients waiting
272 C. Süsal and G. Opelz

for a regraft had HLA antibodies of the IgG, IgM, or IgA isotypes
in their serum (18). At our own center, retransplant recipients cur-
rently make up 14% of the kidney waiting list.
An additional critical issue is antibody development against
allogeneic antigen systems on graft endothelium other than HLA
that are not necessarily detected in routine antibody testing. The
influence of high PRA on graft outcome in transplants between
HLA-identical siblings (19), the anti-MICA reactivity in eluates
from rejected kidney allografts described by Zou et al. (20), the
report by Terasaki and Ozawa who found that lymphocytotoxicity
correlated better with graft outcome than solid phase flow cytom-
etry or ELISA (which exclude non-HLA reactivity) (21), the recent
results by Breimer et al. who observed, in the absence of HLA
antibodies, significantly more rejections in patients with a positive
flow cytometric endothelial crossmatch (22), and the work by
Reinsmoen et al. who found a strong association between the pres-
ence of high antibody binding to Angiotensin type 1 receptor and
antibody-mediated rejection in recipients whose sera did not con-
tain antibody to donor HLA or MICA (23), all indicate that anti-
bodies directed against non-HLA antigens are important. In our
Heidelberg algorithm for transplantation of broadly sensitized
high risk patients, we use additional measures, such as peritrans-
plant apheresis, to reduce such presumed undetected antibody
activity below clinically relevant threshold levels (24).

4.4. Disadvantages Presensitized kidney recipients have two major disadvantages: (a)
Associated with they are less likely to receive a transplant due to a high number of
Presensitization HLA antigens defined as unacceptable because they would cause a
positive pretransplant crossmatch; positive crossmatches occur
especially if the antibodies are directed against a wide range of
HLA specificities (high panel reactive antibodies [PRA]); (b) they
reject their grafts more readily even if the pretransplant crossmatch
is negative, possibly due to the presence of weak donor-specific
antibodies that were overlooked in the crossmatch test. Alternatively,
it has been suggested that high PRA reactivity may be an indicator
of a generally high responder status against foreign alloantigens. It
is an important task of the HLA laboratory to identify a suitable
donor against whom the recipient is not sensitized.
According to Fuggle and Martin, presensitized patients remain,
dependent on the grade of PRA, two or three times longer on the
waiting list than nonsensitized patients (25). Actual data from
the CTS indicate that, despite recent developments in the manage-
ment of presensitized patients, patients who are positive in cyto-
toxic PRA testing as well as patients who are negative in cytotoxic
PRA testing but positive in ELISA testing for HLA class I antibod-
ies continue to show an inferior graft outcome (2, 26).
Positive crossmatches in the complement-dependent lym-
phocytotoxic test were shown in the 1960s to be associated with
 15 Impact of HLA Matching and HLA Antibodies in Organ Transplantation… 273

hyperacute rejection (27) so that performance of a lymphocytotoxic

crossmatch prior to transplantation has become a standard method
for reducing antibody-mediated damage and improving graft out-
come in kidney transplantation. Because heart and liver grafts can-
not be cold-stored as long as the kidney, it is still a matter of debate
whether crossmatching in heart and liver transplantation is useful.
Last year, we analyzed the relationship of positive lymphocytotoxic
T-cell and B-cell crossmatches and flow-cytometric crossmatches
with transplant survival in kidney, heart, and liver transplants from
deceased donors reported to CTS (28). These results are described
below.

5. Analysis of
Positive Kidney,
Heart, and Liver
Transplant Currently, with improved methods of immunosuppression, desen-
Crossmatches sitization, and antibody detection, some “weak” positive cross-
Reported to the match reactions are no longer considered a contraindication at
some centers. Virtual crossmatches gained ground (29) and conse-
Collaborative
quently the mandatory performance of a pretransplant crossmatch
Transplant Study
is not anymore an iron rule. Sometimes, a crossmatch test per-
formed after the transplant operation unexpectedly can turn out
positive. For these reasons, a number of kidney transplantations
were performed and reported to the CTS despite a positive cross-
match against donor T lymphocyte targets. On the other hand,
positive crossmatches against B-cell targets have generally not been
considered a contraindication to transplantation because investiga-
tors disagreed on the clinical relevance of this test (30–33). Some
20 years ago, flow cytometry crossmatching was introduced and
claimed to be superior to the lymphocytotoxicity assay because of
higher technical sensitivity (34). Others claimed that the flow
crossmatch excludes too many patients from transplantation based
on false positive results (35). Therefore, at some centers, comple-
ment-dependent lymphocytotoxic crossmatching and flow cytom-
etry crossmatching was performed in parallel, and the flow
crossmatch was not necessarily considered the determining test.
This resulted in positive flow crossmatches reported to CTS. In
this manner, over the years a sizable number of positive crossmatch
kidney transplants were recorded in the CTS database and became
available for analysis. Most heart and liver transplants reported to
CTS are done without a pretransplant crossmatch test. At some
centers, crossmatches are performed retrospectively for study pur-
poses, and the results have been compiled in the CTS database.
Our analysis showed that, overall, high rates of hyperacute
rejection were not observed with any type of positive crossmatch
or type of organ transplant (28), probably because immunosup-
pressive treatment has improved. When first kidney transplant
274 C. Süsal and G. Opelz

recipients were analyzed, positive lymphocytotoxic T-cell cross-

matches were associated with significantly decreased graft survival
in the period between 1990 and 1999, but not from 2000 to 2007,
suggesting that special selection and treatment measures may have
been effectively developed and applied in recent years. Positive
B-cell crossmatches were associated with inferior graft outcome
during both time intervals.
When kidney retransplants were analyzed, the percentage of
positive crossmatches was approximately twice as high as compared
to the fraction of positive crossmatches in first transplants, empha-
sizing the sensitizing effect of first graft rejection. The effect of
positive T-cell or B-cell crossmatches on graft survival was mark-
edly stronger in retransplants as compared with first grafts, and the
effect was fairly consistent in the two time periods. Positive flow
crossmatches were not associated with statistically inferior out-
comes, neither in first nor in retransplants, suggesting that anti-
body reactivity detectable with the sensitive flow technology is
often so weak that current immunosuppressive regimens are
sufficient for preventing antibody-mediated rejection. Alternatively,
special preconditioning and immunosuppressive treatment meth-
ods may have prevented graft rejections in these patients. The
impact of a positive T- and B-cell cytotoxic and flow cytometric
crossmatch on kidney graft survival in transplantations performed
between 1995 and 2008 is shown in Fig. 2.
In the analysis of heart transplants, the fraction of positive
T-cell crossmatches was approximately eight times higher, and that
of positive B-cell crossmatches approximately two times higher
than that in kidney transplants. Interestingly, both T-cell and B-cell
crossmatches were associated with significantly lower heart trans-
plant outcome in the recent time period from 2000 to 2007, as
compared to the earlier 1990–1999 period, conceivably because
induction treatment for prevention of graft rejection with potent
antibody agents (OKT3 or ATG) has become less common due to
the well recognized increased risk of posttransplantation lymphoma
associated with such treatment (36–38).
In liver transplantation, the fraction of positive crossmatches
was even higher than that in heart transplants during the first time
period (1990–1999) and similar during the recent time period
(2000–2007). Positive T-cell crossmatches in liver transplantation
were statistically significantly associated with impaired graft sur-
vival, especially during the recent analysis period (2000–2007).
Positive B-cell crossmatches did not show a significant association
with outcome.
That the majority of heart and liver transplants continue to be
done without a pretransplant crossmatch test is probably related to
the fact that immediate hyperacute graft rejection is rarely observed
and that the impact of a positive crossmatch is therefore not obvious.
 15 Impact of HLA Matching and HLA Antibodies in Organ Transplantation… 275

Fig. 2. Impact of positive crossmatches on survival of deceased-donor kidney transplants in the time period between 1995 and
2008. Lymphocytotoxic crossmatches using T-cell (left) or B-cell targets at warm incubation temperature (middle) were
analyzed (right). Results obtained with flow cytometry. Numbers of patients studied and P values (log rank) are indicated.
(a) First transplants; (b) retransplants.

Posttransplantation follow-up, however, shows a progressive graft

loss rate in crossmatch-positive recipients during the first 3 months.
The CTS results strongly argue for the performance of pretrans-
plantation T-cell crossmatches in heart and liver transplantation.
Because the crossmatch test can be performed in the period
between declaration of brain death and removal of the donor
organ, prolonged organ preservation times can usually be avoided.
However, in the case of a positive crossmatch finding, a cross-
match-negative back-up patient needs to be selected for transplan-
tation, a procedure that requires logistical preparations.
Pretransplantation characterization of the recipient’s alloantibod-
ies and subsequent virtual crossmatching are further approaches
for reducing the number of positive crossmatch results in heart
and liver transplantation.
276 C. Süsal and G. Opelz

6. Conclusion

In conclusion, despite overall improved graft survival rates in recent

years, the CTS data continue to support organ sharing based on
HLA matching in kidney transplantation. Furthermore, the data
suggest that, similar to kidney transplant recipients, end-stage heart
and liver patients would benefit from HLA antibody testing and
crossmatching.

Acknowledgments

We gratefully acknowledge the generous support of the staff members

at participating CTS centers.

References
1. Mange KC et al (2001) A comparison of the phoma: a collaborative transplant study report.
survival of shipped and locally transplanted Transplantation 90:292–297
cadaveric renal allografts. N Engl J Med 345: 10. Nikkel LE et al (2009) Risk of fractures after
1237–1242 renal transplantation in the United States.
2. Collaborative Transplant Study, University of Transplantation 87:1846–1851
Heidelberg, Germany, Website. www.ctstrans- 11. O’Shaughnessy EA et al (2002) Risk factors
plant.org. Accessed 29 Nov 2010 for fractures in kidney transplantation.
3. Opelz G, Döhler B (2007) Multicenter analysis of Transplantation 74:362–366
kidney preservation. Transplantation 83:247–253 12. Vautour LM et al (2004) Long-term fracture
4. Opelz G, Döhler B (2007) Effect of human risk following renal transplantation: a popula-
leukocyte antigen compatibility on kidney graft tion-based study. Osteoporos Int 15:160–167
survival: comparative analysis of two decades. 13. Patel S et al (2001) Prevalence and causes
Transplantation 84:137–143 of low bone density and fractures in kidney
5. Opelz G, Döhler B (2010) Impact of HLA mis- transplant patients. J Bone Miner Res 16:
matching on incidence of posttransplant non- 1863–1870
Hodgkin lymphoma after kidney transplantation. 14. Opelz G et al (1999) HLA compatibility and
Transplantation 89:567–572 organ transplant survival. Collaborative Trans-
6. Opelz G, Döhler B (2010) Association of mis- plant Study. Rev Immunogenet 1:334–342
matches for HLA-DR with incidence of post- 15. Süsal C, Pelzl S, Opelz G (2003) Strong human
transplant hip fracture in kidney transplant leukocyte antigen matching effect in nonsensi-
recipients. Transplantation 91:65–69 tized kidney recipients with high pretransplant
7. Meier-Kriesche HU et al (2009) A lifetime ver- soluble CD30. Transplantation 76:1231–1232
sus a graft life approach redefines the impor- 16. Süsal C, Döhler B, Opelz G (2009) Presensitized
tance of HLA matching in kidney transplant kidney graft recipients with HLA class I and II
patients. Transplantation 88:23–29 antibodies are at increased risk for graft failure:
8. Opelz G et al (2010) Impact of HLA compati- a Collaborative Transplant Study report. Hum
bility on lung transplant survival and evidence Immunol 70:569–573
for an HLA restriction phenomenon: a collab- 17. Opelz G et al (1981) Lymphocytotoxic anti-
orative transplant study report. Transplantation body responses to transfusions in potential kid-
90:912–927 ney transplant recipients. Transplantation 32:
9. Opelz G, Döhler B (2010) Pediatric kidney 177–183
transplantation: analysis of donor age, HLA 18. El-Awar N et al (2002) Almost all patients who
match, and posttransplant non-Hodgkin lym- are waiting for a regraft of a kidney transplant
 15 Impact of HLA Matching and HLA Antibodies in Organ Transplantation… 277

have anti-HLA antibodies. Transplant Proc crossmatches reported to the Collaborative

34:2531–2532 Transplant Study. Hum Immunol 70:627–630
19. Opelz G, Collaborative Transplant Study 29. Bingaman AW et al (2008) A virtual crossmatch
(2005) Non-HLA transplantation immunity protocol significantly increases access of highly
revealed by lymphocytotoxic antibodies. Lancet sensitized patients to deceased donor kidney
365:1570–1576 transplantation. Transplantation 86:1864–1868
20. Zou Y et al (2006) Detection of anti-MICA 30. Ettenger RB et al (1979) Long-term cadaver
antibodies in patients awaiting kidney trans- allograft survival in the recipient with a positive
plantation, during the post-transplant course, B lymphocyte crossmatch. Transplantation
and in eluates from rejected kidney allografts by 27:315–318
Luminex flow cytometry. Hum Immunol 67: 31. Hourmant M et al (1990) Effect of positive
230–237 crossmatch against donor B lymphocytes in
21. Terasaki PI, Ozawa M (2005) Predictive value kidney transplantation: a prospective one center
of HLA antibodies and serum creatinine in study. Clin Transpl 301–309
chronic rejection: results of a 2-year prospective 32. Bryan CF et al (2007) Renal graft survival is
trial. Transplantation 80:1194–1197 not influenced by a positive flow B-cell cross-
22. Breimer ME et al (2009) Multicenter evalua- match. Clin Transpl 21:72–79
tion of a novel endothelial cell crossmatch test 33. Magee CC et al (2008) Renal transplantation
in kidney transplantation. Transplantation 87: in patients with positive lymphocytotoxicity
549–556 crossmatches: one center’s experience.
23. Reinsmoen NL et al (2010) Anti-angiotensin Transplantation 86:96–103
type 1 receptor antibodies associated with 34. Kerman RH et al (1990) Improved graft sur-
antibody mediated rejection in donor HLA vival for flow cytometry and antihuman globu-
antibody negative patients. Transplantation 90: lin crossmatch-negative retransplant recipients.
1473–1477 Transplantation 49:52–56
24. Morath C et al (2010) An integrative approach 35. Christiaans MH et al (1996) No advantage of
for the transplantation of high-risk sensitized flow cytometry crossmatch over complement
patients. Transplantation 90:645–653 dependent cytotoxicity in immunologically
25. Fuggle SV, Martin S (2008) Tools for human well-documented renal allograft recipients.
leukocyte antigen antibody detection and their Transplantation 62:1341–1347
application to transplanting sensitized patients. 36. Swinnen LJ et al (1990) Increased incidence of
Transplantation 86:384–390 lymphoproliferative disorder after immunosup-
26. Süsal C, Döhler B, Opelz G (2009) Presensitized pression with the monoclonal antibody OKT3
kidney graft recipients with HLA class I and II in cardiac-transplant recipients. N Engl J Med
antibodies are at increased risk for graft failure: 323:1723–1728
a Collaborative Transplant Study report. Hum 37. Opelz G, Henderson R (1993) Incidence of
Immunol 70:569–573 non-Hodgkin lymphoma in kidney and heart
27. Patel R, Terasaki PI (1969) Significance of the transplant recipients. Lancet 342:1514–1516
positive crossmatch test in kidney transplanta- 38. Opelz G, Döhler B (2004) Lymphomas after
tion. N Engl J Med 280:735–739 solid organ transplantation: a Collaborative
28. Opelz G, Döhler B, Süsal C (2009) Analysis Transplant Study report. Am J Transplant
of positive kidney, heart, and liver transplant 24:222–230
 Chapter 16

Screening for Antibodies Against MICA by Luminex

Flow Cytometry
Yizhou Zou and Peter Stastny

Abstract
Antibodies against MICA have been found in organ transplant recipients and were found to be associated
with decreased survival of kidney allografts. The MICA antibody screening assay is a Luminex-based solid
phase immunoassay designed to detect IgG antibodies binding to beads pre-coated with recombinant
preparations of MICA alleles. These beads coated with soluble MICA recombinant proteins including 11
common alleles have been produced in our laboratory and similar preparations have been available from
commercial sources. Here, we describe the procedure of MICA antibody screening with a prepared kit, in
which all the reagents were optimized and standardized. We also review how to document the quality of
single MICA antigen beads using MICA-specific monoclonal antibodies, as well as quality control of the
procedure and data analysis.

Key words: Single antigen MICA beads, Antibodies, Luminex, MICA epitopes, Positive and negative
control sera

1. Introduction

To test for antibodies against MICA antigens (1), pre-diluted posi-

tive and negative control sera and patient sera are added to the
appropriate wells, allowing any antibodies against MICA to bind
to the corresponding beads having attached to MICA recombinant
glycoproteins (2). Any unbound antibody is washed away. A
PE-labeled anti-human IgG antibody is added. A second incuba-
tion allows the PE-labeled anti-IgG antibody to bind to any anti-
MICA (IgG) antibodies that have become attached to the bound
MICA antigens on the beads. Next, any unbound anti-IgG is
washed away. As in other Luminex-based immunoassays (3), beads
with different single antigens can be distinguished by their color
and the specific MICA antibody signal is measured in a Luminex

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_16, © Springer Science+Business Media New York 2012

279
280 Y. Zou and P. Stastny

flow cytometry system (4). This PE signal mean fluoresce intensity

(MFI) is proportional within a range to the amount of MICA anti-
body that is bound. The bead mixture and the goat anti-human
IgG-PE are light sensitive and should be stored in the dark. All
reactions occur at room temperature and in the dark. Store serum
aliquots, wash buffer, and bead mixture on ice until used.

2. Materials

1. Serum Samples.
Freshly obtained serum is the specimen of choice. A sterile clot-
ted blood sample with no anticoagulant is preferred (see Note 1).
The following reagents are calculated for 25 tests:
2. Bead Mixture: (500 mL): We use a mixture of 11 beads coated
with single MICA glycoproteins and two control beads. The
storage buffer is optimized and the preparation is ready for
testing human sera. Do not dilute or concentrate these beads.
Protect the beads from light by keeping exposure to light to a
minimum. It is possible to store at 2–8°C in the dark for 4
weeks or to store them at −80°C in the dark for 6 months.
Ensure the beads are resuspended before dispensing.
3. Sample Dilution Buffer 5× (500 mL): This sample dilution buf-
fer is used to reduce the background signals with human serum.
The sample dilution buffer must be diluted to a working solu-
tion (1×) with 2 mL of PBS pH = 7.4 containing 0.05% Tween-
20 (prepared by user) and can be stored at 2–8°C.
4. Positive Control Serum (40 mL ready to use): This serum or
serum mixture is obtained by combining sera which contain
MICA-specific antibodies to all the MICA antigens on the
beads. Store at 2–8°C (see Note 1) (5). Negative Control
Serum (40 mL ready to use): This serum is obtained from an
individual known to have no antibodies against MICA anti-
gens. Store at 2–8°C (see Note 2).
5. Secondary Antibody-PE: (20 mL) Goat anti-Human IgG con-
jugated to phycoerythrin in PBS. Must be diluted by adding
appropriate amount of sample dilution buffer (1×) before use
(we use 1,630 mL). Store at 2–8°C (see Note 3).
6. Wash Solution: 0.05% Tween-20/PBS (PBST): Not provided
in the kit and should be prepared by the user.
7. Luminex beads with recombinant MICA antigens (Table 1) in
the storage buffer (0.05% Tween-20/PBS plus 0.05% Azide).
8. Luminex 100 System (Luminex, Austin, TX).
 16 Screening for Antibodies Against MICA by Luminex Flow Cytometry 281

Table 1
Luminex beads and MICA antigens attached
to them (see Note 4)

Luminex bead ID Antigens

#02 Negative control (see Note 5)

#03 Positive control (see Note 5)
#04 rMICA*001
#06 rMICA*002
#11 rMICA*004
#17 rMICA*006
#19 rMICA*007
#24 rMICA*008
#26 rMICA*009
#32 rMICA*012
#34 rMICA*017
#41 rMICA*018
#43 rMICA*019

9. Whatman 96-well Plate (Whatman Inc., Piscataway NJ, Cat #

7701-2250).
10. Thermowell Sealing Tape (Corning Inc., Corning NY, Cat#
6569).
11. Eppendorf pipette and tips (Eppendorf, Hauppauge NY).

3. Methods

3.1. The Microbead 1. Collect serum samples to be tested and centrifuge the samples
Assay for MICA for 10 min at 1,300 × g (see Note 6).
Antibody Detection 2. Prepare working solution (1×) from sample dilution buffer
(5×, 500 mL) in a new tube by adding 2 mL of PBS pH7.4,
05% tween-20 (prepared by user).
3. Add 25 mL of the working solution (1×) to each well that is
going to be used for the assay in a 96-well black Whatman tray.
Save remaining buffer for the secondary antibody dilution.
4. Add 15 mL of each centrifuged serum sample to each appropri-
ately assigned well and mix with the preloaded working solu-
tion 1× (see Note 7). Cover the plate with Thermowell sealing
282 Y. Zou and P. Stastny

tape to prevent the entry of light and allow this mixture to

incubate for 20 min at room temperature, shaking it at 150 rpm
in a desk-shaking machine at room temperature.
5. While this mixture is incubating, prepare the bead mixture by
resuspending the beads and mixing them well by vortexing at
room temperature.
6. Add 20 mL of bead mixture to each well and mix by
pipetting.
7. Cover the plate with Thermowell sealing tape (to restrict light
and for further sample mixing) and shake the plate at 150 rpm
for 40 min as step 4 (first incubation).
8. During the first incubation prepare the appropriate volume of
wash solution (0.05% tween-20/PBS, 100 mL).
9. After the first incubation is complete, remove the tray from
the rotator, and gently remove the cover tape from the tray
(make sure that the suspension does not spill to other wells of
the tray. Also, remove any bubbles that have been formed over
the wells by blotting the top of the tray with a paper towel; do
not overturn the tray at this point).
10. Add 120 mL of the prepared wash solution (PBST) to each
well, cover with sealing tape and vortex well.
11. Centrifuge the tray for 5 min at 1,300 × g.
12. After centrifugation, gently remove the sealing tape, invert and
flick the tray. While still overturned, blot onto several paper
towels. Tap the tray gently to remove excess wash solution
(flicking involves throwing the supernatant out of the washed
tray without disturbing the pelleted immune complexes formed
by the Luminex beads). Once the tray has been flicked and
returned to upright position, do not invert the tray for a sec-
ond time before blotting, as this will cause the beads to be
flushed out onto the paper towel (if this has occurred repeat
steps 10 and 11 before blotting).
13. Repeat steps 10–12 three times using 180 mL of wash solution
for a total of at least four washes (see Note 8).
14. During the final washing step, prepare the secondary antibody.
Mix well.
15. Add 60 mL of the secondary antibody solution into each well;
cover with sealing tape and vortex the tray. Place the tray on
the shaker at 150 rpm for a 30-min incubation.
16. After the final wash is flicked out of the tray, resuspend the
contents of the well in 60 mL of PBS 1×.
17. Proceed to the Luminex flow cytometry according to the
instruction from manufacturer. Bead information for a Luminex
template file is provided in Table 1.
 16 Screening for Antibodies Against MICA by Luminex Flow Cytometry 283

18. Determine MICA antibodies and specificities by a computer

program or by manual calculation using the method given in
Table 2.

3.2. Quality Controls 1. Positive control and negative control sera should be included
with each test. The MFI values of each bead with the control
sera indicate if the test is valid (see Note 9).
2. Different reagent lots in kits must never be mixed.
3. Monoclonal antibodies against MICA can be used to assess
conjugation of MICA antigens to the beads.
Current lot Luminex beads are coated with 11 preparations of
single MICA antigens, including MICA*001, *002, *004, *006,
*007, *008, *009, *012, *017, *018, and MICA*019, representing

Table 2
Data analysis form for Luminex

Negative control Positive control Sample MFI Assignment

Cutoff
Bead MICA MFI MFI MFI Raw Adjusted value Score
ID antigen ref. raw (1) MFI ref. raw (2) (3) (4) suggested (5) ± (7) Note

02 Negative Ctrl 188 44 1,339

03 Positive Ctrl 22,934 21,436 4,000
04 MICA*001 99 10,975 985
06 MICA*002 69 10,149 958
11 MICA*004 271 12,319 1,092
17 MICA*006 151 12,830 1,183
19 MICA*007 264 15,730 1,106
24 MICA*008 133 13,314 1,205
26 MICA*009 96 6,152 1,072
32 MICA*012 119 7,038 905
34 MICA*017 62 8,849 938
41 MICA*018 103 8,067 913
43 MICA*019 280 16,914 1,237
(1) = MFI raw data of negative control serum tested
(2) = MFI raw data of positive control serum tested
(3) = MFI raw of sample serum tested
(4) = (3)—the smallest one of raw (3), the smallest one usually is the raw data generated from the negative control beads
if not, find the smallest one generated from 11 single antigen beads. Adjusted MFI is the value after background
removed. Cutoff value listed is used as reference that was calculated from 20 normal serum testing for each single anti-
gen beads
(5) = score by 0-invalid; 1-negative; 2-doubt; 4-weak positive; 6-positive; 8-strong positive
284 Y. Zou and P. Stastny

Fig. 1. Reaction of single MICA antigen beads with MICA-specific monoclonal antibodies. The beads were tested with mAb
6B3 (black bars) or 3.2H3 (gray bars) or normal mouse IgG (mIgG, open bars) on 11 single antigen beads named as
rMICA*001,*002, *004, *006, *007, *008, *009, *012, *017, *018, and *019. Results are given as mean fluorescence
intensities (MFI).

the 11 most common MICA alleles. All of these beads give strong
signals with the monomorphic MICA-specific monoclonal anti-
body 6B3, which is a monoclonal antibody recognizing native
MICA proteins. All are positive also with 3.2H3, except for
MICA*001, which is known not to be recognized by this anti-
body. The mean fluorescence intensities vary between 6,000 and
10,000. Normal mouse IgG and beads without antigen show very
low background (Fig. 1).

3.3. Analysis of Results 1. Data files are generated with the bead numbers and their
median fluorescence intensities (MFI) in spreadsheet format.
The raw MFI data obtained by Luminex flow cytometry are
converted to the relative amounts of MICA antibodies by sub-
tracting the background. The threshold is determined for each
bead using the means of the relative amount plus threefold SD
of IgG binding in serum from normal subjects (n > 20) as
shown in Table 3. Based on the relative amount of each MICA
allele-specific antibody and the threshold for each antigen bead
(see Note 10), scores are assigned as follows: 1 = negative;
2 = doubtful, ³threshold and < threshold + 600; 4 = weak posi-
tive, ³threshold + 600 and < threshold + 1,800; 6 = positive,
³threshold + 1,800 and < threshold + 5,400; 8 = _ strong posi-
tive, ³threshold + 5,400.
 16 Screening for Antibodies Against MICA by Luminex Flow Cytometry 285

Table 3
Threshold values above which results are considered positive
(score ≥ 4)

Luminex bead ID MFI after removing background

#02 Negative control 1339

#03 Positive control 4000
#04 rMICA*001 985
#06 rMICA*002 958
#11 rMICA*004 1092
#17 rMICA*006 1183
#19 rMICA*007 1106
#24 rMICA*008 1205
#26 rMICA*009 1072
#32 rMICA*012 905
#34 rMICA*017 938
#41 rMICA*018 913
#43 rMICA*019 1237

2. MFI generated by each bead with a tested serum minus back-

ground signal is used to interpret the results using the thresh-
old values suggested above. To do this, first we find the lowest
MFI value among the negative control beads and the 11 anti-
gen beads for a given serum. Then subtract this smallest num-
ber from each of the 13 raw MFI values obtained. Finally, the
remaining MFI is used to compare to the threshold values
given in Table 3. Sera with an MFI greater than the threshold
value are considered positive.

3.4. Common 1. There are 14 common serologic patterns of human antibodies

Serologic Patterns against MICA antigens as demonstrated by their differential
of Antibodies Against absorption/reactivity (5). Among these, two broad serologic
MICA patterns of antibodies against MICA groups (MICA-G1 and
MICA-G2, Table 4) are most frequently detected in transplant
recipients. The next frequent serologic patterns of MICA anti-
bodies are MICA-G3, G4, and G5. Those antibodies might
react with public epitopes of MICA antigens. Five serologic
patterns of MICA antibodies (ID6-10, Table 4) may react with
MICA epitopes that recognize shared epitopes among small
groups of MICA alleles. Four serologic patterns of MICA anti-
bodies (ID11-14, Table 4) that react with a single MICA allele
are less frequently found in patient sera (see Note 11).
286 Y. Zou and P. Stastny

Table 4
Common serologic patterns of antibodies against MICA

Reaction with single antigens of MICA

ID Pattern *001 *002 *004 *006 *007 *008 *009 *012 *017 *018 *019

1 MICA-G1 + + + + + +
2 MICA-G2 + + + + +
3 MICA-G3 + + + + + + + + +
4 MICA-G4 + + + + + + + + +
5 MICA-G5 + + + +
6 MICA-0217 + +
7 MICA-469 + + +
8 MICA-0819 +
9 MICA-46919 + + + +
10 MICA-11218 + + +
11 MICA-01 +
12 MICA-04 +
13 MICA-06 +
14 MICA-12 +

4. Notes

1. Serum should be separated from red cells when stored or

shipped. Microbial contaminated, hemolyzed, lipemic, or heat
inactivated sera may give inconsistent test results and should be
avoided.
(a) Re-calcified plasma can also be used.
(b) Unacceptable serum specimens.
i. Serum exposed to excessive heat or repeated freeze–
thaw cycles.
ii. Specimens with bacterial or fungal contamination.
iii. Specimens containing visible aggregates and/or fibrin.
Preclear these specimens by centrifugation at 1,300 × g
for 10 min. Positive control serum produces a high MFI with
each bead.
16 Screening for Antibodies Against MICA by Luminex Flow Cytometry 287

2. Negative control serum gives a low MFI with all beads except
the positive control beads.
3. Since the fluorescence of diluted Ab2-PE may change with
time, make dilution immediately before use. The optimal con-
centration of Ab2-PE was determined by titration. It will be
different with each new batch of second antibody or new
kit lot.
4. MICA single antigen beads were developed in our laboratory.
Commercial MICA single antigen beads are available from
Tepnel (6) and One Lambda (3). Validation of each kit of
beads is needed before use in testing. Serum exchange, parallel
testing, and antibody absorption studies are useful to deter-
mine the sensitivity, specificity, and stability of the MICA anti-
body test kit.
5. Positive control beads (#03) were pre-coated with the purified
normal human IgG. Negative control beads (#02) were pre-
pared in the same way as other beads, but no antigen was
added. High MFI reactivity with bead #03 indicates that the
appropriate anti-human IgG secondary antibody was added to
each of the positive control sera, negative control serum, and
patient serum assay wells.
6. If there are suspended particles on the surface of the serum
after centrifugation, do not include the suspended particles
when adding serum samples for testing.
7. Pipetting accuracy is important, since dilution of patient serum
is performed as part of the assay. Do not pipet 15 mL patient
serum with a 200-mL pipetting device.
8. Complete washing is important and necessary before you add
the secondary antibodies. A common error when using the
assay is incomplete washing. This could result in a low MFI in
the positive control beads, and could cause false negative
results.
9. The positive control serum supplied in the kit should give a
high MFI for each of the antigen beads. The negative control
serum in the kit should not give an MFI of more than 1,000
except the MFI from positive control (beads #03).
10. Each laboratory should determine thresholds for positive tests
similar to the methods described here for defining cutoff
values.
11. A few patient sera will exhibit high background binding. They
may require special treatment in order to lower the background
nonspecific binding so that the assay can be correctly inter-
preted. Antibody absorption and elution can be used to resolve
questionable results in most cases.
288 Y. Zou and P. Stastny

References
1. Bahram S, Bresnahan M, Geraghty DE et al and MICA antibodies on kidney graft survival.
(1994) A second lineage of mammalian major Am J Transplant 7:408–415
histocompatibility complex class I genes. Proc 4. Zou Y, Stastny P, Süsal C et al (2007) Antibodies
Natl Acad Sci 91:6259–6263 against MICA antigens and kidney-transplant
2. Zou Y, Heinemann FM, Grosse-Wilde H et al rejection. N Engl J Med 357:1293–1300
(2006) Detection of anti-MICA antibodies 5. Zou Y, Qin Z, Silveus A et al (2009)
in patients awaiting kidney transplantation, Polymorphisms of MICA recognized by human
during the post-transplant course, and in elu- alloantibodies. Immunogenetics 61:91–100
ates from rejected kidney allografts by 6. Duquesnoy RJ, Mostecki J, Hariharan J et al
Luminex flow cytometry. Hum Immunol 67: (2008) Structurally based epitope analysis of
230–237 major histocompatibility complex class I-related
3. Terasaki PI, Ozawa M, Castro R (2007) Four- chain A (MICA) antibody specificity patterns.
year follow-up of a prospective trial of HLA Hum Immunol 69:826–832
 Chapter 17

HLA Antibody Detection and Characterization by Solid Phase

Immunoassays: Methods and Pitfalls
Andrea A. Zachary, Renato M. Vega, Donna P. Lucas, and Mary S. Leffell

Abstract
Solid phase immunoassays for the detection and characterization of HLA-specific antibodies provide
greatly increased sensitivity, specificity, and time and reagent efficiency, compared to the traditionally used
cell-based methods. Testing is performed using commercially available test kits. The assays are of two gen-
eral types: enzyme-linked immunosorbent assays and multianalyte bead. The types vary in both sensitivity
and equipment requirements.
While these assays afford great improvement over the cell-based assays, they can be confounded by
interference from substances within the serum that result in high background reactivity. The high sensitiv-
ity of the assays also makes them more susceptible to environmental factors and operator variability. The
user must be aware of the capabilities of the various formats, the factors that can affect test results, and lot
to lot variability of any single product. Knowledge of the characteristics of each product and thorough and
accurate analysis of the results are essential to the utility of these assays.

Key words: Enzyme-linked immunosorbent assays, HLA antibody, HLA phenotype panel, Luminex®,
Multianalyte bead assays, Pooled antigen panel, Single antigen panel, Solid phase immunoassays

1. Introduction

For more than 30 years, detection and characterization of HLA-

specific antibodies (HLA-Ab) were performed using cell-based
methods. These methods were hampered by the requirement for a
sufficient quantity of viable lymphocytes, low specificity, and prior
to the use of flow cytometry, low sensitivity. Isolation of soluble
HLA antigens was achieved in the mid 1970s (1), but it was 20
more years before there was success in maintaining the native con-
formation of soluble HLA molecules following fixation to a solid
phase (2, 3). The development of solid phase immunoassays (SPI)

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_17, © Springer Science+Business Media New York 2012

289
290 A.A. Zachary et al.

for HLA-Ab provided enormous advantages for both HLA-Ab

testing for transplantation and the investigation of serologically
defined HLA epitopes. Testing by SPI reduces reactivity of anti-
bodies to non-HLA antigens, thereby increasing specificity,
increases sensitivity significantly, provides evaluation of antibody
strength on a continuous scale facilitating manual analysis and
determination of specificity, and can be partially or fully automated.
Further advantages are that SPI are available as kits from several
manufacturers, allowing laboratories to avoid the laborious task of
preparing tests and providing some measure of consistency among
laboratories. However, optimizing utility of these assays requires
thorough and ongoing quality control. As of this writing, kits are
available for the characterization of antigens encoded by antigens
of the HLA-A, -B, -C, -DRB1, -DRB3-5, -DQA, -DQB, -DPA,
-DPB, and MICA loci. Collectively, these kits test for antibodies to
most, but not all, HLA antigens which would be impossible given
the polymorphism of the HLA system. Additionally, many of the
kits provide the allele identity of the target molecule.
This chapter will discuss the types of assays commercially avail-
able, the test procedures, manufacturer-provided materials and
reagents and those that must be provided by the laboratory, advan-
tages and disadvantages of the various tests available, test capabili-
ties, problems with and shortcomings of the various tests, and
clinical applications. Manufacturers provide detailed instructions
for performance of their tests, and therefore, descriptions of the
assays provided here will be somewhat generic. Specific informa-
tion not included here can be found at the web sites of the various
manufacturers which are listed below in alphabetical order. Note:
since manufacturers may introduce new products or platforms or
modify existing ones, web sites should be checked for such changes
before deciding to use a particular system.
GEN-PROBE: http://www.gen-probe.com/products-services/
ONE LAMBDA: http://www.onelambda.com

1.1. Commercially Kits for the detection and characterization of HLA-Ab are available
Available Assays in three formats: pooled HLA antigens, panels of individual HLA
phenotypes, and panels of single HLA antigens (4, 5). Pooled anti-
1.1.1. Formats
gen kits, as the name implies, consist of a pool or mixture of vari-
ous HLA antigens and generally include the most common HLA
specificities found in several racial groups. The pools may be derived
through isolation of antigens from the cells of hundreds of indi-
viduals or by mixing antigens prepared in transfected cell lines or
baculovirus systems. The pools may be comprised of class I (cI)
antigens only, class II (cII) only, or a mixture of cI and cII if the
platform used can distinguish the individual antigen classes in a
mixture. Phenotype panels are comprised of an array of individual
 17 HLA Antibody Detection and Characterization… 291

cI or cII phenotypes while single antigen panels have the HLA

antigens present as individual, purified antigens.
The pooled antigen test provides a rapid and cost-efficient sys-
tem for determining the presence or absence of HLA-Ab. This test
does not determine specificity. However, it does provide an indica-
tion of relative strength and breadth of antibody making it useful
when monitoring sequential sera from a patient (4). Phenotype
panels provide an opportunity to determine antibody specificity
and strength. These panels frequently include a phenotype that
includes two or more antigens of a potential donor, enhancing the
ability to assess the collective strength of the antibodies against
that donor. Further, each antigen is almost always present in mul-
tiple phenotypes providing a safeguard if one phenotype fails to
react properly. Sera reactive with more than 95% of the panel may
make it difficult to determine, conclusively, the presence of an anti-
body to antigens represented by a limited number of phenotypes.
However, because the phenotypes that all bear the target of an
antibody cluster together when the reactions are listed in descend-
ing order of strength, many antibody specificities can be identified
in very broadly reactive sera. Single antigen panels permit
identification of antibodies that may be masked by other antibodies
or react poorly in phenotype panels. Further, since these panels
often include multiple alleles within an antigen group, it is easier to
detect antibodies reactive with only some alleles within an antigen
group. Although single antigen panels have the ability to detect
most, if not all, HLA-Ab present in a serum, there are several things
to be considered. It is difficult to assess the collective strength of
multiple antibodies in a serum as it has been shown that the collec-
tive strength of the antibodies is not equivalent to the sum of the
strengths of reactivities with the individual antigens (6). Further,
the failure of one antigen may not be recognized which may result
in the failure to detect antibody to that antigen. As expected, the
overall sensitivities of the assays increase with the amount of indi-
vidual antigen on each bead, so that the order of sensitivity is single
antigens > phenoytpes > pooled antigens.

1.1.2. Assay Types, Two general types of assays are available, multiplex or multianalyte
Matrices, and Platforms bead assays and enzyme-linked immunosorbent assays (ELISA) (4,
5, 7). For both types of assay, commercial kits provide the matrix
with bound soluble HLA antigens. The bead assays are performed
on either a conventional flow cytometer or a Luminex® fluoroanalyzer,
depending on the kit. In these assays, polystyrene beads are impreg-
nated with fluorescent dyes and beads bearing different pools of
antigens, phenotypes, or single antigens are differentiated by the
fluorescence of the bead. After incubation with control or test sera,
a fluorochrome-labeled anti-human globulin is added to detect
beads that have antibody bound. When tested in a conventional
flow cytometer, beads of different fluorescence are identified on the
292 A.A. Zachary et al.

FL2 channel while the intensity of fluorescence is read on the FL1

channel. Multiplex beads for use with the Luminex® platform are
incorporated with different combinations of ten dilutions of each of
two different dyes to yield 100 different beads. The Luminex®
fluoroanalyzer has two lasers: one of which identifies the bead and
the other identifies beads with bound antiglobulin. In both assays,
different levels of reactivity can be used to approximate relative anti-
body strength. Depending on the assay format and manufacturer,
these assays utilize 10–20 μL of test serum per test.
ELISA are performed on either microtiter or microtest plastic
plates. The procedure is comparable to that for the bead assays
except that the antiglobulin is bound to an enzyme and test
reactions are detected colorimetrically after addition of a color-
producing substrate. The required serum volume is approximately
300 μL. In general and depending on the cutoff value used for
positivity, the bead-based assays are more sensitive than are the
ELISA (8–10).
The antiglobulin reagent provided with all assays is specific for
human IgG. It is possible to use reagents specific for other classes
of immunoglobulin and for IgG subclasses (11); however, the
titer and specificity of those reagents must be verified by the user.
Assays have also been developed for the detection of the C4d
and C1q components of complement (12–14). However, in our
experience, the values obtained for IgG subclasses and C4d may
be low (15) and may be more applicable to determining relative
quantities.

2. Materials

2.1. Bead-Based Materials Provided (see Note 1)

Assays
1. Analysis software.
2. Beads with bound HLA antigens (see Note 2).
3. Fluorochrome-labeled antiglobulin (varies among manu-
facturers).
4. Negative control serum (varies among manufacturers).
5. Positive control serum (varies among manufacturers).
6. Wash buffer.
Materials Required But Not Provided
1. Adhesive plate sealers.
2. Instrument: flow cytometer or Luminex® fluoroanalyzer,
depending on kit.
3. Sheath fluid.
 17 HLA Antibody Detection and Characterization… 293

4. Calibration beads.
5. Microcentrifuge.
6. Microfuge tubes.
7. Millipore filter plates or microtiter plates.
8. Multiscreen vacuum manifold (kit dependent).
9. Plate holder for centrifuge.
10. Rotator or plate shaker.
11. Various adjustable and multichannel pipettes.
12. Vortex mixer.
13. Computer.

2.2. Conventional Materials Provided

ELISA
1. Microtest plates or well strips for use in microtiter plates,
preloaded with HLA molecules (see Note 3).
2. Enzyme-conjugated anti-human IgG.
3. Antibody diluent.
4. Colorimetric enzyme substrate (varies among manufacturers,
see Note 4).
5. Substrate buffer (varies among manufacturers).
6. Control sera (varies among manufacturers).
7. Plate sealers.
8. Stop reagent.
9. Wash buffer.
Materials Required But Not Provided
1. 37°C water bath or incubator.
2. Deionized water.
3. Tubes for sample and reagent dilution.
4. Microplate washer (optional).
5. Adjustable pipetors and Multichannel pipetor with tips.
6. Plate reader with appropriate filters.†
† Adapted for microtest plates when necessary.

3. Methods

3.1. Sample 1. Obtain blood samples in tubes without anticoagulant.

Preparation 2. Separate and centrifuge serum to remove aggregates. Do not
(See Notes 5–7) use sera that are lipemic, hemolyzed, microbially contaminated,
or heat-inactivated.
294 A.A. Zachary et al.

3. Store sera at 4–8°C for up to 48 h or below 35°C for longer

periods. Serum to be stored longer at 4–8°C should have
sodium azide added to a final concentration of 0.1% to prevent
microbial growth.

3.2. Bead-Based Generic Procedure (see Note 1).

Assays
1. Reagents such as wash solution and PE-IgG must be thor-
oughly mixed prior to use.
2. Pre-wet wells (when using filter plates on the Luminex plat-
form only).
3. Add serum and beads. Manufacturers specify volumes of each.
Bead suspensions should be mixed well periodically to assure
even dispersion of the beads (see Note 2).
4. Incubate 30 min on a shaking platform to allow specific bind-
ing of antibody, when present, to bead-bound antigen. Shaking
is necessary to prevent beads from settling which will reduce
contact with antibody (see Note 8).
5. Discard supernatant fluid (see Note 9).
6. Wash beads (number of washes is manufacturer specific) to
remove unbound immunoglobulin (see Note 10).
7. Discard supernatant fluid and add labeled antiglobulin reagent.
8. Incubate 30 min on a shaking platform.
9. Wash to remove unbound antiglobulin and resuspend beads
(see Note 10).
10. For delayed acquisition with a flow cytometer, add a fixing
solution.
11. Acquire beads (see Notes 11–17).
12. Perform analysis using manufacturer-provided software. Total
time needed for procedure is approximately 4 h excluding data
review and analysis. Times will vary depending on the use of
robotic fluid handling systems, operator experience, and num-
ber of samples tested.
13. In addition to control sera, bead sets have built in control
beads.

3.3. Conventional Generic Procedure (see Note 3)

ELISA
1. Dilute all reagents and sera according to manufacturer’s
instructions. Include a negative serum with each run to estab-
lish background reactivity.
2. Pre-wet test wells, let stand, aspirate fluid. This will rehydrate
the bound antigens.
3. Add pre-diluted test and control serum to wells of microtest
trays or microtiter strips (see Note 4).
 17 HLA Antibody Detection and Characterization… 295

4. Cover tray and incubate 30–45 min at 37°C (GenProbe) or

1 h at 20–25°C (One Lambda) (see Note 18).
5. Remove liquid from wells and wash three times to remove any
unbound immunoglobulin (see Note 19).
6. Add conjugated antiglobulin reagent.
7. Cover tray and incubate 30–45 min at 37° (GenProbe) or
40 min at 20–25°C (One Lambda).
8. Remove liquid from wells and wash to remove unbound anti-
globulin (see Note 20).
9. Add enzyme substrate (see Note 21).
10. Incubate in the dark 30 min at 20–25°C (GenProbe) or
10–15 min at 37°C (One Lambda) to allow the reaction to
equilibrate (see Note 22).
11. Add stop reagent in the same sequence as for substrate. Trays
may be stored in the dark for 30 (GenProbe) or 60 min (One
Lambda) (see Note 23).
12. Read absorbance (optical density) at 405–410 nm (GenProbe)
or 630 nm (One Lambda) (see Note 24).
13. Perform analysis using manufacturer-provided software or
manually utilizing recording sheets provided with kit. The per-
cent PRA is calculated by dividing the number of positive
HLA-Ab-containing wells by the total number of antigen
preparations in the panel. The cutoff value for positivity is cal-
culated as a percentage of the range of reactivity of the pro-
vided serum control (SC) tested in the positive wells minus the
non-specific background of the test serum (or antibody diluent)
tested in the blank wells (One Lambda) or two times the OD
value of the mean of the negative control wells (GenProbe).
The assignment of antibody specificity is aided by ranking the
reactions with test sera in descending order of strength.
Total time needed for procedure is approximately 4 h excluding
data review and analysis. Times will vary depending on the use of
robotic fluid handling systems, operator experience, and number
of samples tested.

4. Notes

1. GenProbe reagents are provided together in a kit and are lot-

specific. One Lambda reagents are sold individually. GenProbe
pooled antigen and phenotype panels must be stored at 4°C.
GenProbe single antigen beads must be stored at −80°C and
can be refrozen up to six times. One Lambda single antigen,
296 A.A. Zachary et al.

pooled antigen, and phenotype beads must be stored at −65°C.

They can be frozen only once, but they may be stored at 4°C
up to 3 months.
2. Beads should be kept in suspension by periodic mixing on a
Vortex mixer while being added to tubes or wells.
3. Reagents and trays should be stored according to manufac-
turer’s instructions. Trays and microwell strips/plates should
be protected from moisture.
4. Be careful not to touch the well bottom. Change pipet tips
regularly to avoid cross contamination.
5. Substances inherent in the serum or external agents such as
therapeutic antibodies may interfere with test results. We have
shown that low values for the internal positive control (approx-
imately <8,500 MFI for single antigen panels and <12,000
MFI for phenotype panels) or high values for the internal neg-
ative control (approximately >150 MFI for all negative con-
trols or >500 MFI for any one control for panels with multiple
negative controls, or >200 MFI for panels with a single nega-
tive control) indicate test interference that may result in
reduced values with HLA targets and that may alter the
specificities detected (16). Substances inherent in the serum
may include high levels of IgM, immune complexes, and anti-
body to plastic, among others. There are two commercial
bead-bound products for the reduction of background in
Luminex® assays: SeraClean™ from GenProbe and Adsorb
Out™ from One Lambda, Inc. For both products, the beads
are added to a serum known to have high background reactiv-
ity, incubated, and then removed, ideally with whatever is caus-
ing high background bound to the beads. The results are
variable and are probably affected by the amount and nature of
interfering factors. Also, it is not possible to know if this pro-
cess also reduces or dilutes HLA-Ab. We have shown that
hypotonic dialysis, dialyzing serum against distilled water, is
effective in eliminating or substantially reducing high back-
ground in nearly all cases. This procedure is based on the dif-
fering solubilities of IgM and IgG in distilled water which
results in the precipitation of IgM but not IgG. It is possible
that this procedure also eliminates some immune complexes.
Dithiothreitol (DTT), a reducing agent, has been reported to
reduce background in sera tested on single antigen beads (17),
but in our experience it actually increases reactivity with the
negative control and is not as effective as hypotonic dialysis in
restoring normal reactivity to test sera in the Luminex® assay.
Interestingly, we have found that sera tested on glass micro-
chips often do not display the high background that they
exhibit with a plastic matrix. Low positive and/or high nega-
tive controls indicate that reactivity with antigen-bearing beads
 17 HLA Antibody Detection and Characterization… 297

has been compromised and accurate interpretation of the

results requires some treatment of the serum.
6. Therapeutic agents may also result in test interference. Agents
shown to impact test results include Thymoglobulin, high dose
IVIg, Eculizumab, and Bortezomib (18, 19). Thymoglobulin
is a polyclonal rabbit serum and can thus be removed by
absorbing the serum with beads coated with an anti-rabbit
immunoglobulin. We have shown that hypotonic dialysis also
resolves some or all of the interference caused by Eculizumab
and Bortezomib.
7. Groups of patients, defined by the type of transplant, may
demonstrate, on average, different levels of background or test
interference.
8. All incubations should be performed in the dark at 20–25°C.
Beads and fluorochrome-labeled antiglobulin are light sensi-
tive and extended exposure may cause photo-bleaching. These
reagents should be protected from light as much as possible.
9. When using filter plates, the vacuum pressure should be no
greater than that required to aspirate samples. If using a vac-
uum manifold, do not exceed 100 mmHg. High vacuum pres-
sure may cause beads to be crushed (misshaped) resulting in
bead failures.
10. Insufficient washing may produce false negative reactions due
to blocking of the antiglobulin by residual immunoglobulin in
the wells. Inconsistent washing may yield inconsistent test bead
and control reactivity.
11. Instruments for bead analysis should be calibrated daily or
before each run if daily runs are not performed.
12. Bead acquisition should occur within the manufacturer’s
specified time frame.
13. Unusually high acquisition times (>20 s/sample) may be
indicative of addition of incorrect bead volumes (concentra-
tions). This may result in false negative or weak test bead
reactivity.
14. If using a filter plate and one or more test wells appear clogged,
scratch the plastic dimple underneath test well.
15. Low bead counts may be a result of instrument clogs and/or
instrument out of calibrated range. Possible solutions are to
sonicate the sample probe and/or re-calibrate instrument.
16. Inclusion of one or more well-characterized sera in a test run
can provide an indication of when test sensitivity is
unacceptable.
17. There is no serum control for Luminex assays; however, high
positive control bead values (>20,000 MFI) and low negative
298 A.A. Zachary et al.

control bead values (<50 MFI), with unusually low test bead
values (<100 MFI) (together) may indicate that serum was not
added to the assay.
18. Test plates must be covered during incubations to prevent
evaporation.
19. Inadequate washing before addition of antiglobulin will result
in reduced specific reactivity due to blocking of the antiglobu-
lin reagent by residual immunoglobulin.
20. Inadequate washing before addition of substrate will increase
non-specific reactivity due to reactivity between free conju-
gated antiglobulin and substrate.
21. Colorimetric enzyme substrate and substrate buffer are light
sensitive and should be mixed immediately prior to use.
22. Time and temperature of the substrate incubation are critical
and failure to adhere to manufacturer’s specifications may lead
to increased background.
23. Optimal reactivity requires that stopping solution be added at
the required time.
24. Plate readers should be calibrated according to manufacturer’s
recommendations or whenever problems arise. If trays are
stored for rereading, OD values will diminish over time.

5. Data Analysis
and Interpretation
Analysis software is available for all products and the manufacturers
also provide calculations for performing analyses manually. The ini-
tial calculations performed by the software are to determine posi-
tivity and negativity. These calculations take into account the
overall reactivity of the assay as determined from the negative con-
trol bead(s) and background noise due to individual bead variabil-
ity. The values for background noise are lot specific. Also, different
assays within and between manufacturers may yield different negative
and positive control bead values. Thresholds for positivity can be
altered by the user. The software can also perform analyses to
determine specificity, correlation coefficients, chi-squared values,
two-by-two analysis data for each specificity, and values normalized
for antigen density. The analysis software can: provide customized
reports designed by the user and lot-specific updates for back-
ground noise; allow the user to view all data for a particular patient;
search by sample, patient, or test run; and generate single patient
or batch reports in a variety of document formats.
Specificities may be analyzed by antigen, allele, or cross-reac-
tive antigen groups.
 17 HLA Antibody Detection and Characterization… 299

We have found that computer-generated interpretation of

results is not sufficient for accurate specificity assignment and that
all results should be examined manually. Attention must be paid to
controls. Very low negative controls may result in low test values
being scored by computer analysis as positive. Low positive con-
trols (MFI < 12,000) or high negative controls (MFI > 400) may
indicate the presence of interfering factors that may result in
reduced test serum values and incorrect specificity determination
(this will be discussed further in Subheading 7).
When using a phenotype panel and listing reactions in
descending order of strength, phenotypes bearing a common target
antigen will be clustered together when the test serum contains
antibody to that antigen. This can provide two types of informa-
tion. First, the strongest reacting phenotypes often contain two or
more antigens that are targets of antibodies in the serum, providing
an estimate of the collective strength of those antibodies. Second,
the test scores of these clustered phenotypes may be low and inter-
preted as negative by the computer but the non-random assortment
of antigens strongly suggests the presence of low-level antibody.
Certain beads or wells may give consistently high results with
negative control serum or consistently low or negative results with
specific antibody. This information should be taken into account
when assigning specificity.
Both the alpha and beta chain of HLA-DQ and HLA-DP mol-
ecules are polymorphic and interpretation of positive reactions
with DQ or DP molecules should take into account the possibility
of specificity to either chain. It may be possible to identify antibody
directed to antigens on the alpha chain by the clustering of anti-
gens with a common alpha chain. Greater assurance of that
specificity is provided by a positive reaction with a molecule com-
prised of the suspect alpha chain and a beta chain found in the
serum donor’s own phenotype. Additionally, a serologic epitope
may depend on the combination of certain alpha and beta chains
(20). Specificity for this type of epitope may be difficult to define
with certainty if the panel contains only one representation of that
alpha-beta combination. Further confirmation may be obtained by
performing a crossmatch with cells bearing the target molecule.
However, this would require that the antibody is strong enough to
yield a positive reaction and that the serum does not contain anti-
bodies to other antigens on the target phenotype.
There is a maximum amount of antibody that can bind to
targets. For sera that contain more antibody than can be bound to
the target, the relative strength of the antibody cannot be assessed
accurately. In such cases, tests of diluted serum may yield better
information about antibody strength, but the user is advised to
note any manufacturer’s cautions or recommendations about
interpreting data from serum diluted beyond the manufacturer’s
recommendations and to interpret data from such dilutions with
caution.
300 A.A. Zachary et al.

6. Quality Control

Each lot of kits should be tested with human serum from a healthy,
non-sensitized male or a pool of sera from such individuals. The
serum should be known to lack HLA-Ab and should give unques-
tionable negative reactions with all or nearly all targets, individual
beads, or wells (except the internal positive control if one is included
in the kit). Higher than expected reactions with any target may
indicate the likelihood of non-specific reactivity and/or higher
than expected reaction strength with that target in the presence of
antibody specific for the target. Reactivity with the negative con-
trol serum also provides information about the different back-
ground reactivities of the targets, although this should be accounted
for by the manufacturer’s software.
Each new lot of kits should be tested with several sera contain-
ing different, well-characterized, HLA-Ab to check the sensitivity
and specificity of the lot.
Each run should include positive and negative control sera to
determine the validity of the run.
Ongoing monitoring of each target should be performed to
assess the possibility of inappropriately low reaction strength or
increased background. The data should also be used to evaluate
lot-to-lot differences in reaction strength.
Positive and negative control targets should be monitored for
consistency and level of reactivity.
Whenever new targets are introduced into a new lot, their
reactivity should be evaluated with antibody specific for the HLA
antigen(s) of that target.

7. Technical Issues

Variability exists among formats, among platforms, among similar

products from different manufacturers, between lots of the same
product from the same manufacturer, between tests run on differ-
ent days, and between tests run by different technologists. We have
already noted that specificity and sensitivity are greater for single
antigen panels than for phenotype panels and for phenotype panels
than for pooled antigen panels. Not surprisingly, single antigen
panels appear to be less robust than do phenotype panels. This may
be due in part to the higher amounts of individual antigen in the
single antigen panels. If one antigen in a phenotype is susceptible
to losing its native conformation, the impact on reaction strength
may be buffered, to some extent by other antigens in the pheno-
type. However, the effect will be more obvious when the affected
 17 HLA Antibody Detection and Characterization… 301

antigen is the only target. Further, the increased sensitivity of the

single antigen panel may make it more susceptible to environmen-
tal variables. In phenotype panels, the effect of distortion of one
antigen is greater when that antigen is in the homozygous state or
when the amount of that antigen has been increased above normal
by the manufacturer.
As noted earlier, we and others have observed differences in
sensitivity and in susceptibility to background among the different
platforms.
The single antigen products from different manufacturers dif-
fer both in the panel composition and in the sensitivity and
specificity of the reactions with different antigens; therefore, use of
more than one product can provide complementary information.
While use of more than one manufacturer’s kit may be cost pro-
hibitive, the cost may be offset by the improved information
obtained from using more than one kit, particularly when accurate
assessment of antibody strength is critical.
We and others have observed significant differences in the sen-
sitivity of different lots of the same product. Recently, we docu-
mented mean increases in sensitivity between consecutive lots of a
single antigen multiplex bead kit of 5,500 MFI for the A locus and
2,400 MFI for the B locus, representing a 125–275% increase in
reactivity (unpublished data). On the one hand, this increased sen-
sitivity could be beneficial if the specificity is high; but on the other
hand, this is a serious problem when monitoring sequential sam-
ples from a single patient. The problem is exacerbated if the labora-
tory is unaware of this change until it becomes apparent after
extended use of the product.
Day-to-day and tech-to-tech variability is inevitable for sero-
logic assays but can lead to incorrect interpretation of antibody
strength. After introducing a robotic liquid handling system into
the laboratory, we performed an analysis of the variability between
sequential duplicate tests with the pooled antigen format per-
formed by a single technologist compared to duplicate tests per-
formed using the robotic device. We tested sera with different
strengths of reactivity. The variability in reaction strength for the
technologist-performed assays was 7.9–22.8 compared to only
4.9–6.6% for the robotic-assisted liquid handling assays (21). Run-
to-run variability must be accounted for when interpreting results.
An example is given in Fig. 1, which shows the strength of HLA-
Abs from sequential serum specimens from the same patient plot-
ted against the strength of the positive control. The left side of the
plot shows an unmistakable increase in HLA-Abs with a much
more limited change in the reaction strength of the positive con-
trol. However, beginning with week 6, the HLA-Ab strength and
that of the positive control are parallel suggesting that apparent
changes in the HLA-Ab strength most likely reflect day-to-day dif-
ferences in test sensitivity. Correct interpretation of changes in
302 A.A. Zachary et al.

Fig. 1. Test serum vs. positive control. Run-to-run variability must be accounted for when interpreting results of HLA anti-
body testing. The strength of HLA-Abs from sequential serum specimens from the same patient is plotted against the
strength of the positive control. The left side of the plot shows an unmistakable increase in HLA-Abs with a much more
limited change in the reaction strength of the positive control. However, beginning with week 6 (dotted vertical line), the
HLA-Ab strength and that of the positive control are parallel suggesting that apparent changes in the HLA-Ab strength most
likely reflect day-to-day differences in test sensitivity.

HLA-Ab is critical when these data are used to determine clinical

treatment.
The amount and condition of antigen is known to vary
among the different HLA antigens (22). Distortion of the HLA
target will yield a false positive reaction with antibody reactive
with a non-native epitope resulting from the distortion of the
molecule and will yield reactions that are incorrectly negative or
weaker than expected with antibody to a native HLA epitope
that is lost.
Differing amounts of antigen will yield different reaction
strengths with a fixed amount of antibody. The amounts of Cw,
DQ, and possibly DP antigens are known to be increased above
that of other antigens in some kits (6). This enhances the ability to
detect antibodies to these antigens, but the reaction strength can-
not be considered comparable to that for other antigens, given the
same amounts of antibody.
If the target bead or well is not completely coated with antigen
or a blocking agent, immune complexes and immunoglobulin can
adhere nonspecifically to the matrix or to the capture antibody if a
capture antibody is used. This could lead to a false positive reac-
tion. While one might anticipate that the reaction strength would
be low, this may not be the only source of non-specific binding.
 17 HLA Antibody Detection and Characterization… 303

8. Applications

The development and use of SPI for HLA-Ab has provided a

significant benefit for transplantation and for further identification
of serologic epitopes. The more specific characterization of HLA-Ab
has enhanced their detection. Applied prior to transplant, this infor-
mation permits the identification of antibodies that are likely to
yield crossmatch tests positive at different sensitivity levels such as
cytotoxicity and flow cytometry. In turn, this permits calculation of
the percentage of the donor population with which a patient will
have a positive crossmatch (23). In the United States, this is utilized
by the United Network for Organ Sharing (UNOS), the adminis-
trator of the National Organ Procurement and Transplantation
Network. Using allele and haplotype frequencies for HLA-A, -B,
-DRB1, and -DQB antigens from thousands of donors from four
racial groups, UNOS has established a program for calculating this
percentage, referred to as the calculated PRA (CPRA). An online
program is maintained by UNOS and can be accessed at http://
optn.transplant.hrsa.gov/resources/professionalResources.
asp?index=78. Transplant programs are required to test patient sera
using at least one SPI and enter unacceptable antigens from which
the patient’s CPRA is determined and incorporated into the alloca-
tion algorithm. Use of the CPRA has decreased the incidence of
failure to transplant the intended recipient because of a positive
crossmatch by 83% (24). Even without the elegant CPRA calcula-
tions, this would not have been possible without the more precise
recognition of HLA-Ab made possible by SPI.
The assessment of HLA-Ab strength and specificity by SPI can
also be used to perform a virtual crossmatch which has multiple
applications (6, 25–31). Prior to transplantation, the virtual cross-
match based on SPI results permits procurement of organs with
minimal permissible cold ischemia times, such as hearts and lungs,
from a broader geographic area than does using data from cytotox-
icity tests. The virtual crossmatch can expedite the process of paired
donation among multiple transplant centers by eliminating the
need for testing geographically distant donors who would have a
positive crossmatch. The virtual crossmatch is also useful for moni-
toring patients following transplantation when cell-based cross-
match tests are not possible because of the presence of humanized
or chimeric lymphocyte reactive therapeutic antibodies (4).
SPI are invaluable to desensitization programs. Reinsmoen
et al. (32) have correlated the results of crossmatch tests with those
of SPI and determined a maximum antibody strength that permits
successful transplantation in patients being desensitized with high-
dose IVIg. In our program, the course of donor-specific antibody
(DSA), assessed by SPI prior to transplant is used to measure treat-
ment efficacy and, in turn, adjust the treatment when indicated (33).
304 A.A. Zachary et al.

Post-transplant monitoring is essential since desensitized patients

are at higher risk for antibody-mediated rejection (AMR) and SPI
allows for rapid detection of changes in DSA strength and of the
development of new DSA.
As has been shown by several groups, post-transplant monitor-
ing of DSA, particularly in patients at higher risk for AMR, pro-
vides an opportunity for early intervention when DSA appears or
increases in strength (34–40). SPI tests are more rapid and more
sensitive and specific than are cell-based assays. With cell-based
assays, the presence of DSA was often not detected until nephrec-
tomy of a failed graft. This may have been due to adsorption of
DSA onto the graft, reducing circulating levels of DSA below that
detectable in the cell-based assays. However, SPI can detect very
low levels of circulating DSA in the presence of an allograft. We
have shown (41, 42) that DSA is always detected when renal biop-
sies are positive for C4d but that not all patients with circulating
DSA have C4d positive biopsies. Thus, early detection of DSA may
be a surrogate for the more invasive biopsy. We have also shown
that pro-inflammatory events can cause increases in the breadth
and/or strength of HLA-Ab (43). Again, early detection of these
changes can be clinically relevant to patients awaiting deceased
donor transplantation and to patients already transplanted.
An unresolved issue is the level of DSA that is clinically rele-
vant. It has been shown that increased episodes of AMR are associ-
ated with DSA levels detectable by ELISA but not with levels
detectable only by bead-based assays (9, 44). However, the risk
incurred with DSA present only at levels detected by the most
sensitive assays is one of current debate, with numerous reports
both supporting (27, 45, 46) and questioning (26, 47–49) a nega-
tive impact associated with low-level antibodies. Since any amount
of DSA indicates sensitization, there is a risk of a clinically relevant
increase in the DSA level; therefore, monitoring of such patients is
appropriate. The frequency of DSA among transplanted patients
was not discernible in the era of cell-based assays. Terasaki has
shown that dysfunction of renal grafts may not occur for months
after the development of DSA (50). Thus, the clinical conundrum
is when to intervene. SPI assessment of DSA is providing valuable
information about the effects of antibodies of different strengths
and different specificities and is likely to yield information that will
lead to improved graft outcomes.
SPI tests of HLA-Ab combined with determination of HLA
alleles have provided information about epitopes shared among
alleles and have provided the opportunity to identify antibodies to
an antigen that appears to be identical in a donor-recipient pair, but
which is encoded by different alleles in the donor and recipient
(51–53). Using SPI one can also identify antibodies to the alpha chains
of DQ and DP antigens. This information is useful to determining
the epitope composition of different HLA antigens and to reconcile
crossmatch results that would otherwise appear anomalous (20).
 17 HLA Antibody Detection and Characterization… 305

9. Considerations
and Recommend-
ations
The review of the capabilities and shortcomings of various formats
and platforms suggests that complete and accurate information
about a patient’s HLA-Abs is best obtained with the use of more
than one assay. While this may appear to be cost prohibitive, by
customizing the use of multiple assays according to the antibodies
of a patient, we have been able to maintain test costs in the lowest
quartile of independent US laboratories.
There has been much recent discussion about two issues sur-
rounding SPI: results reporting and standardization. There is inter-
est in having some quantitative measure of sensitization using test
values expressed as MFIs or OD ratios. As we have noted above,
the values are affected by several factors, including: the antibody
specificity, with antibodies to enhanced antigens yielding higher
values; by the format used, with higher values obtained with single
antigen panels than with phenotype panels; by the sensitivity of the
particular lot of reagents used; by the presence of interference
causing high background and reduced test values; and by the day-
to-day variability inherent in serologic assays. All of these factors
must be taken into consideration for a meaningful interpretation of
the test results. The trend of DSA and the crossmatch reactivity
that would be predicted from the SPI results are likely to be more
meaningful clinically. However, this implies that thorough and
accurate analysis of the test results by the laboratory scientist and
technologist is absolutely necessary to providing such meaningful
interpretation.
Standardization of test results would be extremely useful for
interpreting data from different laboratories. However, there are
several hurdles to overcome to achieve this goal. Most impor-
tantly, at present there are no reference sera with known amounts
of antibody to use as standards. Monoclonal antibodies are not
representative of the polyclonal nature of antibodies present in
most patients and even mixtures of such antibodies would not
replicate the nature of patient’s sera which contains other factors
that could affect test results. Second, while some laboratories treat
sera to remove interfering factors, at present this is not done
universally. Third, the high degree of sensitivity of SPI and its
susceptibility to variability suggests that standardization might
only be achieved by considering ranges of test values. Fourth,
standardization cannot be readily achieved until manufacturer-
provided kits are standardized from lot to lot. The question is,
how much of a problem is this. A perusal of the literature reveals
that multiple variations of the cytotoxicity crossmatch have been
used. Differences in incubation times and temperatures, the use of
an antiglobulin, the source and specificity of the antiglobulin, and
306 A.A. Zachary et al.

the ratio of cells to serum can have a significant impact on test

sensitivity. These same factors will affect the results of flow cyto-
metric crossmatch tests. Further, with flow cytometric testing,
there has been no standardization of the cytometers used, the
antiglobulin reagent used, the reading scale, or the threshold for
positivity. The only data normalization has been with the use of
ratios or molecules of equivalent soluble fluorochromes. In fact,
there are many publications of the results of allogeneic crossmatch
results without results of autologous crossmatches which, when
positive, would affect the interpretation of the allogeneic cross-
match. To date, this has not been raised publicly as an issue as has
standardization of SPI. We believe that the more important prac-
tice is to have results standardized and correlated with clinical out-
comes within a transplant program. However, this does not
eliminate the need for appropriate testing and accurate test inter-
pretation, nor does it preclude longer term efforts to achieve some
standardization in test reporting by inclusion of criteria for the
definition of clinically relevant antibodies. Despite the above noted
issues, the development and use of SPI for testing HLA-Abs has
had a very significant impact that has improved the care of trans-
plant patients, our knowledge of the immunogenicity of HLA
antigens, and is likely to lead to increased understanding of the
humoral response to HLA antigens.

References
1. Oh SK et al (1975) Soluble HL-A antigens in 8. Zeevi A, Girnita A, Duquesnoy R (2006) HLA
serum. I. Isolation and purification. Eur J antibody analysis: sensitivity, specificity and
Immunol 5:161–166 clinical significance in solid organ transplanta-
2. Buelow R et al (1995) Detection of panel- tion. Immunol Res 36:255–264
reactive anti-HLA class I antibodies by enzyme- 9. Phelan D et al (2009) Living donor renal
linked immunosorbent assay or transplantation in the presence of donor-
lymphocytotoxicity. Results of a blinded, con- specific human leukocyte antigen antibody
trolled multicenter study. Hum Immunol detected by solid-phase assay. Hum Immunol
44:1–11 70:584–588
3. Zachary AA et al (1995) Evaluation of HLA 10. Leffell MS, Montgomery RA, Zachary AA
antibodies with the PRA-STAT test. An ELISA (2005) The changing role of antibody testing
test using soluble HLA class I molecules. in transplantation. Clin Transpl 259–271
Transplantation 60:1600–1606 11. Bartel G et al (2007) Determinants of the
4. Zachary AA, Leffell MS (2008) Detecting and complement-fixing ability of recipient presen-
monitoring human leukocyte antigen-specific sitization against HLA antigens. Transplantation
antibodies. Hum Immunol 69:591–604 83:727–733
5. Tait BD (2009) Solid phase assays for HLA 12. Smith JD et al (2007) C4d fixing, Luminex
antibody detection in clinical transplantation. binding antibodies—a new tool for prediction
Curr Opin Immunol 21:573–577 of graft failure after heart transplantation. Am
6. Zachary AA et al (2009) Using real data for a J Transplant 7:2809–2815
virtual crossmatch. Hum Immunol 13. Wahrmann M et al (2009) Clinical relevance of
70:574–579 preformed C4d-fixing and non-C4d-fixing
7. Fuggle SV, Martin S (2008) Tools for human HLA single antigen reactivity in renal allograft
leukocyte antigen antibody detection and their recipients. Transpl Int 22:982–989
application to transplanting sensitized patients. 14. Chin C et al (2011) Clinical usefulness of a
Transplantation 86:384–390 novel C1q assay to detect immunoglobulin G
 17 HLA Antibody Detection and Characterization… 307

antibodies capable of fixing complement in 29. Tambur AR et al (2008) Tailoring antibody

sensitized pediatric heart transplant patients. J testing and how to use it in the calculated
Heart Lung Transplant 30(2):158–163 panel reactive antibody era: the Northwestern
15. Lucas DP et al (2010) Effect of desensitization University experience. Transplantation
on DSA strength and C4D deposition. 86:1052–1059
Transplantation 90:96 30. Bingaman AW et al (2008) A virtual crossmatch
16. Zachary AA et al (2009) Naturally occurring protocol significantly increases access of highly
interference in Luminex assays for HLA- sensitized patients to deceased donor kidney
specific antibodies: characteristics and resolu- transplantation. Transplantation 86:1864–1868
tion. Hum Immunol 70:496–501 31. Appel JZ 3rd et al (2006) Role of flow cytom-
17. Kosmoliaptsis V et al (2010) Improved etry to define unacceptable HLA antigens in
Luminex-based human leukocyte antigen- lung transplant recipients with HLA-specific
specific antibody screening using dithiothre- antibodies. Transplantation 81:1049–1057
itol-treated sera. Hum Immunol 71:45–49 32. Reinsmoen NL et al (2008) Acceptable donor-
18. Houp J et al (2010) Considerations in inter- specific antibody levels allowing for successful
preting solid phase antibody data. Hum deceased and living donor kidney transplanta-
Immunol 71:s18 tion after desensitization therapy.
19. Lucas DP (2007) Removal of interfering sub- Transplantation 86:820–825
stances from sera and cells for screening and 33. Leffell MS, Zachary AA (2009) The role of the
crossmatch. In: Rosenberg JC (ed) ASHI lab- histocompatibility laboratory in desensitiza-
oratory manual, 4.2 edn. Mt. Laurel, New tion for transplantation. Curr Opin Organ
Jersey. American Society for Histocompatibility Transplant 14:398–402
and Immunogenetics 34. Leffell MS, Zachary AA (2010) Antiallograft
20. Tambur AR et al (2010) The complexity of antibodies: relevance, detection, and monitor-
human leukocyte antigen (HLA)-DQ antibod- ing. Curr Opin Organ Transplant 15:2–7
ies and its effect on virtual crossmatching. 35. Riethmüller S et al (2010) Donor-specific anti-
Transplantation 90(10):1117–1124 body levels and three generations of cross-
21. Vega RM et al (2008) Laboratory automation: matches to predict antibody-mediated rejection
coming of age. Hum Immunol 69:s105 in kidney transplantation. Transplantation
22. Warner P (2009) Nebulous humors: defining 90:160–167
alloantibodies in the 21st century. Hum 36. Hachem RR et al (2010) Anti-human leuko-
Immunol 70:623–626 cyte antigen antibodies and preemptive anti-
23. Zachary AA, Braun WE (1985) Calculation of body-directed therapy after lung transplan-
a predictive value for transplantation. tation. J Heart Lung Transplant 29:973–980
Transplantation 39:316–318 37. Zeevi A et al (2009) Emerging role of donor-
24. Cecka JM et al (2010) Calculated PRA: initial specific anti-human leukocyte antigen antibody
results show benefits for sensitized patients determination for clinical management after
and a reduction in positive crossmatches. solid organ transplantation. Hum Immunol
Am J Transplant 11:719–724 70:645–650
25. Stehlik J et al (2009) Utility of virtual cross- 38. Everly MJ, Terasaki PI (2009) Monitoring and
match in sensitized patients awaiting heart treating post transplant human leukocyte anti-
transplantation. J Heart Lung Transplant gen antibodies. Hum Immunol 70:655–659
28:1129–1134 39. Piazza A et al (2006) Post-transplant donor-
26. Ferrari P et al (2010) Virtual crossmatch specific antibody production and graft out-
approach to maximize matching in paired kid- come in kidney transplantation: results of
ney donation. Am J Transplant 11:272–278 sixteen-year monitoring by flow cytometry.
27. Morris GP et al (2010) Virtual crossmatch by Clin Transpl 323–336
identification of donor-specific anti-human 40. Reinsmoen NL, Nelson K, Zeevi A (2004)
leukocyte antigen antibodies by solid-phase Anti-HLA antibody analysis and crossmatch-
immunoassay: a 30-month analysis in living ing in heart and lung transplantation. Transpl
donor kidney transplantation. Hum Immunol Immunol 13:63–71
71:268–273 41. Kraus ES et al (2009) Subclinical rejection in
28. Amico P et al (2009) Utility of the virtual stable positive crossmatch kidney transplant
crossmatch in solid organ transplantation. Curr patients: incidence and correlations. Am J
Opin Organ Transplant 14:656–661 Transplant 9:1826–1834
308 A.A. Zachary et al.

42. Zachary AA et al (2003) Specific and durable 48. Bielmann D et al (2007) Pretransplant risk
elimination of antibody to donor HLA anti- assessment in renal allograft recipients using
gens in renal-transplant patients. Transplant- virtual crossmatching. Am J Transplant
ation 76:1519–1525 7:626–632
43. Locke JE et al (2009) Proinflammatory events 49. Aubert V et al (2009) Low levels of human
are associated with significant increases in leukocyte antigen donor-specific antibodies
breadth and strength of HLA-specific anti- detected by solid phase assay before transplan-
body. Am J Transplant 9:2136–2139 tation are frequently clinically irrelevant. Hum
44. Zachary AA, Leffell MS (2010) Barriers to Immunol 70:580–583
successful transplantation of the sensitized 50. Terasaki PI, Cai J (2008) Human leukocyte
patient. Expert Rev Clin Immunol 6:449–460 antigen antibodies and chronic rejection: from
45. Verghese PS et al (2010) Impaired graft sur- association to causation. Transplantation
vival in pediatric renal transplant recipients 86:377–383
with donor-specific antibodies detected by 51. El-Awar N et al (2009) Epitopes of HLA-A, B,
solid phase assays. Pediatr Transplant C, DR, DQ, DP, and MICA antigens. Clin
14:730–734 Transpl 295–321
46. Gibney EM et al (2006) Detection of donor- 52. Cano P, Fernández-Viña M (2009) Two
specific antibodies using HLA-coated micro- sequence dimorphisms of DPB1 define the
spheres: another tool for kidney risk immunodominant serologic epitopes of
stratification. Nephrol Dial Transplant HLA-DP. Hum Immunol 70:836–843
21:2625–2629 53. Marrari M et al (2010) Human monoclonal
47. Leffell MS, Zachary AA (2010) Anti-allograft antibody reactivity with human leukocyte anti-
antibodies: some are harmful, some can be gen class I epitopes defined by pairs of mis-
overcome, and some may be beneficial. Discov matched eplets and self-eplets. Transplantation
Med 9:478–484 90(12):1468–1472
 Chapter 18

Detection and Characterisation of Alloreactive T Cells

Mandvi Bharadwaj, Nicole A. Mifsud, and James McCluskey

Abstract
T cell alloreactivity is responsible for much of the morbidity and mortality associated with tissue transplan-
tation and graft versus host disease. Immunoassays for ex vivo monitoring and quantitation of alloreactive
T cells are being increasingly utilised to provide valuable information for individualised clinical manage-
ment of transplant recipients. Here we describe detailed methodologies for both traditional and novel
assays utilised for the detection, quantitation, and functional characterisation of alloreactive T cells and
highlight the key advantages and disadvantages of each system.

Key words: Allogeneic, Alloreactivity, 51Cr release assay, Enzyme-linked immunospot, Graft versus
host disease, Intracellular cytokine staining, T cells, T cell receptors, Tetramer staining

1. Introduction

ab T cell receptors (TCR) conventionally respond to foreign

peptide antigens bound to polymorphic molecules encoded by self
major histocompatibility complex (MHC) genes. However, some
T cells can also respond to non-self (allogeneic or allo) peptide-
MHC (pMHC) molecules resulting in alloreactivity. T cell allorec-
ognition is thus defined as the ability of a T cell to recognise
allogeneic MHC molecules, i.e. non-self but from the same species.
Alloreactive T cells play a central role in graft rejection and graft
versus host disease (GVHD) (1).
Historically, 51Cr release assays were specifically developed as a
simple and highly sensitive in vitro assay system to measure the
cytolytic activity of alloreactive T cells (2) expanded in response to
stimulation with donor allogeneic lymphocytes, in mixed lympho-
cyte cultures. However, the mixed lymphocyte culture only per-
mitted a qualitative analysis that confirmed the presence of

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_18, © Springer Science+Business Media New York 2012

309
310 M. Bharadwaj et al.

alloreactive T cells, giving no information on their frequency within

the culture. Subsequent development of limiting dilution analysis
(LDA) (3), wherein multiple replicates of serially diluted effector
cells (patient peripheral blood mononuclear cells [PBMC)) are co-
incubated with allogeneic stimulators (allogeneic PBMC or target
cells) for a minimum of 7 days before being analysed in a 51Cr
release assay (2), was a significant advance, facilitating quantification
of the frequency of specific T cells participating in an immune
response (4–6). A comprehensive description of the LDA is detailed
in a recent review (7). This technique was later adopted for assessing
the in vitro potential of allogeneic mismatches to generate GVHD
or allograft rejection (8).
Technological advance in the past 2 decades has enabled the
development of other novel approaches that cannot only detect,
but also quantify and characterise alloreactive T cells based on phe-
notypic and functional differences in T cell subsets, thus providing
the scientist/clinician with a wide spectrum of immunoassays.
Indeed these are being increasingly utilised for ex vivo monitoring
and quantitation of alloreactive T cells to aid the modulation of
immunosuppressive therapy post-organ transplant. The following
chapter provides detailed methods for the traditional as well as some
more recent assays utilised for the detection and characterisation
of alloreactive T cells, and highlights the advantages and disadvan-
tages of each system.

1.1. 51Chromium (Cr) Despite the increasing reluctance to utilise radioactive assay systems,
Release Assays more than 4 decades after it was first developed, the 51Cr release
assay is still considered as the gold standard for measuring the
cytolytic efficiency of CD8+ T cells for both alloreactivity (8, 9) and
anti-viral T cell responses (10, 11). Host PBMCs or alloreactive T
cells proliferating in an LDA or mixed lymphocyte culture are co-
incubated with 51Cr labelled, allogeneic stimulators (targets), and
the 51Cr released in the culture medium is measured after 4–5 h.
This provides a measurement of the specific lysis of the 51Cr labelled
targets by the alloreactive T cells. Whilst a direct correlation has
been demonstrated between precursor frequencies determined by
an LDA and clinical data on prediction and outcomes of GVHD
(12–15), there are also reports suggesting that the enumeration of
allo-specific T cell precursor frequencies by an LDA may not cor-
relate with human transplant outcomes (16–18). Furthermore,
studies based on anti-viral T cells in Epstein–Barr virus (EBV),
Influenza virus, and Lymphocytic Choriomeningitis virus infection
have highlighted inherent biological and technical problems asso-
ciated with the LDA (12, 19–22). Overall, the LDA may only
assess the precursor frequency of a single dominant alloantigen
rather than accurately reflecting the complexity of the alloresponse
generated by all of the mismatched HLA antigens.
 18 Detection and Characterisation of Alloreactive T Cells 311

1.2. Enzyme-Linked The enzyme-linked immunospot (ELISPOT) assay provides a

Immunospot Assay methodology for the detection and quantitation of activated T cells
(11, 22–25) on the basis of cytokine secreted by each cell. Briefly,
wells of specialized ELISPOT plates are coated with a primary
antibody specific for the cytokine of interest; T cells (PBMCs,
purified T cell subsets or in vitro expanded T cells) are added to the
wells and co-cultured with a stimulus (24) (e.g. allogeneic PBMCs,
target cells, antigenic peptide, or a mitogen) for a relevant time
period (24–48 h). During the incubation period, activated T cells
locally secrete cytokines that are captured by the primary antibody.
A secondary biotinylated antibody is subsequently applied for the
specific detection of the secreted cytokine followed by an enzyme-
conjugated Streptavidin. Finally, a substrate relevant for the enzyme
(conjugated to Streptavidin) is added and the enzyme converts the
substrate into an insoluble coloured product that forms a discrete
spot (see Fig. 1) at the location of the secreting cell. Enumeration
of these spots provides the frequency of the relevant T cells.
As very low frequency T cells (one cell in 100,000), secreting a
wide array of cytokines can be analysed, this technique therefore
has a broad range of clinical applications including vaccine trials
(26) and immune tracking for T cells in infectious disease (11),
cancer (27), autoimmunity (28), and transplantation (29, 30). The
ELISPOT assay has various advantages over the 51Cr release assay,
including (1) non-radioactive quantitation of activated T cells, (2)
ex vivo analysis (independent of clonal expansion), (3) less labour
intensive, (4) shorter assay time, and (5) increased sensitivity. The
recent development of the Fluorospot technique (31), which util-
ises fluorescent detection in lieu of enzymatic detection, has facili-
tated the simultaneous detection of two cytokines in a single test,
allowing differentiation of cells with alternative cytokine profiles as

Fig. 1. Figure shows pictures of three wells of an enzyme-linked immunospot (ELISPOT)

plate containing the negative control (responders alone), test (responders + allogeneic
stimulators), and a positive control (responders + viral peptide or phytohaemagglutinin).
The dark spots in the wells represent cytokine secreting cells. All samples should be
tested in triplicate and analysed under a constant setting on the ELISPOT reader.
312 M. Bharadwaj et al.

well as the co-expression of two cytokines by a single cell (31).

However, as the measurement of the number of spots generated by
each cytokine secreting T cell is assisted by computerised imaging
equipment, the user-defined options including size, gradient, and
intensity thresholds can introduce a degree of subjectivity in the
analysis of the assay and interpretation of results. This can however
be resolved by using appropriate controls and keeping the settings
of the ELISPOT/Fluorospot reader consistent for each assay.

1.3. Peptide/MHC The design and application of soluble MHC tetrameric complexes
Tetramer Staining for investigating anti-viral CD8+ T cells was first described in
1996 (32). Subsequently, tetrameric complexes have revolu-
tionised the quantitation and dissection of antigen-specific T cell
responses (32–35). The generation of pMHC tetramers is
described in a number of publications (7, 23, 32). The HLA/b2m/
peptide tetrameric complex is conjugated to a fluorochrome of
interest, such as phycoerythrin (PE), fluorescein isothiocyanate
(FITC), or allophycocyanin (A-PC), and used as a staining reagent
for flow cytometry. MHC class I tetramers have been extensively
used for the detection of antigen-specific CD8+ T cells in both
viral infections (10) and alloreactive conditions (9). However,
more recently the development of MHC class II tetramers
( 36– 38) have further advanced the analysis of CD4+ T cells in
the fields of autoimmunity (39–41), tumour antigens (42), and
virology (43).
It is noteworthy that CD8+ T cell frequencies quantitated
by tetramer staining have been shown to be higher than those
detected by conventional cytotoxicity assays (44). This is probably
due to the ability of tetramers to detect different T cell pheno-
types including naïve, memory, and effector cells and the ten-
dency for memory T cells to undergo apoptosis upon activation
(activation-induced cell death) resulting in a loss of reactive T cells
in culture (45). Compared to the traditional assays, the pMHC
complexes (1) are quantitative and sensitive (1:50,000), (2) do
not involve radioactivity, (3) require minimal material (blood,
tissue), (4) are a fast approach, (5) permit batch processing for
large numbers, (6) can be used in combination with other cell
surface antibodies used for T cell phenotyping (10, 19, 26, 32,
46, 47), and (7) do not kill the T cells allowing the cells to be
sorted for use in other assays (such as ELISPOT and intracellular
cytokine staining [ICS]). Recent studies investigating the anti-
viral T cells (10) and/or cross-reactive virus-specific T cells recog-
nising specific HLA alloantigens have demonstrated the power in
combining both pMHC tetramers and the ICS assay (10, 48, 49)
(described below). The only major limitation of this assay is that
both the peptide epitope and MHC-restriction elements must be
known for the detection of specific T cells.
 18 Detection and Characterisation of Alloreactive T Cells 313

1.4. Intracellular The activation of T cells initiates the production of cytokines that
Cytokine Staining can be detected and quantitated using ICS (7, 50, 51). Briefly,
T cells are activated through co-incubation with a stimulus for a total
of 6 h. An inhibitor of intracellular transport is added after the first
2 h to arrest the transport of newly synthesised proteins, including
cytokines, from the endoplasmic reticulum into the Golgi. Post-
incubation, the T cells are stained with a cocktail of specific mono-
clonal antibodies (mAbs) for cell surface phenotyping. The T cells
are then paraformaldehyde fixed, permeabilised in mild detergent,
and stained with anti-cytokine antibodies. Phenotypic analysis is
facilitated by flow cytometry.
The T cell frequencies detected by ICS have been shown to
correlate with frequencies obtained by pMHC tetramers in HIV
patients (52) and with ELISPOT in an Influenza study of healthy
individuals (53). Conversely, no correlation was demonstrated
between the frequencies detected by ICS and pMHC tetramers in
a metastatic melanoma study (54), and the enumeration of EBV-
specific T cells and alloreactive T cells by ICS yielded higher fre-
quencies than that obtained by the LDA-based 51Cr release assay
(55). Notwithstanding the discrepancies in the frequency of T cells
enumerated by the different assay systems, ICS has been used for
the detection of cytokine producing CD4+ T cells in the peripheral
blood of individuals with HIV (56) and quantification of activated
allo-specific CD8+ T cells in healthy controls (9, 57) and solid
organ transplant recipients (58). It has various advantages over the
other assays described above, including (1) analysis of large num-
bers of cells, (2) multiparameter analysis using combinations of
intracellular and cell surface antibodies, (3) parallel measurement
of T cell cytokine profiles, and (4) measurement of cytotoxic poten-
tial using a degranulation marker.

2. Materials

All consumables and reagents should be sterile unless otherwise

specified.

2.1. Isolation of PBMC 1. Polypropylene tubes: 50 and 10 mL.

and In Vitro Expansion 2. Polystyrene tubes: 30 mL (V-bottom).
of Allogeneic CD8+
3. Serological pipettes: 10 and 25 mL.
T Cells
4. Polypropylene 1 mL transfer pipettes.
5. Glass Pasteur pipettes.
6. Pipette tips: 0.5–10, 10–200, and 200–1,000 mL.
7. Cell culture plates: 6-well, 12-well, 24-well, 48-well, and
96-well (U-bottom).
314 M. Bharadwaj et al.

8. RPMI (Culture/assay media): RPMI 1640 media (Gibco-

Invitrogen Corporation, USA) supplemented with 150 mM
Non-essential Amino Acids (Gibco-Invitrogen Corporation,
USA), 7.5 mM HEPES (MP Biomedicals, Eschwege,
Germany), 150 mg/mL Streptomycin sulphate (Sigma, MO,
USA), 150 U/mL Benzyl-penicillin (CSL, Australia), 2 mM
L-glutamine (MP Biomedicals, Eschwege, Germany), and
76 mM 2-mercaptoethanol (Sigma, MO, USA). Should be
stored at 2–8°C.
9. Foetal calf/bovine serum (FCS or FBS, heat inactivated for
30 min at 56°C, Millipore).
10. RF10: RPMI (supplemented with antibiotics, non-essential
amino acids, and glutamine, as described in step 8) supple-
mented with 10% heat inactivated FCS/FBS (as described in
step 9).
11. Human recombinant Interleukin-2 (IL-2) (Cetus Corporation,
Emeryville, CA). This should be stored at −20°C in small
aliquots to prevent frequent freeze thawing.
12. RF10 + 20 U/mL of IL-2: RF10 supplemented with human
recombinant IL-2 to achieve a final concentration of 20 U/mL.
(Medium containing IL-2 should be made fresh but can be
stored overnight at 4°C for up to 24 h).
13. Ficoll-Paque PLUS (GE Healthcare), stored at room tempera-
ture in the dark.
14. Irrigation water (Baxter), stored at room temperature.
15. Phosphate Buffered Saline (PBS, Life Technologies), stored at
room temperature.
16. Trypan Blue (Sigma) dilute to 0.2% in PBS, stored at room
temperature.
17. Biohazard class II cabinet.
18. Centrifuge.
19. Aspirator.
20. Pipette gun.
21. Pipettes: 0.5–10, 10–200, and 200–1,000 mL.
22. Bright field microscope (cell counting).
23. Haemocytometer (cell counting).
24. Incubator: 37°C, 5% CO2.
25. Centrifuge.
26. Gamma irradiator.
27. Vortex.
 18 Detection and Characterisation of Alloreactive T Cells 315

2.2. Dissection 1. Polypropylene tube: 10 mL.

and Quantitation 2. Syringe: 1 mL.
of the Alloreactive
3. Pipette tips: 0.5–10, 10–200, and 200–1,000 mL.
CD8+ T Cell Response
4. Cell culture plate: 96-well (U-bottom).
2.2.1. 51Cr Release Assays
5. RPMI and RF10: as described in Subheading 2.1 (steps 8
and 10).
6. 200 mM peptide, stored at −20°C.
7. Na51CrO4 as the 51Cr stock solution (stored in a lead shielded
box in a room dedicated to radioisotopes) (Perkin Elmer,
USA).
8. Triton X-100 (Sigma).
9. LumaPlate-96 containing scintillant (Packard Biosciences-
Perkin Elmer, USA).
10. Plastic plate sealers.
11. Radiation badge.
12. Laminar flow or Biohazard class II cabinet.
13. Centrifuge.
14. Pipettes: 0.5–10, 10–200, and 200–1,000 mL.
15. Multichannel pipette: 10–200 mL.
16. Bright field microscope (cell counting).
17. Haemocytometer (cell counting).
18. Incubator: 37°C, 5% CO2.
19. Lead lined Esky to transport labelled targets.
20. Top count plate reader (Top count Microplate; Packard
Instrument Co, Meriden, CT).
21. Biohazard bag and containers for low level and high level
“hot-waste”.
22. Geiger-Müller counter.

2.2.2. ELISPOT Assay 1. Polypropylene tubes: 1.5 and 10 mL.

2. Serological pipettes: 10 mL.
3. Pipette tips: 0.5–10, 10–200, and 200–1,000 mL.
4. Reagent reservoirs.
5. Aluminium foil.
6. Capture mAb: Anti-IFN-g (suggested clone l-DIK, anti-IFN-g
mouse IgG1 1 mg/mL, from Mabtech) (see Note 1).
7. Detection mAb: Anti-IFN-g-biotin (suggested clone 7-B6-1-
biotin anti-IFN-g mouse IgG1-1 mg/mL from Mabtech).
316 M. Bharadwaj et al.

8. Multiscreen plates (0.45 ~ 1 m) from Millipore Australia Pty.

Ltd (Sterile; 10 plates) (see Note 1).
9. Streptavidin-alkaline phosphatase (Lyophilized powder, 1 mg
protein, Sigma; resuspended in 1 mL of sterile H2O).
10. 5-bromo-4-chloro-3-indolyl phosphate/Nitro blue tetrazolium
(BCIP/NBT) tablet- SigmaFast™ (Sigma).
11. PBS.
12. PBS + 5% FCS.
13. PBS + 0.05% Tween 20 (PBST) (Sigma).
14. 0.1 M sodium bicarbonate (NaHCO3), pH 9.6.
15. 1 mg/mL of Phytohaemagglutinin (PHA).
16. Stock solutions of 200 mM peptides stored at −20°C.
17. RPMI and RF10: as described in Subheading 2.1 (steps 8
and 10).
18. Biohazard class II cabinet.
19. Centrifuge.
20. Pipettes: 0.5–10, 10–200, and 200–1,000 mL.
21. Multichannel pipette: 10–200 mL.
22. Bright field microscope (cell counting).
23. Haemocytometer (cell counting).
24. Incubator: 37°C, 5% CO2.
25. ELISPOT reader (AID Autoimmun Diagnostika GmbH,
Germany).

2.2.3. Intracellular 1. Cell culture plates: 96-well (U-bottom).

Cytokine Staining 2. Polypropylene tubes: 1.5 and 10 mL.
3. Polystyrene tube: 5 mL (non-sterile).
4. Glass Pasteur pipettes.
5. Serological pipettes: 10 mL.
6. Pipette tips: 0.5–10, 10–200, and 200–1,000 mL.
7. Reagent reservoirs.
8. Aluminium foil.
9. RPMI and RF10, RF10 + 20 U/mL IL-2: as described in
Subheading 2.1 (steps 8, 10, and 12).
10. Geneticin (G418, Life Technologies).
11. Hygromycin B (Life Technologies).
12. 200 mM peptide, stored at −20°C.
13. Monoclonal antibodies, stored at +2 to 8°C (see Note 2 and
Table 2).
14. Tetramers, stored at +2 to 8°C (see Note 3 and Table 3).
 18 Detection and Characterisation of Alloreactive T Cells 317

15. 500 mg/mL Monensin (Sigma), stored at +2 to 8°C.

16. 10 mg/mL Brefeldin A (Sigma), reconstituted in 100% methanol
and stored at −20°C.
17. 16% Paraformaldehyde (ProSciTech), stored at room tempera-
ture. Prior to use in assay, dilute to 1% final concentration in
PBS. Mix well and store at +2 to 8°C.
18. 1% Saponin (Sigma), stored at room temperature. Dissolve 0.5 g
Saponin in 50 mL Milli-Q water. Mix well and filter through
0.22-mm disc filter. Prior to use in assay, dilute to 0.3% final con-
centration in PBS. Mix well and store at room temperature.
19. PBS, stored at room temperature.
20. Biohazard class II cabinet.
21. Centrifuge.
22. Incubator: 37°C, 5% CO2.
23. Aspirator.
24. Pipette gun.
25. Pipettes: 0.5–10, 10–200, and 200–1,000 mL. Multichannel:
5–50 and 200–1,000 mL.
26. Bright field microscope (cell counting).
27. Haemocytometer (cell counting).
28. Flow cytometer for FACS.
29. Flow Cytometric software: CellQuest™ (Becton Dickinson),
FlowJo (Treestar Inc., Ashland, OR) (see Note 4).

2.2.4. Tetramer Staining 1. Cell culture plates: 96-well (U-bottom).

2. Polypropylene tubes: 1.5 and 10 mL.
3. Polystyrene tube: 5 mL (non-sterile).
4. Glass Pasteur pipettes.
5. Pipette tips: 0.5–10, 10–200, and 200–1,000 mL.
6. Aluminium foil.
7. RPMI, stored at +2 to 8°C.
8. RPMI and RF10, RF10 + 20 U/mL IL-2: as described in
Subheading 2.1 (steps 8, 10, and 12).
9. Geneticin (G418, Life Technologies).
10. Hygromycin B (Life Technologies).
11. 200 mM peptide, stored at −20°C.
12. Monoclonal antibodies stored at +2 to 8°C (see Note 2 and
Table 2).
13. Tetramers, stored at +2 to 8°C (see Note 3 and Table 3).
14. Biohazard class II cabinet.
15. Centrifuge.
318 M. Bharadwaj et al.

16. Incubator: 37°C, 5% CO2.

17. Pipette gun.
18. Bright field microscope (cell counting).
19. Haemocytometer (cell counting).
20. Pipettes: 0.5–10, 10–200, and 200–1,000 mL. 8-channel: 5–50
and 200–1,000 mL.
21. Flow cytometer for FACS.
22. Flow Cytometric software: CellQuest™ (Becton Dickinson),
FlowJo (Treestar Inc., Ashland, OR).

3. Methods

Carry out all procedures at room temperature unless otherwise

indicated. Handling of human samples and disposal of biohazard
waste should be performed according to the international guide-
lines on biosafety level 2 containment (or similar for each country,
e.g. physical containment level 2 in Australia).
Allo-specific CD8+ T cells can either be detected ex vivo by
ELISPOT or tetramer staining, or after in vitro expansion in a
mixed lymphocyte reaction (MLR), commonly utilised for 51Cr
release assays and ICS. For an MLR, responder (alloreactive)
PBMC are stimulated with irradiated allogeneic PBMC at a 2:1
ratio. Antigen-specific CD8+ T cells are generally used as positive
controls and require stimulation of responder PBMC with irradi-
ated, peptide-pulsed autologous PBMC at a 2:1 responder to
stimulator ratio (see Note 5). Antigenic peptides derived from
common viruses such as Influenza virus, EBV, or Cytomegalovirus
(CMV) (59) with known HLA-restriction are utilised here.
Selection of sterile cell culture plates depends on the seeding density
of PBMC (see Table 1, adapted from Life Technologies (formerly
Invitrogen) GIBCO cell culture system catalogue).

Table 1
Selection of cell culture plates

Surface Seeding density Culture

Culture plates area (cm2) (maximum) volume (mL)

6-well 9.6 30 × 106 10

12-well 3.8 10 × 106 4
6
24-well 2.0 5 × 10 2
6
48-well 0.75 2.5 × 10 1
Adapted from GIBCO cell culture system catalogue
 18 Detection and Characterisation of Alloreactive T Cells 319

Following in vitro expansion in an MLR (see Subheading 3.1),

alloreactive T cells can be detected directly by tetramer staining
if the alloantigen and its HLA restriction are known (9) (see
Subheading 3.2). Alternatively, the total alloresponse can be dis-
sected following a short restimulation with B-lymphoblastoid cell
lines (B-LCL) carrying the allogeneic HLA allotypes or restimula-
tion with cell lines (such as the MHC class I reduced cell line,
C1R) transfected with single HLA class I molecules similar to that
of the allogeneic stimulators. These cell lines are referred to as “reporter
cells” (see Note 6) and facilitate the determination of the contribu-
tion of each mismatched HLA alloantigen to the total alloresponse.
Antigen-specific CD8+ T cells expanded against known viral anti-
gens (positive controls) are restimulated with peptide-pulsed
reporter cells carrying the appropriate HLA. The restimulation of
allo-specific T cells activates these cells, enabling functional charac-
terisation via production or upregulation of both cell surface
molecules and intracellular proteins/cytokines that can be analysed
in an ELISPOT, 51Cr release, or ICS assay.

3.1. Isolation of PBMC 1. Collect peripheral blood in appropriate blood collection tubes
and In Vitro Expansion containing anti-coagulant (see Note 7).
of Allo-Specific CD8+ 2. To a 50 mL tube, add 5–20 mL of peripheral blood and an
T Cells equal volume of RPMI to achieve a 1:2 dilution of the periph-
eral blood.
3. To another 50 mL tube, add 10 mL of Ficoll-Paque PLUS
(see Note 8).
4. Using a 25 mL serological pipette gently overlay the diluted
peripheral blood onto the Ficoll-Paque PLUS reagent using
the slow setting on the pipette gun (see Note 9).
5. Centrifuge for 20 min at 931 × g, room temperature with the
brake setting = OFF (see Note 10).
6. Blood components will be fractionated and plasma can be col-
lected using a 1 mL transfer pipette and stored at −80°C if
required (see Fig. 2).
7. Using a sterile 1 mL transfer pipette, carefully collect lympho-
cyte/monocyte/platelet layer (interface) into a 30 mL
V-bottom tube. Wash cells in a final volume of 20 mL of RPMI.
8. Centrifuge for 15 min at 524 × g, room temperature.
9. A cell pellet should be visible at the bottom of the tube, discard
supernatant by aspiration, leaving approximately 1 mL of resid-
ual volume and resuspend cell pellet.
10. Transfer the cell pellet to a 10 mL tube containing 8 mL of
RPMI and centrifuge for 10 min at 524 × g, room temperature.
Repeat wash by resuspending the cell pellet in 8 mL of RPMI
and centrifuging for 10 min at 524 × g, room temperature.
320 M. Bharadwaj et al.

Fig. 2. Isolation of PBMCs. Peripheral blood is layered over the Ficoll-Paque and, following centrifugation (with brake OFF),
the blood components are separated into (1) plasma, (2) lymphocytes, monocytes, platelets, and (3) granulocytes, erythro-
cytes. The lymphocytes, monocytes, and platelets layer can be extracted, centrifuged to remove platelets with the residual
cells representing the peripheral blood mononuclear cell population.

11. Resuspend the cell pellet in 8 mL of RPMI and centrifuge for

2 min at 931 × g, room temperature (see Note 11).
12. Discard supernatant by aspiration and resuspend the cell pellet
in 1 mL of assay media or PBS.
13. Perform a cell count for viable cells, using 0.2% Trypan Blue.
14. PBMC can be used immediately or cryopreserved.
15. To set up an MLR: Resuspend the responder PBMC in an
appropriate volume of assay media (RF10 + 20 U/mL IL-2)
(see Note 12) to achieve a final concentration of 2 × 106 cells/
mL. Add the right volume of responder cell suspension to a
suitable cell culture plate (see Table 1) and incubate the plate
at 37°C, 5% CO2. For example, to achieve a 2:1 responder to
stimulator ratio, one would add 1 mL of responders (2 × 106
cells/mL) and 1 mL of stimulators (1 × 106 cells/mL) to 1 well
of a 24-well plate.
16. Allogeneic stimulator cells: Resuspend the allogeneic PBMC at
a concentration of 3 × 106 cells/mL and gamma-irradiate at
3,000 Rad (30 cGy) (see Note 13). After irradiation, wash cells
by resuspending in 10 mL of RPMI and centrifuging for 7 min
at 524 × g. Resuspend the cell pellet in assay media
(RF10 + 20 U/mL IL-2) to achieve a final concentration of
1 × 106 cells/mL.
17. Preparation of peptide-pulsed autologous PBMC as positive
control (for virus-specific T cells): Pellet autologous PBMCs
and resuspend in plain RPMI to achieve 1 × 106 cells/100 mL
at 10 mM final concentration of peptide (e.g. 5 mL of 200 mM
peptide stock/100 mL of RPMI). Incubate at 37°C, 5% CO2
for 1 h. Vortex every 15 min. Wash cells by resuspending in
10 mL of RPMI and centrifuging for 7 min at 524 × g.
 18 Detection and Characterisation of Alloreactive T Cells 321

Resuspend the cell pellet in assay media (RF10 + 20 U/mL IL-2)

to achieve a final concentration of 1 × 106 cells/mL.
18. Add the irradiated allogeneic PBMC and peptide-pulsed autol-
ogous PBMC to separate wells containing the responder
PBMC, to achieve a 2:1 ratio of responder to stimulator. Mix
cells gently and thoroughly and incubate the plate at 37°C, 5%
CO2 for 5–7 days.
19. The in vitro expansion of allo-specific or antigen-specific CD8+
T cells is dependent on the kinetics of the responder cell pro-
liferation in response to the stimulus, and also on the cytokine/
protein being measured (see Fig. 3, Note 14).

Fig. 3. In vitro expansion of allo-specific and antigen-specific CD8+ T cells. (a) PBMC from
a healthy individual was stimulated for up to 16 days with allogeneic PBMC that enabled
generation of allo-specific CD8 +T cells against all three mismatched HLA-A1, -B8, and
-B57 allotypes. The maximal CD8+ T cell response, measured by the production of IFN-g,
was observed at 12–13 days post-stimulation. (b) PBMC from a HLA-B8 healthy individual
was stimulated with autologous PBMC pulsed with EBV peptide RAKFKQLL (RAK) for up to
13 days. The maximal response, measured in HLA-B8/RAK tetramer+CD8+ T cells by the
production of IFN-g, was observed at 12–13 days post-stimulation.
322 M. Bharadwaj et al.

3.2. Dissection and 1. Preparation of responder (effector) cells: Count effector cells
Quantitation of the and serially dilute in 2 mL of RF10 in a series of 10 mL tubes
Alloreactive CD8+ to achieve a cell concentrations of 2 × 106, 1 × 106, and 5 × 105
T Cell Response cells/mL. The aim is to achieve multiple replicates of an
effector:target (E:T) ratio of 100:1, 50:1, and 25:1 in the assay.
3.2.1. 51Cr Release Assay
2. Preparation of reporter (target) cells—51Cr labelling: Wear
protective eyewear, gown, radiation badge, and double gloves.
Scan the area in which you are planning to perform 51Cr label-
ling of target cells with a Geiger-Muller counter to assess if
there is a pre-existing background level of radiation. Use the
Geiger counter when using the isotope to monitor spills and
exposure (see Note 15).
(a) Count allogeneic reporter or target cells and aliquot 3 × 106
cells/10 mL tube into two tubes for test sample and posi-
tive (viral) control (traditionally viral controls are target
cells pulsed with 10 mg/mL of peptide antigen derived
from EBV/Influenza/CMV, for 1 h at 37°C).
(b) Centrifuge, both test and viral control tubes for 7 min at
335 × g at room temperature to pellet the cells.
(c) Discard supernatant by aspiration and resuspend the cell
pellet in 100 mL of RPMI.
(d) Label target cells with 51Cr by adding 50–100 mL of
Na51CrO4 to each 10 mL tube using a 1-mL syringe (see
Note 15).
(e) Mix each tube gently and incubate for 1 h at 37°C.
(f) Wash the 51Cr-labelled target cells in a final volume of
10 mL in cold RF10 and centrifuge for 7 min at 800 × g at
room temperature.
(g) Repeat the wash step an additional two times.
(h) Resuspend the 51Cr-labelled target cells in 1 mL of RF10.
(i) Count cells and resuspend to 2 × 104 target cells/mL in
RF10.
3. Add 100 mL of serially diluted responder (effector) cells
(2 × 106, 1 × 106, and 5 × 105 cells/mL—as described in step 1)
to wells of a 96-well U-bottom plate. Cells at each different
concentration (titrated for different E:T ratios) should be
added into three wells of a 96-well U-bottom plate (see Fig. 4).
4. Add 100 mL of 51Cr-labelled target cells (2 × 103 target cells) as
described in step 2 to each well of the 96-well U-bottom
plate.
5. Negative controls (media only): 3–6 wells containing 100 mL
of 2 × 103 target cells/well (without responders) plus an addi-
tional 100 mL RF10/well is used to calculate the spontaneous
release of 51Cr from the target cells.
 18 Detection and Characterisation of Alloreactive T Cells 323

Effectors Effectors Targets Targets

E:T ratio E:T ratio + + Triton X-
100:1 50:1 Media 100

Target 1
Target 2
Target 3
Target 4
Target 5
Target 6
Target 7
Target 8

Fig. 4. Proposed plan for setting up the 51Cr release assay at an E:T ratio of 100:1 and
50:1. Other ratios may be tested similarly in different plates. Target 1–8 represent the
different cell lines (BLCL or transfected C1R) that need to be tested for allogeneic reactivity
to the responders (or effectors), the untransfected parent C1R cell line and a positive control
(reporter cell line pulsed with a viral peptide). All targets are resuspended in the assay media
at a set concentration (e.g. 2 × 104 target cells/mL) and the effectors or responder cells
are serially diluted to achieve an E:T ratio of 100:1, 50:1, and 25:1 (and more if required).

6. Positive controls (1% of Triton X-100): 3–6 wells containing

100 mL of 2 × 103 target cells/well (without responders) plus
100 mL of 1% Triton X-100/well is used to calculate the maxi-
mal release of 51Cr from the target cells (see Note 16).
7. Ensure that each well contains 51Cr-labelled target cells with
effector cells, media (RF10) only, or 1% of Triton X-100 (see
Fig. 4).
8. Centrifuge the plate at 233 × g for 3 min, to pellet cells.
9. Incubate at 37°C, 5% CO2 for 4–5 h.
10. Centrifuge the plate at 233 × g for 3 min, to ensure cells are at
the bottom of the well.
11. Using a multichannel pipette, remove 25 mL of supernatant
from each well and transfer directly to the same position on a
LumaPlate, which already contains scintillant.
12. Leave LumaPlate at room temperature to dry overnight.
13. The 51Cr levels in assay supernatant samples collected on the
LumaPlate are measured on a beta scintillation counter and
the % Specific Lysis is calculated using the following formula
(see Note 17):

(Experimental release − Spontaneous release)

× 100
(Maximal release − Spontaneous release)
324 M. Bharadwaj et al.

14. Using the mean plus 3 standard deviations (SD) cut-off, the
number of negative wells is determined for each dilution.
The relationship between the number of negative wells and the
mean number of precursors are plotted, and a frequency of
alloreactive T cells is obtained (3, 60, 61).

3.2.2. ELISPOT Assay Procedures on day 0 and day 1 are to be performed under sterile
conditions in a Biohazard class II cabinet to maintain sterility of
plate and reagents.
Day 0: Coat Plate
1. Prepare capture mAb as per manufacturer’s instructions. For
example: dilute anti-IFN-g mAb (1-DIK anti-IFN-g mouse
IgG1) to 10 mg/mL in 0.1 M sodium bicarbonate buffer pH
9.6 or 1× PBS and dispense 100 mL to each well of the
ELISPOT plate (see Notes 18 and 19). Incubate plate with lid
overnight (approximately 18 h) at 4°C (refrigerator) on a level
surface.
Day 1:
2. Plate blocking: After overnight incubation, retrieve test plate
from 4°C incubation and wash wells by flicking off coating
mAb solution into discard tray. Then add 200 mL sterile PBS
per well and flick off into discard tray. Repeat washing another
five times. After the final wash add 200 mL per well of sterile
PBS with 5% FCS (blocking solution) to test plate, incubate
for 1 h at 37°C (or at room temperature if to be left longer).
3. Prepare effectors: PBMC (fresh or cryopreserved) or in vitro
expanded alloreactive T cells can be used as effector cells for
assessing the alloreactive response (see Note 20). Perform cell
count and adjust cell concentration to 2 × 106 cells/mL to seed
at a final concentration of 2 × 105 cells/well/100 mL in the
ELISPOT plate (see Note 19). If using in vitro expanded
alloreactive T cells, the cell concentration should be between
1 × 103–5 × 104cells/well.
4. Prepare targets: Allogeneic PBMC depleted of CD3+ T cells
(B cell enriched) are utilised (see Note 21). Relevant B-LCLs or
reporter cell lines (see Subheading 3 introduction) can also be
used although they may give a high background. Seed target
cells in limiting dilution (5 × 104, 2.5 × 104, and 1 × 104 cells/
well) in triplicate in a total volume of 100 mL to appropriate
wells.
5. Positive control: A final concentration of 10 mg/mL of either
phytohemagglutinin (PHA) or a known anti-viral peptide
(EBV, CMV, or Influenza A) restricted to the same HLA mol-
ecules expressed by the effector cells, in a total volume of 5 mL
serves as positive control.
 18 Detection and Characterisation of Alloreactive T Cells 325

6. Negative control: Replicate wells with effector cells only

(without targets) and targets only (without effectors) serve as
negative control.
7. The ELISPOT plate is incubated for 20–24 h at 37°C, 5% CO2
(see Note 22).
Day 2:
8. After incubation remove cells by flicking plate over a discard
tray in the biosafety cabinet. Wash wells with 200 mL/well of
filtered 0.05% PBST (PBS containing 0.05% Tween 20; see
Note 23) using a multichannel pipette (can use a wash bottle
with a blunt nozzle positioned very close to the well). Wash an
additional two times.
9. Change pipette tips and repeat three washes with 200 mL/well
of filtered PBS (no Tween) (see Note 24). Dry the ELISPOT
plate by gently blotting inverted plate on a stack of tissues
or paper towels just before adding the next reagent.
10. Dilute biotinylated detection mAb (7-B6-1 anti-IFN-g mouse
IgG1) in filtered PBS to a final concentration of 1 mg/mL in a
10-mL tube.
11. Dispense 100 mL of detection mAb to each well of ELISPOT
plate and incubate for 4 h at room temperature on a level surface,
in darkness.
12. After 4 h incubation, wash ELISPOT plate six times, as
described in steps 8 and 9. This step can be performed over the
sink, rather than within the biosafety cabinet.
13. Dilute stock Streptavidin-alkaline phosphatase (SAP) (1 mg/
mL) in filtered PBS to a final concentration of 10 mg/mL in a
10-mL tube.
14. Dispense 100 mL (1 mg) of SAP to each well of ELISPOT plate
and incubate for 2 h at room temperature (see Note 25).
15. Prepare substrate for the SAP by dissolving one BCIP/NBT
tablet in 10 mL of sterile H2O, filter through syringe filter
(0.2 mm ministart filter, Sartorius) (see Note 26). The substrate
should be prepared no more than 15 min prior to use as the
solution is light sensitive and, once prepared, it should be stored
in the dark/wrapped in foil.
16. Wash ELISPOT plate (from step 14) six times as described in
step 8 and dispense 100 mL/well of BCIP/NBT substrate
solution. Wrap plate in aluminium foil and incubate at room
temperature, on a flat surface. The plate is usually left for
10–60 min, although in some cases it may take longer for full
development.
17. When the spots are visible (see Fig. 1), stop colour develop-
ment by washing in tap water.
326 M. Bharadwaj et al.

18. Tease away the bottom plastic covering gently and wash the
bottom of the ELISPOT plate. Allow the plates to dry over-
night in open air on bench top. Once dry, the plate is ready to
be analysed using an ELISPOT reader.
19. Results are presented as mean values of spots detected in triplicate
wells containing effectors plus stimulator cells or antigen, after
subtracting the response of wells with responder cells or effector
cells alone (normally less than 10 spots per 200,000 cells).

3.2.3. Intracellular Before an ICS is conducted, individual mAbs (to be used for phe-
Cytokine Staining Assay notyping and functional analysis) should be titrated to determine
amounts for optimal staining (see Table 2). Prior to the assay,
alloreactive/responder T cells derived from an MLR are restimu-
lated by specific reporter cells (i.e. transfected cell lines or B-LCLs),
and the antigen-specific CD8+ T cells are restimulated by appropri-
ate reporter cells pulsed with the relevant peptide. Reporter cells
for alloreactive T cells should carry HLA allotypes corresponding
to the allogeneic stimulator (PBMC utilised for the initial stimula-
tion). For example, for dissecting T cells from an MLR in which
responder PBMC were expanded with an allogeneic stimulator
carrying: HLA-A*0101, A*0201; HLA-B*0801, B*5701, the
following reporter cell lines would be required:
Transfected cell line panel: C1R parental, C1R.A*0101,
C1R.A*0201, C1R.B*0801, C1R.B*5701.
B-LCL panel: Cell lines expressing individual HLA allotypes-
A*0101, A*0201, B*0801, or B*5701.
1. Calculate the total number of T cells and reporter cells required
for the assay (for the test sample, negative controls, and com-
pensation controls) considering that you will require 2 × 105
T cells/100 mL of assay media (RF10 + 20 U/mL of IL-2) per
well and 1 × 105 reporter cells/100 mL of assay media

Table 2
Selection of monoclonal antibodies used for characterising
cell phenotype and function

Antibody Flurochrome Clone Isotype Dilution Manufacturer

CD3 APC SK7 IgG1 1:20 Becton Dickinson

CD4 PE SK3 IgG1 1:20 Becton Dickinson
CD8 PE-Cy5 HIT8a IgG1 1:20 Becton Dickinson
CD107a FITC H4A3 IgG1 1:20 Becton Dickinson
IFN-g APC B27 IgG1 1:1,000 Becton Dickinson
TNF-a PE MAb11 IgG1 1:20 Becton Dickinson
 18 Detection and Characterisation of Alloreactive T Cells 327

(RF10 + 20 U/mL of IL-2) per well. Each test sample should be

assessed in duplicate or triplicate. In a 96-well U-bottom plate:
(a) Test: Dispense 2 × 105 cells/100 mL/well of responder T
cells and 1 × 105 cells/100 mL/well of transfected C1R cell
line or B-LCL.
(b) Negative control: Dispense 2 × 105 cells/100 mL/well of
responder T cells and add 100 mL of assay media.
(c) Negative control: Dispense 1 × 105 cells/100 mL/well of
transfected C1R cell line or B-LCL and add 100 mL of
assay media.
(d) FACS compensation control: Dispense 2 × 105
cells/100 mL/well of responder T cells and add 100 mL of
assay media into multiple wells for FACS compensation.
For example, for 4 fluorochromes you will require 5 wells:
Unstained, FL1 (FITC), FL2 (PE), FL3 (PE-Cy5, PerCP),
and FL4 (APC) (see Note 27).
2. Add 10 mL/well of CD107a FITC (see Note 28) and mix
carefully to avoid excess bubbles.
3. After 1 h incubation at 37°C, 5% CO2, add 1.46 mL/well of
500 mg/mL Monensin (5 mM final concentration, see Note 29).
4. After 2 h incubation at 37°C, 5% CO2, add 23 mL/well of
100 mg/mL Brefeldin A (10 mg/mL final concentration, see
Note 30).
5. Incubate for 4 h at 37°C, 5% CO2.
6. Centrifuge plate for 5 min at 335 × g, room temperature.
7. Discard supernatant by flicking off in a single, sharp motion.
8. Resuspend cells in 20 mL of antibody mix/well (antibody mix
should contain all the desired antibodies for surface staining,
e.g. anti-CD3, CD8, etc.). The following panels are included:
(a) Allo-specific CD8+ T cells are stained with antibody mix
containing anti-CD8 (and other desired antibodies for
surface staining of phenotypic markers) diluted in PBS
(see Table 2).
(b) Antigen-specific CD8+ T cells are stained with tetramer
(see Table 3) and antibody mix with anti-CD8 (and other
antibodies, see Table 2) diluted in PBS.
(c) Compensation controls: Responder T cells in PBS only
(unstained) and T cells with individual FL1 to FL4 anti-
bodies diluted in PBS (see Note 31).
9. Wrap plate in foil and incubate for 30 min at 4°C in the dark.
10. Wash cells with 150 mL/well of PBS.
11. Centrifuge plate for 5 min at 335 × g, room temperature.
12. Discard supernatant by flicking off in a single, sharp motion.
328 M. Bharadwaj et al.

Table 3
Selection of pMHC tetramers used to quantify and dissect antigen-specific T cells

Tetramer HLA restriction Virus Protein Sequence Dilution Manufacturer

B8/RAK B8 EBV BZLF1 RAKFKQLL 1:50 The University of Melbourne

B8/FLR B8 EBV EBNA3A FLRGRAYGL 1:50 The University of Melbourne
A1/YSE A1 CMV pp65 YSEHPTFTSQY 1:40 Monash University
A2/NLV A2 CMV pp65 NLVPMVATV 1:100 Monash University
B7/TPR B7 CMV pp65 TPRVTGGGAM 1:40 Monash University
B8/DAN B8 CMV pp65 DANDIYRIF 1:40 Monash University

13. Resuspend cells in 100 mL/well of 1% Paraformaldehyde (16%

stock diluted in PBS, see Note 32).
14. Wrap plate in foil and incubate for 20 min on bench at room
temperature.
15. Wash cells with 100 mL/well of PBS.
16. Centrifuge plate for 5 min at 335 × g, room temperature.
17. Discard supernatant by flicking off in a single, sharp motion.
18. Resuspend cells in 50 mL/well of anti-IFN-g (and/or mAb for
other desirable cytokines) in 0.3% Saponin (1% stock diluted in
PBS, see Note 33).
19. Wrap plate in foil and incubate for 1 h at room temperature in
the dark or overnight at 4°C in the dark.
20. Wash cells with 100 mL/well of PBS.
21. Centrifuge plate for 5 min at 335 × g, room temperature.
22. Discard supernatant by flicking off in a single, sharp motion.
23. Resuspend cells in 60–100 mL/well of PBS.
24. Transfer cells into 5 mL polystyrene FACS tube. (This step is
unnecessary, if a flow cytometer with a 96-well plate reader is
available; as samples can be acquired directly from the plate).
25. Acquire by FACS.
26. Analyse data using flow cytometric software package (e.g.:
CellQuest™, FlowJo).
27. Gating strategies for allo-specific CD8+ T cells and antigen-
specific CD8+ T cells are shown in Figs. 5 and 6 respectively.

3.2.4. Tetramer Staining Following generation of in vitro expanded allo-specific or antigen-

specific T cells (see Subheading 3.1), the use of pMHC tetrameric
complexes in combination with mAb specific for cell surface markers
 18 Detection and Characterisation of Alloreactive T Cells 329

Fig. 5. Gating strategy for allo-specific CD8+ T cells. Identification of lymphocytes is achieved using a forward scatter (FSC)
vs. side scatter (SSC) dot plots. From the lymphocyte population, the percentage of CD8+ T cells can be measured using a
CD8 (FL3 channel) vs. SSC dot plot. Functional analysis of the CD8+ T cells can be quantitated using a CD107a (cytolytic
potential; FL1 channel) vs. IFN-g (cytokine production; FL4 channel).

Fig. 6. Gating strategy for antigen-specific CD8+T cells. Identification of lymphocytes is achieved using a FSC vs. SSC dot
plots. From the lymphocyte population, the percentage of CD8+ T cells can be measured using a CD8 (FL3 channel) vs. SSC
dot plot. Antigen-specific CD8+T cells are identified using a HLA-B8/FLRGRAYGL (FLR) Tetramer (FL2 channel) vs. SSC.
Functional analysis of the CD8+ T cells can be quantitated using a CD107a (cytolytic potential; FL1 channel) vs. IFN-g
(cytokine production; FL4 channel).

or intracellular proteins can be used for powerful quantitation and

characterisation of T cell subsets. The number of T cells needed
for the analysis depends on the frequency of T cell subsets being
investigated. For example, low frequency cells (»1%) require higher
numbers of cells (»1 × 106 cells) to be stained, in order to generate
sufficient data to analyse (minimum 10,000 events) using flow
cytometry. Individual mAbs and/or pMHC tetramers should be
titrated to determine optimal staining amounts prior to use in the
assay.
1. To a 96-well U-bottom plate or 5 mL polystyrene FACS tube,
add 2 × 105/well of T cells.
2. Add responder cells (PBMC or T cells) in separate wells or
5 mL polystyrene FACS tubes for FACS compensation.
Example: Unstained FL1, FL2, FL3, and FL4.
330 M. Bharadwaj et al.

Fig. 7. Detection of circulating EBV-specific CD8+ T cells using pMHC tetramers. Ex vivo PBMC isolated from a HLA-B8
healthy individual were co-incubated with an anti-CD8 antibody in the presence of either the HLA-B8/RAK or HLA-B8/FLR
tetramer. The corrected frequencies, following the subtraction of the background value, for both RAK- and FLR-specific
CD8+ T cells detected were 1.1 and 0.6%, respectively.

3. Centrifuge plate or tubes for 5 min at 335 × g, room temperature.

4. Discard supernatant by flicking off in a single, sharp motion for
the plate or aspirate for the tubes.
5. Resuspend cells in 20 mL/well in antibody mix containing cell
surface mAb and pMHC tetramer diluted in PBS. Compensation
controls: Unstained cells in PBS only and FL1 to FL4 (tetramer
or antibody diluted in PBS, see Note 31).
6. Wrap plate/tubes in foil and incubate for 30 min at 4°C in the
dark.
7. Wash cells with 150 mL/well or 1 mL/tube of PBS.
8. Centrifuge plate for 5 min at 335 × g, room temperature.
9. Discard supernatant by flicking off in a single, sharp motion for
the plate or aspirate for the tubes.
10. If combining tetramer staining with intracellular detection of
proteins, follow steps 13–25 of Subheading 3.2.3.
11. A gating strategy for antigen-specific CD8+ T cells is shown in
Fig. 7.

4. Notes

1. ELISPOT reagents: The capture and secondary antibodies will

vary depending on the cytokine that needs to be detected.
Here we describe reagents obtained from (Mabtech, Sweden)
 18 Detection and Characterisation of Alloreactive T Cells 331

for an IFNg ELISPOT. Detection kits and individual antibodies

are available for an array of cytokines (TNFa, Granzyme B,
Perforin, IL2, IL10, IL4, IL5, and others) from Mabtech and
other manufacturers such as PharMingen, Endogen, Woburn,
MA, USA; Cell Sciences, Norwood, MA, USA; R & D Systems
etc. ELISPOT plates are also available from different manufac-
turers (Mabtech, Millipore), and come with different mem-
branes custom made to increase specificity and/or sensitivity of
the assay. Plates with polyvinylidene difluoride (PVDF) mem-
branes such as the ELIIP and MSIP plates from Mabtech
require pre-treatment with ethanol.
2. Monoclonal antibodies (mAb) can be purchased from an array
of manufacturers. In our methodologies antibodies were pur-
chased from Becton Dickinson Biosciences. Regardless of the
manufacturer, each mAb should be titrated to determinate
optimal dilution (range 1:10 to 1:1,000) for use in the assay,
without compromising both specificity and sensitivity. A loss
of detectable signal, or a high background staining are clear
indicators that these dilutions are below or above the desired
mAb amounts, respectively.
3. HLA class I tetramers can either be purchased from manufac-
turers including National Institutes of Health, Tetramer Core
Facility, USA, or ProImmune, UK, or if you have the available
technology and reagents can be produced “in-house” as were
those used in our assays from either The University of
Melbourne or Monash University. The optimal amount of
tetramer used in an assay must also be evaluated using a similar
titration range (1:10 to 1:500) as discussed in Note 2 for mAbs.
4. The method described uses CellQuest™ (Becton Dickinson)
for acquisition of samples and FlowJo (Treestar Inc., Ashland,
OR) for analysing the data. The softwares can vary depending
on the Flow cytometer used for acquiring the samples.
Generally, Cell Quest™ software is used on the FACSCalibur,
and FACSDiva™ (Becton Dickinson) on a FACSCanto. The
software for both Cell Quest™ and FACSDiva™ is suitable for
data acquisition, as well as analysis.
5. Generation, expansion, and detection of allo-specific T cells
requires prior knowledge of HLA mismatches between
responder and stimulator and HLA-restricted molecule, whilst
expansion and detection of antigen-specific T cells requires the
knowledge of the viral epitope (minimal antigen and its HLA
restriction).
6. Target cell lines (B-LCLs and transfected parental lines: C1R,
721.221, K562) are cultured in RF10. Depending on method-
ology utilised some transfected cell lines require additional
antibiotic selection [Geneticin (G418) or Hygromycin B, Life
Technologies] and the optimal antibiotic concentration varies
332 M. Bharadwaj et al.

between cell lines. The type of antibiotic and its optimal

concentration is determined at the time of transfection of
the parental cell line.
7. Peripheral blood samples collected in the anticoagulant sodium
or lithium heparin should be stored at room temperature and
the blood should be processed within 24 h of collection.
8. Ficoll-Paque PLUS: Layering of anticoagulant-treated blood
onto Ficoll-Paque PLUS followed by centrifugation enables
differential migration of cell types and results in the formation
of distinct layers (see Fig. 2). Ficoll-Paque Plus is stored at
room temperature and protected from direct light.
9. When overlaying blood onto Ficoll-Paque PLUS reagent, it is
very important to hold the end of the serological pipette towards
the top of the 50 mL tube and, holding the tube at a 45° angle,
trickle the very first 1–5 mL of diluted blood down the lower
side of the tube. This will ensure that the layer of diluted blood
will remain intact and undisturbed. Once the first 5 mL has
been overlayed, the rate of loading can be increased. Do not let
the samples stand for too long as cells will start to sediment
through the interface between the blood and the Ficoll-Paque
PLUS, which may affect your lymphocyte recovery.
10. It is very important that the brake setting is OFF, otherwise the
Ficoll-Paque PLUS layer will be disturbed and separation of
blood components will not be achieved. However, if this does
occur, then mix the entire sample and repeat overlay step in a
new 50 mL tube. The g values provided herewith are quoted
for Beckman Coulter Allegra X-12-R centrifuge for which,
2,000 rpm = 931 × g, 1,500 rpm = 524 × g, 1,200 rpm = 335 × g,
and 1,000 rpm = 233 × g.
11. Wash to remove residual platelets. This generally only requires
one wash, but two washes may be needed for heavily con-
taminated preparations (i.e. if supernatant looks cloudy).
Additionally, if heavy red blood cell (RBC) contamination is
observed, the cells may require treatment with a commercial
RBC lysis buffer as per the manufacturer’s instructions.
12. Assay media: RF10 + 20 U/mL IL-2, media conditions have
been optimised for expansion of either allo-specific or antigen-
specific CD8+ T cells. IL-2 cytokine mediates its effects by
binding to IL-2 receptors, expressed by lymphocytes. In this
in vitro culture system, the IL-2/IL-2R interaction then stim-
ulates the growth, differentiation, and survival of allo-specific
or antigen-specific CD8+ T cells.
13. Gamma irradiation: Stimulator cells are gamma irradiated to
ensure that either the MLR or auto-stimulation for antigen-
specific T cells is uni-directional. Stimulator cells are unable to
undergo proliferation due to DNA damage caused by effects
of the ionising radiation.
 18 Detection and Characterisation of Alloreactive T Cells 333

14. Bulk T cell cultures should be observed every day. It is

paramount that adequate cell density is maintained to promote
cell–cell contact and interactions essential for T cell activation
and expansion. Media colour changes is an indicator of prolif-
eration. It is recommended that sub-culturing of bulk T cell
cultures (split 1:2 only) is not performed until at least day 3 in
order to maximise priming events. Replenish media every 3–4
days if not actively sub-culturing the T cells. The number of
days for in vitro expansion of T cells may vary depending on
(1) stimulator potency, (2) kinetics of cell proliferation, and
(3) kinetics of protein to be measured.
15. Safety Precautions: 51Cr is a strong gamma-emitter. A lead
shield should be used at all times between the 51Cr source and
user when labelling the targets. Wear protective eyewear, radia-
tion badge, gown, and double gloves. Make sure there is no
spill. Record any spills and contain them according to institu-
tional procedures prescribed for 51Cr spill containment. Discard
the syringe in a yellow syringe disposal bin reserved for radio-
active waste. When using the isotope, use a Geiger counter to
monitor spills and exposure. Dispose off high-level 51Cr waste
in a lead bucket surrounded by a lead shield in a designated
area. 51Cr-labelled target cells should be transported in an esky
lined with lead.
16. Take care not to splash 1% Triton X-100 into any of the sur-
rounding wells as this will lyse all targets and confound results
for target-specific lysis. It might be easier to leave 1% Triton
X-100 addition until the very end of the assay just before incu-
bating the plate.
17. The mean spontaneous lysis for target cells in culture medium
should always be less than 10%, and the variation about the
mean specific lysis, less than 5%.
18. ELISPOT plates may need to be pre-treated as per manufac-
turers instructions and buffer used for diluting the antibodies
may also vary between suppliers.
19. Avoid touching the bottom of the wells of the ELISPOT plate
with the pipette tips as this may disrupt the membrane or give
the impression of artificial spots. Dispense down the side of the
wells with the tips touching the walls.
20. To use cryopreserved PBMC, retrieve cell vials from liquid
nitrogen storage. Pre-fill 10 mL yellow capped tubes with
8 mL RPMI or RF10. Thaw vials in warm tap water (37°C) in
beaker and transfer cells to 10 mL tubes containing RPMI or
RF10. Centrifuge 10 mL tubes immediately for 5 min at
1,026 × g. Remove supernatant and resuspend pellet in 2–5 mL
RF10 and incubate for 2 h at 37°C, 5%CO2 before use.
21. CD3+ T cells can be separated from the whole blood using
Rosette sep (StemCell Biotechnologies, CellSystems GmbH,
334 M. Bharadwaj et al.

St Katherinen, Germany) as recommended by the manufacturer.

Alternatively, Dynabeads® CD3 (Cat. no. 111.51D, Invitrogen)
can be used to deplete CD3+ T cells directly from whole blood,
buffy coat, or PBMCs.
22. The incubation time for the responder cells with the antigen in
the ELISPOT assay can vary depending on the cytokine to be
detected. For example, for IL-2 or IFN-g, the recommended
incubation time is 20–24 h whilst for IL-10, a longer incuba-
tion period of up to 48 h may be required.
23. Caution! Tween 20 may be harmful and should be treated as
hazardous. Avoid inhalation and contact with skin and eyes.
24. Between and after final wash steps (in steps 8, 9, 12, and 16 of
ELISPOT assay), do not let the plate dry. Leave last wash solu-
tion (PBS) in the plate until the reagent for the next step has
been prepared and ready to be used.
25. Caution! Streptavidin-alkaline phosphatase can be hazardous.
Avoid inhalation and skin contact.
26. Caution! BCIP/NBT is an irritant causing irritation to eyes,
respiratory system, and skin. Avoid contact to these areas of
the body.
27. Number of fluorochromes (FL) would depend on the pheno-
typic and functional markers to be assessed and the tetramers
utilised.
28. CD107a, also known as Lysosomal-associated membrane
protein 1 (LAMP1), is a marker of degranulation on
lymphocytes.
29. Monensin, isolated from Streptomyces cinnamonensis, blocks
intracellular protein transport.
30. Brefeldin A is a lactone antibiotic produced by fungal organ-
isms such as Penicillium brefeldianum. Brefeldin A interferes
with protein transport from the endoplasmic reticulum (ER)
to the Golgi apparatus by inhibiting transport in Golgi, which
leads to proteins accumulating inside the endoplasmic
reticulum.
31. Compensation controls are important to select markers of pro-
teins that have relatively high density on surface of cells, not
rare events.
32. Paraformaldehyde is used to fix the cells prior to permeabilisa-
tion to stabilise the integrity of the cell.
33. Saponin is a surfactant, extracted from the Quillaja Bark. It
enhances penetration of proteins and other macromolecules
through cell membranes. Saponin is diluted to 0.3% to ensure
permeability but not potent haemolysis of the cells.
 18 Detection and Characterisation of Alloreactive T Cells 335

Acknowledgements

The authors would like to thank Ms Patricia Therese Illing, The

University of Melbourne, Australia, for her help in proofreading
the manuscript.

References
1. Burrows SR, Khanna R, Burrows JM, Moss DJ dynamics of antigen-specific CD8+ cytotoxic T
(1994) An alloresponse in humans is domi- lymphocytes following primary Epstein–Barr
nated by cytotoxic T lymphocytes (CTL) cross- virus infection. Blood 98(8):2588–2589
reactive with a single Epstein–Barr virus CTL 12. Kaminski E, Hows J, Man S, Brookes P,
epitope: implications for graft-versus-host dis- Mackinnon S, Hughes T et al (1989) Prediction
ease. J Exp Med 179(4):1155–1161 of graft versus host disease by frequency analy-
2. Brunner KT, Mauel J, Cerottini JC, Chapuis B sis of cytotoxic T cells after unrelated donor
(1968) Quantitative assay of the lytic action of bone marrow transplantation. Transplantation
immune lymphoid cells on 51-Cr-labelled allo- 48(4):608–613
geneic target cells in vitro; inhibition by isoan- 13. Schwarer AP, Jiang YZ, Deacock S, Brookes
tibody and by drugs. Immunology 14(2): PA, Barrett AJ, Goldman JM et al (1994)
181–196 Comparison of helper and cytotoxic antirecipi-
3. Lefkovits I (1972) Induction of antibody- ent T cell frequencies in unrelated bone mar-
forming cell clones in microcultures. Eur J row transplantation. Transplantation 58(11):
Immunol 2(4):360–366 1198–1203
4. Lefkovits I, Waldmann H (1984) Limiting dilu- 14. Roelen DL, Stobbe I, Young NT, van Bree SP,
tion analysis of cells of the immune system I. Doxiadis II, Oudshoorn M et al (2001)
The clonal basis of the immune response. Permissible and immunogenic HLA-A mis-
Immunol Today 5:265–268 matches: cytotoxic T-cell precursor frequencies
5. Waldmann H, Lefkovits I (1984) Limiting dilu- reflect graft survival data. Hum Immunol
tion analysis of cells of the immune system II: 62(7):661–667
what can be learnt? Immunol Today 5:295–298 15. Dolezalova L, Vrana M, Dobrovolna M,
6. Sharrock CE, Kaminski E, Man S (1990) Loudova M, Cukrova V, Vitek A et al (2002)
Limiting dilution analysis of human T cells: a Cytotoxic T lymphocyte precursor frequency
useful clinical tool. Immunol Today 11(8): analysis in the selection of HLA matched unre-
281–286 lated donors for hematopoietic stem cell trans-
7. Hickling JK (1998) Measuring human plantation: the correlation of CTLp frequency
T-lymphocyte function. Expert Rev Mol Med with HLA class I genotyping and aGVHD
1:1–20 development. Neoplasma 49(1):26–32
8. Kaminski E, Hows J, Brookes P, Mackinnon S, 16. Fussell ST, Donnellan M, Cooley MA, Farrell
Hughes T, Avakian O et al (1989) Alloreactive C (1994) Cytotoxic T lymphocyte precursor
cytotoxic T-cell frequency analysis and HLA frequency does not correlate with either the
matching for bone marrow transplants from incidence or severity of graft-versus-host dis-
HLA matched unrelated donors. Transplant ease after matched unrelated donor bone mar-
Proc 21(1 Pt 3):2976–2977 row transplantation. Transplantation
9. Macdonald WA, Chen Z, Gras S, Archbold JK, 57(5):673–676
Tynan FE, Clements CS et al (2009) T cell 17. Eberspacher ML, Otto G, Herfarth C, Kabelitz
allorecognition via molecular mimicry. D (1994) Frequency analysis of donor-reactive
Immunity 31(6):897–908 cytotoxic T lymphocyte precursors in in allograft
10. Bharadwaj M, Thammanichanond D, Aitken recipients. Lack of correlation with clinical out-
CK, Moneer S, Drummer HE, Tracy S et al come. Transplantation 57(12):
(2009) TCD8 response in diverse outcomes of 1746–1752
recurrent exposure to hepatitis C virus. 18. Wang XN, Taylor PR, Skinner R, Jackson GH,
Immunol Cell Biol 87(6):464–472 Proctor SJ, Hedley D et al (2000) T-cell fre-
11. Bharadwaj M, Burrows SR, Burrows JM, Moss quency analysis does not predict the incidence
DJ, Catalina M, Khanna R (2001) Longitudinal of graft-versus-host disease in HLA-matched
336 M. Bharadwaj et al.

sibling bone marrow transplantation. Trans- 30. Sherritt MA, Bharadwaj M, Burrows JM,
plantation 70(3):488–493 Morrison LE, Elliott SL, Davis JE et al (2003)
19. Tan LC, Gudgeon N, Annels NE, Hansasuta P, Reconstitution of the latent T-lymphocyte
O’Callaghan CA, Rowland-Jones S et al (1999) response to Epstein–Barr virus is coincident
A re-evaluation of the frequency of CD8+ T with long-term recovery from posttransplant
cells specific for EBV in healthy virus carriers. J lymphoma after adoptive immunotherapy.
Immunol 162(3):1827–1835 Transplantation 75(9):1556–1560
20. Lalvani A, Brookes R, Hambleton S, Britton 31. Gazagne A, Claret E, Wijdenes J, Yssel H,
WJ, Hill AV, McMichael AJ (1997) Rapid effec- Bousquet F, Levy E et al (2003) A Fluorospot
tor function in CD8+ memory T cells. J Exp Med assay to detect single T lymphocytes simultane-
186(6):859–865 ously producing multiple cytokines. J Immunol
21. Murali-Krishna K, Altman JD, Suresh M, Methods 283(1–2):91–98
Sourdive DJ, Zajac AJ, Miller JD et al (1998) 32. Altman JD, Moss PA, Goulder PJ, Barouch
Counting antigen-specific CD8 T cells: a DH, McHeyzer-Williams MG, Bell JI et al
reevaluation of bystander activation during viral (1996) Phenotypic analysis of antigen-specific
infection. Immunity 8(2):177–187 T lymphocytes. Science 274(5284):94–96
22. Kuzushima K, Hoshino Y, Fujii K, Yokoyama 33. McMichael AJ, O’Callaghan CA (1998) A new
N, Fujita M, Kiyono T et al (1999) Rapid look at T cells. J Exp Med 187(9):1367–1371
determination of Epstein–Barr virus-specific 34. Ogg GS, McMichael AJ (1998) HLA-peptide
CD8(+) T-cell frequencies by flow cytometry. tetrameric complexes. Curr Opin Immunol
Blood 94(9):3094–3100 10(4):393–396
23. Hernandez-Fuentes MP, Warrens AN, Lechler 35. Miles JJ, Thammanichanond D, Moneer S,
RI (2003) Immunologic monitoring. Immunol Nivarthi UK, Kjer-Nielsen L, Tracy SL et al
Rev 196:247–264 (2011) Antigen-driven patterns of TCR bias
24. Miyahira Y, Murata K, Rodriguez D, Rodriguez are shared across diverse outcomes of human
JR, Esteban M, Rodrigues MM et al (1995) hepatitis C virus infection. J Immunol 186(2):
Quantification of antigen specific CD8+ T cells 901–912
using an ELISPOT assay. J Immunol Methods 36. James EA, LaFond R, Durinovic-Bello I, Kwok
181(1):45–54 W (2009) Visualizing antigen specific CD4+ T
25. Czerkinsky C, Andersson G, Ekre HP, Nilsson cells using MHC class II tetramers. J Vis Exp
LA, Klareskog L, Ouchterlony O (1988) (25) pii:1167
Reverse ELISPOT assay for clonal analysis of 37. Landais E, Romagnoli PA, Corper AL, Shires J,
cytokine production. I. Enumeration of Altman JD, Wilson IA et al (2009) New design
gamma-interferon-secreting cells. J Immunol of MHC class II tetramers to accommodate
Methods 110(1):29–36 fundamental principles of antigen presentation.
26. Elliott SL, Suhrbier A, Miles JJ, Lawrence G, J Immunol 183(12):7949–7957
Pye SJ, Le TT et al (2008) Phase I trial of a 38. Vollers SS, Stern LJ (2008) Class II major his-
CD8+ T-cell peptide epitope-based vaccine for tocompatibility complex tetramer staining:
infectious mononucleosis. J Virol 82(3): progress, problems, and prospects. Immunology
1448–1457 123(3):305–313
27. Scheibenbogen C, Romero P, Rivoltini L, Herr 39. Raddassi K, Kent SC, Yang J, Bourcier K,
W, Schmittel A, Cerottini JC et al (2000) Bradshaw EM, Seyfert-Margolis V et al (2011)
Quantitation of antigen-reactive T cells in Increased frequencies of myelin oligodendro-
peripheral blood by IFNgamma-ELISPOT cyte glycoprotein/MHC class II-binding
assay and chromium-release assay: a four-centre CD4 cells in patients with multiple sclerosis. J
comparative trial. J Immunol Methods 244 Immunol 187(2):1039–1046
(1–2):81–89 40. Laughlin E, Burke G, Pugliese A, Falk B,
28. Pelfrey CM, Rudick RA, Cotleur AC, Lee JC, Nepom G (2008) Recurrence of autoreactive
Tary-Lehmann M, Lehmann PV (2000) antigen-specific CD4+ T cells in autoimmune
Quantification of self-recognition in multiple diabetes after pancreas transplantation. Clin
sclerosis by single-cell analysis of cytokine pro- Immunol 128(1):23–30
duction. J Immunol 165(3):1641–1651 41. Bonvalet M, Wambre E, Moussu H, Horiot S,
29. Bestard O, Nickel P, Cruzado JM, Kwok WW, Louise A et al (2011) Comparison
Schoenemann C, Boenisch O, Sefrin A et al between major histocompatibility complex
(2008) Circulating alloreactive T cells corre- class II tetramer staining and surface expres-
late with graft function in longstanding renal sion of activation markers for the detection of
transplant recipients. J Am Soc Nephrol 19(7): allergen-specific CD4(+) T cells. Clin Exp
1419–1429 Allergy 41(6):821–829
 18 Detection and Characterisation of Alloreactive T Cells 337

42. Ayyoub M, Dojcinovic D, Pignon P, Raimbaud cally infected adults and children. J Exp Med
I, Schmidt J, Luescher I et al (2010) Monitoring 192(12):1819–1832
of NY-ESO-1 specific CD4+ T cells using 53. Asemissen AM, Nagorsen D, Keilholz U,
molecularly defined MHC class II/His-tag- Letsch A, Schmittel A, Thiel E et al (2001)
peptide tetramers. Proc Natl Acad Sci USA Flow cytometric determination of intracellular
107(16):7437–7442 or secreted IFNgamma for the quantification of
43. Cusick MF, Yang M, Gill JC, Eckels DD (2011) antigen reactive T cells. J Immunol Methods
Naturally occurring CD4+ T-cell epitope vari- 251(1–2):101–108
ants act as altered peptide ligands leading to 54. Whiteside TL, Zhao Y, Tsukishiro T, Elder EM,
impaired helper T-cell responses in hepatitis C Gooding W, Baar J (2003) Enzyme-linked
virus infection. Hum Immunol immunospot, cytokine flow cytometry, and
72(5):379–385 tetramers in the detection of T-cell responses to
44. Sun Y, Iglesias E, Samri A, Kamkamidze G, a dendritic cell-based multipeptide vaccine in
Decoville T, Carcelain G et al (2003) A system- patients with melanoma. Clin Cancer Res
atic comparison of methods to measure HIV-1 9(2):641–649
specific CD8 T cells. J Immunol Methods 55. Koehne G, Smith KM, Ferguson TL, Williams
272(1–2):23–34 RY, Heller G, Pamer EG et al (2002)
45. McMichael AJ, Ogg G, Wilson J, Callan M, Quantitation, selection, and functional charac-
Hambleton S, Appay V et al (2000) Memory terization of Epstein–Barr virus-specific and
CD8+ T cells in HIV infection. Philos Trans R alloreactive T cells detected by intracellular
Soc Lond B Biol Sci 2355(1395):363–367 interferon-gamma production and growth of
46. Ogg GS, Jin X, Bonhoeffer S, Dunbar PR, cytotoxic precursors. Blood 99(5):1730–1740
Nowak MA, Monard S et al (1998) Quantitation 56. Waldrop SL, Pitcher CJ, Peterson DM, Maino
of HIV-1-specific cytotoxic T lymphocytes and VC, Picker LJ (1997) Determination of
plasma load of viral RNA. Science antigen-specific memory/effector CD4+ T cell
279(5359):2103–2106 frequencies by flow cytometry: evidence for a
47. Callan MF, Tan L, Annels N, Ogg GS, Wilson novel, antigen-specific homeostatic mechanism
JD, O’Callaghan CA et al (1998) Direct visual- in HIV-associated immunodeficiency. J Clin
ization of antigen-specific CD8+ T cells during Invest 99(7):1739–1750
the primary immune response to Epstein–Barr 57. Mifsud NA, Purcell AW, Chen W, Holdsworth
virus In vivo. J Exp Med 187(9):1395–1402 R, Tait BD, McCluskey J (2008) Immu-
48. Mifsud NA, Nguyen TH, Tait BD, Kotsimbos nodominance hierarchies and gender bias in
TC (2010) Quantitative and functional diversity direct T(CD8)-cell alloreactivity. Am J
of cross-reactive EBV-specific CD8+ T cells in a Transplant 8(1):121–132
longitudinal study cohort of lung transplant 58. Allen KJ, Mifsud NA, Williamson R, Bertolino
recipients. Transplantation 90(12):1439–1449 P, Hardikar W (2008) Cell-mediated rejection
49. Nguyen TH, Sullivan LC, Kotsimbos TC, results in allograft loss after liver cell transplan-
Schwarer AP, Mifsud NA (2010) Cross- tation. Liver Transpl 14(5):688–694
presentation of HCMV chimeric protein enables 59. Elkington R, Walker S, Crough T, Menzies M,
generation and measurement of polyclonal T Tellam J, Bharadwaj M et al (2003) Ex vivo
cells. Immunol Cell Biol 88(6):676–684 profiling of CD8+-T-cell responses to human
50. Pala P, Hussell T, Openshaw PJ (2000) Flow cytomegalovirus reveals broad and multispecific
cytometric measurement of intracellular cytok- reactivities in healthy virus carriers. J Virol
ines. J Immunol Methods 243(1–2):107–124 77(9):5226–5240
51. Prussin C, Metcalfe DD (1995) Detection of 60. Taswell C (1981) Limiting dilution assays for
intracytoplasmic cytokine using flow cytometry the determination of immunocompetent cell
and directly conjugated anti-cytokine antibod- frequencies. I. Data analysis. J Immunol
ies. J Immunol Methods 188(1):117–128 126(4):1614–1619
52. Goulder PJ, Tang Y, Brander C, Betts MR, 61. Strijbosch LW, Buurman WA, Does RJ, Zinken
Altfeld M, Annamalai K et al (2000) PH, Groenewegen G (1987) Limiting dilution
Functionally inert HIV-specific cytotoxic T assays. Experimental design and statistical anal-
lymphocytes do not play a major role in chroni- ysis. J Immunol Methods 97(1):133–140
 Chapter 19

Detection of Allo-HLA Cross-Reactivity by Virus-specific

Memory T-Cell Clones Using Single HLA-Transfected
K562 Cells
Lloyd J. D’Orsogna, Ellen M.W. van der Meer-Prins, Yvonne M. Zoet,
Dave L. Roelen, Ilias I.N. Doxiadis, and Frans H.J. Claas

Abstract
The ability to directly measure virus-specific lymphocytes using fluorochrome-labeled tetrameric complexes
has proven a great advancement for the transplantation field. Viral peptide/HLA tetrameric complexes allow
the rapid generation of virus-specific clones using single cell sorting apparatus, permitting the determination
of alloreactivity from a single TCR with known specificity. When combined with new target “detector” cells
called single HLA antigen-transfected K562 cells (SALs), the human alloresponse can for the first time be
examined specifically and reliably. Here we describe a method for detection of “heterologous immunity”
from virus-specific memory T-cells using single HLA expressing cell lines as allogeneic targets.

Key words: Heterologous immunity, Memory T-cells, Viral immunity, Alloreactivity, SALs

1. Introduction

The mechanisms by which alloreactive memory T-cells are gener-

ated in nonsensitized individuals have begun to be elucidated. It is
generally accepted that a very high level of cross-reactivity is an
essential feature of the T-cell receptor, e.g., memory T-cells that
have been generated as a result of a previous viral infection can sub-
sequently respond to a second unrelated infection. However, it has
only recently been shown that alloreactivity from virus-specific
memory T-cells is far more common than predicted; 45% of virus-
specific T-cell clones were found to be allo-HLA cross-reactive (1).
Detection of alloreactivity from virus-specific memory T-cells
should first be performed by screening against a panel of cells
expressing most common HLA molecules, such as a panel of EBV

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_19, © Springer Science+Business Media New York 2012

339
340 L.J. D’Orsogna et al.

LCLs as described elsewhere (1). However, cross-reactivity should

be confirmed using a single HLA-transfected cell type. This will
not only confirm the individual HLA molecule recognized but also
exclude the possibility of the clone recognizing viral peptides pre-
sented on allo-HLA.
Single HLA-transfected K562 cells (SALs) are a new sensitive
and specific tool to detect alloresponses from virus-specific T-cells
(2). SALs express only the transfected HLA class I molecule (3),
unlike C1R cells that may have low expression of other HLA mol-
ecules. For example, the previously described alloreactivity of the
EBV EBNA3A specific T-cell against HLA-B*4402 has been
confirmed using SALs (2). Furthermore, the same T-cell clone was
also found to recognize HLA-B*5501 expressing SALs (2), sug-
gesting that SALs may be a more sensitive target than other cells
expressing multiple HLA molecules. Therefore, detection of HLA-
specific alloresponses from virus-specific T-cells in vitro is now fea-
sible in the routine laboratory.
This technique to define cross-reactive T-cells may lead to
important future improvements in donor selection and monitor-
ing. The ability to define public TCR responses that give specific
allo-HLA cross-reactivity may assist in the definition of (un) accept-
able mismatches. If in vitro tools can more specifically predict
which transplant recipients are at risk for rejection and which
patients are predisposed to tolerance, the immunosuppressive regi-
men could be adjusted accordingly. We also hypothesize that anti-
viral therapy may decrease the proportion and activation status of
these alloreactive T-cells.

2. Materials

2.1. Virus-specific 1. Viral peptide/HLA tetrameric complex of interest (see Note 1).
T-Cell Cloning Using 2. PBMCs from donor known to have viral peptide/HLA
Single Cell Sorting tetramer complex binding T-cells in the peripheral blood (or at
least known to be serologically positive for virus of interest)
(see Notes 2–3).
3. Iscoves Modified Dulbecco Medium (IMDM) with 1%
l-glutamine.
4. Fetal Calf Serum (FCS) made to 1 and 10% solutions in
IMDM.
5. Human Serum (HS) made to 5% HS and 5% FCS in IMDM.
6. Phytohaemagglutinin (PHA) 800 ng/mL.
7. Interleukin-2 (IL-2).
8. Freshly collected human PBMCs to be used as feeder cells
(see Note 4).
 19 Detection of Allo-HLA Cross-Reactivity... 341

9. Sterile 96-well round bottom plates.

10. Sterile 24-well flat bottom plates.
11. FACS sorting apparatus, e.g., FACSAriaII.

2.2. Generation 1. K562 cell line.

of Single HLA- 2. Plasmid (pLNCX, pCDNA3.0, resistant to neomycin (G418),
Transfected K562 Cells pEAK10 resistant to puromycin, with the HLA-cDNA con-
Lines (SALs) struct of interest (10 μg of plasmid cDNA per transfection))
(obtained from 13th international histocompatibility working
group) (3, 4).
3. IMDM supplemented with 1% l-glutamine.
4. 10% FCS in IMDM supplemented with penicillin/streptomycin.
5. G418 (200 μg/mL) (see Note 5).
6. Gene pulser (Biorad).
7. Dynabeads Sheep anti-Mouse Immunoglobulin—Dynal.
8. w6/32 (anti-MHC class I antibody), w6/32-PE, w6/32-
FITC (Abcam, United Kingdom).
9. Anti-IgG antibody (Goat anti-Mouse-FITC) (BD, the
Netherlands).
10. Puromycin (0.5 μg/mL) (Invitrogen).
11. Hygromycin (50 μg/mL) (Clontech).
12. Sterile Gen pulser 0.4-cm cuvettes (Biorad).
13. FACS calibur flow cytometer.
14. FACS tubes.
15. Solution of 0.1% bovine serum albumin (BSA) in phosphate
buffered saline (PBS).
16. HLA-specific Monoclonal Antibodies (Leiden University Medical
Centre (LUMC), Leiden, the Netherlands) (see Note 6).
17. Rabbit anti-Human IgG FITC.
18. Rabbit anti-Human IgM FITC.
19. Dimethyl sulfoxide (DMSO).
20. Incubator.
21. PBS/1% paraformaldehyde.
22. 15-mL falcon tubes (BD biosciences).
23. 75-cm2 culture flask (Corning).
24. Dynal magnet.

2.3. IFN g ELISA 1. Virus-specific T-cell clone—see Subheading 3.1.

2. SALs transfected with the HLA molecule of interest—see
Subheading 3.2.
342 L.J. D’Orsogna et al.

3. Nontransfected K562 cells.

4. Viral peptide for which the clone is specific (positive control),
and a control viral peptide (see Note 7).
5. IMDM supplemented with 1% l-glutamine.
6. Freshly prepared T-cell culture medium—see Subheading 3.1,
step 9.
7. PHA 800 ng/mL.
8. Interleukin-2 (IL-2).
9. Sterile 96-well round bottom plates.
10. Human IFN-γ Elisa kit (U-Cytech, Netherlands).

2.4. Cytotoxicity Assay 1. Virus-specific T-cell clone—see Subheading 3.1.

2. SALs transfected with the HLA molecule of interest—see
Subheading 3.2.
3. Nontransfected K562 cells.
4. Viral peptide for which the clone is specific (positive control),
and a control viral peptide (see Note 7).
5. IMDM supplemented with 1% l-glutamine.
6. Freshly prepared T-cell culture medium—see Subheading 3.1,
step 9.
7. PHA 800 ng/mL.
8. Interleukin-2 (IL-2).
9. Sodium Chromate (1 mCi/mL (370mBq)).
10. 1% Triton X-100.
11. 96-well round bottom plates.

3. Methods

3.1. Virus-specific CD8 1. Perform all steps at 4°C (see Note 8).
T-Cell Cloning Using 2. Ensure cell donor for single cell sorting is serologically positive
Single Cell Sorting for the virus and has a population of the relevant tetramer
binding T-cells (see Notes 2 and 9).
3. Count the ficoll hypaque separated PBMCs ensuring you have
at least 1 × 106 cells. Wash twice in medium containing 1%
FCS/IMDM by spinning down the cells at 700 G for 8 min.
4. After second wash, resuspend the pellet in the residual medium
and transfer to a FACS tube (see Note 9).
5. Add 1 μL tetramer-PE to the FACS tube on ice (see Note 1).
6. Add markers CD45Ra/CD4/CD14-FITC (see Note 10).
 19 Detection of Allo-HLA Cross-Reactivity... 343

7. Incubate on ice for 30 min.

8. Wash PBMCs twice in 1% FCS/IMDM medium as for
step 3.
9. Place up to 5 million cells for sorting in 1 mL 10% FCS/
IMDM without phenol red.
10. Prepare T-cell Medium—5% FCS/5% Human serum in IMDM
medium with 1% penicillin/streptomycin, 3 mg l-Glutamine
and IL-2 100 IU/mL (see Note 11).
11. Prepare T-cell feeder mix for stimulation of sorted virus-specific
T-cells in 96 well plate—To the T-cell medium add 0.5 million
irradiated feeder cells, 0.05 × 106 EBV LCLs, and 2 μL PHA
per mL of required feeder mix (approximately 10 mL per plate
is required) (see Notes 11–13).
12. Add 100 μL feeder mix per well to 96-well round bottom
plates(s).
13. Perform single cell sorting into preprepared 96-well round
bottom plate (with feeder mix) based on tetramer positive gate
using single cell sorting apparatus. After sorting place 96-well
plates into incubator at 37°C (see Notes 14–16).
14. Add further 100 μL T-cell medium per well on day 7.
15. At day 14 check the plates for cell growth. Select the growing
clones and transfer the contents to individual wells in a 24-well
plate to which 1 mL feeder cell mix has already been added.
Continue to culture at 37°C.
16. Check daily for growth and if wells are full split the culture
into a separate well containing 1 mL of T-cell culture
medium.
17. After day 6 confi rm tetramer positivity of T-cell clones
by resuspending and taking 100 μL aliquot for FACS
analysis.
18. Determine Vβ IOtest TCR usage using Vβ kit (Beckman
Coulter) and/or TCR PCR using primers designed to cover
the complete TCR repertoire.
19. Continue to culture cells. Freeze interesting clones if required
for later use (freeze before day 10 since last stimulation).
20. These T-cell clones are now ready for use in IFNγ ELISA assays
(see Subheading 3.3) or cytotoxicity assays (see Subheading 3.4).
However defrosted T-cell clones can also be used but must be
restimulated prior to use. 1 × 106 defrosted T-cell clone should
be restimulated in T-cell medium with 5 million feeder cells,
0.5 × 106 EBV LCLs and 2 μL/mL PHA in a T75 flask. IFNγ
ELISA is usually performed 9–10 days after stimulation and
cytotoxicity 7–8 days after stimulation.
344 L.J. D’Orsogna et al.

3.2. Generation 1. Culture K562 until you have 10 × 106 cells per transfection in
of Single HLA- IMDM supplemented with 10% FCS and penicillin/streptomycin.
Transfected K562 Cells 2. Spin down the cells and wash two times in IMDM/10%FCS.
Lines (SALs)
3. Count the cells and resuspend the cells in IMDM/10%FCS at
a concentration of 10 × 106 cells per mL.
4. Mix 1 mL of the cells with 10 μg of the plasmid DNA contain-
ing the HLA-cDNA construct of the HLA specificity
required.
5. Put 1 mL of the cell/DNA mixture in a 0.4-cm cuvette and
store on ice for a minimum of 5 min.
6. Mix and pulse (960 μF, 230 V).
7. Store on ice for 15 min.
8. Add a few drops of IMDM/10%FCS/pen/strep medium to
the cuvette and transfer the cells to a culture flask containing
9 mL of the same medium.
9. Culture the cells for 2 days at 37°C.
10. Add selection antibiotic at the correct final concentration (see
manufacturers protocol) (see Note 5).
11. Culture for another week at 37°C. If culture flasks become full
then split flasks and add new medium.
12. Perform a flow cytometry test to see whether there are HLA-
positive cells (incubate 105 cells with w6/32-PE for 30 min on
ice, wash twice, take up in 100 μL PBS/1% paraformaldehyde,
read in FACS calibur and analyze) (see Note 17).
13. If positive cells are present, separate positive cells from nega-
tive cells (described in steps 14–31) (see Notes 18 and 19).
14. Take 20–25 mL of your cultured cells (leave some in the flask
and continue culturing them).
15. Spin down, remove supernatant, and resuspend the cell pellet.
16. Add 0.5 mL unlabeled-w6/32 antibody and vortex.
17. Incubate on ice for 30 min.
18. Wash three times with ice cold medium (IMDM/10%FCS) in
a 15-mL tube.
19. Resuspend the cells and add 30 μL Sheep anti-Mouse dynabeads.
20. Incubate on a rollerbench in 4°C room for 30 min.
21. Put the tube in the magnet, add 10 mL ice cold medium.
22. Wait for 5 min.
23. Discard the medium with the nonbound cells.
24. Repeat from step 21 twice.
25. Add 10 mL ice cold medium and resuspend the cells bound to
the beads, put them in a small culture flask, and culture at 37°C.
 19 Detection of Allo-HLA Cross-Reactivity... 345

26. Wait for the cells to grow and expand. Check daily (see Notes
19 and 20).
27. When the cells have expanded, remove the beads as follows.
28. Put the cells in a 15-mL tube within the dynal magnet.
29. Wait for 5 min.
30. Collect the medium with the cells into a culture flask leaving
the beads in the tube.
31. Add 10 mL of fresh medium (10%FCS/IMDM) to the tube.
32. Repeat twice from step 28 and on the last removal of beads
transfer the cells to a culture flask, add 10% of fresh culture
medium.
33. Culture at 37oC.
34. After a few days growth, test the class I expression of the cells
using HLA-specific monoclonal antibodies and FACS analysis
(step 12).
35. If there still are cells with no expression at all, repeat the bead
sorting (repeat from step 14 onwards).
36. If the cells all have a good expression, they can be used in the
IFNγ assay (see Subheading 3.3) or cytotoxicity assay (see
Subheading 3.4). Alternatively the cells can be frozen in sev-
eral aliquots for subsequent use in these assays.
37. Test the cells with a flow cytometry test against a panel of anti-
HLA monoclonal antibodies to confirm the HLA expression
of the K562 cell (see Note 17).

3.3. IFN g ELISA 1. Spin down the virus-specific T-cell clone (generated using
Using Responder method in Subheading 3.1), resuspend and count (see Notes
Virus-specific T-Cell 21 and 22).
Clones and SAL 2. Make a T-cell clone solution of 0.1 × 106 T-cells/mL in T-cell
Stimulator Cells medium (see Subheading 3.1, step 10).
3. Dilute the SALs and the K562 line in T-cell medium to a con-
centration of 0.25 × 106 cells/mL.
4. Add 50 μL/well of the T-cell clone solution to a sterile 96-well
plate.
5. Add 100 μL/well of the different SALs in duplicate to the
T-cell clone. Include:
(a) A SAL transfected with the restricting HLA molecule
loaded with the viral peptide that is recognized by the
clone as a positive control. To load peptides onto SALs
first incubate harvested cell pellet with 100 μg/mL of
peptide for 30 min at room temperature, then wash twice
and resuspend to required concentration of 0.25 × 106
cells/mL.
346 L.J. D’Orsogna et al.

(b) A SAL transfected with the restricting HLA molecule

loaded with the control peptide that is not recognized by
the clone as a negative control (see Notes 23–25).
(c) A nontransfected K562 as a negative control.
(d) As background control also add only T-cell medium to
some of the wells without SALs.
6. Incubate for 24 h at 37°C.
7. Transfer 120 μL supernatant from each well by pipetting to a
new 96-well plate and store the samples at −20°C till use in
IFNγ ELISA (see Notes 26 and 27).
8. Thaw the supernatants and use in an IFN-γ ELISA according
to the manufacturer’s protocol.

3.4. Cytotoxicity Assay 1. Spin down the SALs and the nontransferred K562 cell line (see
Using Effector Subheading 3.2, step 36) direct from the culture medium or
Virus-specific T-Cell following thawing. Do not resuspend. The cell pellets must
Clones and SAL first be labeled with chromium (see Note 28). Include:
Target Cells (a) A SAL transfected with the restricting HLA molecule
loaded with the viral peptide that is recognized by the
clone as a positive control (see Notes 23–25).
(b) A SAL transfected with the restricting HLA molecule
loaded with the control peptide that is not recognized by
the clone as a negative control.
(c) A nontransfected K562 as a negative control.
2. Add the appropriate amount of sodium chromate to the cell
pellets (see Note 29)
3. Incubate the cells in a 37°C water bath for 1 h.
4. Meanwhile spin down, resuspend, and count the T-cell clone
and make the necessary dilutions of the clone (see Note 30).
Add 100 μL/well to a 96-well plate.
5. Wash the chromium-labeled SAL and K562 cells three times
with 4 mL IMDM + 1%FCS.
6. Add 100 μL of the labeled target cells (concentration of 5,000
cells/100 μL) to each of the wells already containing the T-cell
clone.
7. Make a control plate for spontaneous and maximum release.
For each target make 3 wells with 100 μL TCM and 3 wells
100 μL 1% Triton X-100. Add 100 μL target suspension per
well.
8. Spin the plates for 1 min at 550 × g.
9. Incubate the plates for 4 h at 37°C, 5%CO2.
10. Spin the plates for 1 min at 550 × g.
11. Harvest the supernatants.
 19 Detection of Allo-HLA Cross-Reactivity... 347

12. Measure the chromium release and calculate the specific lysis
(see Note 31).

4. Notes

1. We recommend using PE-labeled tetrameric complexes for

brighter cell staining. However, other conjugates may be used.
The tetramers used were produced in-house. However, a num-
ber of companies (e.g., ProImmune, United Kingdom) have a
range of available tetramers.
2. Prior to attempting sorting we recommend screening first for
the presence of viral peptide/HLA tetramer complex binding
T-cells within the peripheral blood of the donor. Ensuring the
donor has a high proportion of the virus-specific T-cells of
interest in the peripheral blood increases the probability of suc-
cessful sorting and cloning. Viral seropositivity does not guar-
antee the donor blood will contain T-cells with the viral
specificity of interest.
3. PBMCs should be prepared by Ficoll hypaque gradient cen-
trifugation. Freshly defrosted PBMCs are also suitable for sin-
gle cell sorting.
4. Cells from a single random volunteer donor are suitable irre-
spective of HLA typing. 5 × 106 donor cells are required for
every 1 × 106 virus-specific T-cells to be stimulated.
5. A G418 resistance gene has been added to some of the plasmid
constructs. SALs generated with such constructs are also cul-
tured with G418 added to the culture medium to provide a
selection advantage and facilitate growth of the transfectant.
6. The antibodies were obtained in-house from the Department
of Immunohematology and Blood Transfusion, LUMC, the
Netherlands. They are available commercially on request at
[email protected].
7. Peptides were produced in-house at LUMC, the Netherlands.
These are available for purchase; however, a number of compa-
nies will synthesize peptides on request.
8. Perform all steps prior to sorting at 4°C. When tetramers are
used at room temperature they are capable of activating the
virus-specific T-cells for sorting.
9. Do not forgot positive and negative control tubes for set up of
single cell sorting apparatus.
10. It is most convenient to use a pool of FITC-labeled conjugates
which will mark the cells which are not of interest. For example,
if CD8 memory T-cells are to be sorted then CD45Ra/CD4/
348 L.J. D’Orsogna et al.

CD19-FITC will exclude naïve cells, CD4 T-cells, and B-cells.

In this case the CD8 T-cells to be sorted would be PE positive
(tetramer positive) and FITC negative.
11. We recommend using fresh T-cell medium for every T-cell
sorting or stimulation (Maximum 2 weeks old).
12. Only freshly collected PBMCs are suitable as feeder cells.
Previously frozen cells are not suitable.
13. EBV LCLs are used to provide additional nonspecific stimula-
tion to the sorted T-cells. However, they are not definitely
required for successful stimulation. The HLA expression of the
EBV LCLs is not important.
14. Virus-specific T-cell clones are used to confirm that the allore-
activity and viral specificity are mediated by the same TCR.
Virus-specific T-cell lines can also be generated by sorting mul-
tiple tetramer complex binding T-cells per well (e.g., 25 cells/
well). T-cell lines are useful for screening purposes and may
contain T-cells with the same viral specificity but different TVR
Vb usage to that of the single cell sorted clones.
15. For single cell sorted wells expect growth at day 12–14.
16. The TCR will often be absent from the cell surface for the first
5 days after sorting/stimulation.
17. Flow cytometry tests for HLA should always be performed on
ice to avoid capping.
18. When you have a low number of HLA-positive cells (before
sorting) use PE-labeled w6/32, it is more sensitive than FITC-
labeled w6/32.
19. If NO positive cells are present consider repeating procedure
from Subheading 3.2, step 1. If repeatedly negative seek spe-
cialist advice.
20. Usually there is evidence of cell growth by day 3–4. SALs will
continue to grow as long as they are split and supplied with
fresh medium. You may wish to freeze aliquots for future use.
21. Virus-specific T-cell clones should be used in IFNγ ELISA
assays at day 9–10 after stimulation. Earlier use of the T-cell
clone may be associated with ongoing IFNγ production from
the T-cells.
22. The cells survive better if the wash medium contains protein
(e.g., 1% FCS). Instead of IMDM you can also use other media
for harvesting and washing, e.g., RPMI.
23. Leaving SALs at 26°C for two nights increases HLA
expression.
24. IFNγ stimulation probably does increase HLA expression.
25. K562 cells do not have functional CIITA gene product (5).
Therefore, K562 cells will not express HLA class II molecules.
 19 Detection of Allo-HLA Cross-Reactivity... 349

26. For easy transfer of your supernatants, make the layout of your
culture plate the same as the layout of the ELISA plate and
leave wells open for your standard dilutions and blank.
27. When the supernatants have to be stored for a longer period
(>1 month), it is better to store them in small siliconized
vials.
28. Cytotoxicity assays are best performed at day 7–8 after stimula-
tion of the T-cell clone.
29. The half life off sodium chromate is short, be sure to use the
correct amount according to the manufacturers instructions.
30. The effector target ratios commonly used are 30:1, 10:1, 1:1,
and 0.1:1. Therefore, given that there are 5,000 target cells/
well then the number of effector cells to be added are 150,000,
50,000, 5,000, and 500 (each in 100 μL), i.e., cell concentra-
tions 1.5 × 106, 0.5 × 106, 0.05 × 106, and 0.005 × 106/mL
respectively.
31. % specific lysis = ((test release–spontaneous release) (maximum
release–spontaneous release)) × 100%. For more details on cal-
culations of % specific lysis, see ref. (2).

References
1. Amir A, D’Orsogna L, Roelen D, van Loenen HLA-specific antibodies. Hum Immunol 66:
M, Haagedoorn R, de Boer R et al (2010) Allo- 519–525
HLA reactivity of viral specific memory T-cells 4. Mulder A, Zoet Y, Eijsink C. Single MHC anti-
is common. Blood 115:3146–3157 gen expressing cell lines for the definition of
2. D’Orsogna LJA, Amir A, Zoet Y, van der Meer- monoclonal antibody specificities. In: Hansen
Prins P, van der Slik A, Kester M et al (2009) JA(ed) Histocompatibility testing 2002, HLA
New tools to monitor the impact of viral infec- 2002, Munksgaard, 2005
tion on the alloreactive T-cell repertoire. Tissue 5. van der Stoep N, Biesta P, Quinten E, van den
Antigens 74:290–297 Elsen P (2002) Lack of IFNγ mediated induc-
3. Zoet Y, Eijsink C, Kardol M, Franke-van Dijk tion of the class II transactivator (CIITA)
M, Wilson G, de Paus R et al (2005) The through promoter methylation is predomi-
single antigen expressing lines (SALs) con- nantly found in developmental tumor cell Lines.
cept: an excellent tool for screening for Int J Cancer 97:501–507
 Chapter 20

Separation and Cryopreservation of Lymphocytes

from Spleen and Lymph Node
Gabriella Tassone and Samantha J. Fidler

Abstract
Spleen and lymph node retrieved post-mortem from deceased organ donors are a rich source of lympho-
cytes. Storage of lymphocytes separated from these sources can be valuable where post-transplant testing
(crossmatching) is required. DNA extraction from stored lymphocytes also allows further genetic testing
where required, for example additional HLA typing not performed at the time of transplant for donor-
specific antibody monitoring. Methods for the isolation and freezing of such cells is described.

Key words: Lymphocytes, Spleen, Lymph node, Cryopreservation

1. Introduction

Peripheral blood is readily available as a source of lymphocytes for

serological HLA typing and histocompatibility testing, however
for some individuals such as deceased organ donors, high volumes
of peripheral blood are not available and re-bleeds are not possible.
Alternative sources, such as spleen and lymph node, can be utilised
for bulk separation of lymphocytes for immediate use and or cryo-
preservation. The Ficoll-Hypaque density gradient separation
technique used for peripheral blood lymphocyte separation can
also be utilised for spleen and lymph node preparations (1). A cell
suspension is layered over the density gradient. On centrifugation,
the denser erythrocytes and granulocytes pass through the gradi-
ent and pellet at the bottom of the tube. Platelets remain in the
supernatant, while lymphocytes remain at the plasma/gradient
interface. The interface is removed and washed to remove any
platelet contamination. The lymphocytes may then be used imme-
diately or cryopreserved for future use (2, 3).

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_20, © Springer Science+Business Media New York 2012

351
352 G. Tassone and S.J. Fidler

2. Materials

2.1. Isolation of Spleen 1. RPMI medium 1640 (Gibco Invitrogen) (see Note 1).
and Lymph Node Cells 2. Sections of fresh lymph node or spleen (see Note 2).
3. Ficoll-Paque (GE Healthcare Bioscience) (see Note 1).
4. Sterile Petri dishes (Sarstedt) (see Note 1).
5. Disposable sterile scalpels (Swann- Morton).
6. 30-mL round bottom capped tubes (Sarstedt).
7. 30-mL syringes (Becton Dickinson).
8. Cannulae (Maerski Medical).
9. Size 7 and/or size 9 sterile plastic pasteur pipettes (Samco
Scientific).
10. Centrifuge (Beckman TJ-6 or equivalent).
11. 50-mL Cellstar tubes (Greiner).
12. 20% Foetal calf serum (FCS)/RPMI (see Note 3).

2.2. Cryopreservation 1. Lymphocytes prepared in Subheading 3.1.

of Separated 2. 20% FCS/RPMI (see Note 3).
Lymphocytes
3. 15% Dimethyl sulphoxide (DMSO)/RPMI (see Note 3).
4. Nunc cryo tube vials (Thermo Scientific).
5. Neubauer cell counting chamber (see Note 4).
6. Light microscope.

2.3. Thawing Frozen 1. 10% FCS/RPMI (see Note 3).

Lymphocytes for 2. FCS (Gibco).
Histocompatibility
3. 15-mL round bottomed tubes (Sarstedt).
Testing
4. 1.5-mL microtubes (Sarstedt).
5. Microfuge (Hetlich Mikro 22R or equivalent).
6. 37°C water bath (Thermoline Scientific or equivalent).
7. Citrated phosphate-buffered saline (PBSC) (see Note 5).

3. Methods

3.1. Isolation of Spleen 1. Place an approximate 2 cm3 section of spleen onto a sterile
and Lymph Node Cells Petri dish and using a sterile scalpel, trim the capsule from the
spleen and discard.
2. In a separate Petri dish, place the lymph nodes and remove any
fatty tissue from around the nodes.
 20 Separation and Cryopreservation of Lymphocytes from Spleen and Lymph Node 353

3. Using sterile forceps and scalpels, gently tease spleen or lymph

node cells into approximately 5 mL of RPMI placed in the
Petri dish (see Note 6).
4. Using a sterile pasteur pipette, place the cell suspension into
2 × 30-mL round bottom capped tubes. Wash the Petri dish
out with RPMI to get all of the cell suspension. Avoid picking
up pieces of tissue.
5. Spin the tubes at 1,800 × g for 30 s (see Note 7).
6. Transfer 15 mL aliquots of cell suspension (supernatant) into
an appropriate number of 50-mL CellStar tubes.
7. Dilute the transferred supernatant 1:2 with RPMI, making up to
a volume of 30 mL. Underlay 10 mL of sterile Ficoll-paque using
a cannula attached to a 30-mL syringe (see Notes 8 and 9).
8. Centrifuge the cell suspension/gradient for 30 min at
4,000 × g.
9. Using a sterile pasteur pipette, carefully remove the cells at the
gradient interface and transfer to clean 30-mL round bot-
tomed capped tubes. Wash by filling the tube to 30 mL with
RPMI and centrifuging at 1,500 × g for 10 min.
10. Carefully pour out and discard the RPMI and resuspend the
remaining cell pellet in 5 mL of 20% FCS/RPMI.

3.2. Freezing 1. Determine the lymphocyte concentration using a Neubauer

of Isolated chamber (see Notes 4 and 10).
Lymphocytes 2. Adjust the cell concentration to 20 × 106/mL by the addition
for Storage or removal of 20% FCS/RPMI as described in Note 11.
3. Place the tube of cell suspension onto wet ice to begin the
cooling process. Leave to cool for a few minutes before pro-
ceeding to step 4.
4. Add an equal volume of 15% DMSO/RPMI to the cell sus-
pension drop-wise with gentle constant mixing (see Note 12).
5. Transfer 1 mL aliquots of cell suspension into an appropriate
number of Nunc tubes and freeze at −80°C overnight or for at
least 1.5 h before transferring into liquid nitrogen (see Note 13).

3.3. Thawing Frozen 1. Place a 1 mL Nunc containing frozen cells prepared in

Lymphocytes in Subheading 3.2 in 37°C water bath to thaw (see Note 14).
Preparation for 2. Immediately add 0.5 mL FCS drop-wise and transfer the cells
Dynabead Separation to a 1.5-mL microtube (see Note 15).
for Serological HLA 3. Microfuge for 1 min at 1,000 × g.
Class I or II Typing
4. Remove the supernatant by tipping off slowly taking care not
to disturb the cell pellet and lose cells, and resuspend cells in
1 mL 10% FCS/RPMI.
354 G. Tassone and S.J. Fidler

5. Incubate at 37°C in a waterbath for 10 min, mix by gentle

inversion after 5 min.
6. Centrifuge the cell suspension at 1,000 × g for 1 min, remove
the supernatant by tipping off slowly.
7. Resuspend in 1 mL 10% FCS/RPMI, transfer to a 15-mL
round bottomed tube and repeat step 6.
8. Slowly add 8 mL PBSC. Cells are ready for T- and B- cell sep-
aration—refer to Chapter 22 (see Note 16).

4. Notes

1. Use sterile equipment and reagents throughout this procedure.

Perform in a Class II Biosafety cabinet.
2. Spleen and lymph node should be collected fresh and asepti-
cally by the surgical retrieval team with appropriate donor con-
sent. The material should be placed in RPMI at room
temperature and expeditiously transported to the laboratory
for processing. The specimens should be processed within 12 h
of donor retrieval. After this time, cells may still be extracted
but viability may not be sufficient for use in crossmatching or
any assay requiring viable cells. However, these cells can still be
used as a source of DNA.
3. 10% FCS/RPMI, 20% FCS/RPMI, and 15% DMSO/RPMI
should be made in a sterile cabinet using aseptic techniques.
4. Coulter cell counter or equivalent may be used to determine
cell numbers if available.
5. pH 7.4 PBS can be made easily in-house or bought commer-
cially from suppliers, such as MP Biomedicals and Dako. PBSC
is PBS pH7.4 + 0.6% Tri-sodium citrate, i.e. 12 g Tri-sodium
citrate in 2 L filtered PBS.
6. To extract the cells hold the tissue firmly with the forceps. With
the scalpel use minimal pressure to gently tease the cells into
the RPMI. From time to time, wash the tissue with RPMI
using a pasteur pipette. A cloud of cells should be seen coming
into the media. Do not allow the media to become too over-
loaded with cells, add more RPMI as required so that the
medium looks opaque.
 20 Separation and Cryopreservation of Lymphocytes from Spleen and Lymph Node 355

Sample introduction point Cover glass

Counting 0.1 mm sample
chambers depth
Cover glass
mounting support

Fig. 1. The Neubauer chamber, indicating sample introduction point, and counting chamber.

7. This centrifugation will spin down any tissue debris but is short
and gentle enough to leave the cells in suspension. This step
can be repeated if necessary.
8. Remove the syringe plunger from the barrel and discard. Place
a cannula on the end (where a hypodermic needle is usually
placed) and carefully place the cannula into the cell suspension
so the end sits on the bottom of the tube. Pour 10 mL of den-
sity gradient into the syringe barrel and lift the syringe slightly
to allow flow of gradient.
9. If the tubes are accidently knocked or dropped resulting in
mixing of the gradient/cell suspension interface, the samples
can be recovered by underlaying the gradient/cell mix with a
further 10 mL of density gradient.
10. Cell concentration determination using a Neubauer chamber
(see Fig. 1):
● Clean lens thoroughly prior to use.
Ensure cover slip is also thoroughly cleaned.
● Place cover slip over the counting surface. Add the cell
suspension using a pipette. (The area under the cover slip
fills by capillary action).
● Place the counting chamber on the microscope stage and
bring the counting grid into focus at low power (40×).
● There are 9 large squares, each of which contains 16
medium squares, each of which contains 25 small squares.
One large square has 0.1 μL volume (see Fig. 2).
● Count the number of lymphocytes in 5 of the medium
squares and multiply by 5 (=number of cells/0.1 μL).
356 G. Tassone and S.J. Fidler

Fig. 2. Diagrammatic representation of the Neubauer counting chamber. One large square
have 0.1 μL volume.

11. Remove FCS/RPMI by centrifuging the cell suspension at

1,500 × g for 10 min. Resuspend the cell pellet in an appropriate
volume of FCS/RPMI to adjust to the correct concentration.
12. DMSO minimises cell damage in the freezing process by grad-
ually replacing water within the cells.
13. The cooling rate should be approximately −1°C per minute
and can be assisted by first placing the Nunc tubes in a cryo
container, such as “Mr Frosty” from Nalgene and placing the
container in the −80°C freezer, or by using a controlled rate
freezer.
14. Cells should be thawed as rapidly as possible as prolonged
exposure to DMSO will compromise cell viability.
15. DMSO is toxic to lymphocytes at room temperature so rapid
washing of the cells is essential.
16. You can at this stage check the cell count (see Note 9) and
viability. This can be prudent if you are suspicious about the
 20 Separation and Cryopreservation of Lymphocytes from Spleen and Lymph Node 357

storage conditions of the cryopreserved cells or the viability of

the cells prior to freezing. However, if you have a limited num-
ber of cells stored, you may wish to proceed to T-and B-cell
separation and check the viability and cell count following
separation.

References

1. Bøyum A (1968) Isolation of mononuclear 3. Lorentzen D (2000) Cell preservation. ASHI

cells and granulocytes from human blood. Laboratory Manual 4th Edition, vol 1
Scand J Clin Lab Invest Suppl 97:77–87
2. Strong DM (2000) Cryopreservation of lym-
phocytes in bulk. ASHI Laboratory Manual
4th Edition, vol 1
 Chapter 21

Crossmatching by Complement-Dependent
Lymphocytotoxicity
Samantha J. Fidler

Abstract
The presence of preformed donor-specific HLA antibodies detected by Complement-dependent cyto-
toxicity (CDC) crossmatch assay is associated with a high incidence of hyperacute or accelerated rejection
and remains one of the gold standard tests pre-transplant. The standard CDC crossmatch detects IgG1,
IgG3, and IgM antibody, i.e. complement fixing, bound to the native viable cell surface of lymphocytes.
The crossmatch can be enhanced with the addition of anti-human-globulin to detect non-complement
fixing antibodies (IgG2 and IgG4), and sensitivity can be improved with prolonged incubation times.

Key words: Crossmatch, Complement-dependent cytotoxicity, Alloantibody, Autoantibody, IgG, IgM

1. Introduction

Since the landmark paper by Patel and Terasaki (1) was published
in the late 1960s, a positive (T cell) lymphocytotoxic crossmatch
has been considered a contraindication to renal transplantation.
The crossmatch assay detects high levels of recipient antibodies
against donor HLA antigens, and a positive crossmatch is associ-
ated with hyperacute and accelerated acute rejection, and long-
term graft failure. The complement-dependent lymphocytotoxicity
(CDC) crossmatch remains to this day the gold standard as the
final pre-transplant check for donor-specific anti-HLA antibodies.
Since the 1960s a number of modifications, such as the AHG-
CDC crossmatch and treatment of sera to remove IgM reactivity,
have been made to improve sensitivity and specificity. The barrier
of a positive crossmatch to transplantation has somewhat dimin-
ished in the current era of desensitisation protocols and effective
treatment for rejection episodes. However, it is clear that the

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_21, © Springer Science+Business Media New York 2012

359
360 S.J. Fidler

presence of class I anti-HLA antibodies pre-transplant is associated

with poor graft survival and there is increasing evidence that class
II anti-HLA antibodies are associated with an increase in rejection
episodes and poor long-term survival, particularly in the presence
of a positive B cell crossmatch (2, 3).
Serum samples from the recipient are incubated with separated
T and B lymphocytes from a potential organ donor. If donor reac-
tive antibodies are present, addition of complement results in
cell death via activation of the complement cascade. A positive T cell
crossmatch is usually due to antibodies against HLA class I and is
a contraindication to transplantation. A positive B cell crossmatch
in the absence of a positive T cell crossmatch may be due to anti-
bodies against HLA class II or weak antibodies against HLA class I.
A positive B cell crossmatch is not considered a contraindication to
transplantation in all centres, but rather a risk factor for transplan-
tation (4). Positive crossmatches due to IgM antibodies are not
usually considered a contraindication to transplantation as many
are due to the presence of harmless IgM autoantibody. IgM reac-
tivity can be removed by dithiothreitol (DTT) treatment or heat
treatment of sera.

2. Materials

2.1. Separation of T 1. Anticoagulated blood samples (ACD) (see Note 1).

and B Lymphocytes 2. Human B cell whole blood (WB) positive selection EasySep
beads—includes positive selection cocktail (StemCell
Technologies Product Number 18184 HLA).
3. Human CD3 WB positive selection EasySep beads—includes
positive selection cocktail (StemCell Technologies Product
Number 18081HLA).
4. 15-mL round bottomed tubes.
5. Magnetic Separation Device-EasySep magnet.
6. Sterile phosphate buffered saline (PBS) pH 7.4-Filtered and
stored at 4°C.
7. Carboxyfluorescein diacetate (CFDA) staining solution (stored
at 4°C wrapped in aluminium foil) (see Notes 2 and 3).
8. McCoys 5A tissue culture medium + 5% heat-inactivated fFoetal
calf serum (HIFCS).
9. Haemoglobin/ethidium bromide staining solution (stored at
4°C wrapped in aluminium foil) (see Notes 2 and 3).
10. 1× RBC lysis buffer (Made by diluting 10× RBS lysis buffer)
(see Note 4).
11. Hamilton syringes—1 and 5-μL dispensing.
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 361

12. 10-, 20-, 100-, 200-, and 1,000-μL pipette—Gilson or

equivalent.
13. Pipette tips (10, 20, 100, 200, 1,000 μL).
14. 60-Well Terasaki trays (containing 5 μL mineral oil per well).
15. RoboSep-automated cell separator (see Note 5).
16. 22°C incubator (see Note 6).
17. Inverted fluorescence microscope.

2.2. Standard CDC 1. 1 and 5-μL single dispense pipettes (e.g., Hamilton, Robbins).
Crossmatch 2. 60- or 72-well microlymphocytotoxicity (Terasaki) trays
(containing 5 μL mineral oil per well).
3. Recipient sera (see Notes 7 and 8).
4. Donor lymphocytes—refer to Subheading 2.1.
5. Fluorescent stain (Haemoglobin/ethidium bromide or
alternative).
6. Pooled rabbit complement—refer to Subheading 3.5:
Evaluation of complement—(PelFreeze) (see Note 9).
7. Citrated PBS pH 7.4.
8. PBS pH 7.4.
9. AB Negative serum (see Note 10)—negative control.
10. Anti-thymocyte globulin (ATG) (see Note 11)—positive
control.
11. B cell positive control (see Note 12).
12. Inverted fluorescence microscope.
13. 37°C waterbath.
14. 22°C incubator (see Note 6).
15. 1- and 5-μL automatic dispensers (e.g. Greiner, OLI, Robbins).
16. 1% Kaolin (Sigma).

2.3. AHG-CDC As Subheading 2.2 plus.

Crossmatch
1. Anti-human globulin (AHG) (Biotest).
2. RPMI Media (Sigma).

2.4. Heat Treatment As Subheading 2.2 plus.

of Sera to Remove
1. 400-μL capped tubes (Beckman).
IgM Reactivity
2. 60°C waterbath or thermal cycler.
3. Microfuge.
4. IgM control serum (see Note 13).
5. IgG control serum (see Note 14).
362 S.J. Fidler

2.5. Complement As Subheading 2.2 plus.

Evaluation
1. Well-characterised antisera (see Note 15).

3. Methods

3.1. Separation of T 1. Into a 15-mL round bottomed tube add 4 mL ACD blood and
and B Lymphocytes 4 mL of 1× RBC lysis buffer. Mix by gently inverting.
2. Add 200 μL of positive selection cocktail (Human B cell WB
positive selection for B cells or Human CD3 WB positive selec-
tion for T cells) to the whole blood/lysis mixture. Incubate at
room temperature for 15 min (see Note 6).
3. Mix the EasySep beads to ensure uniform suspension by
pipetting up and down vigorously at least five times (DO NOT
VORTEX).
4. Add 200 μL of the mixed EasySep beads to the respective
tubes. Incubate for 10 min at room temperature.
5. Add 2 mL PBS pH 7.4 and mix gently by pipetting up and
down.
6. Remove cap from tube and place on EasySep magnet. T cells
require 5 min, B cells require 10 min on the magnet.
7. Pick up magnet and in one continuous motion pour contents
out of the 15-mL tube, leave inverted for 2–3 s (DO NOT
SHAKE OR BLOT ANY DROPS REMAINING ON THE
MOUTH OF THE TUBE).
8. Remove tube from magnet and add 10 mL PBS pH 7.4. Place
on magnet for another 5 min (both T and B cells).
9. Pick up magnet and in one continuous motion pour contents
out leave inverted for 2–3 s as per step 7 (DO NOT SHAKE
OR BLOT ANY DROPS REMAINING ON THE MOUTH
OF THE TUBE).
10. Remove the 15-mL tube from magnet add 0.5 mL CFDA
solution drop wise directly onto the cells in the 15-mL tube and
incubate for 5 min at room temperature (mix the tube gently by
hand from time to time during this period).
11. Place the 15-mL tube into the magnet and fill the tube to top
of magnet with PBS pH 7.4. Incubate for 2 min.
12. Pick up magnet and in one continuous motion pour contents
out leave inverted for 2–3 s (DO NOT SHAKE OR BLOT
ANY DROPS REMAINING ON THE MOUTH OF THE
TUBE).
13. Fill CTS tube to top of magnet with PBS pH 7.4 for 2 min.
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 363

Fig. 1. Diagrammatic representation of the 60-well Terasaki tray.

14. Pick up magnet and in one continuous motion pour contents

out of the 15-mL tube, leave inverted for 2–3 s (DO NOT
SHAKE OR BLOT ANY DROPS REMAINING ON THE
MOUTH OF THE TUBE).
15. Resuspend cells by adding fresh 5% HIFCS/McCoy’s (1 mL
for class I and 0.5 mL class II) to the 15-mL tube.
16. Using a Hamilton syringe spot 1 μL of cell suspension in
triplicate onto an oiled Terasaki tray.
17. Add 5 μL Haemoglobin/ethidium bromide stain to each well
(see Note 4). Allow cells to settle for approximately 20 min.
18. View the cells under the fluorescent microscope to determine
cell concentration and viability. Concentration can be adjusted
by addition or removal of 5% FCS/McCoys (see Note 16).
19. Hold the cells at room temperature for use in the crossmatch
assay (see Note 17).

3.2. Standard CDC 1. Add 1 μL of ATG dilutions and AB negative sera to an oiled
Crossmatch Terasaki tray using a Hamilton syringe (see Note 18, Fig. 1).
3A—ATG50, 3B—AB Neg, 3C—ATG500, 3D—ATG200,
3.2.1. Tray Set-Up
3E—AB Neg, 3F—B Cell Pos, 4A—ATG400, 4B—ATG25,
4C—AB Neg, 4D—ATG100, 4E—ATG800, 4F—ATG1000.
2. Add 1 μL recipient serum to wells 1D, 1E, and 1F and 1 μL
AB negative serum to wells 1A, 1B, and 1C using a Hamilton
syringe. Historical peak sera can be added as appropriate in
triplicate to wells 2A to 2F (see Note 8). Trays should be made
at least in duplicate for both T and B cell crossmatch, i.e. at
least four trays in total (see Note 19).

3.2.2. Crossmatch Set-Up 1. Add 1 μL T or B cell suspension (refer to Subheading 3.1)

using an automated cell dispenser to avoid cross-contamination.
Ensure all wells are thoroughly mixed (see Note 20).
2. Incubate the trays for 45 min at 22°C (or room temperature—
see Note 6) (see Note 21).
364 S.J. Fidler

Table 1
Cytotoxicity scores

Percentage cell death Score Conventional score

0 A 1
1–2 B 1
3–5 C 1
6–10 D 1
11–15 1 1
15–25 2 1
25–35 3 2
35–45 4 2/4
45–55 5 4
55–65 6 4/6
65–75 7 6
75–85 8 8
85–100 9 8
Empty well or unreadable X 0
Insufficient cell numbers 0 0

3. Following incubation, add 5 μL of previously screened pooled

rabbit complement (see Subheading 3.5) to each well using an
automated dispenser to avoid cross-contamination.
4. Incubate the trays for 45 min at 22°C (or room temperature—
see Note 6) (see Note 21).
5. Following complement incubation, add 5 μL Haemoglobin/
ethidium bromide stain to each well. Allow cells to settle for
approximately 20 min.
6. Score the percentage cell death semi-quantitatively under the
fluorescent microscope according to Table 1 (see Notes 22
and 23).

3.2.3. Interpretation 1. A positive or negative crossmatch is determined by comparing

of Crossmatch Results the crossmatch score with the background reactivity, i.e. the
AB negative control serum as shown in Fig. 2 (see Notes 24
and 25).
2. Results should be interpreted according to Table 2.
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 365

Fig. 2. (a) Autologous crossmatch (The circle denotes B cell crossmatch results, the triangle denotes T cell crossmatch
results). Crossmatches are read according to Table 1. The AB serum is read as a background control, i.e. contains no HLA
antibodies. The autologous crossmatch should be interpreted in relation to the crossmatch results of this control. The AB
Neg is read as a carry-over control—this AB serum is located next to strongly positive ATG controls. The ATG controls can
be used to monitor the sensitivity of the assay. The laboratory should determine its own QC criteria for example an assay
might only pass QC if the ATG 400 dilution crossmatch result is >AB serum control. In this example, the autologous T and
B cell crossmatches are negative. (b) Allogeneic crossmatch (the circle denotes B cell crossmatch results, the triangle
denotes T cell crossmatch results). Crossmatches are read according to Table 2. The AB serum is read as a background
control, i.e. contains no HLA antibodies. The allogeneic crossmatch should be interpreted in relation to the crossmatch
results of this control. Other controls on this crossmatch tray follow the same criteria as those in the autologous cross-
match (a). In this example, both the T and B cell allogeneic crossmatches are positive.
366 S.J. Fidler

3. Results are no longer considered purely as an indication/

contraindication to transplantation; but are reviewed with other
immunological data to identify level of risk (Tables 3 and 4)
(see Note 26).

Table 2
Interpretation of CDC crossmatch results

Autologous Autologous Allogeneic Allogeneic

T cell B cell T cell B cell
crossmatch crossmatch crossmatch crossmatch
result results results results Interpretation

Negative Negative Negative Negative Low-risk transplant, not a contraindication

Negative Negative Negative Positive Moderate-risk transplant, not an absolute
contraindication
Suggests donor-specific HLA class II antibodies
Negative Negative Positive Positive High-risk transplant, possibility of hyperacute
or accelerated rejection, contraindication
Negative Negative Positive Negative Potential high-risk transplant, however unusual
result for HLA antibodies as the B cell
crossmatch should be positive. Warrants
further investigation for non-HLA antibody
Negative Positive Any result Any result Removal of IgM required prior to crossmatch
Positive Positive Any result Any result Removal of IgM required prior to crossmatch

Table 3
Interpretation of T cell crossmatch results taking account of the presence or absence
of donor-specific antibodies (DSA)

Allogeneic T-cell crossmatch

Positive—not reduced Positive—reduced by heat

DSA (MFI) Negative by heat or DTT or DTT
>8,000 These results are not an These results are a These results are not an
(strong) absolute contraindication to contraindication to absolute contraindication to
transplantation; however, transplantation transplantation; however,
the the presence of donor- the presence of donor-
specific HLA antibodies specific HLA antibodies
detected by Single Antigen detected by Single Antigen
Bead assay increases the risk Bead assay increases the risk
of rejection episodes of rejection episodes
(continued)
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 367

Table 3
(continued)

Allogeneic T-cell crossmatch

Positive—not reduced Positive—reduced by heat

DSA (MFI) Negative by heat or DTT or DTT
2,000–8,000 These results are not an These results are a These results are not an
(moderate) absolute contraindication to contraindication to absolute contraindication
transplantation; however, transplantation to transplantation; however,
the presence of donor- the presence of donor-
specific HLA antibodies specific HLA antibodies
detected by Single Antigen detected by Single Antigen
Bead assay may increase the Bead assay may increase the
risk of rejection episodes risk of rejection episodes
500–2,000 These results are not a These results are a These results are not a
(weak) contraindication to contraindication to contraindication to
transplantation. The clinical transplantation transplantation. The clinical
significance of weak significance of weak
donor-specific HLA donor-specific HLA
antibodies detected by antibodies detected by
Single Antigen Bead assay is Single Antigen Bead assay is
unknown unknown
<500 These results are not a These results are a These results are not a
(negative) contraindication to contraindication to contraindication to
transplantation transplantation transplantation

Table 4
Interpretation of B cell crossmatch results taking account of the presence
or absence of DSA

Allogeneic B-cell crossmatch

Positive—not reduced Positive—reduced by

DSA (MFI) Negative by heat or DTT heat or DTT
>8,000 These results are not an These results are a contraindi- These results are not an
(strong) absolute contraindication cation to transplantation absolute contraindica-
to transplantation; tion to transplantation;
however, the presence of however, the presence
donor-specific HLA of donor-specific HLA
antibodies detected by antibodies detected by
Single Antigen Bead Single Antigen Bead
assay increases the risk of assay increases the risk
rejection episodes of rejection episodes
(continued)
368 S.J. Fidler

Table 4
(continued)

Allogeneic B-cell crossmatch

Positive—not reduced Positive—reduced by

DSA (MFI) Negative by heat or DTT heat or DTT
2,000–8,000 These results are not an These results are not an These results are not an
(moderate) absolute contraindication absolute contraindication absolute contraindica-
to transplantation; to transplantation; however, tion to transplantation;
however, the presence of the presence of donor- however, the presence
donor-specific HLA specific HLA antibodies of donor-specific HLA
antibodies detected by detected by Single Antigen antibodies detected by
Single Antigen Bead Bead assay in the presence Single Antigen Bead
assay may increase the of a positive B cell cross- assay may increase the
risk of rejection episodes match is associated with an risk of rejection
increased risk of rejection episodes
episodes
500–2,000 These results are not a These results are not an These results are not a
(weak) contraindication to absolute contraindication contraindication to
transplantation. The to transplantation; however, transplantation. The
clinical significance of the risk of rejection clinical significance of
weak donor-specific HLA episodes may be increased weak donor-specific
antibodies detected by HLA antibodies
Single Antigen Bead detected by Single
assay is unknown Antigen Bead assay is
unknown
<500 These results are not a These results are not an These results are not a
(negative) contraindication to absolute contraindication contraindication to
transplantation to transplantation. The transplantation
presence of a positive B cell
crossmatch in the absence
of DSA is uncertain but
may increase the risk of
rejection episodes

3.3. AHG-CDC Follow Subheading 3.2.1 for tray set up.

Crossmatch
3.3.1. Tray Set Up

3.3.2. Crossmatch Method 1. Follow Subheading 3.2.2 to step 3 (see Note 27).
2. Wash trays with RPMI. Add 10 μL RPMI to each well.
Centrifuge for 1 min at 1,000 rpm. Quickly flick out the RPMI
in one swift movement.
3. Repeat step 2 two more times, then add 1 μL AHG to each
well. Leave for 1 min.
4. Follow Subheading 3.2.2 from step 4 until the end.
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 369

3.3.3. Interpretation As per Subheading 3.2.3.

of Results

3.4. Heat Treatment See Note 28.

of Sera to Remove
IgM Reactivity

3.4.1. Preparation of Sera 1. Pre-warm the waterbath to 60°C (see Note 29).
2. Aliquot 200 μL of patient crossmatch sera and control sera
(AB Neg, IgG, and IgM controls) into suitably labelled
Beckman tubes.
3. Place the tubes into the waterbath ensuring that the sera are
fully submerged (see Note 30).
4. Incubate the tubes for exactly 13 min (see Note 31).
5. Remove the samples from the waterbath and immediately
microfuge at 10,000 rpm for 1 min.
6. Remove the supernatant into a clean Beckman tube, taking
care not to disturb the gelatinous protein precipitate.

3.4.2. Tray Set-Up 1. Add ATG dilutions as per step 1 of Subheading 3.2.1.
2. Add 1 μL AB Negative serum to wells 1A and 1B and 1 μL
heat treated (HT) AB negative serum to wells 1C and 1D using
a Hamilton syringe.
3. Add 1 μL IgG control serum to wells 1E and 1F and 1 μL HT
IgG control serum to wells 2E and 2F.
4. Add 1 μL IgM control serum to wells 2C and 2D and 1 μL
HT IgM control serum to wells 2A and 2B.
5. Add 1 μL patient serum to wells 5A and 5B and 1 μL HT
patient serum to wells 5C and 5D. Historical peak sera can be
added as appropriate in duplicate to wells 5E to 6F (see Note 8).
Trays should be made in duplicate for both T and B cell cross-
match, i.e. four trays in total.

3.4.3. Crossmatch Method The standard CDC crossmatch as described in Subheading 3.2.2,
or the AHG crossmatch as described in Subheading 3.3.2 may be
performed.

3.4.4. Interpretation 1. A positive or negative crossmatch is determined by comparing

of Results the crossmatch score with the background reactivity, i.e. the
AB negative control serum as shown in Fig. 3 (see Notes 24
and 25).
2. Results should be interpreted according to Tables 5 and 6.
370 S.J. Fidler

Fig. 3. Suggested layout for complement evaluation titration.

Table 5
Interpretation of heat-treated T cell crossmatch results

Allogeneic T cell crossmatch result

Interpretation
Neat (non-heat treated) Heat treated of crossmatch Conclusion
Positive Positive Positive IgG alloantibodies detected. High risk,
possibility of hyperacute or accelerated
rejection, contraindication
Positive Weak positive Indeterminate Possible IgG alloantibodies and IgM
autoantibodies detected. Requires further
investigation
Positive Negative Negative IgM autoantibodies detected. Not a
contraindication

Table 6
Interpretation of heat-treated T cell crossmatch results

Allogeneic T cell crossmatch result

Interpretation
Neat (non-heat treated) Heat treated of crossmatch Conclusion
Positive Positive Positive IgG alloantibodies detected. Moderate risk,
not a contraindication
Positive Weak positive Indeterminate Possible IgG alloantibodies and IgM
autoantibodies detected. Requires further
investigation
Positive Negative Negative IgM autoantibodies detected. Not a
contraindication
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 371

3. Results are no longer considered purely as an indication/contrain-

dication to transplantation; but are reviewed with other immuno-
logical data to identify level of risk (Tables 3 and 4) (see Note 26).

3.5. Complement Pooled rabbit serum is used as a source of complement (see Note 9).
Evaluation

3.5.1. Tray Set-Up 1. Serially dilute the complement N/2 with McCoys medium
(see Note 32).
2. Serially dilute antisera with well-characterised reactivity N/2
in AB negative serum (see Note 33).
3. Add 1 μL of serum to a Terasaki tray from Neat to 1/512 (see
Fig. 3).

3.5.2. Crossmatch Set-Up 1. Add 1 μL T or B cell suspension (refer to Subheading 3.1) using
an automated cell dispenser to avoid cross-contamination (see
Note 34). Ensure all wells are thoroughly mixed (see Note 20).
2. Add 5 μL of diluted complement according to the format in
step 3. Incubate at room temperature for 45 min (see Note 21).
3. Add 5 μL ethidium bromide/haemoglobin stain and allow to
rest for 20 min (see Note 3).
4. Score the percentage cell death under the fluorescent micro-
scope according to Table 1 (see Notes 22 and 23).

3.5.3. Interpretation of 1. A positive or negative crossmatch is determined by comparing the

Results crossmatch score with the background reactivity, i.e. the AB nega-
tive control serum as shown in Fig. 4 (see Notes 24 and 25).

Fig. 4. (a) Autologous crossmatch (the circle denotes B cell crossmatch results, the triangle denotes T cell crossmatch
results). Crossmatches are read according to Table 1. The AB serum is read as a background control, i.e. contains no
372 S.J. Fidler

Fig. 4. (continued) HLA antibodies. The autologous crossmatch should be interpreted in relation to the crossmatch results
of this control. The IgG control should not reduce with heat treatment (HT). Reduction may be due to over-incubation of
serum or incubation at the incorrect temperature (>63°C) leading to denaturation of IgG antibody. The IgM control should
reduce with HT. Failure to reduce may be due to under-incubation of serum or incubation at the incorrect temperature
(<60°C). In this example, the autologous T and B cell crossmatches are positive, and reduce with HT indicating the pres-
ence of IgM autoantibody. (b) Allogeneic crossmatch (the circle denotes B cell crossmatch results, the triangle denotes T
cell crossmatch results). Crossmatches are read according to Table 2. The AB serum is read as a background control, i.e.
contains no HLA antibodies. The allogeneic crossmatch should be interpreted in relation to the crossmatch results of this
control. The IgG control and IgM controls follow the same criteria as in the autologous crossmatch (a). In this example,
the allogeneic T and B cell crossmatches are positive, and reduce with HT consistent the presence of IgM autoantibody.

Fig. 5. (a) Example of crossmatch results of a homozygous cell against neat antiserum where the circle is the current comple-
ment, the triangle is evaluation complement 1, the rectangle is evaluation complement 2 and the star is evaluation complement
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 373

Fig. 5. (continued) 3. Evaluation complement 2 has low reactivity and should be discarded. Complement 1 appears to have
reactivity similar to the current complement. Complement 3 has increased reactivity, this may be due to high background
or stronger complement activity (see (b)). (b) Example of crossmatch results of a negative cell against neat antiserum
where the circle is the current complement, the triangle is evaluation complement 1, the rectangle is evaluation comple-
ment 2 and the star is evaluation complement 3. Complement 5 has high reactivity and should be discarded. Complement
1 appears to have consistent reactivity with the current complement.

2. Crossmatch scores of the current and evaluation complement

should be plotted to compare reactivity (see Fig. 5).
3. Evaluation complement can be immediately discarded if it
consistently has high reactivity with cells which are negative for
the antigen against which the antiserum is reactive, i.e. high
background/non-specific reactivity.
4. Evaluation complement can be discarded if it consistently has
low reactivity (compared to the current batch) with cells that
are homozygous or heterozygous for the antigen against which
the antiserum is reactive, i.e. low reactivity.
5. Complement should be re-tested at delivery to ensure shipment
has not caused any reduction in reactivity.

4. Notes

1. Heparin blood is also suitable for cell extraction by this method;

however, EDTA blood is unsuitable. The volume of blood
required is determined by the number of crossmatch tests
(trays), but as a guideline 2 × 10 mL of blood is sufficient for
approximately 10 B cell crossmatch trays and 20 T cell cross-
match trays.
2. Solutions containing fluorescent dyes must be kept in the dark
to prevent quenching.
374 S.J. Fidler

3. Other stains are available, such as ethidium bromide/Indian

Ink; eosin/formalin and other commercial stains.
4. RBS lysis buffer can be stored up to 3 months at 1× concentration
at 4°C.
5. This process can be fully automated with a walk-away robotic
separation system. However, the process can be easily per-
formed manually where a robotic system is not available.
6. An incubator is not necessary if the room temperature of the
laboratory can be regulated to 18–22°C and is carefully
monitored.
7. Serum is obtained from fully clotted blood samples (no
anticoagulant) and should be stored at 4°C for short intervals
(1–2 days) and at −20°C for longer periods. Freeze/thaw cycles
should be avoided. Incomplete clotting of the blood sample can
cause fibrin clots and clumping of lymphocytes and can be
rectified by incubation with 1% Kaolin at 37°C for 1 h. The
serum is frozen and thawed to remove the clot. Approximately
5 μL of serum is required per crossmatch test (tray), i.e. 20 μL
of serum is required for 2 × T cell and 2 × B cell crossmatches.
8. Historical peak sera should be selected according to laboratory
protocol but in general should be the serum with the highest
antibody reactivity (highest PRA). If the recipient has had a
previous transplant, then two historical sera should be chosen:
one with the highest PRA, and one immediately following the
previous graft failure if available.
9. Alternative suppliers of Rabbit complement include CSix, GTI
Diagnostics and BAG.
10. Blood group AB Negative serum from untransfused Male
donors is used as a negative control. It is often pooled from
three or more donations in order to increase volume and decrease
QC frequency. The sera must be screened for the presence of
lymphocyte antibodies prior to use by the most sensitive screen-
ing method available in the laboratory and at the very minimum
by CDC crossmatch.
11. We use ATG serially diluted to determine assay sensitivity
(1/25, 1/50, 1/100, 1/200, 1/400, 1/800, 1/1,000,
1/1,200). An alternative IgG control can be obtained com-
mercially or by pooling sera positive for IgG HLA antibodies.
However, this control must be carefully and thoroughly tested
to ensure that a positive crossmatch will be consistently
obtained against all cells tested.
12. A positive control specific for B cells should be sought. We find
a 1/10,000 dilution of Rituximab to be a good quality control
which shows no T cell reactivity.
21 Crossmatching by Complement-Dependent Lymphocytotoxicity 375

13. An IgM control which is strongly reactive neat, but reduces

completely with heat-treatment (or DTT treatment) should be
sought. This can be obtained commercially, e.g. anti-lymphocyte
serum IgM (ALSM) from OneLambda Inc. Alternatively pooled
IgM antisera may used. The control must be carefully and thor-
oughly tested to ensure it reacts with all cells (T and B cells of
any HLA type) and reduces completely and consistently with
heat-treatment, i.e. it contains no IgG antibody.
14. An IgG control which is strongly reactive neat, but does not
reduce with heat-treatment (or DTT treatment) should be
sought. This can be obtained commercially e.g. anti-lymphocyte
serum IgG (ALSG) from OneLambda Inc. Alternatively pooled
IgG antisera may used. The control must be carefully and
thoroughly tested to ensure it reacts with all cells (T and B cells
of any HLA type) and does not reduce with heat-treatment.
15. Sera fully characterised by CDC crossmatch are used in evalu-
ation and should be monospecific for a particular HLA antigen
(class I or II) if possible. The corresponding cells used in the
complement evaluation crossmatch should include (1) cells
homozygous for the antigen against which the antiserum
reacts; (2) cells heterozygous for the antigen against which the
antiserum reacts; and (3) cells negative for the antigen against
which the antiserum reacts.
16. With experience, cell numbers can be estimated manually
under the fluorescent microscope. 100–150 cells per well
equate to a cell concentration of 1 × 105/mL. Alternatively, a
Neubauer chamber can be used to count the cells more accu-
rately. The concentration can be adjusted to 1 × 105/mL by the
addition/removal of cell medium.
17. Cells should be kept at room temperature for crossmatch. If a
delay of over a few hours should occur cells may be stored at
4°C, however they should be brought back to room tempera-
ture before use. Cells extracted by this method are not suitable
for freezing.
18. Trays can be pre-prepared up to 6 months in advance and
stored at −20°C. Ensure the trays are fully enclosed and prefer-
ably vacuum packed in plastic bags prior to freezing to avoid
ice crystal formation in the trays and subsequent diluting of
controls on thawing. Trays must be thawed to room tempera-
ture before use.
19. If more than one donor is to be crossmatched (i.e. n donors),
the total number of trays required should be at least 4 × n.
20. Trays can be mixed using static by rubbing a latex gloved hand
across the bottom of the tray. Alternatively, the trays may be
centrifuged for a few seconds to bring the cell/serum mixture
to the bottom of the tray well.
376 S.J. Fidler

21. Incubation times can be increased/decreased according to the

complement used (see Subheading 3.4) and the sensitivity
required. B cells are more sensitive than T cells and therefore a
shorter incubation time with complement is often required.
However, use of a weaker complement for B cell crossmatching
may compensate the difference in reactivity, and with careful
selection the incubation times of T and B cell crossmatches can
be synchronised.
22. Live cells are stained with CFDA and appear green under the
fluorescent microscope, dead cells take up ethidium bromide
and appear red. It is not necessary to count individual cells in
each well but rather given an average 2,000 cells are added
per well for lower cytotoxicity an approximation of dead cell
numbers can be made and an estimate of proportion of total
cells determined. For higher cytotoxicity, it is easier to estimate
live cell numbers.
23. We find that by reading standard crossmatch trays in a sensitive
manner, distinguishing between lower differences in cytotoxicity
than the conventional scoring system, we can achieve cross-
match results with the same sensitivity as an AHG-enhanced
crossmatch. We have found with experience cytotoxicity can
be reliably scored by different readers to an average ±1 cyto-
toxicity score.
24. Some laboratories read the negative serum as background and
visually subtract this from all other readings. We recommend
that each of the wells are scored independently and according to
the scale in Table 1, and the results interpreted in relation to
the negative control as shown in Fig. 2.
25. Many laboratories categorise crossmatches according to cross-
match score in terms of strength of reactivity, i.e. weak positive,
moderate positive, strong positive.
26. Crossmatch data should be interpreted in combination with
donor-specific antibody data. We use the Luminex Single
Antigen Bead Assay to derive our antibody results and therefore
Table 3 may need to be tailored for your own requirements
depending on local MFI cut-offs for positivity or the antibody
detection method used.
27. The AHG crossmatch is usually only performed with T cells.
28. DTT is still widely used for the removal of IgM from sera but
has the distinct disadvantage that dilution of the sera is required
and weak IgG alloantibodies may become negative by dilution
and not reduction. In addition, the assay is poorly standardised
between laboratories so exchange of crossmatch trays for local
or national exchange of organs is problematic.
 21 Crossmatching by Complement-Dependent Lymphocytotoxicity 377

29. The precise temperature is critical for this assay. We prefer to use
a waterbath as the temperature can be constantly monitored
and adjusted as necessary.
30. Place tubes in polystyrene floats so that the serum is fully sub-
merged in the waterbath. These floats can be easily made by
cutting the bottom from polystyrene cups and making small
holes in which the tubes are held snugly.
31. The precise timing and incubation temperature should be vali-
dated in your laboratory to ensure IgM antibody is reduced
and IgG reactivity is not affected.
32. Perform the evaluation in parallel with the current batch of
complement. It is reasonable to request evaluation samples
from several different commercial companies for comparison.
33. Well-characterised class I and class II antisera previously tested
by CDC crossmatch should be selected. Monospecific sera, or
sera against well-defined epitopes present in several antigens
(e.g. Bw4 or Bw6) are ideal for this evaluation.
34. A number of cells should be chosen with the following charac-
teristics: (1) homozygous for the antigen against which the
antiserum is reactive; (2) heterozygous for the antigen against
which the antiserum is reactive; and (3) negative for the antigen
against which the antiserum is reactive.

References

1. Patel R, Terasaki PI (1969) Significance of the in positive B-cell crossmatches by Luminex

positive crossmatch test in kidney transplanta- predict late graft loss. Am J Transplant 8:
tion. N Engl J Med 280:735–739 2335–2342
2. Nickerson P, Pochinco D, Karpinski M et al 4. Gebel HM, Bray RA, Nickerson P (2003) Pre-
(2004) Class II donor specific antibodies are transplant assessment of donor-reactive. HLA
pathogenic in primary renal allografts. Am J specific antibodies in renal transplantation:
Transplant 4:257 contraindication vs risk. Am J Transplant 3:
3. Eng HS, Bennett G, Humphreys I et al (2008) 1488–1500
Anti-HLA donor-specific antibodies detected
 Chapter 22

The Lymphocyte Crossmatch by Flow Cytometry for Kidney

Transplantation
Jonathan Downing

Abstract
The flow cytometric lymphocyte crossmatch is a standard technique for evaluating the compatibility of
potential kidney transplant recipients and donors. Recipient serum is incubated with donor lymphocytes
and the latter are analysed in a flow cytometer for the presence of bound IgG antibodies. An increase in
the level of IgG binding compared to a negative control indicates the presence of donor-specific antibodies
which may lead to deleterious graft function. Described here is a method to perform T and B lymphocyte
crossmatching in the same tube to detect IgG donor-reactive antibodies.

Key words: Flow cytometry, Lymphocyte crossmatch, Kidney transplantation, Human leukocyte
antigen, Pronase treatment

1. Introduction

The flow cytometric lymphocyte crossmatch (FCXM) was first

described by Garovoy et al. (1), and has since become a standard
test performed prior to kidney transplantation. The FCXM exam-
ines potential donor lymphocytes for recipient antibody binding by
passing the cells one at a time through a laser and measuring the
amount of a fluorescently labelled secondary anti-Ig antibody.
With the use of labelled monoclonal antibodies differentiation can
be made between IgG and IgM antibodies and between reactivity
towards T and B lymphocytes (2).
A positive complement-dependent cytotoxicity (CDC) cross-
match in the presence of CDC detectable donor-specific HLA anti-
bodies is a contraindication to transplant, and its link to hyperacute
rejection is well established (3). However, it was recognised that a
proportion of transplant recipients experienced early graft loss
despite a negative CDC crossmatch test. The FCXM is able to

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_22, © Springer Science+Business Media New York 2012

379
380 J. Downing

detect HLA antibodies of low titre or of a non-complement binding

nature that may be responsible for this early graft loss. Numerous
studies have since described a link between positive T and B lym-
phocyte FCXM and graft rejection and, or loss in both primary and
re-graft recipients and in highly sensitised patients (4–12). Despite
these studies, controversy has persisted as to the clinical relevance
of a pre-transplant T or B cell positive FCXM test.
During the interpretation of the crossmatch result, it is essen-
tial to determine the nature of the antibody causing the reactivity.
The most clinically relevant antibodies are donor reactive, HLA-
specific, IgG antibodies (10, 12). Irrelevant positive reactions can
be caused by antibodies reacting with self epitopes shared by the
donor cells; Fc receptors that bind antibodies in a non-specific
manner; or antibodies against non-HLA antigens. Thus, extensive
characterization of patient HLA antibody must always occur along-
side the crossmatch test.
The era of fluorescent bead solid phase HLA antibody detec-
tion has led to a rethink of FCXM relevance. It is now reasonable
to consider the strength of positivity of FCXM in relation to quan-
titative, precise detection of donor-specific antibody (DSA) and
assess a degree of risk of rejection for each individual patient and
donor pair (12–14). Thus, a positive FCXM result even in the pres-
ence of DSA may not be considered as a contraindication to trans-
plant. Rather it allows the clinician to stratify risk for each patient.

2. Materials

1. 2 × 10 mL anti-coagulated peripheral blood sample on donor

and patient (see Note 1).
2. 1 × 10 mL clotted peripheral blood sample on patient.
3. Phosphate buffered saline (Invitrogen, Carlsbad, CA, USA)
(see Note 2).
4. Ficoll-Paque™ Plus (GE Healthcare Biosciences, Uppsala,
Sweden).
5. 50-mL plastic centrifuge tubes (Thermo-Fisher, Waltham,
MA, USA).
6. 15-mL plastic centrifuge tubes (Greiner-Bio-one, Monroe,
NC, USA).
7. RPMI 1640 (Invitrogen) with 20% v/v foetal bovine serum
(Invitrogen).
8. Pronase/Protease XIV (Sigma, St. Louis, MO, USA).
9. 3-mL flow cytometry tubes (Becton Dickinson Labware,
Franklin Lakes, NJ, USA).
 22 The Lymphocyte Crossmatch by Flow Cytometry for Kidney Transplantation 381

10. Negative control serum—male blood group AB blood donor

serum tested for the absence of HLA antibodies (see Note 3).
11. Positive control serum—human serum known to contain broad
specificity IgG HLA antibodies (see Note 4).
12. PBS-Azide: 1 L PBS containing 10% sodium azide (Scharlau
Chemicals, Sentmenant, Spain).
13. Anti-human IgG conjugated to fluorescein isothiocyanate
(FITC) (Jackson Immunoresearch, West Grove, PA, USA)
(see Note 5).
14. Anti-CD3 conjugated to peridinin chlorophyll protein
(PERCP) (Becton Dickinson, San Jose, CA, USA).
15. Anti-CD19 conjugated to phycoerythrin (PE) (Becton
Dickinson).
16. Three colour flow cytometer (Becton Dickinson FACSCalibur).

3. Methods

3.1. Isolation Peripheral blood lymphocytes (PBL) are the main target of the
of Peripheral Blood crossmatch test. Anti-coagulated blood samples should be stored
Mononuclear Cells at room temperature and should be no older than 48 h before
isolating the lymphocytes, where possible.
1. Transfer 10 mL of anti-coagulated peripheral blood into two
labelled 50-mL centrifuge tubes.
2. Dilute with PBS to a total volume of 30 mL.
3. Underlay diluted blood with Ficoll-Paque™ Plus taking care
not mix the blood-Ficoll interface.
4. Centrifuge for 10 min at 1,600 × g.
5. Carefully aspirate the mononuclear cell layer at the Ficoll—
plasma interface from each 50-mL centrifuge tube into clean
labelled 15-mL centrifuge tubes and top up to 15 mL with
PBS.
6. Centrifuge for 3 min at 650 × g.
7. Decant the supernatant and re-suspend the cell button in
15 mL of PBS.
8. Centrifuge for 3 min at 650 × g.
9. Decant the supernatant.
10. Pool and re-suspend the cell buttons from the two 15-mL cen-
trifuge tubes into 1 mL of RPMI cell medium (see Note 6).
11. Adjust the mononuclear cell count to 8 × 106 cells/mL (see
Note 7).
382 J. Downing

3.2. Cleavage of Fc T and B lymphocyte cell surface Fc receptors can bind IgG molecules
Receptors Using non-specifically and lead to a high anti-IgG FITC measurement on
Pronase the cell. This can lead to difficulty in discriminating between nega-
tive and positive crossmatch results. Cleavage of Fc receptors using
Pronase at can lead to a reduction in the incidence of false negative
crossmatch due to the presence of excess irrelevant IgG bound to
Fc receptors (15) (see Note 8).
1. Add 0.5 mL of 1 mg/mL Pronase to 1 mL of donor or recipient
cells.
2. Incubate for 30 min at 37°C.
3. After incubation top up to 15 mL with RPMI.
4. Centrifuge for 3 min at 650 × g.
5. Decant the supernatant and re-suspend the cell button in
0.5 mL RPMI.
6. Re-adjust the mononuclear cell count to 8 × 106 cells/mL.

3.3. Crossmatch PBLs should be prepared from the donor sample for the allo-cross-
Procedure match and from the patient sample where an auto-crossmatch test
is required (see Note 9).
1. Label 7× 3-mL flow cytometry tubes in the following sequence:
3 negative control serum vs. donor cells; 3 patient serum vs.
donor cells; 1 positive control serum vs. donor cells. Additionally,
if required, label a series of flow tubes for an auto control cross-
match to include a negative control serum vs. patient cells and
at least two replicates of patient serum vs. patient cells (see
Note 10).
2. In each tube, add 50 μL of appropriate serum to 50 μL of
appropriate cells, mix by gentle vortexing and incubate at room
temperature for 30 min.
3. After incubation, wash each tube by adding approximately
2 mL of cold PBS-azide (see Note 11) and centrifuging for 2 min
at 1,800 × g. Decant the supernatant and gently re-suspend the
cell button. Alternatively an automated cell washer can be
employed.
4. Repeat the wash process twice. After the third wash remove as
much supernatant as possible and gently re-suspend the cell
button (see Note 12).
5. Add 50 μL of diluted anti-human-IgG FITC to each tube, mix
by gentle vortexing, and incubate for 15 min in the dark at 5°C
(see Note 13).
6. Wash once as described in step 3 above.
7. To each tube add 2 μL each of anti-CD3 and anti-CD19
reagents (see Note 14). Mix by gentle vortexing and incubate
for 15 min in the dark at 5°C.
 22 The Lymphocyte Crossmatch by Flow Cytometry for Kidney Transplantation 383

8. Wash once as described in step 3 above. Re-suspend the cells in

200 μL of PBS-azide. The tubes are ready for flow analysis (see
Note 15).

3.4. Flow Cytometric The flow cytometer instrument must be set up correctly before
Analysis use. The manufacturer’s daily start-up and shut-down procedures
must be followed and a daily calibration using commercial stan-
dard reagents performed. In addition, optimisation (see Note 16)
and colour compensation must be performed (see Note 17).
1. Collect data on 1,200 B lymphocyte events as follows, see
Fig. 1 (see Note 18).
2. Use the cytometer acquisition and analysis software to create a
1,024 channel 4-decade log plot of forward scatter (FSC) vs.
side scatter (SSC) (Fig. 1a).
3. Draw a polygon gate around the lymphocyte population and
use this gate to create a dot plot to show FL2 (PE) vs. FL3
(PerCP) to discriminate the B and T lymphocytes, respectively
(Fig. 1b) (see Note 19).
4. Draw a rectangular gate around each of the lymphocyte popu-
lations, and use these to plot an FL1 (FITC) vs. cell count
histogram for each population (Fig. 1c, d). Draw a histogram

Fig. 1. Flow cytometer analysis histograms. Flow analysis histograms created using Becton Dickinson Cell Quest Pro analysis
software. Lymphocytes are displayed by plotting forward scatter (FSC) against side scatter light on a 1,024 channel log-
scale plot (a). A gate is drawn around the lymphocytes and gated data used to populate a CD19-PE against CD3-PERCP
plot to distinguish the B and T population, respectively (b). Rectangular gates are drawn around each lymphocyte popula-
tion and gated data used to plot anti-IgG FITC on a log scale against count on a linear scale for T lymphocytes (c) and B
lymphocytes (d).
384 J. Downing

marker around the main peak of the B lymphocyte FITC

histogram (see Note 20).
5. Obtain the median FITC channel value from each histogram,
in the case of the B lymphocytes this will be the M1 median
value (see Note 21).

3.5. Interpretation To determine whether a crossmatch result is negative or positive,

the median FITC channel in the patient serum—donor lympho-
cyte reactions is compared to that of the negative control serum—
donor lymphocyte reactions. The following method is recommended
as it uses median FITC values simply obtained from the analysis
software of the cytometer and applies a statistical method to deter-
mine whether the result is in the normal range expected if no
donor-reactive antibody is present. It must be made clear, however,
that there are a number of alternative methods (see Note 22).
1. For each lymphocyte population, subtract the mean of the
median FITC channel values of the negative serum vs. donor
lymphocyte replicates from the mean of the median FITC
channel values of the patient serum vs. donor lymphocyte rep-
licates to give the median channel shift (MCS).
2. An absolute MCS value can be established as a positive thresh-
old by testing the negative control serum used in the assay
against a panel of lymphocytes from 30 healthy individuals
and deriving the standard deviation (SD) of the median FITC
value for each of the T and B lymphocytes. A threshold of the
mean median FITC value plus three SD of the MCS can then
calculated. This will give the threshold above which a patient
result can be described as having a positive crossmatch (see
Notes 23 and 24).

4. Notes

1. Common anti-coagulants, such as ethylenediaminetetracetic

acid (EDTA), acid citrate dextrose or heparin, are all suitable
for the collection of blood used to obtain lymphocytes for flow
cytometry analysis.
2. Phosphate buffered saline can be purchased as a solution or as
tablets to be dissolved in a specified volume of water.
3. The negative control serum is a human serum that has been
shown to contain no reactivity to T or B lymphocytes. This
serum can be sourced locally from male blood group AB blood
donors or can be obtained commercially from suppliers, such
as Invitrogen or One Lambda (Canoga Park, CA, USA). The
negative control serum should be evaluated for lack of reactivity
22 The Lymphocyte Crossmatch by Flow Cytometry for Kidney Transplantation 385

before used routinely. This should consist of testing to confirm

lack of HLA antibodies using a solid phase method, and by
testing by flow cytometry using the above FCXM method with
lymphocytes from a number of individuals to show lack of
reactivity.
4. The positive control serum is required to establish that correct
binding of the anti-IgG FITC secondary antibody has occurred
during the assay. A pool of locally available highly sensitised
transplant patients is a common source for this positive con-
trol. Pools of up to ten individuals known to possess strong
IgG HLA antibodies can be made and aliquots stored at −30°C
for continued used. Each pool should be evaluated for reactiv-
ity before use, by testing by flow cytometry using the above
FCXM assay. Alternatively, a commercial source of a monoclo-
nal antibody, such as w6/32 (Abcam, Cambridge, MA, USA)
could be used.
5. The fluorophore-labelled antibodies, anti-human-IgG-FITC,
anti-CD19-PE, and anti-CD3-PERCP fluoresce at particular
light wavelengths so that they can be detected by the flow
cytometer fluorescence (FL) detectors, FL1, FL2, and FL3,
respectively. These detectors collect signals at 530, 585, and
>670 nm, respectively, in a BD FACSCalibur instrument. The
fluorophores emit at 519, 578, and 675 nm for FITC, PE and
PERCP, respectively. Thus, if a particular labelled antibody is
not available, any replacement needs to emit in the correct
range for it to be collected by a specific FL detector. The anti-
CD19 and anti-CD3 antibodies are required to discriminate
between B and T lymphocytes and this can give valuable infor-
mation as to the nature of the specificity and strength of patient
antibody binding. It is possible to perform the crossmatch
without these, but no B and T lymphocyte discrimination will
be possible.
6. If the cell buttons are quite large, increase the re-suspension
volume. This will ease the process for adjusting the cell count
to the optimum.
7. It is important to adjust the cells to the correct optimum con-
centration. Too few cells will make interpretation of the results
difficult. With too many cells, there is a risk of diluting any
allo-antibody to levels below detection. In addition, an exceed-
ingly high cell event rate can lead to inaccurate data capture.
8. The pronase cleavage of FC receptors could be considered as
an optional step and is not performed by all laboratories.
Pronase can potentially reduce the incidence of false negative
results by reducing background fluorescence caused by irrele-
vant binding of IgG to Fc receptors (15). The requirement for
pronase treatment is difficult to predict as the incidence of
386 J. Downing

background can vary depending on patient serum and donor

lymphocytes and it is recommended that this step be carried
out with all crossmatch tests. In addition, for patients under-
going treatment with Rituximab, the anti-CD20 monoclonal
antibody, binding of Rituximab will lead to high levels of bind-
ing of FITC. Using pronase will cleave the anti-CD20 and
hence remove Rituximab binding.
9. An auto crossmatch result will aid interpretation where the allo
crossmatch result is positive for either T or B lymphocyte, but
no donor-specific HLA antibody can be detected in the patient.
If the auto crossmatch is also positive and of a similar magni-
tude, this can indicate non-HLA auto-reactive antibodies that
are also able to bind to donor lymphocytes. It must be stressed
though, that this interpretation can only be made in the pres-
ence of precise determination of patient HLA antibody using
sensitive solid phase methodology.
10. Each laboratory will have its own requirements for the number
of replicate reactions performed during the crossmatch.
Flexibility may be required in the event of low serum or cell
preparation volumes.
11. Sodium azide is typically included in the wash buffer to pre-
vent the modulation and internalisation of surface antigens
which can produce a loss of fluorescence intensity.
12. Careful removal of most of the wash supernatant is important
at this stage to ensure that the volume of FITC that the cells are
incubated in the following stage is optimum and consistent.
13. The optimum dilution of anti-IgG FITC should be determined
in each laboratory by testing a range of dilutions against known
positive and negative serum/cell combinations. An optimum
dilution will give the best differentiation between the negative
and positive reactions.
14. The anti-CD3 and anti-CD19 can be combined in equal vol-
umes before addition to the cells. This reduces the number of
low volume reagent dispensing steps required. Thus, 4 μL of the
combined reagent can be added to each crossmatch tube.
15. At this stage, the cells should be kept in the dark until ready for
analysis. This should be performed within an hour of the end
of the assay. Alternatively, a cell fixing reagent such as para-
formaldehyde can be used to preserve the cells for a number
of days.
16. The flow cytometer must be set up before analysis of the cross-
match assay in order to optimally record the amount of patient
IgG antibody binding to cells. Using lymphocytes prepared in
the same manner as the crossmatch assay, adjust the FSC amp
gain and SSC voltage to display the lymphocytes appropriately
on the scale of a log-scale FSC vs. SSC plot. When prepared in
22 The Lymphocyte Crossmatch by Flow Cytometry for Kidney Transplantation 387

the manner described above, lymphocytes should represent the

densest population recorded and thus be easily identifiable.
Adjust the FSC threshold in order to exclude the collection of
data on unwanted small particles and debris.
17. At some stage prior to analysis, colour compensation should be
performed in order to optimise the instrument settings for use
with the lymphocyte crossmatch assay. Compensation is
required where the emission spectra of fluorophores used in an
assay overlap, as is the case for example with FITC and PE and
with PE and PERCP. Compensation ensures that fluorescence
emitted by FITC is not detected by the FL2 photomultiplier as
a PE signal. Compensation is performed by individually pre-
paring lymphocytes with the fluorophores to be used in the
assay and then using the flow cytometer acquisition software
compensation controls to ensure there is minimal overlap
between the FL detectors. It is recommended that compensa-
tion be performed using lymphocytes in addition to any manu-
facturer’s calibration reagents. This is because lymphocytes
have different optical properties to these reagents. Once com-
pensation is performed, the saved instrument settings should
be used when subsequently analysing lymphocytes from the
crossmatch test. Compensation does not need to be performed
before each analysis, but should be carried approximately 6
monthly to monitor wear of the cytometer optics or in the
event that changes are made to the optics.
18. As B lymphocytes are usually less numerous than T lympho-
cytes, setting a minimum number of B lymphocyte events is
important to ensure enough data is collected to be able to
make an accurate interpretation. Generally, enough T cell
events will be collected under these conditions. In the case of
samples with low B lymphocyte counts, where 1,200 events
cannot be obtained, a time limit should be placed on each
acquisition such that all 200 μL of the sample is acquired at a
high flow rate. It is useful to have a minimum T and B lympho-
cyte count, for example, 2,000 T and 350 B, so that accurate
interpretation can be made. If the counts obtained fall below
this the crossmatch should be repeated. With very low cell
counts, natural variation in FITC level can introduce inaccu-
racy that may lead to false interpretation of the test result.
19. The T and B lymphocyte population should show as clearly
defined clusters of events segregated to high on the PERCP
axis for T lymphocytes and high on the PE axis for B lympho-
cytes. Quadrant markers are placed on the plot to indicate the
areas of negative and positive cell populations (see Fig. 1b).
If either of these populations or the negative population (which
should appear low on both axis) appear outside of the expected
areas, this may indicate poor compensation set up of the cytometer
388 J. Downing

or contamination from other sources of fluorescence that emit

at the same wavelength, e.g. ethidium bromide. Failure to
recognise this problem could lead to a false positive result.
Note also that the count scale for B lymphocytes is lower than
that of T lymphocytes in Fig. 1c, d, so that the peak is appro-
priately sized.
20. B lymphocyte preparations may display low numbers of events
with high levels of bound FITC that manifest as very small
peaks towards the right of the x axis of an FITC vs. cell count
histogram derived from the B lymphocyte gate. This is due to
non-specific IgG binding that can be disregarded from the
analysis by placing a histogram marker (M1 in Fig. 1d) sym-
metrically around the main peak and deriving the median FITC
value from this marker. FITC values outside of the marker will
not be included in the median value, and this has the effect of
lowering the overall median value but potentially lowering the
signal-to-noise ratio.
21. The FITC histograms should show roughly symmetrical peaks
with a clearly defined population based around a single maxi-
mum FITC level. Poorly defined peaks will lead to less accurate
definition of the median FITC level and may be caused by low
cell counts, inadequate mixing of cells, serum or fluorescent
labels during the assay, or by the labelling of abnormal donor
or patient lymphocyte subpopulations.
22. Each laboratory must establish its own criteria for a threshold
for a positive crossmatch and there are a variety of methods
available to do this. The method described above utilises an
absolute cut-off of MCS calculated by testing the negative
control serum against a panel of healthy cells to determine the
normal range of background FITC binding. An alternative is
to calculate the SD of the FITC channel values of the negative
control replicates derived from the crossmatch. If the patient
FITC channel value exceeds the negative value plus three SD
this can be described as a positive crossmatch. Another way of
comparing the patient and negative channel values is to calcu-
late the relative fluorescence by dividing the mean of patient
median FITC channel value by that of the negative control,
and a cut-off value of 1.5 or 2 used to indicate a positive result.
A further method for comparing the patient and negative val-
ues is to employ calibration beads to convert the arbitrary
FITC channel values into molecules of equivalent soluble
fluorochrome (MESF). In this process, fluorescence from com-
mercially available combinations of precisely quantitated
solutions and microbead suspensions is measured in the cytom-
eter. The results are used to create a standard curve which can
be used to convert any cytometer reading into an MESF value.
 22 The Lymphocyte Crossmatch by Flow Cytometry for Kidney Transplantation 389

Thus, median FITC values can be converted into MESF values

and the shift of patient MESF from negative MESF applied to
a threshold level in a similar way as MCS described above.
MESF methods are particularly useful when comparing data
from crossmatch tests performed at different times or in different
centres.
23. The use of a fixed threshold, however it is determined, is prob-
lematic when trying to interpret a result that is close to this
threshold. Whether or not a particular result relates to the
presence of donor-specific HLA antibody is not always well
reflected by a fixed threshold. It might be useful to have a second
threshold to indicate weak positive reactions, for example, the
MCS of the negative plus two SD. A threshold of the negative
median plus two SD relates to the fact that 95% of “normal”
negative reactions, i.e. those reactions that do not show
increased specific FITC binding will fall within this threshold.
Only 5% of reactions that exceed the threshold will be due to
“normal” negative reactions, all others being defined as true
positive due to increased FITC binding. Using three SD
assumes that only 0.3% of negative reactions will exceed this
level and that 99.7% of reactions above this level are of a true
positive nature and hence is a more stringent definition of a
positive result.
24. By testing the negative control serum against a range of differ-
ent lymphocyte samples it is possible to establish ranges of
reactivity that can be used as acceptance criteria for the assay.
For example, it is desirable to have a minimum and maximum
acceptable median FITC value for the negative serum against
any particular donor lymphocyte. If the negative serum median
FITC is too high, this can lead to a false negative interpretation
as a large amount of “background” is subtracted from the
patient FITC value during calculation of the MCS. When
the negative serum is tested against 30 healthy individuals, the
lowest and highest FITC value obtained for each lymphocyte
population can be used as the lower and upper range. If, dur-
ing a crossmatch test, these ranges are exceeded, then the
crossmatch may be repeated or the interpretation questioned.

References

1. Garovoy MR et al (1983) Flow cytometry anal- 3. Patel R, Terasaki PI (1969) Significance of the
ysis: a high technology crossmatch technique positive crossmatch test in kidney transplanta-
facilitating transplantation. Transplant Proc tion. N Engl J Med 280:735–739
15:1939–1944 4. Chapman JR et al (1985) Analysis of flow
2. Bray RA, Lebeck LK, Gebel HM (1989) The cytometry and cytotoxic crossmatches in renal
flow cytometric crossmatch: dual-colour analy- transplantation. Transplant Proc 17:2480–2481
sis of T and B cell reactivities. Transplantation 5. Ogura K et al (1993) The significance of a posi-
48:834–840 tive flow cytometry crossmatch test in primary
390 J. Downing

kidney transplantation. Transplantation 56: for renal transplantation. Exp Mol Pathol 85:
294–298 59–63
6. Sridhar NR et al (1992) Evaluation of 12. Gebel HM, Bray RA, Nickerson P (2003) Pre-
flowcytometric crossmatching in renal allograft transplant assessment of donor-reactive. HLA-
recipients. Nephron 62:262–266 specific antibodies in renal transplantation:
7. Lazda VA (1994) Identification of patients at contraindication vs. risk. Am J Transplant
risk for inferior renal allograft outcome by a 3:1488–1500
strongly positive B cell flow cytometry cross- 13. Taylor CJ et al (2009) Back to the future: appli-
match. Transplantation 57:964–969 cation of contemporary technology to long-
8. Talbot D et al (1995) Flow cytometric cross- standing questions about the clinical relevance
matching in renal transplantation—the long of human leukocyte antigen-specific alloanti-
term outcome. Transpl Immunol 3:352–355 bodies in renal transplantation. Hum Immunol
9. Graff RJ et al (2009) The role of positive flow 70:563–568
cytometry crossmatch in late renal allograft 14. Riethmuller S et al (2010) Donor-specific anti-
loss. Hum Immunol 70:502–505 body levels and three generations of cross-
10. Lefaucher C et al (2008) Clinical relevance of matches to predict antibody mediated rejection
preformed HLA donor-specific antibodies in in kidney transplantation. Transplantation
kidney transplantation. Am J Transplant 90:160–167
8:324–331 15. Vaida S et al (2001) Improved flow cytometric
11. Delgado JC, Eckels DD (2008) Positive B-cell detection of HLA alloantibodies using pronase.
only flow cytometric crossmatch: implications Transplantation 71:422–428
 Chapter 23

Overview of the Killer Cell Immunoglobulin-Like

Receptor System
Raja Rajalingam

Abstract
Natural killer (NK) cells are more than simple killers and have been implicated in control and clearance of
malignant and virally infected cells, regulation of adaptive immune responses, rejection of bone marrow
transplants, and autoimmunity and the maintenance of pregnancy. Human NK cells largely use a family of
germ-line encoded killer cell immunoglobulin-like receptors (KIR) to respond to the perturbations from
self-HLA class I molecules present on infected, malignant, or HLA-disparate fetal or allogenic transplants.
Genes encoding KIR receptors and HLA class I ligands are located on different chromosomes, and both
feature extraordinary diversity in the number and type of genes. The independent segregation of KIR and
HLA gene families produce diversity in the number and type of KIR-HLA gene combinations inherited in
individuals, which may determine their immunity and susceptibility to diseases. This chapter provides an
overview of NK cells and their unprecedented phenotypic and functional diversity within and between
individuals.

Key words: NK cells, Innate immunity, HLA, KIR, Polymorphism, Immune genes

1. Natural Killer
Cells
Natural killer (NK) cells are the third population of lymphocytes
defined by CD3−, CD56+ cell surface phenotype, and they represent
5–25% of the mononuclear cell fraction of normal human periph-
eral blood (1, 2). NK cells share several features with CD8+ cytolytic
T lymphocytes (CTL) in their development, morphology, cell-
surface phenotypes, killing mechanism, and cytokine production
(3). Compared to T and B lymphocytes, NK cells are larger and
contain preformed cytolytic granules (granzymes and perforin) as
well as an exponential amount of constitutively expressed tran-
scripts for interferon-g (IFN-g) and certain cytokines that activate
other immune cell types. The fundamental function of NK cells is

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_23, © Springer Science+Business Media New York 2012

391
392 R. Rajalingam

to provide the first line of defense by inherently responding to

infection and tumor transformation without any prior sensitiza-
tion, and therefore NK cells are considered to be the integral com-
ponent of innate immunity (4, 5). Recent reports suggest that
similar to B and T cells, NK cells exhibit many features normally
associated with adaptive immunity. These include the expansion of
pathogen-specific cells, the generation of long-lasting “memory”
cells that persist after cognate antigen encounter, and the ability to
mount an enhanced secondary recall response to rechallenge (6).
On the basis of adhesion molecule CD56 expression level, two
subsets of NK cells have been distinguished in human peripheral
blood (7). Over 90% belong to the CD56dim subset that expresses
high levels of CD16 and KIR receptors, and represent the most
differentiated and mature NK cell type, and have superior cytotox-
icity capacity (1, 8). The remaining 10% are the CD56bright pheno-
type (CD16−, KIR+/−) representing the precursors of CD56dim
subset, and have greater ability to produce proinflammatory cytok-
ines (9). In contrast to blood, the CD56bright cells are the dominant
NK cell subset in human lymph nodes and they interact with den-
dritic cells (10, 11). During pregnancy, the placenta houses unique
subsets of NK cells that contribute to successful implant and func-
tion of the placenta by secreting angiogenetic factors, such as vas-
cular endothelial growth factor (VEGF) (12). Gut-associated
lymphoid tissues harbor a unique NK cell subset that specializes in
production of interleukin (IL)-22, a proinflammatory cytokine
that mediates host defense against extracellular pathogens (13).
NK cells are therefore an unexpectedly heterogeneous population
involving both innate and adoptive immune responses and successful
pregnancies, which altogether promote human health and survival.

2. Natural Killer
Cell Receptors
NK cells have a highly specific and complex target cell recognition
receptor system arbitrated by the integration of signals triggered by a
multitude of inhibitory and activating receptors, which trigger cyto-
toxicity and the secretion of chemokines and cytokines (14, 15).
Unlike T and B lymphocytes, NK cells do not express receptors that
require somatic gene rearrangements to generate receptor diversity
and specificity. Instead, NK cells express a wide array of conventional
germline-encoded receptor families with inhibitory or activating
functions, including (i) killer cell immunoglobulin-like receptors
(KIR), (ii) killer cell lectin-like receptors (KLR) such as CD94/
NKG2, (iii) leukocyte immunoglobulin-like receptors (LILR), and
(iv) natural cytotoxicity receptors (NCR), such as NKp46, NKp44,
and NKp30, and 2B4 (2, 16–18). Most of these receptors are
expressed in stochastic, variegated combinations of activating and
inhibitory receptors, resulting in many subsets of functionally distinct
23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 393

NK cells (19, 20). In addition to the receptor–ligand-mediated

regulation, the lytic potential of NK cells and their ability to produce
IFN-g is enhanced by type I IFNs (IFN-a and IFN-b), IL-2, IL-18,
and IL-15 that are secreted by dendritic cells, macrophages as well as
pathogen-infected tissue (21).
Because NK cells circulate in a state that can spontaneously
deliver effector function, it is critical that they do not attack sur-
rounding healthy cells. To prevent such detrimental autoreactivity,
NK cells express an array of inhibitory receptors recognizing self-
HLA class I molecules (Fig. 1). Abundant expression of four

Fig. 1. Natural killer (NK) cell response against healthy and unhealthy cells. NK cells
express both inhibitory and activating receptors. Inhibitory receptors recognize self HLA
class I molecules and trigger signals that stop the natural lytic function of NK cells. By
expressing normal levels of multiple HLA class I molecules, the healthy cells are resistant
to NK cell attack (a). Downregulation of HLA class I expression due to tumor transforma-
tion or viral infection relieves the inhibitory influence on NK cells (b), permitting NK cells
to lyse these unhealthy target cells, a phenomenon first described as the “missing-self”
hypothesis. Upon transformation or infection, the unhealthy cells express ligands for
activating receptors, which could be either “induced-self” (MICA, MICB, ULBP), “altered-
self” (HLA-class I loaded with foreign peptide), and/or “non-self” (pathogen-encoded mol-
ecules). Upon recognizing these ligands, the activating receptors trigger signals that
augment NK cell lysis of unhealthy targets.
394 R. Rajalingam

distinct HLA class I molecules (HLA-A, -B, -C, and -E) on normal
healthy cells provide ligands for a variety of inhibitory receptors of
NK cells, and consequently are resistant to NK cell attack.
Downregulation of HLA class I expression due to certain viral
infections, neoplastic transformations or other forms of stress,
relieves the inhibitory influence on NK cells, permitting NK cells
to eliminate these unhealthy target cells, a phenomenon originally
described as the “missing-self” hypothesis (22). In addition to the
“missing-self” mechanism, the expression of ligands for activating
receptors on target cell surface might also contribute to NK cell
attack (Fig. 1). The ligands identified for activating receptors are
either “induced-self” that are structurally related to HLA class I
molecules (such as, major histocompatibility complex [MHC] class
I-like chains A and B [MICA and MICB] and unique long binding
proteins [ULBP]), “altered-self” (HLA class I molecule loaded
with foreign peptide) or pathogen encoded “non-self” (molecules
associated with infection and tumor transformation).

2.1. Killer Cell Compared to other NK cell receptors, the KIR receptors are con-
Immunoglobulin-Like sidered to be the key receptors that control the development and
Receptors function of human NK cells (16, 23–26). The following rationale
substantiates the importance of KIR receptors: (i) KIR receptor
system is totally absent in rodents, the basic model for clinical
research, and therefore the mouse and human NK cell biology do
not converge in many crucial aspects (27, 28); (ii) species compari-
son suggests that the KIR system originated in mammals, and
evolved rapidly to keep up with species-specific evolution and adap-
tation, and thus only a small minority of KIR genes is shared by
humans and chimpanzees (29); (iii) KIR receptors are encoded by
a polygenic and polymorphic gene family that produces remarkable
diversity within individuals and in populations in the number and
type of genes (30, 31); (iv) KIR receptors recognize an array of
polymorphic HLA class I molecules that evolved rapidly to mount
effective adaptive immune response against rapidly evolving viruses
and pathogens; (v) genes encoding KIR receptors and HLA class I
ligands are located on different chromosomes, and their indepen-
dent segregation results in variable KIR-HLA combinations in
individuals, which may contribute to the individual’s health and
disease. Consistent with this theory, combinations of certain KIR-
HLA genes have been associated with diseases as diverse as autoim-
munity, infection, cancer, and reproductive failure (32, 33).
Fourteen KIR receptors triggering either inhibition (3DL1-3,
2DL1-3, 2DL5) or activation (3DS1, 2DS1-5), or both (2DL4)
have been identified in humans (Fig. 2) (34–36). KIR receptors
possess two or three extracellular Ig-like domains involved in ligand
binding and either a long or short cytoplasmic tail involved in sig-
naling. KIRs are named on the basis of their number of Ig domains
and type of cytoplasmic tail (37). The first digit following the
23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 395

Fig. 2. Killer cell immunoglobulin-like receptors (KIR). Fourteen distinct KIR receptors have
been characterized in humans that comprise either two (2D) or three (3D) extracellular
Ig-like domains and either a long (L) or short (S) cytoplasmic tail. Six KIR receptors are
activating types and the remaining KIR are inhibitory types. The ITIM motifs in the cyto-
plasmic tails of inhibitory KIRs are shown as boxes, and positively charged residues in the
transmembrane regions of activating KIRs are shown as circles. The inhibitory KIR recep-
tors bind to distinct HLA class I allotypes and the ligands for most activating KIR receptors
are unknown.

KIR acronym corresponds to the number of Ig-like domains in the

molecule and the “D” denotes “domain.” The “D” is followed by
either an “L” indicating a “Long” cytoplasmic tail, an “S” indicat-
ing a “Short” cytoplasmic tail or a “P” for pseudogenes (2DP1 and
3DP1). The final digit indicates the number of the gene encoding
a protein with this structure. Thus, KIR2DL1, KIR2DL2, and
KIR2DL3 all encode receptors having two extracellular Ig-like
domains and long cytoplasmic tails, while KIR3DS1 encodes a
receptor having three extracellular Ig-like domains and a short cyto-
plasmic tail. The long tails contain immune-receptor tyrosine-based
inhibitory motifs (ITIMs) that trigger inhibitory signals. The short
cytoplasmic tails lack ITIM but possess a positively charged residue
in the transmembrane domain that interacts with the signaling
adaptor DAP12 (38). DAP12 contains immune-receptor tyrosine-
based activation motifs (ITAM), which trigger activating signals
upon the short-tailed KIR receptor recognizing a ligand.
396 R. Rajalingam

2.2. Ligands for KIR KIR receptors recognize specific motifs of HLA class I molecules,
Receptors which are the products of highly polymorphic genes of the MHC
located on chromosome 6 (39–41) (Figs. 2 and 3). KIR binds in a
nearly orthogonal orientation across the a1 and a2 helices of HLA
class I molecules, and the KIR footprint on HLA overlaps with but
is distinct from that of the T cell receptor (42). The six loops of D1
and D2 interdomain hinge region of KIR2DL as well as KIR3DL
form the HLA class I binding site, while the membrane-distal D0
domain of KIR3D enhances HLA class I ligand binding (43). Since
the KIR binds to the polymorphic peptide binding groove of HLA
class I molecules, it is constantly subjected to changes driven by
rapidly evolving viruses. Therefore, the KIR binding motifs are
often variable between individuals and their distribution differs
substantially between populations (39, 44, 45).
HLA-C is the dominant HLA class I locus that provides ligands
for many KIR receptors (Fig. 2). Amino acid residues 76 and 80 on
HLA-C allotypes determine KIR binding epitopes (46–48) (Fig. 3).
All HLA-C allotypes carry valine (V) at position 76, while position

Fig. 3. KIR receptor binding region of HLA class I molecule. Locations of amino acids (position 73, 76–83, and 90) in the a1
domain of HLA class I molecules (shown in dark shade) involved in KIR receptor binding.
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 397

80 displays a dimorphism of either asparagine (N) or lysine (K).

Nearly half of the HLA-C allotypes (Cw2, Cw4, Cw5, Cw6, and
Cw15) carry a K80 (conventionally termed C2 epitope) that binds
inhibitory receptor KIR2DL1 (49–51). The remaining HLA-C
allotypes (Cw1, Cw3, Cw7, and Cw8) carry N80 (termed C1
epitope) and bind inhibitory receptors KIR2DL2 and 2DL3. Two
exceptionally diverse HLA-B allotypes, B46 and B73 that both
have V76 and N80 motifs are good ligands for KIR2DL2/3 (52).
In addition to C1 binding, the KIR2DL2/3 can also interact with
several C2 allotypes, notably C*0501 and C*0202 (52). The
inhibitory signals triggered by the 2DL2/3-C1 interaction is rela-
tively weaker as compared to those triggered by the 2DL1-C2
interaction (52, 53). The activating KIR receptor KIR2DS1 has
similar Ig-like domains to inhibitory KIR2DL1 and also binds
HLA-C2, but with reduced avidity (54). KIR2DS2 is an activating
receptor that has similar Ig-like domains to inhibitory KIR2DL2
but has no detectable avidity for HLA-C1 (53). KIR2DS4 is the
oldest and most prevalent activating receptor and has unique ligand
specificity for subsets of HLA-C allotypes carrying C1 or C2
epitopes and HLA-A11 (55).
Since the C1 and C2 are nonoverlapping subsets of HLA-C
allotypes, individuals can be either C1 homozygous, C2 homozy-
gous or C1/C2 heterozygous. Contrarily, only a subset of HLA-A
(HLA-A23, 24, 25, and 32) and HLA-B (40% of the known B
allotypes) molecules that carry a Bw4 epitope in a similar region
(residues 77–83) function as ligands for the KIR3DL1 receptor
(56–58) (Fig. 3). KIR3DL2 binds to only HLA-A3 and A11 allo-
types and the strength of this interaction is highly sensitive to the
bound peptide sequence (59–61). The KIR2DL4 receptor binds to
the extravillous trophoblast-specific HLA-G molecule and induces
rapid IFN-g production that promotes vascularization of the mater-
nal decidua, which provides the placenta with blood and the growing
fetus with gases and nutrients (62–65). In addition to its activation
function, the KIR2DL4 receptor carries a single ITIM motif in its
cytoplasmic tail and exhibits an inhibitory function (66, 67). The
ligand specificities for KIR2DS2, 2DS3, 2DS5, 3DS1, and 2DL5
remain elusive. Epidemiological studies implicate HLA-Bw4
specificity for KIR3DS1 (68).

3. Extraordinary
Diversity of KIR
Receptor System
The unique feature of the KIR system that distinguishes it from
other types of NK cell receptors is its substantial diversity, which is
contributed by various factors such as individual-specific KIR gene
content, nucleotide sequence polymorphism of each KIR gene, and
stochastic and variegated expression of KIR receptor repertoires on
individual NK cell clones.
398 R. Rajalingam

3.1. Gene Content KIR receptors are encoded by a family of tightly clustered genes on
Diversity of KIR Gene the leukocyte receptor complex (LRC) that spans a region of about
Complex 150 kb on chromosome 19q13.4 (34, 35). The number and type
of KIR genes vary substantially between haplotypes (Fig. 4). Over
30 distinct KIR haplotypes with distinct gene content have been
characterized to date by sequencing genomic clones and haplotype
segregation analysis in families (34, 69–74). The most commonly
occurring haplotype in most human populations is conventionally
called the “group-A haplotype,” which carries a fixed gene content
comprising KIR3DL3-2DL3-2DP1-2DL1-3DP1-2DL4-3DL1-
2DS4-3DL2 (Fig. 4; haplotype 1) (31, 34). Remaining KIR hap-
lotypes are collectively referred to as “group-B haplotypes,” which
have variable gene content comprising several genes that are not
part of the A haplotype (Fig. 4; haplotypes 2–22). Particularly,
KIR2DS1, 2DS2, 2DS3, 2DS5, 2DL2, 2DL5, and 3DS1 are asso-
ciated only with group-B haplotypes, and thus B haplotypes gener-
ally encode more activating KIR receptors than the A haplotype
that encodes a single activating receptor, KIR2DS4 (34, 69, 73, 74).
Despite such extraordinary gene-content diversity, four KIR genes,
2DL4, 3DL2, 3DL3, and 3DP1, are present on virtually all KIR

Fig. 4. KIR haplotypes differ by gene content. Map of KIR haplotypes as determined by family segregation analysis. Haplotype
1 represents group-A KIR haplotype and the remainder group-B haplotypes. The framework genes, present in all haplotypes
are shown in dark boxes; genes encoding activating KIR are in white boxes ; and those for inhibitory receptors are in gray
boxes. The KIR2DP1 and 3DP1 are pseudogenes that do not express a receptor. Maps are not drawn to scale.
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 399

haplotypes and therefore have been referred to as “framework”

genes (34). Inheritance of paternal and maternal haplotypes com-
prising different KIR gene content haplotypes (A + A, A + B, or
B + B) generates extraordinary diversity between humans. For
example, homozygotes for group-A haplotypes have only seven
functional KIR genes (Fig. 5; sibling-1), while heterozygotes for
group-A and group-B haplotypes may have all 14 functional KIR
genes (Fig. 5; sibling-2). Data from various populations revealed
over 240 genotypes that differ in their KIR gene content, of which
only a single genotype, suggestive of the homozygote for group-A
haplotype, occurs in all 51 studied populations (75, 76) (http://
www.allelefrequencies.net/). The gene content for the ten KIR
genotypes that occur most frequently in populations are listed
in Fig. 6.
A stretch of 14 kb enriched with L1 repeats upstream of
KIR2DL4 divides the KIR haplotype into two halves (34) (Figs. 4
and 7). The centromeric half is delimited by 3DL3 at the 5¢-end and
3DP1 at the 3¢-end, while the telomeric half is delimited by 2DL4
at the 5¢-end and 3DL2 at the 3¢-end. These four framework genes

Fig. 5. KIR gene content differs between individuals. The number and type of KIR genes can be substantially variable
between individuals, even within a family. (a) Is the gel picture indicating the extreme difference in the KIR gene content of
two siblings. Sibling-1 has fewer genes and sibling-2 has all known genes. Arrows indicate internal positive control bands
specific to an invariant gene. The pseudogenes KIR2DP1 and 3DP1 are not typed. (b) Shows the possible KIR haplotype
composition of each sibling.
400 R. Rajalingam

Fig. 6. Ten most commonly occurring KIR genotypes in populations. Genotyping studies revealed a significant ethnic difference
in the distribution of KIR genotypes. Listed are the ten most frequently occurring KIR genotypes in 51 distinct human popu-
lations. The presence of a gene is indicated by a shaded box while the absence of a gene is indicated by a white box.

Fig. 7. Cetromeric and telomeric halves of KIR haplotypes. A stretch of 14 kb enriched with L1 repeats upstream of KIR2DL4
divides the KIR haplotype into two halves. The centromeric half is delimited by 3DL3 and 3DP1 while the telomeric half is
delimited by 2DL4 and 3DL2. Multiple reciprocal meiotic recombination events between 3DP1 and 2DL4 shuffled the
centromeric (c) and telomeric (t) motifs, and thus diversify gene content KIR haplotypes across individuals and populations.
Most of the KIR gene content haplotypes published to date can be explained by the recombination of these 10 centromeric
and 10 telomeric gene content motifs. The framework genes, present in all haplotypes are shown in dark boxes; genes
encoding activating KIR are in white boxes; and those for inhibitory receptors are in gray boxes. The KIR2DP1 and 3DP1
are pseudogenes that do not express a receptor. Letter “A” in gene-content motif identification indicates parts of group-A
haplotypes while “B” indicates parts of group-B haplotypes.

are present on all KIR haplotypes and thus occur in 100% of all
populations (75). In contrast, the existence of the other 12 KIR
genes is considerably variable. The inhibitory receptors KIR2DL2
and 2DL3 segregate as alleles of a single locus at the centromeric
half. Similarly, the inhibitory KIR3DL1 and activating KIR3DS1
behave as alleles of the same locus at the telomeric half. Almost all
23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 401

haplotypes contain these two loci, such that virtually every one has
either 2DL2 or 2DL3, and 3DL1 or 3DS1 within their KIR genome.
KIR2DL1, 2DL2, 2DL3, and 2DS2 are specific to the centromeric
half while KIR3DL1, 3DS1, 2DS1, and 2DS4 are specific to the
telomeric half. Three KIR genes, 2DL5, 2DS3, and 2DS5, are
found in both centromeric and telomeric locations (70, 74). For
genes within each half, there is significant linkage disequilibrium,
but much less for genes in the two different halves (73, 77).
Multiple reciprocal recombination events at the center of the
KIR complex, between 3DP1 and 2DL4, presumably diversify
gene content for KIR haplotypes across individuals and popula-
tions (75, 78). Most of the KIR gene content haplotypes published
to date can be explained by the recombination of 10 centromeric
and 10 telomeric gene content motifs (Figs. 4 and 7). For example,
haplotype 1 listed in Fig. 4 is a recombinant of the centromeric half
cA01 and telomeric half tA01; similarly, haplotype 8 is a recombinant
of the centromeric half cB02 and telomeric half tB02. The recipro-
cal recombination also results haplotypes carrying both group-A
and group-B haplotype-specific motifs. For instance, haplotype 6
listed in Fig. 4 is a recombinant of the centromeric half cA01 and
telomeric half tB04, and haplotype 9 is a recombinant of the cen-
tromeric half cB01 and telomeric tA01. Recombination events
also are reported outside the region between 3DP1 and 2DL4.
These are generally nonallelic crossovers generating several unusual
haplotypes, including truncated haplotypes that are missing some
framework genes (70, 77) or elongated haplotypes that contain
duplicated genes (79–81). It is important to remark here that the
current genotyping methods used to detect the presence and
absence of individual KIR gene will not identify the copy number
of each KIR gene within a given individual. Therefore, more pre-
cise typing methods, such as a yet to be developed quantitative
real-time PCR assay are necessary to investigate the KIR gene copy
number and their impact on human disease.
All human populations have both group-A and B haplotypes,
but their distribution varies considerably across distinct popula-
tions. Individuals carrying both A and B haplotypes are more com-
mon in Caucasians and Africans, and therefore A and B haplotypes
are approximately equally distributed in these populations (31,
75). In contrast, the prevalence of inhibitory haplotypes (group-A)
dominates over the activating haplotypes (group-B) in one popula-
tion and vice versa in others. Individuals carrying homozygous
group-A KIR haplotypes (AA genotypes) are common in Northeast
Asians (Chinese, Japanese, and Koreans) (72, 82, 83). Conversely,
individuals carrying AB or BB genotypes are common in the natives
of America (84, 85), Australia (86), and India (87–89). The NK
cells from AA homozygous individuals can express a maximum of
four inhibitory KIRs (2DL1, 2DL3, 3DL1, and 3DL2) and one
activating KIR (2DS4). In contrast, individuals carrying AB or BB
402 R. Rajalingam

genotypes can express a maximum of six inhibitory KIRs (2DL1-3,

2DL5, 3DL1, and 3DL2) and 2–6 activating KIRs (3DS1, 2DS1-5).
NK cells of group-B haplotype carriers express more activating
KIR receptors and respond more vigorously to pathogens. These
data suggest that the aboriginal populations of India, Australia,
and America acquired activating KIR genes to survive the environ-
mental challenges during their extensive prehistoric migrations
from Africa (87).

3.2. Nucleotide In addition to gene content diversity, each KIR gene itself exhibits
Polymorphism considerable nucleotide sequence polymorphism (37, 73, 77,
of KIR Genes 90, 91). Allelic sequence variants of KIR genes are named in a
fashion analogous to that used for HLA alleles (37) (Fig. 8). After
the gene name, an asterisk is used as a separator before a numerical
allele designation. The first three digits of the numerical designa-
tion indicate alleles that differ in the sequences of their encoded
proteins. The next two digits distinguish alleles that differ only by
synonymous (noncoding) differences within the coding sequence.
The final two digits distinguish alleles that only differ by substitu-
tions in an intron, promoter, or other noncoding region of the
sequence.
To date, 614 KIR nucleotide sequences encoding 321 distinct
KIR proteins have been deposited in IPD-KIR database (Release
2.4.0, 15 April 2011), the database that provides a centralized
repository for human KIR sequences (http://www.ebi.ac.uk/ipd/
kir/). Inhibitory KIR receptors are more polymorphic than the
activating KIR receptors. The highest allelic polymorphisms are
reported for 3DL1, 3DL2, and 3DL3 loci, which is probably due
to the subjective sequencing analyses of large cohorts by certain
laboratories that focus on these loci (92–95).
The functional consequence of sequence polymorphism has
been studied in great detail for the KIR3DL1 locus. The amino

Fig. 8. Nomenclature of KIR alleles. Allelic sequence variants of KIR genes are named in a fashion analogous to that used for
HLA alleles. After the gene name, an asterisk is used as a separator before a numerical allele designation. The first three digits
of the numerical designation indicate alleles that differ in the sequences of their encoded proteins. The next two digits distin-
guish alleles that only differ by synonymous (noncoding) differences within the coding sequence. The final two digits distinguish
alleles that only differ by substitutions in an intron, promoter, or other noncoding region of the sequence.
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 403

acid substitutions that distinguish allelic diversity of 3DL1 are

enriched in polymorphic HLA-Bw4 ligand-binding Ig-like domains
compared to the cytoplasmic tail region, and such diversity appears to
be the result of natural selection (92). The sequence polymorphism
of 3DL1 is known to influence their expression, ligand binding
and cytolytic and cytokine secretion functions (53, 71, 93, 96, 97).
Some KIR sequences are null alleles that carry either a premature
stop codon or one or more nucleotide substitution/deletion/
insertion(s) that alter the reading frame of protein synthesis, which
affect their cell surface expression. The most frequently encoun-
tered unexpressed alleles belong to KIR2DS4 locus, which is the
most polymorphic among the activating KIR gene exhibiting 13
distinct protein variants. Nine (2DS4*003, *004, *006, *007,
*008, *009, *010, *012, and *013) have a 22 bp deletion in exon-5,
which shifts the reading-frame and results in a premature stop
codon causing this receptor to be unexpressed at the cell surface
(70, 98). Individuals homozygous for group-A haplotypes carrying
2DS4-deletion alleles will not express any activating KIR receptors
as they carry only 2DS4 as an activating KIR gene. Frequencies
of individuals with zero activating KIR genotypes vary between
ethnic populations (17% of Caucasians, 9% of Hispanics, 10% of
Africans, and 11% of Asians) (70, 98, 99).
The other frequently occurring unexpressed alleles belong to
KIR2DL4 locus. Based on the presence of a sequence of 9 or 10
adenine residues at the end of exon 6, all 46 alleles of 2DL4 can be
divided into two groups, 9A and 10A (100). The deletion of one
adenine in group 9A alleles results in a frame-shift and the produc-
tion of either a protein with a truncated cytoplasmic tail or one
lacking the transmembrane region, both of which are not expressed
at the cell surface. In contrast, the 10A alleles, which can be further
divided into the 10A-A and 10A-B subgroups, encode receptors
that may be expressed at the cell surface (101). Sequence variation
in the promoter region is implicated with the lack of 2DL5B
expression (102). Certain amino acid residues that distinguish
allelic polymorphism are known to be largely responsible for the
intracellular retention of 3DL1*004 (103) and 2DL2*004 (104).
The synergistic combination of allelic polymorphism and variable
gene content individualize KIR genotypes to an extent where
unrelated individuals almost always have different KIR genotypes
(77). This level of diversity likely reflects a strong pressure from
pathogens on the human NK cell response.

3.3. KIR Receptor In addition to the gene content variation and sequence polymor-
Repertoire Diversity phism, the level of mRNA expression varies among KIR genes and
of NK Cell Clones alleles. For instance, the KIR3DL3 transcripts are detected at low
levels in peripheral blood when compared to other KIR genes
(24, 105). Further, alternatively spliced mRNA are reported for
most KIR genes, and such isoforms can hinder the cell surface
404 R. Rajalingam

Fig. 9. Sequence polymorphism of KIR receptors. The number of nucleotide sequences (black bars) and their predicted
protein variants (gray bars) characterized to date for each KIR locus are shown. This data was extracted from IPD-KIR
database (http://www.ebi.ac.uk/ipd/kir/stats.html; Release 2.4.0; April 2011) that provides a centralized repository for
human KIR sequences. Null alleles identified for each locus are given in the parenthesis.

receptor expression (106). Although mRNA expressions have been

detected for all KIR genes (except for pseudogenes KIR2DP1 and
3DP1), due to the lack of specific antibodies, the cell surface
expression of KIR3DL3, 2DS3, and 2DS5 proteins is not confirmed
(Fig. 9).
The KIR receptors are clonally expressed on mature NK cells
in a stochastic manner such that each NK cell clone within a given
individual does not express the entire set of KIR genes present in
that individual’s genome, but rather only a portion of the genes in
a apparently random combination (20, 107). Combinatorial diver-
sity of KIR expression yields a broad range of functionally distinct
NK cell clones, which presumably are critical for a rapid and sensi-
tive detection of reduced HLA class I expression on target cells.
The NK cell clone stably maintains its acquired KIR receptor rep-
ertoire throughout subsequent cell divisions. Although the KIR
repertoire is largely determined by genetically defined factors
(108), there is increasing evidence that epigenetic mechanisms
modulate KIR receptor expression patterns. The methylation state
of cytosines within CpG islands at the promoter regions of KIR
loci is implicated in perpetual expression of KIR receptors on NK cell
clones (105, 109, 110). Furthermore, the presence of the cognate
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 405

HLA class I ligand is shown to increase the frequency of NK cells

expressing the specific inhibitory KIR receptor (71), but these
findings are challenged by another study (111). Finally, a recent
study indicates an unexpected and strong epistatic influence on the
expression of one KIR receptor (KIR2DL1) by other KIRs (2DL2
and 2DS2) (112).
The majority of NK cells in peripheral blood express at least
one inhibitory receptor for self-MHC class I and are functionally
competent to recognize and eliminate target cells that have down-
regulated their respective MHC class I ligands (20, 107).
Additionally, a subpopulation of developmentally immature NK
cells exists that lacks inhibitory receptors for self-MHC class I and
is generally hyporesponsive to target cells that are deficient in MHC
class I expression (113–115). Developmental interactions of the
KIRs with HLA class I trigger the acquisition of functional compe-
tence in a process called “licensing” (114, 116), “arming” (117),
or “education” (113). Therefore, individuals carry a minimum of
one inhibitory KIR-HLA gene pair, crucial for the development of
functional NK cells (99). Consistent with this, the NK cells from
MHC-deficient mice and humans are shown to be defective in target
killing (118, 119). Expression of progressively higher numbers of
inhibitory KIRs for self-HLA-B and HLA-Cw molecules has been
correlated with an increased effector capacity (120). In summary,
interactions of KIR to HLA class I ligands set the threshold of NK
cell capacity as well as control the NK cell response.
Unlike T and B cells, NK cells use a multiple receptor recogni-
tion strategy, whereby an individual NK cell can be triggered
through various receptors independently or in combination,
depending on the ligands presented by the target cell in a given
encounter (20, 25, 107). If a given NK cell uses both inhibitory
and activating receptors to recognize the target, the balance
between these disparate signals determines the action of that NK
cell (18). In addition to NK cells, KIR receptors are expressed on
a subset of T lymphocytes, in particular on gd T cells and memory/
effector CD8+ T cells indicating that the KIRs can regulate the
antigen-specific T cell immune response (121, 122).

4. Diversity
of KIR-HLA
Compound
Genotypes and KIR and HLA loci are both highly polymorphic and they map to
Relavance in distinct human chromosomes (Chromosomes 19 and 6, respec-
Disease tively). KIR and HLA gene families segregate independently, yield-
ing many individuals who express KIR receptors for which they
lack HLA class I ligands, and vice versa, thus creating human diver-
sity in the number and type of KIR-HLA inherited gene pairs (99).
Nevertheless, at the population level there is compelling evidence
406 R. Rajalingam

for coevolution of interacting KIR-HLA pairs (123, 124). Although

most individuals have all four well-defined inhibitory KIR recep-
tors (3DL1, 3DL2, 2DL1, and 2DL2/3), we found only a subset
that expresses all relevant HLA class I ligands, HLA-Bw4 (3DL1),
HLA-A3/11 (3DL2), HLA-C2 (2DL1), and HLA-C1 (2DL2/3)
(99). The majority of Caucasians, Hispanics, and African Americans
carry either two or three inhibitory KIR-HLA combinations.
Interestingly, one out of five individuals in these populations
carries only a single receptor–ligand pair, KIR2DL3 + HLA-C1.
Environmental invectives, such as viral infections, affecting HLA-C
expression in individuals carrying this single KIR-HLA combina-
tion can break self-tolerance and may endorse autoimmunity.
Furthermore, all 759 individuals tested in our study had at least
one of the four inhibitory KIR-HLA class I ligand gene pairs (99).
Similar studies of KIR-HLA gene combinations are warranted in
populations enriched with either group-A or group-B KIR haplo-
types, such as Chinese, Japanese, and natives of Australia, America,
and India, which will enlighten how the epistatic interactions
between KIR and HLA involved in the survival of modern human
populations.
The interaction between KIR2DL3 and HLA-C1 is weak and
signals triggered by this low affinity combination are shown to be
overridden by activating signals compared to those triggered
through the stronger inhibitory interactions of KIR2DL1 + HLA-
C2 (53). The clinical implication of such strength variation in KIR-
HLA-mediated inhibitions is evidenced from two lines of research.
Firstly, the poorly inhibiting KIR2DL3 + HLA-C1 gene combina-
tion more readily resolves infection with hepatitis C virus than
infected individuals of other genotypes (125). Secondly, the stron-
ger inhibitory interaction of KIR2DL1 + HLA-C2 was shown to
affect placental development causing many common reproductive
disorders, such as preeclampsia, recurrent miscarriage, and fetal
growth restriction (63, 126, 127). Moreover, these pregnancy
disorders were found to be less frequent in mothers that pos-
sessed the telomeric end of the group-B KIR haplotype, which
contains KIR3DS1, 2DL5A, 2DS1, and 2DS5 (127) (Fig. 7,
telomeric half tB02).
Although the cell surface expression and ligands for activating
KIR receptors have not been clearly defined, a series of genetic
epidemiological data have revealed the association of distinct acti-
vating KIR in antiviral immunity, autoimmune diseases, and cancer
progression (32, 33, 128). In these models, the activation signals
were proposed to prevail over HLA-dependent inhibition that pre-
sumably exacerbates NK cell response (18). KIR2DS2 or 2DS1
were found to be strongly associated with most of the autoimmune
conditions. A recent study with Japanese, a highly homogeneous
population appropriate for studying genetic associations, found a
strong association between the telomeric KIR3DS1-2DL5A-
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 407

2DS1-2DS5 gene content motif and patients with Vogt-Koyanagi-

Harada (VKH) disease (129).
In contrast to the autoimmune conditions, the presence of
activating receptor KIR3DS1 slowed down the AIDS progression
in HIV-1 infected patients, compared to patients lacking KIR3DS1,
thus indicating a potential anti-HIV effect of KIR3DS1 (68). The
KIR3DS1 receptor was also shown to be protective to hepatitis-C
virus (HCV) infection (125, 130). Furthermore, KIR3DS1 and
KIR2DS1 genes protect from developing a severe form of recur-
rent respiratory papillomatosis (RRP), a rare disease of the larynx
and upper airway caused by human papilloma viruses (HPV)-6/11
(131). Nonetheless, the presence of KIR3DS1 was shown to be
associated with cervical neoplasia progression to cervical cancer, a
tumor that is strongly associated with another HPV strain, HPV-
16/18 (132). The difference in clinical disease expression induced
by different HPV strains in 3DS1 carriers could be interaction of
3DS1 with putative HPV strain-specific ligands leading to either
killing of HPV-6/11-infected cells keeping the host from develop-
ing RRP, or leading to inappropriate tissue-specific hyperrespon-
siveness promoting growth of cervical cancer, vs. benign respiratory
papillomas.
Taken together, the genotypes encoding a dominant inhibi-
tory KIR receptor repertoire is likely protective against autoimmu-
nity but susceptible for infection and reproductive disorders, while
the genotypes encoding a dominant activating KIR receptor reper-
toire are presumably detrimental to autoimmunity and cancer
growth but instrumental against viral infection. Greater under-
standing of KIR-HLA diversity and functional distinctions between
NK cell subsets are needed to strengthen the ability to harness the
power of NK cells for therapeutic aims.

5. Alloreactive
NK Cells in
Hematopoietic
Stem Cell HLA-matched allogeneic hematopoietic stem cell transplantation
Transplantation (HSCT) is an effective treatment for hematologic malignancies,
including leukemia, lymphoma, and inherited hematopoietic stem
cell diseases (133). Donor T cells in the allograft are critical for pro-
moting engraftment and eradicating malignant cells. Unfortunately,
alloreactive T cells also cause graft-versus-host disease (GVHD),
which is an attack on recipient tissues, primarily the gastrointestinal
tract, liver, and skin. T cell depletion prevents GVHD but increases
the risk of graft rejection and leukemic relapse.
Following allogeneic HSCT, the progeny of the donor stem
cells repopulate the entire hematopoietic system of the recipient.
NK cells are the first lymphocyte population to appear in peripheral
blood shortly after HSCT. KIR receptor repertoires of NK cells
408 R. Rajalingam

reconstituted from the donor hematopoietic precursors are

consistently of donor type (108). The donor-derived NK cells can
be alloreactive if their inhibitory KIRs do not see a relevant HLA
class I ligand that was present in the donor. Such alloreactive
NK cells greatly contribute many potential benefits, including
decreased rates of GVHD, decreased rates of graft rejection
mediated by NK lysis of host T cells, decreased relapse, improved
engraftment mediated by NK-cell release of hematopoietic cytok-
ines, and enhanced immune reconstitution and decreased infec-
tious complications mediated by NK-cell antiviral activity. Such
beneficial NK cell alloreactivity, which can be predicted from the
differences in KIR-binding HLA class I ligands between donor and
recipient based on their HLA class I type, was first described for
HLA haploidentical transplantation by the use of an extensively T
cell-depleted graft in acute myeloid leukemia (AML) patients (134)
and later investigated in other transplantation settings (135, 136).
The patients with myeloid malignancies were more responsive
to treatment than those with lymphoid malignancies. Several studies
have suggested that adult acute lymphoid leukemia (ALL) is not as
susceptible to KIR ligand-mismatched haploidentical allogeneic
HSCT (134, 137–139). However, alloreactive NK cell-mediated
effects can impact childhood ALL (140). Beneficial effects from
KIR-ligand mismatch have not been seen in the T cell-replete
setting (141, 142).
In addition to the recipient lacking HLA class I ligands for the
donor-derived NK cells, expression of activating KIR receptors on
donor NK cells is also shown to influence HSCT outcome.
Compared to donors with AA genotypes (express one or no acti-
vating KIR), the Bx genotype (express 1–6 activating KIRs) donors
were shown to contribute significantly superior relapse protection
and improved disease-free survival for AML patients (143). Gene
content motif analyses further reveal that the centromeric and telo-
meric B haplotype-specific motifs both contribute to relapse pro-
tection and improved survival, but centromeric B homozygosity
(Cen-B/B) has the strongest independent effect (143). Further
studies are required to determine if the clinical benefit conferred
by Cen-B/B is caused by a single KIR gene (such as 2DS2, 2DS3,
or KIR2DL2) or by the combination of specific KIR genes.
Recent studies reveal the strongest clinical impact of telomeric
B haplotype-specific activating KIR genes (144, 145). Compared
with KIR3DS1-negative donors, a donor with KIR3DS1 was shown
to be associated with lower-grade II–IV acute GVHD, but not with
relapse (144). Furthermore, grade II–IV acute GVHD, overall
mortality, and transplantation-related mortality all decreased as the
number of copies of donor KIR3DS1 increased, with the lowest
failure rate occurring among patients homozygous for donor
KIR3DS1 (144). Functional experiments reveal that the activating
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 409

KIR2DS1 plays a substantial role in mediating alloreactivity and

confers an advantage in the ability of NK cell alloreactivity to kill
dendritic cells and T cell blasts (145). In summary, knowing the
KIR genotype of the donor, and HLA types of both the donor and
recipient, it is possible to predict the degree of KIR-HLA interac-
tions that may determine an enhanced ability to limit GVHD and
improve engraftment for certain leukemias.

Acknowledgment

This work was supported by start-up funds from the UCLA

Department of Pathology and Laboratory Medicine.

References
1. Caligiuri MA (2008) Human natural killer 12. Manaster I, Mandelboim O (2008) The unique
cells. Blood 112:461–469 properties of human NK cells in the uterine
2. Vivier E et al (2011) Innate or adaptive mucosa. Placenta 29(suppl A):S60–S66
immunity? The example of natural killer cells. 13. Colonna M (2009) Interleukin-22-producing
Science 331:44–49 natural killer cells and lymphoid tissue
3. Colucci F et al (2003) What does it take to inducer-like cells in mucosal immunity.
make a natural killer? Nat Rev Immunol Immunity 31:15–23
3:413–425 14. Bryceson YT, Long EO (2008) Line of attack:
4. Lee SH et al (2007) Keeping NK cells in NK cell specificity and integration of signals.
highly regulated antiviral warfare. Trends Curr Opin Immunol 20:344–352
Immunol 28:252–259 15. Lanier LL (2003) Natural killer cell receptor
5. Smyth MJ et al (2002) New aspects of signaling. Curr Opin Immunol 15:308–314
natural-killer-cell surveillance and therapy of 16. McQueen KL, Parham P (2002) Variable
cancer. Nat Rev Cancer 2:850–861 receptors controlling activation and inhibi-
6. Sun JC et al (2010) Immune memory tion of NK cells. Curr Opin Immunol 14:
redefined: characterizing the longevity of nat- 615–621
ural killer cells. Immunol Rev 236:83–94 17. Moretta L et al (2000) Human NK-cell recep-
7. Cooper MA et al (2001) Human natural killer tors. Immunol Today 21:420–422
cells: a unique innate immunoregulatory role 18. Lanier LL (2005) NK cell recognition. Annu
for the CD56 (bright) subset. Blood Rev Immunol 23:225–274
97:3146–3151 19. Lanier LL (2008) Up on the tightrope: natu-
8. Freud AG, Caligiuri MA (2006) Human ral killer cell activation and inhibition. Nat
natural killer cell development. Immunol Rev Immunol 9:495–502
214:56–72 20. Valiante NM et al (1997) Functionally and
9. Jacobs R et al (2001) CD56bright cells differ structurally distinct NK cell receptor reper-
in their KIR repertoire and cytotoxic features toires in the peripheral blood of two human
from CD56dim NK cells. Eur J Immunol donors. Immunity 7:739–751
31:3121–3127 21. Biron CA (2001) Interferons alpha and beta
10. Fehniger TA et al (2003) CD56bright natural as immune regulators—a new look. Immunity
killer cells are present in human lymph nodes 14:661–664
and are activated by T cell-derived IL-2: a 22. Ljunggren HG, Karre K (1990) In search of
potential new link between adaptive and the ‘missing self’: MHC molecules and NK cell
innate immunity. Blood 101:3052–3057 recognition. Immunol Today 11:237–244
11. Cooper MA et al (2004) NK cell and DC 23. Lanier LL (1998) NK cell receptors. Annu
interactions. Trends Immunol 25:47–52 Rev Immunol 16:359–393
410 R. Rajalingam

24. Long EO et al (2001) Inhibition of natural 41. Klein J, Sato A (2000) The HLA system. First
killer cell activation signals by killer cell immu- of two parts. N Engl J Med 343:702–709
noglobulin-like receptors (CD158). Immunol 42. Boyington JC et al (2000) Crystal structure
Rev 181:223–233 of an NK cell immunoglobulin-like receptor
25. Moretta A et al (2001) Activating receptors in complex with its class I MHC ligand.
and coreceptors involved in human natural Nature 405:537–543
killer cell-mediated cytolysis. Annu Rev 43. Khakoo SI et al (2002) The D0 domain of
Immunol 19:197–223 KIR3D acts as a major histocompatibility
26. Vilches C, Parham P (2002) KIR: diverse, complex class I binding enhancer. J Exp Med
rapidly evolving receptors of innate and adaptive 196:911–921
immunity. Annu Rev Immunol 20:217–251 44. Prugnolle F et al (2005) Pathogen-driven
27. Lanier LL (2008) Evolutionary struggles selection and worldwide HLA class I diversity.
between NK cells and viruses. Nat Rev Curr Biol 15:1022–1027
Immunol 8:259–268 45. Parham P, Ohta T (1996) Population biology
28. Parham P (2003) Immunogenetics of killer- of antigen presentation by MHC class I mol-
cell immunoglobulin-like receptors. Tissue ecules. Science 272:67–74
Antigens 62:194–200 46. Colonna M et al (1992) Alloantigen recogni-
29. Abi-Rached L et al (2010) Human-specific tion by two human natural killer cell clones is
evolution and adaptation led to major qualita- associated with HLA-C or a closely linked
tive differences in the variable receptors of gene. Proc Natl Acad Sci USA 89:
human and chimpanzee natural killer cells. 7983–7985
PLoS Genet 6:e1001192 47. Mandelboim O et al (1996) Protection from
30. Parham P (2005) MHC class I molecules and lysis by natural killer cells of group 1 and 2
KIRs in human history, health and survival. specificity is mediated by residue 80 in human
Nat Rev Immunol 5:201–214 histocompatibility leukocyte antigen C alleles
31. Uhrberg M et al (1997) Human diversity in and also occurs with empty major histocom-
killer cell inhibitory receptor genes. Immunity patibility complex molecules. J Exp Med
7:753–763 184:913–922
32. Khakoo SI, Carrington M (2006) KIR and 48. Mandelboim O et al (1997) The binding site
disease: a model system or system of models? of NK receptors on HLA-C molecules.
Immunol Rev 214:186–201 Immunity 6:341–350
33. Bashirova AA et al (2006) The killer 49. Colonna M et al (1993) HLA-C is the inhibi-
immunoglobulin-like receptor gene cluster: tory ligand that determines dominant resis-
tuning the genome for defense. Annu Rev tance to lysis by NK1- and NK2-specific
Genomics Hum Genet 7:277–300 natural killer cells. Proc Natl Acad Sci USA
34. Wilson MJ et al (2000) Plasticity in the orga- 90:12000–12004
nization and sequences of human KIR/ILT 50. Wagtmann N et al (1995) Molecular clones of
gene families. Proc Natl Acad Sci USA 97: the p58 NK cell receptor reveal
4778–4783 immunoglobulin-related molecules with
35. Trowsdale J (2001) Genetic and functional diversity in both the extra- and intracellular
relationships between MHC and NK receptor domains. Immunity 2:439–449
genes. Immunity 15:363–374 51. Winter CC, Long EO (1997) A single amino
36. Steffens U et al (1998) Nucleotide and amino acid in the p58 killer cell inhibitory receptor
acid sequence alignment for human killer cell controls the ability of natural killer cells to dis-
inhibitory receptors (KIR), 1998. Tissue criminate between the two groups of HLA-C
Antigens 51:398–413 allotypes. J Immunol 158:4026–4028
37. Marsh SG et al (2003) Killer-cell immuno- 52. Moesta AK et al (2008) Synergistic polymor-
globulin-like receptor (KIR) nomenclature phism at two positions distal to the ligand-
report, 2002. Tissue Antigens 62:79–86 binding site makes KIR2DL2 a stronger
38. Lanier LL (2009) DAP10- and DAP12- receptor for HLA-C than KIR2DL3.
associated receptors in innate immunity. J Immunol 180:3969–3979
Immunol Rev 227:150–160 53. Winter CC et al (1998) Direct binding and
39. Marsh SG et al (2011) Nomenclature for fac- functional transfer of NK cell inhibitory recep-
tors of the HLA system, 2010. Tissue Antigens tors reveal novel patterns of HLA-C allotype
75:291–455 recognition. J Immunol 161:571–577
40. Klein J, Sato A (2000) The HLA system. Second 54. Stewart CA et al (2005) Recognition of
of two parts. N Engl J Med 343:782–786 peptide-MHC class I complexes by activating
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 411

killer immunoglobulin-like receptors. Proc 68. Martin MP et al (2002) Epistatic interaction

Natl Acad Sci USA 102:13224–13229 between KIR3DS1 and HLA-B delays the
55. Graef T et al (2009) KIR2DS4 is a product of progression to AIDS. Nat Genet 31:429–434
gene conversion with KIR3DL2 that intro- 69. Uhrberg M et al (2002) Definition of gene
duced specificity for HLA-A*11 while dimin- content for nine common group B haplotypes
ishing avidity for HLA-C. J Exp Med of the Caucasoid population: KIR haplotypes
206:2557–2572 contain between seven and eleven KIR genes.
56. Gumperz JE et al (1995) The Bw4 public Immunogenetics 54:221–229
epitope of HLA-B molecules confers reactiv- 70. Hsu KC et al (2002) Killer Ig-like receptor
ity with natural killer cell clones that express haplotype analysis by gene content: evidence
NKB1, a putative HLA receptor. J Exp Med for genomic diversity with a minimum of six
181:1133–1144 basic framework haplotypes, each with multi-
57. Cella M et al (1994) NK3-specific natural ple subsets. J Immunol 169:5118–5129
killer cells are selectively inhibited by Bw4- 71. Yawata M et al (2006) Roles for HLA and
positive HLA alleles with isoleucine 80. J Exp KIR polymorphisms in natural killer cell rep-
Med 180:1235–1242 ertoire selection and modulation of effector
58. Thananchai H et al (2007) Cutting edge: function. J Exp Med 203:633–645
allele-specific and peptide-dependent interac- 72. Whang DH et al (2005) Haplotype analysis of
tions between KIR3DL1 and HLA-A and killer cell immunoglobulin-like receptor genes
HLA-B. J Immunol 178:33–37 in 77 Korean families. Hum Immunol
59. Pende D et al (1996) The natural killer cell 66:146–154
receptor specific for HLA-A allotypes: a novel 73. Middleton D et al (2007) KIR haplotype con-
member of the p58/p70 family of inhibitory tent at the allele level in 77 Northern Irish
receptors that is characterized by three immu- families. Immunogenetics 59:145–158
noglobulin-like domains and is expressed as a 74. Pyo C et al (2010) Different patterns of evo-
140-kD disulphide-linked dimer. J Exp Med lution in the centromeric and telomeric
184:505–518 regions of group A and B haplotypes of the
60. Dohring C et al (1996) A human killer human killer cell Ig-like receptor locus. PLoS
inhibitory receptor specific for HLA-A1,2. One 5:e15115
J Immunol 156:3098–3101 75. Yawata M et al (2002) Variation within the
61. Hansasuta P et al (2004) Recognition of HLA- human killer cell immunoglobulin-like recep-
A3 and HLA-A11 by KIR3DL2 is tor (KIR) gene family. Crit Rev Immunol
peptide-specific. Eur J Immunol 34:1673–1679 22:463–482
62. Rajagopalan S, Long EO (1999) A human 76. Ashouri E et al (2009) KIR gene content
histocompatibility leukocyte antigen (HLA)- diversity in four Iranian populations.
G-specific receptor expressed on all natural Immunogenetics 61:483–492
killer cells. J Exp Med 189:1093–1100 77. Shilling HG et al (2002) Allelic polymor-
63. Moffett-King A (2002) Natural killer cells and phism synergizes with variable gene content
pregnancy. Nat Rev Immunol 2:656–663 to individualize human KIR genotype.
64. Rajagopalan S et al (2006) Activation of NK J Immunol 168:2307–2315
cells by an endocytosed receptor for soluble 78. Norman PJ et al (2009) Meiotic recombina-
HLA-G. PLoS Biol 4:e9 tion generates rich diversity in NK cell recep-
65. Kikuchi-Maki A et al (2003) KIR2DL4 Is an tor genes, alleles, and haplotypes. Genome
IL-2-regulated NK cell receptor that exhibits Res 19:757–769
limited expression in humans but triggers 79. Martin AM et al (2000) The genomic organi-
strong IFN-gamma production. J Immunol zation and evolution of the natural killer
171:3415–3425 immunoglobulin-like receptor (KIR) gene
66. Faure M, Long EO (2002) KIR2DL4 cluster. Immunogenetics 51:268–280
(CD158d), an NK cell-activating receptor 80. Williams F et al (2003) Multiple copies of
with inhibitory potential. J Immunol 168: KIR 3DL/S1 and KIR 2DL4 genes identified
6208–6214 in a number of individuals. Hum Immunol
67. Ponte M et al (1999) Inhibitory receptors 64:729–732
sensing HLA-G1 molecules in pregnancy: 81. Ordonez D et al (2008) Duplication, muta-
decidua-associated natural killer cells express tion and recombination of the human orphan
LIR-1 and CD94/NKG2A and acquire p49, gene KIR2DS3 contribute to the diversity
an HLA-G1-specific receptor. Proc Natl Acad of KIR haplotypes. Genes Immun 9:
Sci USA 96:5674–5679 431–437
412 R. Rajalingam

82. Jiang K et al (2005) Distribution of killer cell predicts a soluble KIR molecule analogous to
immunoglobulin-like receptor genes in the the KIR1D molecule observed in the rhesus
Chinese Han population. Tissue Antigens monkey. Tissue Antigens 60:254–258
65:556–563 99. Du Z et al (2007) Receptor-ligand analyses
83. Yawata M et al (2002) Predominance of define minimal killer cell Ig-like receptor
group A KIR haplotypes in Japanese associ- (KIR) in humans. Immunogenetics 59:1–15
ated with diverse NK cell repertoires of KIR 100. Witt CS et al (2000) Detection of KIR2DL4
expression. Immunogenetics 54:543–550 alleles by sequencing and SSCP reveals a com-
84. Gendzekhadze K et al (2006) High KIR mon allele with a shortened cytoplasmic tail.
diversity in Amerindians is maintained using Tissue Antigens 56:248–257
few gene-content haplotypes. Immunogenetics 101. Goodridge JP et al (2003) KIR2DL4
58:474–480 (CD158d) genotype influences expression
85. Ewerton PD et al (2007) Amazonian and function in NK cells. J Immunol 171:
Amerindians exhibit high variability of KIR 1768–1774
profiles. Immunogenetics 59:625–630 102. Vilches C et al (2000) Gene structure and
86. Toneva M et al (2001) Genomic diversity of promoter variation of expressed and nonex-
natural killer cell receptor genes in three pop- pressed variants of the KIR2DL5 gene.
ulations. Tissue Antigens 57:358–362 J Immunol 165:6416–6421
87. Rajalingam R et al (2008) Distinct diversity of 103. Pando MJ et al (2003) The protein made
KIR genes in three southern Indian popula- from a common allele of KIR3DL1
tions: comparison with world populations (3DL1*004) is poorly expressed at cell sur-
revealed a link between KIR gene content and faces due to substitution at positions 86 in Ig
pre-historic human migrations. domain 0 and 182 in Ig domain 1. J Immunol
Immunogenetics 60:207–217 171:6640–6649
88. Rajalingam R et al (2002) Distinctive KIR 104. VandenBussche CJ et al (2006) A single poly-
and HLA diversity in a panel of north Indian morphism disrupts the killer Ig-like receptor
Hindus. Immunogenetics 53:1009–1019 2DL2/2DL3 D1 domain. J Immunol 177:
89. Kulkarni S et al (2008) Comparison of the 5347–5357
rapidly evolving KIR locus in Parsis and natives 105. Trompeter HI et al (2005) Three structurally
of India. Immunogenetics 60:121–129 and functionally divergent kinds of promoters
90. Garcia CA et al (2003) Human KIR sequences regulate expression of clonally distributed
2003. Immunogenetics 55:227–239 killer cell Ig-like receptors (KIR), of
KIR2DL4, and of KIR3DL3. J Immunol
91. Rajalingam R et al (2001) Identification of
174:4135–4143
seventeen novel KIR variants: fourteen of
them from two non-Caucasian donors. Tissue 106. Dohring C et al (1996) Alternatively spliced
Antigens 57:22–31 forms of human killer inhibitory receptors.
Immunogenetics 44:227–230
92. Norman PJ et al (2007) Unusual selection on
the KIR3DL1/S1 natural killer cell receptor 107. Raulet DH et al (2001) Regulation of the
in Africans. Nat Genet 39:1092–1099 natural killer cell receptor repertoire. Annu
Rev Immunol 19:291–330
93. Gardiner CM et al (2001) Different NK cell
surface phenotypes defined by the DX9 anti- 108. Shilling HG et al (2002) Genetic control of
body are due to KIR3DL1 gene polymor- human NK cell repertoire. J Immunol
phism. J Immunol 166:2992–3001 169:239–247
94. Jones DC et al (2006) Nature of allelic 109. Santourlidis S et al (2002) Crucial role of
sequence polymorphism at the KIR3DL3 DNA methylation in determination of clon-
locus. Immunogenetics 58:614–627 ally distributed killer cell Ig-like receptor
expression patterns in NK cells. J Immunol
95. Hou L et al (2007) Seventeen novel alleles 169:4253–4261
add to the already extensive KIR3DL3 diver-
sity. Tissue Antigens 70:449–454 110. Chan HW et al (2003) DNA methylation
maintains allele-specific KIR gene expression
96. O’Connor GM et al (2007) Functional poly- in human natural killer cells. J Exp Med
morphism of the KIR3DL1/S1 receptor on 197:245–255
human NK cells. J Immunol 178:235–241
111. Andersson S et al (2009) KIR acquisition
97. Carr WH et al (2005) KIR3DL1 polymor- probabilities are independent of self-HLA
phisms that affect NK cell inhibition by class I ligands and increase with cellular KIR
HLA-Bw4 ligand. J Immunol 175:5222–5229 expression. Blood 114:95–104
98. Maxwell LD et al (2002) A common 112. Schonberg K et al (2011) Analyses of HLA-
KIR2DS4 deletion variant in the human that C-specific KIR repertoires in donors with
 23 Overview of the Killer Cell Immunoglobulin-Like Receptor System 413

group A and B haplotypes suggest a ligand- mediated by fetal HLA-C2. J Clin Invest
instructed model of NK cell receptor aquisi- 120:4102–4110
tion. Blood 117:98–107 128. Terme M et al (2008) Natural killer cell-
113. Anfossi N et al (2006) Human NK cell educa- directed therapies: moving from unexpected
tion by inhibitory receptors for MHC class I. results to successful strategies. Nat Immunol
Immunity 25:331–342 9:486–494
114. Kim S et al (2005) Licensing of natural killer 129. Levinson RD et al (2010) Killer cell immuno-
cells by host major histocompatibility com- globulin-like receptor gene-cluster 3DS1-
plex class I molecules. Nature 436:709–713 2DL5-2DS1-2DS5 predisposes susceptibility to
115. Cooley S et al (2007) A subpopulation of Vogt-Koyanagi-Harada syndrome in Japanese
human peripheral blood NK cells that lacks individuals. Hum Immunol 71:192–194
inhibitory receptors for self MHC is develop- 130. Lopez-Vazquez A et al (2005) Protective
mentally immature. Blood 110:578–586 effect of the HLA-Bw4I80 epitope and the
116. Yokoyama WM, Kim S (2006) How do natu- killer cell immunoglobulin-like receptor 3DS1
ral killer cells find self to achieve tolerance? gene against the development of hepatocel-
Immunity 24:249–257 lular carcinoma in patients with hepatitis C
117. Raulet DH, Vance RE (2006) Self-tolerance virus infection. J Infect Dis 192:162–165
of natural killer cells. Nat Rev Immunol 131. Bonagura VR et al (2010) Activating killer
6:520–531 cell immunoglobulin-like receptors 3DS1 and
118. Bix M et al (1991) Rejection of class I MHC- 2DS1 protect against developing the severe
deficient haemopoietic cells by irradiated form of recurrent respiratory papillomatosis.
MHC-matched mice. Nature 349:329–331 Hum Immunol 71:212–219
119. Furukawa H et al (1999) Tolerance of NK and 132. Carrington M et al (2005) Hierarchy of resis-
LAK activity for HLA class I-deficient targets tance to cervical neoplasia mediated by com-
in a TAP1-deficient patient (bare lymphocyte binations of killer immunoglobulin-like
syndrome type I). Hum Immunol 60:32–40 receptor and human leukocyte antigen loci.
J Exp Med 201:1069–1075
120. Yu J et al (2007) Hierarchy of the human
natural killer cell response is determined by 133. Appelbaum FR (2003) The current status of
class and quantity of inhibitory receptors for hematopoietic cell transplantation. Annu Rev
self-HLA-B and HLA-C ligands. J Immunol Med 54:491–512
179:5977–5989 134. Ruggeri L et al (2002) Effectiveness of donor
121. Chang C et al (1995) Molecular characteriza- natural killer cell alloreactivity in mismatched
tion of human CD94: a type II membrane hematopoietic transplants. Science 295:
glycoprotein related to the C-type lectin 2097–2100
superfamily. Eur J Immunol 25:2433–2437 135. Giebel S et al (2003) Survival advantage with
122. Uhrberg M et al (2001) The repertoire of KIR ligand incompatibility in hematopoietic
killer cell Ig-like receptor and CD94:NKG2A stem cell transplantation from unrelated
receptors in T cells: clones sharing identical donors. Blood 102:814–819
alpha beta TCR rearrangement express highly 136. Davies SM et al (2002) Evaluation of KIR
diverse killer cell Ig-like receptor patterns. ligand incompatibility in mismatched unre-
J Immunol 166:3923–3932 lated donor hematopoietic transplants. Killer
123. Single RM et al (2007) Global diversity and immunoglobulin-like receptor. Blood 100:
evidence for coevolution of KIR and HLA. 3825–3827
Nat Genet 39:1114–1119 137. Kawase T et al (2009) HLA mismatch combi-
124. Gendzekhadze K et al (2009) Co-evolution nations associated with decreased risk of
of KIR2DL3 with HLA-C in a human popu- relapse: implications for the molecular mecha-
lation retaining minimal essential diversity of nism. Blood 113:2851–2858
KIR and HLA class I ligands. Proc Natl Acad 138. Clausen J et al (2007) Impact of natural killer
Sci USA 106:18692–18697 cell dose and donor killer-cell immunoglobulin-
125. Khakoo SI et al (2004) HLA and NK cell like receptor (KIR) genotype on outcome fol-
inhibitory receptor genes in resolving hepati- lowing human leucocyte antigen-identical
tis C virus infection. Science 305:872–874 haematopoietic stem cell transplantation. Clin
Exp Immunol 148:520–528
126. Hiby SE et al (2004) Combinations of mater-
nal KIR and fetal HLA-C genes influence the 139. Willemze R et al (2009) KIR-ligand incom-
risk of preeclampsia and reproductive success. patibility in the graft-versus-host direction
J Exp Med 200:957–965 improves outcomes after umbilical cord blood
transplantation for acute leukemia. Leukemia
127. Hiby SE et al (2011) Maternal activating 23:492–500
KIRs protect against human reproductive failure
414 R. Rajalingam

140. Pende D et al (2009) Anti-leukemia activity registry. Biol Blood Marrow Transplant 12:
of alloreactive NK cells in KIR ligand- 876–884
mismatched haploidentical HSCT for pediat- 143. Cooley S et al (2010) Donor selection for
ric patients: evaluation of the functional role natural killer cell receptor genes leads to supe-
of activating KIR and redefinition of inhibi- rior survival after unrelated transplantation
tory KIR specificity. Blood 113:3119–3129 for acute myelogenous leukemia. Blood 116:
141. Lowe EJ et al (2003) T-cell alloreactivity 2411–2419
dominates natural killer cell alloreactivity in 144. Venstrom JM et al (2010) Donor activating
minimally T-cell–depleted HLA-non-identical KIR3DS1 is associated with decreased acute
paediatric bone marrow transplantation. Br GVHD in unrelated allogeneic hematopoietic
J Haematol 123:23–326 stem cell transplantation. Blood 115:
142. Farag SS et al (2006) The effect of KIR ligand 3162–3165
incompatibility on the outcome of unrelated 145. Sivori S et al (2011) Natural killer cells
donor transplantation: a report from the cen- expressing the KIR2DS1-activating receptor
ter for international blood and marrow trans- efficiently kill T-cell blasts and dendritic cells:
plant research, the European blood and implications in haploidentical HSCT. Blood
marrow transplant registry, and the Dutch 117:4284–4292
 Chapter 24

KIR Typing by Non-Sequencing Methods: Polymerase-Chain

Reaction with Sequence-Specific Primers
David Ordóñez, Manuela Moraru, Natalia Gómez-Lozano,
Elisa Cisneros, and Carlos Vilches

Abstract
The killer-cell immunoglobulin-like receptors (KIR), which enable NK cells to detect allogeneic target
cells and abnormalities in the expression of self-HLA molecules, are encoded by genes that display extensive
copy number variation. These variations in the KIR genotype are relevant for multiple aspects of human
health, including therapy of cancer. PCR with sequence-specific primers (SSP) is simplest and most widely
used among techniques for studying KIR genotypes. Here, we present a protocol that details the critical
steps of a method for KIR genotyping by PCR-SSP.

Key words: Alloreactivity, Copy number variation, Electrophoresis, Genotyping, Killer-cell immuno-
globulin-like receptors, Hemopoietic transplant, HLA, Natural killer cells, PCR, Quality assurance

1. Introduction

1.1. Background Killer-cell immunoglobulin-like receptors (KIR) provide NK cells

with the capacity of surveying normal expression of MHC class I
molecules, frequently altered in intracellular infection and tumours
(1). Human KIR are encoded in chromosome 19 by a gene com-
plex that displays extensive copy number variation—most com-
monly, an individual lacks one or more KIR genes, and, conversely,
some chromosomes carry KIR-gene duplications (2–4). Certain
arrangements of KIR genes, or haplotypes, are more common
than others, their frequencies varying in different populations and
ethnic groups. Variability of KIR genotypes influences human
health in several ways, and is consistent with selective pressures
having favoured functional diversification of KIR in manners and
degrees that resemble those of their HLA ligands.

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_24, © Springer Science+Business Media New York 2012

415
416 D. Ordóñez et al.

1.2. Role of KIR NK cells kill target cells that lack HLA ligands for the inhibitory
Genotyping in the receptors of the NK cell, as it often happens in tumours and infected
Clinical Context cells. Killing can also occur when an NK cell meets an allogeneic
target that lacks one or more of the polymorphic HLA epitopes
1.2.1. Exploiting NK-Cell
that “educated” or “licensed” the NK cell. This phenomenon has
Alloreactivity Against
been proposed to mediate beneficial effects on patient survival, and
Cancer
on graft vs. host and leukaemia-relapse rates after hemopoietic
transplant from non-HLA identical donors (5). Similarly, adoptive
transfer of allogeneic NK cells is a promising form of immuno-
therapy against cancer. In both clinical settings, selection of the
appropriate donors requires identification of the HLA epitopes
recognised by the inhibitory KIR 2DL1, 2DL2/2DL3, 3DL1,
and 3DL2 (lysine-80 of HLA-C; asparagine-80 of HLA-C; the
Bw4 motif of HLA-B and HLA-A; and alleles HLA-A*03 and
-A*11, respectively). Furthermore, because NK-cell alloreactivity
is based on the presence in the allogeneic NK cells of certain inhib-
itory KIR, optimal donor selection for transplant or adoptive
NK-cell therapy should include KIR genotyping (and, if possible,
surface KIR phenotyping by flow cytometry) to ensure that the
necessary receptors are included in the donor KIR repertoire.
Independently of the predictable NK-cell alloreactivity deter-
mined by mismatch between the HLA epitopes of donors and
recipients, some transplant centres have observed that patients
transplanted from donors carrying certain combinations of KIR
genes have better outcomes than patients who received hemopoi-
etic transplants from donors that lacked those genes (6). According
to this, KIR genotyping should be part of the protocol for donor
selection.

1.2.2. Other Clinical KIR polymorphisms can influence multiple aspects of human
Conditions health, including at least: (a) reproduction and its complications;
(b) chronic infection; (c) inflammatory disease and autoimmunity;
(d) solid organ transplant; and (e) tumours. In none of these clinical
settings is KIR genotyping yet unanimously accepted as a part of
decision making for patient care. Further research is needed to
understand how KIR participate in the physiopathology of those
processes, which should allow for predictive models of clinical out-
comes based on the patient KIR genotype, and also for therapies
based on a more comprehensive knowledge of NK-cell physiology.
In addition, subsequent studies will likely identify new diseases in
which risk or clinical course is influenced by the diversity of KIR
genotypes.

1.3. KIR Typing Using The diversity of KIR genotypes was discovered by means of a poly-
a PCR-SSP Method merase-chain reaction with sequence-specific primers (PCR-SSP)
technique that detected distinctive motifs of each KIR gene (7).
Different PCR-SSP methods are still most widely used for exploring
the genetic diversity of KIR. Here, we provide a detailed protocol for
performing a PCR-SSP method for KIR genotyping that amplifies
 24 KIR Typing by Non-Sequencing Methods… 417

short DNA fragments (8). The method comprises 16 reactions,

one of them multiplexed, that enable quick and simple detection
of all KIR genes and pseudogenes.

2. Materials

2.1. Special Equipment 1. Electronic repetitive pipettors (e.g. EDP-1 or -3 from Rainin)
can be programmed to dispense multiple aliquots of the desired
volume after a single aspiration step. They are of extraordinary
help for dispensing primer mixes and liquid wax into typing
sets; DNA:buffer:Taq mixes for PCR setup; and the loading
dye to processed reactions before electrophoresis.
2. Thermal cycler with a 96-well block. We have used models
PTC-100 and PTC-200 from MJR, and models 2720 and
Veriti 96-Well from Applied Biosystems.
3. Electrophoresis system. A horizontal electrophoresis chamber
with a tray that accommodates six 17- or 18-well combs enables
genotyping six DNAs (96 reactions) in a single run (16 reac-
tions and 1 or 2 size markers per row). Moreover, multichan-
nel pipette-compatible combs greatly speed up the loading of
a full gel (see Notes 1 and 2).
4. A multichannel pipettor in the 0.5–10.0 mL range (e.g. CAPP)
for gel loading (see Note 3).

2.2. Consumables 1. Strips of eight attached 0.2-mL PCR microtubes are the most
and Reagents convenient format. The primer set for typing one DNA will
comprise two such strips (16 reactions).
2. The oligonucleotides primers used in this method have been
described previously (8). Standard quality (desalted) oligonu-
cleotides are sufficient.
3. Polymerases: regular Taq polymerase is fine; hot start is not
necessary (see Note 4). Proof-reading polymerases with 3¢–5¢
exonuclease activity must not be used for PCR-SSP.
4. Other speciality reagents: Chill-out red Liquid Wax (Bio-Rad)
for storage of aliquoted primer mixes; store at room tempera-
ture or 4°C.

3. Methods

3.1. Prepare Primer 1. These primer sets (Tables 1 and 2), designed in 2007 (8),
Mixes detected all 168 KIR alleles of the KIR-IPD database release
1.4.0 (http://www.ebi.ac.uk/ipd/kir/). Current release of
the database (2.3.0) includes 614 alleles: 608 are still detected
418 D. Ordóñez et al.

Table 1
Chart for preparing primer mixes for KIR genotyping

Forward primer, 100 mM Reverse primer, 100 mM

Mix Gene Water (mL) IPC283 (mL) Name Volume (mL) Name Volume (mL)
1 KIR2DL1 925.0 50 Fa517a 12.5 Rc621 12.5
2 KIR2DL2 825.0 50 Fcon750 62.5 Rt854 62.5
3 KIR2DL3 825.0 50 Fcon1254 62.5 Rt1375 62.5
4 KIR3DL1 762.5 50 Ft624 62.5 Rt697 62.5
Ftt624 62.5
5 KIR3DL2 932.5 50 Fg864 12.5 Rc962 5.0
6 KIR2DS1 762.5 50 Fg621 62.5 Rcon682 62.5
Fa621b 62.5
7 KIR2DS2 825.0 50 Fa546 62.5 Rcon621 62.5
a
8 KIR2DS3 887.5 50 Ft803 31.25 Ra925 31.25
9 KIR2DS4 825.0 50 Fat781 62.5 Rca877 62.5
10 KIR2DS5 825.0 50 Fc551 62.5 Rcon662 62.5
11 KIR3DS1 762.5 50 Fg624 62.5 Rt697 62.5
Fg624b 62.5
12 KIR2DP1 918.8 50 Fc567 15.62 Rdel674 15.62
13 KIR3DL3 700.0 50 Fg510 62.5 Rta669b 62.5
KIR3DX1 Fma920 62.5 Rdel967b 62.5
14 KIR2DL4 825.0 50 Ftgc157b 62.5 Rga250b 62.5
15 KIR2DL5 887.5 50 Fag843 31.25 Rc953 31.25
16 KIR3DP1 856.3 IPC608: 50 mL Fi2–89 31.25 Rg287 31.25
Fa–97b 31.25
a
See Note 5

by the correct primer mix; none is predicted to cross-react with

a wrong mix; and six new alleles are missed—2DL1*013N,
2DL2*009, 2DL3*010 and *017, 2DS3*004, and 3DL1*054
(see Note 5).
2. Dissolve all your lyophilised primers at 100 mM in 10 mM
Tris–HCl pH 8.0 (e.g. if you receive 78.6 nmol of a given
primer, add 786 mL solvent) (primers will be more stable at pH
8 than dissolved in H2O).
3. Prepare a primer mix for the 283-bp internal positive control
(IPC) by combining 720 mL H2O, 40 mL 100 mM FDRA360,
 24 KIR Typing by Non-Sequencing Methods… 419

Table 2
Oligonucleotide primers for KIR typing by PCR-SSP (8)
(primers sequences reproduced with permission from
John Wiley and Sons)

Primer name Sequence

Fa–97b gtacgtcaccctcccatgatgta
Fi2–89 gtgtggtaggagccttag
Ftgc157b tcaggacaagcccttctgc
Fg510 aatgttggtcagatgtcag
Fa517 gttggtcagatgtcatgtttgaa
Fa546 tgcacagagaggggaagta
Fc551 agagaggggacgtttaacc
Fc567 cgacactttgcacctcac
Fg621 tctccatcagtcgcatgag
Fa621b tctccatcagtcgcatgaa
Fg624 catcggttccatgatgcg
Fg624b catcagttccatgatgcg
Ft624 ccatyggtcccatgatgct
Ftt624 tccatcggtcccatgatgtt
Fcon750 aaaccttctctctcagccca
Fat781 ggttcaggcaggagagaat
Ft803 cttgtcctgcagctcct
Fag843 atctatccagggaggggag
Fg864 catgaacgtaggctccg
Fma920 tttctgtgggccgtgcaa
Fcon1254 agaccctcaggaggtga
Rga250b ggacagggaccccatctttc
Rg287 gaaaacggtgtttcggaatac
Rc621 cctgccaggtcttgcg
Rcon621 ccctgcaaggtcttgca
Rcon662 ctgatagggggagtgagt
Rta669b gcygacaactcatagggta
Rdel674 gggagctgacaactgatg
Rcon682 ggtcactgggagctgac
(continued)
420 D. Ordóñez et al.

Table 2
(continued)

Primer name Sequence

Rt697 ccacgatgtccagggga
Rt854 gccctgcagagaacctaca
Rca877 ctggaatgttccgtkgatg
Ra925 gcatctgtaggttcctcct
Rc953 catagggtgagtcatggag
Rc962 gaccacacgcagggcag
Rdel967b gtcactgggggcttatag
Rt1375 caggagacaactttggatca
FDRA360 gaggtaactgtgctcacgaacagc
RDRA595 ggtccataccccagtgcttgagaag
RDRA633 cacgttctctgtagtctctggg

and 40 mL 100 mM RDRA595 (each primer will be at 5 mM).

This volume will be enough for preparing all primer mixes
1–15 described in Table 1. Label it IPC283.
4. For primer mix 16, you will use 50 mL of an oligonucleotide
pair for a 608 bp IPC. You can prepare it in advance (e.g.
45 mL H2O, 2.5 mL 100 mM FDRA360, and 2.5 mL 100 mM
RDRA633, or scaled-up volumes), and label it IPC608; or add
its components directly when you prepare primer mix 16.
5. Prepare primer mixes 1–16 according to Table 1. Store all
primer mixes and stocks at −20°C or below.

3.2. Prepare Sets A stock of ready-to-use primer mixes, pre-aliquoted into two
of Ready-to-Use 8-tube strips, will allow you to determine quickly the KIR geno-
PCR Primers in Strips types of any number DNA samples as soon as you receive them.
1. Prepare an appropriate number of 0.2 mL, 8-tube strips
(e.g. 24–48, see Note 6) and set them in 96-well trays (empty
tip cases are handy). Mark the left side of each strip with one
dot or line—this tube will hold reaction 1. Mark an identical
number of strips with two dots or lines—the labelled tube will
be reaction 9.
2. Thaw primer mixes 1–8, and homogenise them by inversion.
3. Using an electronic repetitive pipettor, dispense 4 mL primer
mix 1 in tube 1 of all strips. Follow with reactions 2–8.
4. Using an electronic repetitive pipettor, overlay the mixes with
5–7 mL Chill-out red Liquid Wax (see Note 7 and Fig. 1).
 24 KIR Typing by Non-Sequencing Methods… 421

5. Cap one of every eight strips, stack groups of eight strips, and
lay them into one or more boxes (again, empty cases of 1,000-
mL tip racks are handy) (Fig. 2 and video in http://www.
yotube.com/watch?v=8wU55yu-BgA). You can hold the strips

Fig. 1. A few microliters of liquid wax laid over the aliquoted primer mixes avoid their
evaporation during long-term storage in the freezer.

Fig. 2. Strips of primer mixes, overlaid with liquid wax, stacked for frozen storage.
422 D. Ordóñez et al.

together with small rubber bands, which will avoid accidental

loss of their frozen content during storage and manipulation.
Label the boxes “KIR typing, strip 1”, and store them frozen.
6. Repeat steps 2–5 for primer mixes 9–16.

3.3. Prepare PCR 1. If you purchase individual dNTPs, make a 20 mM mix (e.g.
Solutions 100 mL H2O, and 100 mL each 100 mM dATP, dCTP, dGTP,
and dTTP). Store everything at −20°C or below. Do not leave
nucleotide-containing solutions at room temperature for long
periods.
2. Prepare one or more vials of a 2×-concentrated PCR solution
(enough for 12 DNAs): add 923 mL H2O, 260 mL 10× Taq
polymerase buffer (Mg-free), 104 mL 50 mM MgCl2 (see
Note 8), and 13 mL 20 mM dNTPs. Mix well by inversion,
make 12 single-use, 100 mL aliquots, and store them at −20°C
or below.

3.4. Setup a Typing 1. Thaw the DNAs and a corresponding number of 2× PCR-
Assay solution aliquots, and strips with pre-dispensed primer mixes.
2. Add 20 mL DNA (containing 0.2–3.0 mg, see Note 9) and
1.2 mL Taq (5 U/mL) to a 100 mL aliquot of 2× PCR solution,
and mix thoroughly. Prepare as many similar mixes as DNAs
you have to analyse (six is a convenient number).
3. Using a repetitive pipettor, dispense 6 mL aliquots of this solu-
tion onto each primer mix, and cap the strips tightly. Make
sure that reagents have fallen to the tube bottoms, transfer
the strips to the thermocycler, and start cycling (denaturation at
95°C, 2 min; 10 cycles of 94°C, 10 s, and 65°C, 40 s; and 20
cycles of 94°C, 20 s, 61°C, 20 s, and 72°C, 30 s).
4. Prepare a 2% agarose gel in your preferred electrophoresis
buffer.
5. When thermocycling has finished, add 10 mL of a 2×-concen-
trated loading dye (see Note 10) to the wall of each tube using
an electronic repetitive pipettor, and let the dye fall to the bot-
tom. The same pipette tip can be used for all reactions.
6. Load 10 mL of each reaction in the agarose gel, using an
8-channel pipettor (see Notes 11–13 and video in http://
www.youtube.com/watch?v=8wU55yu-BgA). Leave an empty
well between strips 1 and 2, and load in it ~300 ng of a size
marker that covers the range 100–600 bp. Save the remaining
reaction volume for KIR2DS4 subtyping (see Note 14) or for
possible reanalysis.
7. Set the power supply to constant 200 V (or 9 V/cm if your
chamber is shorter than 20 cm) with no current limit, and let
your samples run 10–15 min (bromophenol blue will migrate
~1.5–2.0 cm).
 24 KIR Typing by Non-Sequencing Methods… 423

8. Stain the gel 15–20 min in ethidium bromide, or, preferably, a

non-toxic alternative (e.g. GelRed from Biotium or SYBR Safe
from Invitrogen), reveal by UV exposure, and photograph.

3.5. Quality Control 1. Contamination tests. Before you use your 2× PCR solution
and KIR-typing sets, you should verify that they are not con-
taminated with DNA. To that end, run one typing assay in
which you substitute 20 mL H2O for the DNA, and make sure
that you see no amplification bands (other than possible primer
multimers). If you prepare single-use aliquots of all reagents
(namely, the 2× PCR solution, and the primer-mix strips), you
only have to test periodically that your Taq polymerase is not
contaminated. To that end, you can perform a single-tube
PCR using the 283 bp IPC primers: these will detect contami-
nation with genomic DNA or with any product derived from
the 16 reactions of the KIR typing set.
2. Reference samples. It is advisable to control that the method is
working properly on reference DNAs with known KIR geno-
types (9). You can obtain reference DNAs from another lab, or
purchase them from a repository of well-characterised cell lines
(e.g. http://www.hpacultures.org.uk/collections/ecacc.jsp).
In addition, you can test your proficiency by participating in a
DNA exchange, such as those organised by UCLA (http://
www.hla.ucla.edu/cellDna.htm).

3.6. Interpretation 1. Acceptability of results.

Hints, Potential Errors (a) IPC. Each reaction should yield one or more PCR prod-
ucts. If one or more mixes yield no amplification product
of the correct size, those particular reactions must not be
interpreted (i.e. they are neither negative nor positive). An
IPC band will appear in most or all reactions: in mixes
1–15, it will migrate just below 300 bp (or above if you
prefer to see your gels with the wells in the bottom); and
at approximately 600 bp in mix 16. In the presence of a
strong specific band, the IPC band may disappear, which
shall not be considered a reaction failure, but a positive
result.
(b) KIR-gene-specific bands: size. All specific bands will have
smaller sizes than their IPCs in this method, which helps
avoid falsely negative results in partially degraded DNA
(8). A reaction will be considered positive when you see a
neat band of the correct size, which varies according to the
primer mix (see Table 3 and Note 15). If no bands other
than the IPC are seen, or if any other bands have the wrong
size (an indication that the PCR conditions need optimi-
zation), the reaction is negative. Small-size specific ampli-
cons (e.g. those derived from KIR2DS1, 96 bp, and
424 D. Ordóñez et al.

Table 3
Amplicon sizes (8)

Mix Gene Amplicon length (bp)

1 KIR2DL1 142
2 KIR2DL2 142
3 KIR2DL3 156
4 KIR3DL1 108/109
5 KIR3DL2 131
6 KIR2DS1 96
7 KIR2DS2 110
8 KIR2DS3 158
9 KIR2DS4 (canonical) 133
KIR2DS4 (exon 5-deleted) 111
10 KIR2DS5 147
11 KIR3DS1 107
12 KIR2DP1 141
13 KIR3DL3 196
KIR3DX1 88
14 KIR2DL4 131
15 KIR2DL5 147
16 KIR3DP1 (canonical) 279
KIR3DP1 (exon 2-deleted) 398
1–15 Internal positive control 283
16 Internal positive control 608

KIR3DX1, 88 bp) must not be confused with primer mul-

timers that migrate with just faster mobility.
(c) KIR-gene-specific bands: intensity. Specific amplicons
should be of greater or similar intensity than the IPC band;
weak amplicons of the correct size should not be seen.
Should such weak reactivities appear, they must not be
considered definitely positive or negative, and they are an
indication to repeat that test, and, in case they occur fre-
quently, to adjust the amplification conditions (reagents
stability and concentration, polymerase performance, ther-
malcycler calibration, etc.).
 24 KIR Typing by Non-Sequencing Methods… 425

2. Knowledge on the organisation of KIR genes will also help

troubleshoot your results (see Note 16):
(a) The genes KIR3DL3 and KIR3DX1 are seemingly present
in all humans.
(b) KIR3DL2, in the telomeric end of the KIR complex, and
two genes in its middle, KIR3DP1 and KIR2DL4, are also
very well conserved. Exceptions are: deletions of the KIR-
complex central region, and haplotypes carrying chimeras
of KIR3DL1 and KIR3DL2, which will lack the latter gene.
Both traits are more common in Black populations.
(c) Some pairs of genes are normally inherited as alleles of the
same conserved locus: KIR3DL1 and KIR3DS1;
KIR2DL3 and KIR2DL2; KIR2DS1 and KIR2DS4.
Therefore, the vast majority of haplotypes contain one or
both members of those pairs. However, the aforemen-
tioned deletions of the central KIR genes often affect
KIR3DL1/KIR3DS1; and haplotypes that contain a
KIR3DL1/KIR3DL2 chimera could lack the intervening
KIR2DS1/KIR2DS4 locus.
(d) Some genes are infrequently seen in the absence of certain
others: KIR2DS2 and KIR2DL2; KIR2DL1 and
KIR2DP1; KIR2DL3 and the latter two genes (which, in
turn, are not uncommon in the absence of KIR2DL3);
KIR3DL1 and KIR2DS4; KIR2DL5 and either KIR2DS3
or KIR2DS5; KIR3DS1 and the KIR2DL5-KIR2DS3/
S5-KIR2DS1 cluster. Again, exceptions to these associa-
tions, derived from recombinant haplotypes created by
unequal crossing-over events, are not rare in some
populations.

3.7. Other KIR Typing 1. Variations of PCR-SSP.

Methods (a) The PCR-SSP technique can discriminate between
sequences that differ by a single nucleotide substitution;
therefore, it can be used, not only for detecting presence
or absence of a KIR gene, but also for allelic typing (9).
(b) PCR-SSP typing comprises multiple amplification reactions,
and uses more DNA than methods based on a single PCR.
To minimise the use of DNA, whole genome amplification
can be performed before KIR typing by PCR-SSP (10).
(c) Multiplexing (i.e. including multiple primer pairs that will
produce distinguishable amplicons from two or more KIR
genes in the same reaction) also helps reduce reagent and
DNA expense by limiting the number of reactions, but it
needs extra optimisation to avoid preferential amplification
of certain genes. Our method includes one multiplexed
reaction (number 13, for KIR3DL3 and KIR3DX1), and
426 D. Ordóñez et al.

complete KIR typing methods based on this principle

have been reported (11–13).
(d) PCR-SSP can be adapted to real-time technology in order
to eliminate post-PCR processing and reduce DNA
expense (14).
(e) The PCR-SSP principle can also be applied to cDNA to
determine the KIR genes expressed by an NK- or T-cell
clone or subpopulation (7, 15). Lack of IPC bands in
RT-PCR-SSP techniques makes more difficult (in com-
parison with assays on genomic DNA) to detect possible
failure of individual reactions within an assay.
2. Probe hybridization of KIR amplicons has also been used to
determine KIR geno- and allotypes. Both the classical PCR-
SSOP (hybridization of labelled probes to immobilised PCR
products) and the fluorescent bead-attached probe formats
have been reported (16–18), and they facilitate the analysis of
high numbers of samples.
3. Other non-sequencing methods for analysis of KIR
polymorphism.
(a) High resolution melting analysis after real-time PCR can
discriminate between amplicons differing at a single nucle-
otide and has been used to type for KIR2DS4, KIR2DL4,
and KIR3DL2 alleles (19).
(b) Differences in the spatial conformation (thus, the electro-
phoretic mobility) of slightly different sequences are the
basis for sequence-specific conformational polymorphism
analysis (SSCP). This technique, in combination with cap-
illary electrophoresis in an automated DNA sequencer has
been applied to allelic typing of KIR2DL4 alleles (20).
(c) MALDI-TOF mass spectrometry is a powerful, high-
throughput technique that exploits small differences in
the molecular mass to discriminate between similar DNA
sequences. Mass spectrometry has been used for both
generic and allelic typing of KIR genes, and for identifying
novel KIR alleles (21).

4. Notes

1. Appropriate electrophoresis systems can be obtained from a

few providers (e.g. 96Q High Throughput PCR Sub System,
from Denville Scientific; horizontal systems from CBS Scientific;
SUB25C from Hoefer; and Sub-Cell Model 192 from Bio-Rad).
A power supply should provide constant voltage of 200 V and
support 100–200 mA (current varies with the electrophoresis
 24 KIR Typing by Non-Sequencing Methods… 427

chamber; you will need twice as much current—not voltage—if

you plan to run two chambers in parallel).
2. Make sure that the gel will fit in your UV-transilluminator
screen.
3. Do not use higher volume multichannel pipettors (e.g. 5.0–
50.0 mL), they will spill your PCR products.
4. We use BioTaq from Bioline (UK), or Ecotaq from a local
provider, both with an NH4-based buffer (10×: 160 mM
(NH4)2SO4, 670 mM Tris–HCl (pH 8.8 at 25°C), 0.1% Tween-
20). For other brands and 10× buffers you may need to adjust
the amount of Taq and MgCl2 concentration (see Note 3).
With some KCl-buffer formulations, the IPC band may be
amplified preferentially over the specific bands (which, in
extreme situations, will result in false negatives).
5. Substitution of a new primer (Fy803, cttgtcctgmagctccy) for
Ft803 in mix 8 would suffice for detecting all KIR2DS3 alleles.
Adding a new primer (Fa517b, gttggtcagatgtcatgttttaa) to mix
1 (besides Fa517 and Rc621) should ensure detection of
2DL1*013N (you may not want to detect a null KIR2DL1
allele in certain cases, though). Detecting the four new other
alleles would require substantial modifications to the method
design.
6. Primer-strips preparation is a tedious task; once you set to it,
you may want to prepare a generous amount of strips. We
strongly discourage preparing a primer set for just one or a few
DNAs: in the long term, you will waste an enormous amount
of time and pipette tips, and your primer stocks will suffer an
unnecessarily high number of freeze-thaw cycles, favouring
both degradation and contamination. Preparing 48 typing sets
from primer mixes will take ~30–40 min.
7. You need not change the tip if you dispense the wax on the
tube wall. The wax (liquid at room temperature and solid
below 10°C) forms a red meniscus on top of the primer mixes
(Fig. 1), and minimises evaporation during long-term storage
in the freezer. If you plan to use all your strips in less than 2–4
weeks, you need not use wax.
8. The 2× concentrated PCR solution serves several purposes: (a)
it avoids repeated reagent thawing and freezing, (b) reduces
contaminations, and (c) increases reproducibility by minimis-
ing pipetting steps and errors. If your 10× Taq polymerase buf-
fer contains 15 mM MgCl2, add only 26 mL 50 mM MgCl2
(instead of 104 mL) and increase water to 1,001 mL. Volumes
can be scaled up using devices appropriate for the volumes to
be pipetted.
9. We adjust DNA concentration at ~50–150 ng/mL, but the
method can work with good quality DNA as diluted as 10 ng/mL.
428 D. Ordóñez et al.

Within that DNA concentration range (10–150 ng/mL) you

need not adjust the 20 mL DNA volume in the mix. We have
analysed DNAs isolated from peripheral blood, spleen and lymph
node lymphocytes, paraffin-embedded tissue, and mouth wash,
using salting-out methods, DNAzol (MRC), and a Maxwell-16
machine (Promega).
10. Prepare a 2×-concentrated loading dye by diluting a 6×- or
10×-concentrated loading dye with an appropriate volume of
H2O or electrophoresis buffer (e.g. 0.67 mL of a 6× loading
dye, and 1.33 mL H2O; 160 mL of this solution are needed
for each KIR typing). Increasing the reaction volume right
before electrophoresis facilitates aspiration of the sample with
the multichannel pipettor, and loading it into the gel wells.
If you did not use wax, just add 1.0–1.5 mL of the 6×- or
10×-concentrated loading dye to each tube, and load 5 mL into
each well.
11. You need not remove the red wax, just aspirate the reaction
underneath. The electrophoresis buffer will get somewhat
oleaginous, and you will have to change it periodically, but it
does not affect migration. Should you feel uncomfortable in the
beginning with pipetting below the wax meniscus, reduce the
amount of wax to 3 mL in Subheading 3.2, step 4, harden
the wax by putting the strips on ice, and poke the pipette tips
through the thin hardened wax film.
12. Make sure that the ends of the tips inserted in the multichan-
nel pipettor are well aligned—any misaligned tip will add
difficulty to the loading process.
13. You need not change the pipette tips after each loading step.
Instead, wash them by pipetting two or three times in the elec-
trophoresis buffer before aspirating the next strip. The degree
of contamination between samples with this procedure is unde-
tectable (unless you want to re-amplify a PCR product for any
reason), and you will save a considerable amount of time (and
pipette tips).
14. Canonical and exon-5-deletion carrying KIR2DS4 alleles can
be distinguished from each other by their different amplicon
sizes (133 bp vs. 111 bp, respectively). This is easiest if you run
in parallel reaction 9 of multiple DNAs, which facilitates com-
paring amplicon sizes. To this end, load the remaining 5 mL of
reaction 9 in a 2.0–2.5% agarose gel and allow a longer run
(e.g. 25 min at 200 V) before staining and photographing.
This can also be done in a smaller electrophoresis chamber at
lower voltage.
15. Amplicon sizes in reactions 9 and 16 allow for discrimination
between major structural allotypes of KIR2DS4 and KIR3DP1,
respectively (6).
 24 KIR Typing by Non-Sequencing Methods… 429

16. None of the following indications on KIR-gene organisation is

an absolute rule, and you will probably meet occasional excep-
tions when you type multiple individuals. However, you should
pay attention to any exceptions because they may indicate either
that you have found some interesting genotypes, or that there is
a typing error. In either case, some extra work is warranted before
accepting unusual genotypes (e.g. revising the results, repeating
the test with the same or an alternative method, checking reagents
integrity, optimising the typing procedure, et cetera). Our initial
approach to deal with unusual KIR genotypes is, first, to repeat
the test, and, second, to use a previously published PCR-SSP
method that uses different primer combinations (7).

Acknowledgements

This work was supported by grant SAF2010-22153-C03-03 from

the Spanish Ministerio de Ciencia e Innovación. DO was supported
by a grant from “Fundación LAIR”. NGL is supported by a grant
from Instituto de Salud Carlos III (CP09/182). The authors have
no conflicts of interest.

References

1. Wagtmann N et al (1995) Molecular clones of 7. Uhrberg M et al (1997) Human diversity in

the p58 NK cell receptor reveal immunoglobu- killer cell inhibitory receptor genes. Immunity
lin-related molecules with diversity in both the 7:753–763
extra- and intracellular domains. Immunity 8. Vilches C et al (2007) Facilitation of KIR geno-
2:439–449 typing by a PCR-SSP method that amplifies
2. Vilches C, Parham P (2002) KIR: diverse, rap- short DNA fragments. Tissue Antigens
idly evolving receptors of innate and adaptive 70:415–422
immunity. Annu Rev Immunol 20:217–251 9. Gómez-Lozano N, Vilches C (2002)
3. Gómez-Lozano N et al (2005) The silent Genotyping of human killer-cell immunoglob-
KIR3DP1 gene (CD158c) is transcribed and ulin-like receptor genes by polymerase chain
might encode a secreted receptor in a minority reaction with sequence-specific primers: an
of humans, in whom the KIR3DP1, KIR2DL4 update. Tissue Antigens 59:184–193
and KIR3DL1/KIR3DS1 genes are duplicated. 10. Chainonthee W et al (2010) Improved KIR
Eur J Immunol 35:16–24 gene and HLA-C KIR ligand sequence-specific
4. Ordonez D et al (2008) Duplication, mutation primer polymerase chain reaction genotyping
and recombination of the human orphan gene using whole genome amplification. Tissue
KIR2DS3 contribute to the diversity of KIR Antigens 76:135–143
haplotypes. Genes Immun 9:431–437 11. Sun JY et al (2004) Development of a multiplex
5. Velardi A et al (2009) Natural killer cell allorec- PCR-SSP method for Killer-cell immunoglobu-
ognition of missing self in allogeneic hematopoi- lin-like receptor genotyping. Tissue Antigens
etic transplantation: a tool for immunotherapy 64:462–468
of leukemia. Curr Opin Immunol 21: 12. Ashouri E et al (2009) A novel duplex SSP-
525–530 PCR typing method for KIR gene profiling.
6. Cooley S et al (2010) Donor selection for natural Tissue Antigens 74:62–67
killer cell receptor genes leads to superior survival 13. Kulkarni S, Martin MP, Carrington M (2010)
after unrelated transplantation for acute myelog- KIR genotyping by multiplex PCR-SSP.
enous leukemia. Blood 116:2411–2419 Methods 612:365–375
430 D. Ordóñez et al.

14. Alves LG, Rajalingam R, Canavez F (2009) A 18. Nong T et al (2007) KIR genotyping by reverse
novel real-time PCR method for KIR genotyp- sequence-specific oligonucleotide methodology.
ing. Tissue Antigens 73:188–191 Tissue Antigens 69(suppl 1):92–95
15. Thompson A et al (2006) An improved RT-PCR 19. Gonzalez A et al (2009) Killer cell immuno-
method for the detection of killer-cell immuno- globulin-like receptor allele discrimination by
globulin-like receptor (KIR) transcripts. high-resolution melting. Hum Immunol 70:
Immunogenetics 58:865–872 858–863
16. Crum KA et al (2000) Development of a PCR- 20. Witt CS, Martin A, Christiansen FT (2000)
SSOP approach capable of defining the natural Detection of KIR2DL4 alleles by sequencing and
killer cell inhibitory receptor (KIR) gene SSCP reveals a common allele with a shortened
sequence repertoires. Tissue Antigens 56: cytoplasmic tail. Tissue Antigens 56:248–257
313–326 21. Houtchens KA et al (2007) High-throughput
17. Halfpenny IA et al (2004) Investigation of killer cell immunoglobulin-like receptor geno-
killer cell immunoglobulin-like receptor gene typing by MALDI-TOF mass spectrometry
diversity: IV. KIR3DL1/S1. Hum Immunol with discovery of novel alleles. Immunogenetics
65:602–612 59:525–537
 Chapter 25

Killer Cell Immunoglobulin-Like Receptors (KIR ) Typing

by DNA Sequencing
Lihua Hou, Minghua Chen, Noriko Steiner, Kanthi Kariyawasam,
Jennifer Ng, and Carolyn K. Hurley

Abstract
DNA sequencing is a powerful technique for identifying allelic variation within the natural killer cell
immunoglobulin-like receptor genes. Because of the relatively large size of the KIR genes, each locus is
amplified in two or more overlapping segments. Sanger sequencing of each gene from a preparation
containing one or two alleles yields a sequence that is used to identify the alleles by comparison with a
reference database.

Key words: Natural killer cell, Killer immunoglobulin-like receptor, DNA sequencing, Alleles

1. Introduction

The human killer cell immunoglobulin-like receptors (KIR) are

encoded by 14 genes: KIR2DL1, KIR2DL2, KIR2DL3, KIR2DL4,
KIR2DL5, KIR2DS1, KIR2DS2, KIR2DS3, KIR2DS4, KIR2DS5,
KIR3DL1, KIR3DL2, KIR3DL3, KIR3DS1 (1). These genes
likely arose from gene duplications and unequal crossing over since
they share extensive sequence homology. Each gene is divided into
8–9 exons that encode the signal peptide, two or three extracellular
domains, stem, transmembrane region, and cytoplasmic tail. The
genes are about 9–16 kb in length. The number of KIR loci pre-
sent varies among individuals. For example, some individuals might
carry only seven of the 14 KIR genes while other individuals might
carry 12 of the 14 KIR genes. A clear understanding of the KIR
gene system will be important to understand the basis for the strat-
egies described in this chapter and to correctly interpret the
sequencing results.

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_25, © Springer Science+Business Media New York 2012

431
432 L. Hou et al.

1.1. Overview This protocol describes the amplification and sequencing of each
of Methods KIR gene from genomic DNA. The polymerase chain reaction is
used to obtain two or more overlapping amplicons covering all or
most of each gene (Fig. 1). The nucleotide sequences of the exons
carried by each amplicon are determined using Sanger sequencing
(2) with primers that anneal in the introns and flank each exon.
Both alleles of a locus, if present, are sequenced concurrently and
the allele assignments made by comparison to a KIR reference data-
base. Some loci (KIR2DL1, KIR2DL2, KIR2DL3, KIR2DL5,
KIR2DS4) require special steps in order to obtain unambiguous
sequences as described in Table 1. An initial survey of the KIR genes
present or absent in the sample using sequence-specific priming will
provide the information necessary to determine the additional steps
required to obtain allele assignments.

1.2. Use of Methods The impact of genetic variation in the KIR gene complex on the
in Clinical Practice functional activity of NK cells is yet to be fully understood. The
presence of specific KIR genes has been associated with susceptibil-
ity or resistance to infectious and autoimmune diseases and to
malignancy (1, 3). In hematopoietic progenitor cell transplanta-
tion for acute myelogenous leukemia, a decreased frequency of
relapse and infection has been noted in transplants with donors
carrying haplotypes with increased numbers of activating KIR

Fig. 1. Amplification of overlapping amplicons covering the KIR2DL1 coding region sequence. KIR genes have eight to nine
exons. PCR amplification primers are designed to generate two or more overlapping amplicons. The figure shows the
three amplicons, A, B, and A2, that cover the coding sequence of the KIR2DL1 gene. If the sample does not contain KIR2DS1,
the laboratory needs only to generate the A and B amplicons for sequencing as described in Table 1. Amplicon A will allow the
sequence determination from nucleotide 11 of exon 1 through nucleotide 632 of exon 5; amplicon B will cover nucleotide 332
in exon 4 through the last nucleotide of exon 9. If the sample contains the KIR2DS1 gene, the laboratory will perform instead
three amplifications generating amplicons A, B, and A2. The A2 amplicon will contain only KIR2DL1 and will provide the
sequence covering nucleotide 1 in exon 1 through nucleotide 330 of exon 4. The A amplicon which contains DNA from both
KIR2DL1 and KIR2DS1 genes will provide sequence information covering the region where the A2 antisense and the B sense
primers anneal, i.e., around nucleotide 331 in exon 4. The small arrows under the exons denote the positions of sequencing
primers that anneal in the introns and that provide the sequence of both sense and antisense DNA strands for each exon.
Tables 2 and 3 list the amplification and sequencing primers for all the KIR loci and describe their annealing sites.
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 433

Table 1
Summary of amplification protocols for 15 KIR Locia

Locus Specific amplification or allele isolation protocol required

KIR2DL1 Amplicon A—General PCR in Subheading 3.2 with genomic DNA. If KIR2DS1 is
present, it will coamplify with this amplicon. When KIR2DS1 is present, amplicon A
should be characterized to obtain DNA sequence in the area where the antisense A2
and sense B primers anneal (i.e., in region of nucleotide 331)
Amplicon A2—General PCR in Subheading 3.2 with genomic DNA. In cells carrying
KIR2DS1, coamplification of KIR2DS1 is eliminated in this additional reaction. This
amplification is not required if the cell does not carry KIR2DS1
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR2DL2 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—General PCR in Subheading 3.2 with genomic DNA with the following
exception. For those cells shown to carry KIR2DL1 or KIR2DS2 or both
KIR2DL2 and KIR2DL3, use haplotype-specific extraction with probe KIR2DL2-
999T as described in Subheading 3.4 prior to general PCR in Subheading 3.2 to isolate
KIR2DL2
Amplicon C— General PCR in Subheading 3.2 with genomic DNA with the following
exception. For those cells shown to carry KIR2DL1 or KIR2DS2 or both
KIR2DL2 and KIR2DL3, use haplotype-specific extraction with probe KIR2DL2-
999T as described in Subheading 3.4 prior to general PCR in Subheading 3.2
Amplicon D—If KIR2DL2 and KIR2DL3 are both present in the cell, perform nested
PCR with these primers using the amplicon B template to eliminate the highly
homologous KIR2DL3 gene as described in Subheading 3.3
KIR2DL3 Amplicon A—If the cell is KIR2DL2 negative, follow the general PCR protocol in
Subheading 3.2 beginning with Bc1I digested genomic DNA as described in
Subheading 3.5. Cleavage of KIR2DP1 with the restriction enzyme BclI eliminates its
coamplification. If the cell is KIR2DL2 positive, follow the general PCR protocol in
Subheading 3.2 beginning with haplotype-specific extraction with the KIR2DL3-
1316T probe as described in Subheading 3.4
Amplicon B1—If the cell is KIR2DL2 negative, follow the general PCR protocol in
Subheading 3.2 beginning with the Bc1I digested genomic DNA as described in
Subheading 3.5. Cleavage of KIR2DP1 with the restriction enzyme BclI eliminates its
coamplification. If the cell is KIR2DL2 positive, do not prepare the B1 amplicon but
instead use the amplicon B2 primers described below
Amplicon B2—If the cell is KIR2DL2 positive, use this primer pair and follow the
general PCR protocol in Subheading 3.2 beginning with haplotype-specific extraction
with the KIR2DL3-1316T probe as described in Subheading 3.4. If the cell is
KIR2DL2 negative, do not prepare the B2 amplicon but instead use the amplicon B1
primers described above
Amplicon C1—General PCR in Subheading 3.2 with genomic DNA
Amplicon C2—General PCR in Subheading 3.2 with genomic DNA. Together, the
information provided by the C1 and C2 amplicons produces more robust sequence
results
Amplicon D—This is a nested PCR of amplicon A required to clarify the sequence in this
region
KIR2DL4 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon A2—General PCR in Subheading 3.2 with genomic DNA. This amplicon will
allow characterization of exon 1
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
(continued)
 Table 1
(continued)

Locus Specific amplification or allele isolation protocol required

KIR2DL5 Amplicon A—General PCR in Subheading 3.2 with genomic DNA. The A amplicon
includes 254 bp of the 5¢ upstream region
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
Amplicon A*001+—Use this primer pair with genomic DNA to clarify results for cells
that carry more than two alleles of KIR2DL5
Amplicon B*002+—Use this primer pair with genomic DNA to clarify results for cells
that carry more than two alleles of KIR2DL5
KIR2DS1 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR2DS2 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR2DS3 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR2DS4 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
Amplicon C—In those cells with both full and deletion alleles, an exon 5 nested PCR is
performed using amplicon B as a template (see Subheading 3.3). Cloning as described
in Subheading 3.6 is used to separate alleles for sequencing in these samples
KIR2DS5 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR3DL1 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—Perform the long template PCR protocol described in Subheading 3.7 with
genomic DNA
Amplicon M—General PCR in Subheading 3.2 with genomic DNA. This amplicon
overlaps the sequences of Amplicon A and Amplicon B
KIR3DL2 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon A2— General PCR in Subheading 3.2 with genomic DNA. This amplicon will
allow characterization of exon 1
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR3DL3 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon A2—General PCR in Subheading 3.2 with genomic DNA. This amplicon will
allow characterization of exon 1
Amplicon B—General PCR in Subheading 3.2 with genomic DNA
KIR3DS1 Amplicon A—General PCR in Subheading 3.2 with genomic DNA
Amplicon B—Perform the long template PCR protocol described in Subheading 3.7
with genomic DNA
a
Samples will differ in their requirement for the strategies listed in this table depending on the KIR genes present in
each sample. Once the KIR genes present and absent are evaluated by an initial assay (as described in Chap. 24), the
laboratory should use this table to select the methods required to obtain DNA for sequencing. For example, to obtain
the allele assignments of KIR2DL1: If a cell carries KIR2DL1 and not KIR2DS1, two PCR amplifications are per-
formed to yield KIR2DL1 amplicon A (yielding the sequence of nucleotide 10 through nucleotide 632) and KIR2DL1
amplicon B (nucleotide 332 through the last nucleotide of exon 9). These two overlapping amplicons are subsequently
sequenced to identify the KIR2DL1 alleles. However, if the cell carries both KIR2DL1 and KIR2DS1, amplicon A will
include both KIR2DL1 and KIR2DS1 which makes it difficult to interpret the sequence data. In this case, it is necessary
to perform an additional amplification of KIR2DL1 generating amplicon A2 (nucleotide 1 through nucleotide 330)
which does not include KIR2DS1. Because the antisense primer generating amplicon A2 anneals at nucleotide 331
which is the annealing site of the sense primer for amplicon B, the A2 amplicon does not provide a clear assessment of
the sequence in the region of nucleotide 331. This information is provided by amplicon A
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 435

genes (4, 5). Less is known about the impact of KIR allelic
polymorphism on the immune response. Allelic variation alters the
level of protein expression and the affinity of ligand binding as
demonstrated for KIR2DL2/KIR2DL3 (6) and KIR3DL1 (7, 8).
For example, in HIV infection, allotypic variation of KIR3DL1
influences disease progression and levels of the pathogen in plasma
(9). Thus, as we learn more about their impact, identification of
KIR alleles may be used to predict the response of an individual to
a disease or to therapy and to select optimal stem cell donors for
patients with some malignancies.

2. Materials

Use reagent grade water (e.g., UltraPure™ distilled water,

Invitrogen, Carlsbad, CA, USA) unless noted. Storage conditions
of commercial reagents are indicated by the vendor.

2.1. DNA Preparation 1. Whole blood drawn into a standard blood tube containing the
anti-coagulant acid citrate dextrose (ACD) (see Note 1).
2. QIAamp® DNA Blood Mini Kit (QIAGEN, Valencia, CA,
USA): The kit contains buffers AL, AW1, AW2, protease
and solvent for protease, spin columns, collection tubes, and
instruction manual. The buffers in the kit, AW1 and AW2, are
provided as concentrates. When opening a new bottle, add
the appropriate amount of 96–100% ethanol (as written on the
label). To reconstitute the protease, add the supplied solvent
to the protease powder and invert the bottle several times to
mix. Store for 2 months at 4°C after preparation.
3. 96–100% ethanol.
4. Phosphate buffered saline (PBS).
5. 1.5 mL microcentrifuge tubes.
6. Pipettor (5–200 mL) and tips.
7. Heat block or water bath at 56°C.
8. Vortex mixer.
9. Centrifuge capable of holding 1.5 mL tubes with a maximum
speed of 20,000 × g (14,000 rpm).

2.2. Polymerase Chain 1. Genomic DNA prepared as described in Subheading 3.1.

Reaction 2. Positive and negative control genomic DNA (National Marrow
Donor Program Cell Repository, Minneapolis, MN, USA;
http://www.cibmtr.org/samples/) (see Note 2).
3. Taq polymerase and buffer: Platinum Taq DNA Polymerase
High Fidelity 5 units/mL with 10× High Fidelity PCR Buffer
(Invitrogen, Carlsbad, CA, USA).
436 L. Hou et al.

4. 50 mM MgSO4 (Invitrogen) according to Table 2.

5. 10 mM dNTP mixture (Roche, Mannheim, Germany).
6. KIR locus PCR primers: 10 mM of each oligonucleotide primer
in water, store at −20°C. Table 1 describes the primer sets
needed based on the presence or absence of specific KIR genes
in the sample. Primers are listed in Table 2 (see Note 3).
7. Dimethyl sulfoxide (DMSO) (Sigma-Aldrich, St.Louis, MO).
8. 5 M betaine solution (Sigma-Aldrich).
9. Reagent grade water.
10. 1-kb DNA ladder (e.g., Tracklt™1Kb Plus DNA ladder,
Invitrogen) (see Note 4).
11. Agarose (e.g., UltraPure™ Agarose, Invitrogen).
12. 10× TBE buffer (e.g., UltraPure™ 10× TBE buffer, Invitrogen)
diluted with deionized water at an operational resistivity of
18.2 MW cm−1 at 25°C to 1×.
13. Ethidium bromide solution (10 mg/mL) (Invitrogen) (see
Note 5).
14. 5× sucrose cresol (0.04% cresol red in 30% sucrose) gel loading
solution.
15. Agencourt AMPure kit (Beckman Coulter, Beverly, MA, USA).
16. 70% ethanol in water (e.g., Warner-Graham Company,
Cockeysville, MD, USA).
17. 1.5-mL sterile disposable tubes (Fisher Scientific, Dallas,
TX, USA).
18. Semi-skirted PCR tray (Fisher Scientific, Dallas, TX, USA).
19. Tape seals (One Lambda, Canoga Park, CA, USA).
20. Single channel and multichannel (8 or 12 channel) pipettors
(0.5–200 mL) and tips.
21. Thermal cycler (e.g., model 2720, Applied Biosytems, Foster
City, CA, USA).
22. Vortex mixer.
23. Flat bed slab gel unit (tray 11.9 cm (length) × 11.5 cm (width))
and power supply (e.g., RunOne™ Electrophoresis Unit, Embi
Tec, San Diego, CA, USA).
24. UV transilluminator.
25. Gel photography system.
26. Agencourt SPRIPlate 96R magnet plate (Beckman Coulter).
27. Centrifuge capable of holding 1.5 mL tubes and plates with a
maximum speed of 20,000 × g (14,000 rpm) (e.g., model 5424
(for tubes) and model 5804 (for plates with A-2-deep well
plate rotor), Eppendorf, Hauppauge, NY, USA).
Table 2
KIR locus-specific polymerase chain reaction amplification primersa and conditions

Annealing
sites: sense/ Amplicon PCR reaction PCR reaction components
antisenseb size (bp) conditions (50 mL)

Annealing
temp
(°C)—initial/ DMSO/
secondary Extension MgSO4 Betaine
KIR locus Amplicon Sense primer Antisense primer cycles time (min) (mL) (mL) Taq
KIR2DL1 A TGTAAAACGACGGC CAAGCAGTGGG 10T/633G 5,605 64/61 5 2.0 2.5/− High
CAGTGGCAGCAC TCACTTGAC Fidelity
CATGTCGCTCT
A2 ATAACATCCTG GGGTCACTGGG 5UTR/331G 3,825 66/64 5 1.5 3.0/− High
TGCGCTGCT AGCTGACAC Fidelity
B ACTCACTCCC TGTTGACTCCC 331G/3UTR 10,282 62/59 10 2.0 −/− High
CCTATCAGG TAGAAGACG Fidelity

KIR2DL2 A TCTCAGCACA GCCCTGCAGA 5UTR/505T 5,382 62/58 7 2.0 2.0/− High

GACAGCACC GAACCTACA Fidelity
B CCATGATGGG TCAATGCCTGCATCG 246A/IN6 5,348 60/57 5 3.5 −/10 High
GTCTCCAAA AAGGTTTCT Fidelity
C TCACCCACTGA TGTTGACTCCC 708T/3UTR 5,228 62/58 7 2.0 2.0/− High
ACCAAGCTCT TAGAAGACG Fidelity
D AATGCCTCTTCT CTCTCCTCTGGGTC 375A/IN5 568 62/57 1.5 - −/10 Taq/10×
CCTCCAGGTCTA TCTCCTGACCG Nested Ex5 PCR
buffer
with
MgCl2
(continued)
Table 2
(continued)

Annealing
sites: sense/ Amplicon PCR reaction PCR reaction components
antisenseb size (bp) conditions (50 mL)

Annealing
temp
(°C)—initial/ DMSO/
secondary Extension MgSO4 Betaine
KIR locus Amplicon Sense primer Antisense primer cycles time (min) (mL) (mL) Taq
KIR2DL3 A TGTAAAACGACGGC GCCCTGCAGAG 10A/505C 5,385 62/58 5 3.0 −/10 High
CAGTGGCAGCAC AACCTACG Fidelity
CATGTCGCTCA
B1 GTTCTGTTACTC CTCTCCTCTG 325T/IN5 2,131 62/58 5 3.0 −/10 High
ACTCCCCCT GGTCTCTCC Fidelity
TGACCG
B2 CGTTCTGCACAG 194A/IN5 2,262 62/58 5 3.0 −/10 High
AGAAGGGAAc Fidelity
C1 TCAAGACAGTGGG CTTCGTGAGAC IN6/809G 3,344 62/58 5 3.0 −/10 High
CGTCACATACA TTACTTT TTT Fidelity
TGTTGC
C2 ACACCTGCATGT GCAGGAGACAAC 746G/1024T 879 62/58 5 3.0 −/10 High
TCTGATTGG TTTGGATCA Fidelity
D AGCAAGGGGAAGCC CCAATGACAA IN2/IN4 419 62/57 1.5 - −/10 Taq/10×
TCACTCATTC TGAGAATG PCR
Nested Ex4
buffer
with
MgCl2
KIR2DL4 A CACCCACGGTC CCCTTTCSCTG 28C/IN6 5,378 64/57 6 2.0 2.0/− High
ATCATCC TTGGAGTGT Fidelity
A2 TCCTGGCAGCAG GGAAAGAGCC 5UTR/581G 2,564 64/57 5 2.0 2.0/− High
AAGCTGCACC GAAGCATC Fidelity
B CATGTTCTAGG TGGGCTAAGCA 666T/3UTR 5,420 64/57 6 2.0 2.0/− High
AAACCCTTCT AAGGAGTGT Fidelity
KIR2DL5 A ATCTTGTGTTC TCATAGGGTGA 5UTR/589C 3,274 64/62 5 2.0 2.0/− High
GGGAGGTTG GTCATGGAG Fidelity
B GAGGGGAGGG GGAAGAGCGAT 491C/3UTR 6,193 64/62 7 2.0 2.0/− High
CCCATGAACC CCCCTAAGA Fidelity
A*001+ CTCCCGTGATGTGG TCATAGGGTG 5UTR/589C 3,109 64/62 5 2.0 2.0/− High
TCAACATGTAAA AGTCATGGAG Fidelity
B*002+ CTCCCATGATGTA TCATAGGGTG 5UTR/589C 3,109 64/62 5 2.0 2.0/− High
GTCAACATGTAAG AGTCATGGAG Fidelity
KIR2DS1 A GGCAGCACCAT GCATCTGTAGG 10A/576T 5,540 64/60 7 1.5 1.0/− High
GTCGCTCA TCCCTCCA Fidelity
B TCTCCATCAGT GGGTGTCTTG 272R/3UTR 10,227 64/60 10 2.0 1.0/− High
CGCATGAR GGCCTCTC Fidelity
KIR2DS2 A ATCCTGTGCGCTG CACGCTCTCT 5UTR/418T 5,239 62/58 7 1.5 2.0/− High
CTGAGCTGAG CCTGCCAA Fidelity
B CTTCTGCACAGAGA TTATGCGTATGACA 197A/893A 10,253 62/58 10 1.5 2.0/− High
GGGGAAGTA CCTCCTGAT Fidelity
(continued)
Table 2
(continued)

Annealing
sites: sense/ Amplicon PCR reaction PCR reaction components
antisenseb size (bp) conditions (50 mL)

Annealing
temp
(°C)—initial/ DMSO/
secondary Extension MgSO4 Betaine
KIR locus Amplicon Sense primer Antisense primer cycles time (min) (mL) (mL) Taq
KIR2DS3 A ATCCTGTGCGCTGC GCATCTGTAG 5UTR/576A 5,919 64/61 7 2.0 −/− High
TGAGCTGAG GTTCCTCCT Fidelity
B GACATGTAC TTATGCGTATGAC 485C/888G 8,427 60/57 10 2.0 −/− High
CATCTATCCAC ACCTCCTGA Fidelity
TGGTCC
KIR2DS4 A CATGTCGCTCA ACACTCTCACCT 20T/360G 5,122 64/58 7 2.0 −/− High
TGGTCATCAT ATGATCACC Fidelity
B ATCCTGCAA TTATGCGTATGACAC 153G/893A 10,299 64/58 10 1.5 −/− High
TGTTGGTCG CTCCTGAT Fidelity
C CGCAGTGACCC GTGACGGAAACAA 360G/642T 1,875 62/57 1.5 - −/10 Taq/10×
TCTGGACATGc GCAGTGGA PCR
Nested Ex 5
buffer
with
MgCl2
KIR2DS5 A CCATCATGATCTT CCTCCGTGGGTG 35C/563A 4,541 62/58 5 2.0 −/− High
TCTTTCCAGC GCAGGGT Fidelity
B CATTGATGGGGT TTATGCGTATGACACC 248G/888G 10,188 62/58 10 2.0 −/− High
CTCCAAGGG TCCTGATGGTCC Fidelity
 KIR3DL1 A TGTCKRCACC TAGGTCCCTGC 5UTR/560T 3,454 60/57 5 1.5 2.0/− High
GGCAGCACC AAGGGCAA Fidelity
B CCATCGGTCC GACAACTTTGGA 560T/1303Y 10,365 60/57 11 - −/− Expand
CATGATGCT TCTGGGCTY Long/
Buffer 3
M CAARCCCTTCCT GAGAGAGAAG 100T/659C 3,265 60/57 5 1.5 2.0/− High
GTCTGCCT GTTTCTCATATG Fidelity
KIR3DL2 A GTCGTCAGCA TGCATCCAAGG 30C/IN6 8,706 60/57 8 1.5 2.0/− High
TGGCGTGC CTTCCACC Fidelity
A2 TGTCTGCACCG GACCACACG 5UTR/898C 5,421 60/57 5 2.0 −/10 High
GCAGCACC CAGGGCAG Fidelity
B TCACATCTC GGCTGTTGTC IN5/1362T 7,693 60/57 8 1.5 2.0/− High
TCCTGTCCCG TCCCTAGAAA Fidelity
KIR3DL3 A TTTCCAGGGTTCT TGACCCTCAG 49G/799A 4,415 62/60 5 3.0 2.0/− High
TCTTGCTGG CACYGCAGT Fidelity
A2 TGTCTGCACCG CCGACAACTC 5UTR/605T 3,361 62/60 5 3.0 −/10 High
GCAGCACC ATAGGGTA Fidelity
B CCCGGAGCTTG AGAAGACAAC 756T/3UTR 6,569 58/54 7 3.0 −/− High
TTTGACATT TTTGGATCTGC Fidelity
KIR3DS1 A TGTCKRCACCG CTGTGACCAT 5UTR/A337 2,116 60/57 3 1.0 1.5/− High
GCAGCACC GATCACCAT Fidelity
B GGCAGAATAT AGAGCGATGCC 235G/3UTR 12,324 60/57 11 - −/− Expand
TCCAGGAGG CTAAGATGA Long/
Buffer 3
a
Some of the primers have been previously described (19–23)
b
UTR, untranslated region and/or other 5¢ or 3¢ noncoding sequences; IN, intron. The designations such as 10T/633G indicate the nucleotide at the annealing site of the 3¢ end of the
sense/antisense primers. Position 1 is defined as the first nucleotide of the ATG codon in exon 1 according to the IPD/KIR database (http://www.ebi.ac.uk/ipd/kir/). The numbering of
KIR2DS4 is based on an allele that does not contain the deletion
c
Primer sequence is not identical to KIR gene sequence; a substitution was added to avoid the primer from self-annealing
442 L. Hou et al.

2.3. Nested PCR for 1. AMPure-purified amplicons: KIR2DL2 amplicon B, KIR2DL3

KIR2DL2 Amplicon amplicon A, and KIR2DS4 amplicon B. Table 1 describes the
B, KIR2DL3 Amplicon use of nested PCR to either isolate the product of a specific
A, and KIR2DS4 gene or to clarify the sequence in a specific area.
Amplicon B 2. Taq DNA Polymerase 5 units/mL (Roche, Mannheim,
Germany) with 10× PCR Buffer with MgCl2 (Roche).
3. 10 mM dNTP mixture (Roche).
4. KIR locus PCR primer solutions for nested PCR: 10 mM of
each oligonucleotide primer in water. Primers are listed in
Table 2.
5. Reagent grade water.
6. 5 M betaine solution (Sigma-Aldrich).
7. Supplies and equipment described in Subheading 2.2.

2.4. Isolation of 1. Genomic DNA carrying KIR2DL2 or KIR2DL3. Table 1

KIR2DL2 and KIR2DL3 describes the use of HaploPrep to isolate a specific gene seg-
by HaploPrep ment for sequencing in those samples containing a second
gene sharing extensive sequence homology with the gene being
characterized.
2. HaploPrep™ Kit (QIAGEN, Valencia, CA, USA) with hybrid-
ization buffer H.
3. KIR locus HaploPrep probes 2DL2-999T and 2DL3-1316T,
100 mM of each probe in 1× Tris–EDTA (TE) buffer
(Invitrogen), stored at −20°C.
4. Reagent grade water.
5. Heating block with heated lid at 95°C (e.g., TruTemp DNA
Microheating System, Robbins Scientific, Sunnyvale, CA, USA)
(see Note 6).
6. BioRobot EZ1 (QIAGEN) with HaploPrep card and manual.

2.5. Restriction 1. Genomic DNA from cells carrying KIR2DL3. Table 1 describes
Enzyme Digestion for the use of restriction enzyme digestion to eliminate a highly
the KIR2DL3 Locus homologous gene when present in the sample.
2. Restriction endonuclease BclI (15 U/mL) and 10× NE Buffer
3 (New England BioLabs, Ipswich, MA, USA).
3. Reagent grade water.
4. Phenol:chloroform:isoamyl alcohol 25:24:1,V/V/V (e.g.,
UltraPure™ phenol:chloroform:isoamyl alcohol, Invitrogen)
(see Note 7).
5. 3M sodium acetate (Sigma-Aldrich).
6. 70% ethanol in water (Warner-Graham Company) at −20°C.
7. Heating block at 50°C.
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 443

8. −20°C freezer.
9. Supplies and equipment described in Subheading 2.2.

2.6. KIR2DS4 1. Nested PCR amplicon of KIR2DS4 from Subheading 2.3.

Allele Isolation Table 1 describes the use of cloning to separate alleles in specific
by Cloning KIR2DS4 heterozygous samples.
2. TOPO TA Cloning Kit (Invitrogen) including SOC medium
and instruction manual.
3. LB agar plates containing 50 mg/mL ampicillin.
4. 40 mg/mL X-gal (5-bromo-4-chloro-3-indolyl-b-D-galacto-
pyranoside) in dimethylformamide.
5. 100 mM isopropyl b-D-1-thiogalactopyranoside (IPTG) in water.
6. Reagent grade water.
7. Sterile toothpicks.
8. 1.5 mL sterile disposable tubes (Fisher Scientific, Dallas, TX,
USA).
9. 37°C Shaking and non-shaking bacterial incubators.
10. Centrifuge capable of holding 1.5 mL tubes with a maximum
speed of 20,000 × g (14,000 rpm) (e.g., model 5424 (for
tubes), Eppendorf, Hauppauge, NY, USA).
11. Heating block at 42°C and 94°C.

2.7. Long Template 1. Genomic DNA from samples carrying KIR3DL1 or KIR3DS1.
PCR for KIR3DL1/ Table 1 summarizes the strategies used to obtain amplicons for
KIR3DS1 Amplicon B specific KIR genes.
2. Expand Long Template PCR System with Taq DNA poly-
merase and 10× Expand Long Template buffer 3 (Roche,
Mannheim, Germany).
3. 10 mM dNTP mixture (Roche, Mannheim, Germany).
4. KIR locus PCR primer solutions for KIR3DL1 and KIR3DS1
B amplicons: 10 mM of each oligonucleotide primer in water.
Primers are listed in Table 2.
5. Reagent grade water.
6. Supplies and equipment described in Subheading 2.2.

2.8. DNA Sequencing 1. Amplified DNA purified with AMPure from Subheading 3.2.
2. BigDye Terminator v1.1 diluted 1:1 with 5× sequencing buf-
fer (Applied Biosystems, Foster City, CA, USA).
3. KIR locus sequencing primers: 1.5 mM of each oligonucleotide
primer in water. Store at −20°C (see Table 3) (see Note 8).
4. DMSO.
5. Agencourt CleanSEQ kit (Beckman Coulter, Beverly, MA,
USA).
 444

Table 3
DNA sequencing primers for KIR loci

KIR Sequence covers Use with

locus Primer Sequence (5¢–3¢) Strand Nucleotide positiona exonb amplicon
L. Hou et al.

2DL1 2DL1-SEQ-E1R GGCCCATCACTCCATCTCT Antisense 167–185 Exon 1 A/A2

2DL1-SEQ-E2F CAAGACTCACAGCCCAGTG Sense 917–935 Exon 2 A/A2
2DL1-SEQ-E2R GGAGGCAAGGTCAGAAATGT Antisense 1161–1180 Exon 2 A/A2
2DL1-SEQ-E4F GAYGCCTTCTRAACTCACAAC Sense 3450–3470 Exon 4 A/A2
2DL1-SEQ-E4R AAGTCCTRGATCATTCACTC Antisense 3825–3844 Exon 4 A
2DL1-SEQ-E5F AAGATCCTCCCTGAGGAAAC Sense 5277–5296 Exon 5 A/B
2DL1-SEQ-E5R AGGCTCTAGGATCATAGGACA Antisense 5634–5654 Exon 5 B
2DL1-SEQ-E6F GCCTTTCTTTATGCCAATGT Sense 8488–8507 Exon 6 B
2DL1-SEQ-E6R TGTCAGAGCTGTGAGATGCT Antisense 8887–8906 Exon 6 B
2DL1-SEQ-E7F ATCTGGGTGCTTGTCCTAA Sense 12951–12969 Exon 7 B
2DL1-SEQ-E7R AGGGACCATCCTGTTTGTGA Antisense 13252–13271 Exon 7 B
2DL1-SEQ-E89F AAATGAGGACCCAGAAGTGC Sense 13580–13599 Exons 8, 9 B
2DL1-SEQ-E89R TGTTGACTCCCTAGAAGACG Antisense 13987–14006 Exons 8, 9 B
2DL2 2DL2-SEQ-E1R GGCCCATCACTCCATCTCT Antisense 129–147 Exon 1 A
2DL2-SEQ-E2F CAAGACTCACAGCCCAGTG Sense 861–879 Exon 2 A
2DL2-SEQ-E2R TTGAGCACCCCAGTCTAACC Antisense 1170–1189 Exon 2 A
2DL2-SEQ-E4F GACACCTTCTAAACTCACAAC Sense 3382–3402 Exon 4 A
2DL2-SEQ-E4R AAGTCGTGGATCATTCACTC Antisense 3754–3773 Exon 4 A
2DL2-SEQ-E5F1 GGTCATAGAGCAGGGGAGTG Sense 5136–5155 Exon 5 A
2DL2-SEQ-E5F2 AATGCCTCTTCTCCTCCAGGTCTA Sense 5209–5233 Exon 5 D
2DL2-SEQ-E5R TCTCTGCATCTGTCCATGCT Antisense 5602–5621 Exon 5 A/B/D
2DL2-SEQ-E6F CCCAGGGCCCAATATTAGAT Sense 8681–8700 Exon 6 B
2DL2-SEQ-E6R TCAATGCCTGCATCGAAGGTTTCT Antisense 9193–9217 Exon 6 B/C
2DL2-SEQ-E7F ATCTGGGTGCTTGTCCTAA Sense 12993–13011 Exon 7 C
2DL2-SEQ-E7R AGGGACCATCCTGTTTGTGA Antisense 13294–13313 Exon 7 C
2DL2-SEQ-E89F AAATGAGGACCCAGAAGTGC Sense 13622–13641 Exons 8, 9 C
2DL2-SEQ-E89R GGAGACAACTTTGGATCTGGA Antisense 13976–13996 Exons 8, 9 C
2DL3 2DL3-SEQ-E1R GGCCCATCACTCCATCTCT Antisense 129–147 Exon 1 A
2DL3-SEQ-E2F CAAGACTCACAGCCCAGTG Sense 861–879 Exon 2 A
2DL3-SEQ-E2R TTGAGCACCCCAGTCTAACC Antisense 1170–1189 Exon 2 A
2DL3-SEQ-E4F GACACCTTCTAAACTCACAAC Sense 3382–3402 Exon 4 A/D
25

2DL3-SEQ-E4R CCAATGACAATGAGAATG Antisense 3731–3748 Exon 4 A/D

2DL3-SEQ-E4F-218T TTAAGGACACTTTGCACCTCAT Sense 3542–3563 Exon 4 A/D
2DL3-SEQ-E4R-282T TAGCATCTGTAGGTCCCTGCA Antisense 3627–3647 Exon 4 A/D
2DL3-SEQ-E4F-166C TGGTCAGATGTCAGGTTT C Sense 3493–3511 Exon 4 A/D
2DL3-SEQ-E5F1 GGTCATAGAGCAGGGGAGTG Sense 5136–5155 Exon 5 A/B1/B2
2DL3-SEQ-E5F2 AATGCCTCTTCTCCTCCAGGTCTA Sense 5209–5233 Exon 5 A/B1/B2
2DL3-SEQ-E5R TTCTCTCTGCATCTGTCCATG Antisense 5608–5628 Exon 5 B1/B2
2DL3-SEQ-E5R-618A AGTTTGACCACTCGTAT Antisense 5480–5496 Exon 5 B1/B2
2DL3-SEQ-E6F TGAACCAACCTCAAAGATTTCC Sense 8698–8719 Exon 6 C1
2DL3-SEQ-E6R TTCTACCTCCCCAGGTTT C Antisense 8860–8878 Exon 6 C1
2DL2/3-SEQ-E7F ATCTGGGTGCTTGTCCTAA Sense 12993–13011 Exon 7 C1/C2
2DL3-SEQ-E7R CCCACATGGCCCTGAGC Antisense 11966–11982 Exon 7 C2
2DL3-SEQ-E89F TGCTTATGAAATGAGGGCCC Sense 12336–12355 Exons 8, 9 C2
2DL3-SEQ-E89R AGGGCTCAGCATTTGGAAG Antisense 12683–12701 Exons 8, 9 C2
2DL4 2DL4-SEQ-E1R CATCCTCACCACTCACTTGC Antisense 126–145 Exon 1 A2
2DL4-SEQ-E2F GGCTCAGGAGGAAAGGGTAG Sense 177–196 Exon 2 A/A2
2DL4-SEQ-E2R CAGGCCTTCCCATGGTCAG Antisense 374–392 Exon 2 A/A2
2DL4-SEQ-E3F GGGGAGAATCTTCTGAGCAC Sense 1063–1082 Exon 3 A/A2
2DL4-SEQ-E3R CACCAGAAGCTCTGGGACTC Antisense 1469–1488 Exon 3 A/A2
2DL4-SEQ-E5F AGAGCAGGGCAGTGAGTTCT Sense 2217–2236 Exon 5 A/A2
2DL4-SEQ-E5R TCCACATCTGTCCATGCTTC Antisense 2677–2696 Exon 5 A
2DL4-SEQ-E6F CCAGGGCCCAACATTAGATA Sense 5074–5093 Exon 6 A
2DL4-SEQ-E6R ATCACAGAGCTGGCAGGTG Antisense 5316–5334 Exon 6 A/B
2DL4-SEQ-E7F CCTGGCAACCAAGAAATGAG Sense 9400–9419 Exon 7 B
2DL4-SEQ-E7R AGACTTTCCTGCCAGTGAGG Antisense 9663–9682 Exon 7 B
2DL4-SEQ-E89F CCCCCTGTGTGTTGGTATCT Sense 9965–9984 Exons 8, 9 B
2DL4-SEQ-E89R TAAGCAAGAGACAGGCACCA Antisense 10519–10538 Exons 8, 9 B
Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing

(continued)
445
 446

Table 3
(continued)

KIR Sequence covers Use with

locus Primer Sequence (5¢–3¢) Strand Nucleotide positiona exonb amplicon
L. Hou et al.

2DL5 2DL5-SEQ-E1F ATCTTGTGTTCGGGAGGTTG Sense 5UTR (−274)–(−256) 5¢ noncoding region A

2DL5-SEQ-E1R AACTCCACCTCCAGGCCTAT Antisense 101–120 Exon 1 A
2DL5-SEQ-E2F ACCAAGACTCACAGCCCAGT Sense 706–725 Exon 2 A
2DL5-SEQ-E2R TCCCTCCTGTTTCAGGAAAAT Antisense 873–893 Exon 2 A
2DL5-SEQ-E3F GGGGAGAATCTTCTGAGCACT Sense 1510–1529 Exon 3 A
2DL5-SEQ-E3R TGCTCTGGGATTCAGGAAGT Antisense 1908–1927 Exon 3 A
2DL5-SEQ-E5F GGGAGCTGTGACAAGGAAGA Sense 2697–2716 Exon 5 A/B
2DL5-SEQ-E5R AGCAGGAAGCTCCTCAGCTA Antisense 3088–3107 Exon 5 B
2DL5-SEQ-E6F GCCATGAACCAACCTCAAAG Sense 5131–5150 Exon 6 B
2DL5-SEQ-E6R CTGAGCCAATGCTTGAATCC Antisense 5321–5340 Exon 6 B
2DL5-SEQ-E7F GCTGGCAACCAAGAAATGAG Sense 7950–7969 Exon 7 B
2DL5-SEQ-E7R ACCAGTGTGCTCCCATCCT Antisense 8187–8205 Exon 7 B
2DL5-SEQ-E89F CCCTTCCAGCTGTTTTGATG Sense 8562–8581 Exons 8, 9 B
2DL5-SEQ-E89R TGATGCCTTCAGATTCCAGC Antisense 9010–9029 Exons 8, 9 B
2DS1 2DS1-SEQ-E1R GGCCCATCACTCCATCTCT Antisense 470–488 Exon 1 A
2DS1-SEQ-E2F CAAGACTCACAGCCCAGTG Sense 1220–1238 Exon 2 A
2DS1-SEQ-E2R GGAGGCAAGGTCAGAAATGT Antisense 1464–1483 Exon 2 A
2DS1-SEQ-E4F GAYGCCTTCTRAACTCACAAC Sense 3753–3773 Exon 4 A
2DS1-SEQ-E4R AATTCCTGGATCATTCACTC Antisense 4128–4147 Exon 4 A/B
2DS1-SEQ-E5F AAGGGAGCTGTGACAAGGAA Sense 5581–5600 Exon 5 A/B
2DS1-SEQ-E5R TCTGCATCTGTCCATGCTTC Antisense 6008–6027 Exon 5 B
2DS1-SEQ-E6F GCCTTTCTTTATGCCAGTGTC Sense 8785–8805 Exon 6 B
2DS1-SEQ-E6R CTGAGTCAACGCCTGAATCC Antisense 9166–9185 Exon 6 B
2DS1-SEQ-E7F CCAATCAAGAAATGCGAGACA Sense 13295–13315 Exon 7 B
2DS1-SEQ-E7R CAGGGGAAGGGAATCTGGT Antisense 13609–13620 Exon 7 B
2DS1-SEQ-E89F TCCCCCTGTTTGTTGGTATC Sense 13882–13901 Exons 8, 9 B
2DS1-SEQ-E89R AAGGGCGAGTGATTTTTCTCT Antisense 14155–14175 Exons 8, 9 B
2DS2 2DS2-SEQ-E1R GGCCCATCACTCCATCTCT Antisense 129–147 Exon 1 A
2DS2-SEQ-E2F CAAGACTCACAGCCCAGTG Sense 745–763 Exon 2 A
2DS2-SEQ-E2R GGAGGCAAGGTCAGAAATGT Antisense 989–1008 Exon 2 A
2DS2-SEQ-E4F AAGGGGAAGCCTCACTCATT Sense 3216–3235 Exon 4 A
25

2DS2-SEQ-E4R GCCCAATGACAATGAGAATG Antisense 3614–3633 Exon 4 A/B

2DS2-SEQ-E5F TGAAGAGAGATGGGGTGGAG Sense 4977–4996 Exon 5 A/B
2DS2-SEQ-E5R CTCTCTGCATCTGTCCATGC Antisense 5491–5510 Exon 5 B
2DS2-SEQ-E6F CAGAGTGTTGGCCATGAACC Sense 8486–8505 Exon 6 B
2DS2-SEQ-E6R CTGAGTCAACGCCTGAATCC Antisense 8686–8705 Exon 6 B
2DS2-SEQ-E7F CCAATCAAGAAATGCGAGACA Sense 12818–12838 Exon 7 B
2DS2-SEQ-E7R CAGGGGAAGGGAATCTGGT Antisense 13143–13161 Exon 7 B
2DS2-SEQ-E89F CCTCCGAGCTCTTTTGTTGA Sense 13427–13446 Exons 8, 9 B
2DS2-SEQ-E89R TTATGCGTATGACACCTCCTGAT Antisense 13633–13655 Exons 8, 9 B
2DS3 2DS3-SEQ-E1R AGGCCTATATCTCCACCTCTG Antisense 88–108 Exon 1 A
2DS3-SEQ-E2F GCCTGGCTACCAAGACTCAC Sense 1247–1266 Exon 2 A
2DS3-SEQ-E2R AGAGACTCCCCGACAGGACT Antisense 1443–1462 Exon 2 A
2DS3-SEQ-E4F GGAAGCCTCACTCAATCCAG Sense 3739–3758 Exon 4 A
2DS3-SEQ-E4R CCTCCAAGTCCTGGATCATT Antisense 4165–4184 Exon 4 A
2DS3-SEQ-E5F AAGGGAGCTGTGACAAGGAA Sense 5581–5600 Exon 5 A
2DS3-SEQ-E5R TCTGCATCTGTCCATGCTTC Antisense 6008–6027 Exon 5 A/B
2DS3-SEQ-E6F CCCAGGGCCCAATATTAGAT Sense 8969–8988 Exon 6 B
2DS3-SEQ-E6R GGTGGAAGACAGGGGTACAA Antisense 9229–9248 Exon 6 B
2DS3-SEQ-E7F TCAATCAAGAAATGCGAGACA Sense 13321–13341 Exon 7 B
2DS3-SEQ-E7R CACACCCACGTGCTAACATC Antisense 13556–13575 Exon 7 B
2DS3-SEQ-E89F TCCCCCTGTTTGTTGGTATC Sense 13882–13901 Exons 8, 9 B
2DS3-SEQ-E89R TTATGCGTATGACACCTC Antisense 14141–14158 Exons 8, 9 B
(continued)
Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing
447
 448

Table 3
(continued)
L. Hou et al.

KIR Sequence covers Use with

locus Primer Sequence (5¢–3¢) Strand Nucleotide positiona exonb amplicon
2DS4 2DS4-SEQ-E1R CAGGCCCATATCTCCACCT Antisense 91–109 Exon 1 A
2DS4-SEQ-E2F GGGCTGGCTATCAAGACTCA Sense 2222–2241 Exon 2 A
2DS4-SEQ-E2R TCCCGTTTCAGGAAAATCC Antisense 2396–2414 Exon 2 A
2DS4-SEQ-E4F AGGCTCACTCATTCCAGGTG Sense 4736–4755 Exon 4 A
2DS4-SEQ-E4R TTACAACCACCTGGGTCTCC Antisense 5174–5193 Exon 4 A/B
2DS4-SEQ-E5F GGGAGCTGTGACAAGGAAGA Sense 6610–6630 Exon 5 B/C
2DS4-SEQ-E5R CATGCTGCGTCTTCTCTCTG Antisense 7025–7044 Exon 5 B
2DS4-SEQ-E6F GGCCATGAACCAAACTCAAA Sense 10016–10035 Exon 6 B
2DS4-SEQ-E6R CAGGCGTACAATGTCAGAGC Antisense 10236–10256 Exon 6 B
2DS4-SEQ-E7F GTGGTTACCTGCCAATCAAGA Sense 14327–14347 Exon 7 B
2DS4-SEQ-E7R ATCCTGCTGGTGAGGAACAC Antisense 14592–14611 Exon 7 B
2DS4-SEQ-E89F AAATGAGGACCCAGAAGTGC Sense 14927–14946 Exons 8, 9 B
2DS4-SEQ-E89R TTATGCGTATGACACCTCCTGAT Antisense 15153–15175 Exons 8, 9 B
2DS5 2DS5-SEQ-E2R AGACTCCCTGACAGGACTTC Antisense 1613–1632 Exon 2 A
2DS5-SEQ-E4F AGCCTCACTCAATCCAGGTG Sense 3915–3934 Exon 4 A
2DS5-SEQ-E4R ACCTGTGATCACGATGTCCA Antisense 4273–4292 Exon 4 A/B
2DS5-SEQ-E5F CAGAGCAGGGGAGTGAGTTC Sense 5731–5750 Exon 5 A/B
2DS5-SEQ-E5R AGCAGGAAGCTCCTCAGCTA Antisense 6159–6178 Exon 5 B
2DS5-SEQ-E6F CCCAGGGCCCAATATTAGAT Sense 9145–9164 Exon 6 B
2DS5-SEQ-E6R GGTGGAAGACAGGGGTACAA Antisense 9405–9424 Exon 6 B
2DS5-SEQ-E7F GCTAGGTCTCCCACCATTTG Sense 13440–13459 Exon 7 B
2DS5-SEQ-E7R ATCCTGCCTGTGAGGAACAC Antisense 13752–13771 Exon 7 B
2DS5-SEQ-E89F TCCCCCTGTTTGTTGGTATC Sense 14059–14079 Exons 8, 9 B
2DS5-SEQ-E89R TTATGCGTATGACACCTC Antisense 14318–14335 Exons 8, 9 B
3DL1 3DL1-SEQ-E1R CTCCACTTCAGGCCCATAAC Antisense 138–157 Exon 1 A
3DL1-SEQ-E2F CAAGACKCACAGCCCAGTG Sense 953–971 Exon 2 A
3DL1-SEQ-E2R TGGAGCACCCTAGTCTCACC Antisense 1262–1281 Exon 2 A
3DL1-SEQ-E3F GAGAATCTTCTGGGCACTGG Sense 1739–1758 Exon 3 A
25

3DL1-SEQ-E3R ATTCAGGAGGTGGGACAGTG Antisense 2126–2145 Exon 3 A/M

3DL1-SEQ-E4F ACCCTCACTCATTCCAGGTG Sense 3136–3155 Exon 4 A/M
3DL1-SEQ-E4R AAGTCCTRGATCATTCACTC Antisense 3555–3574 Exon 4 A/B/M
3DL1-SEQ-E5F1 GGTCATAGAGCAGGGGAGTG Sense 4970–4989 Exon 5 B
3DL1/2-SEQ-E5F2 CTTCTCTCTCAGCCCAGC Sense 5080–5097 Exon 5 B
3DL1-SEQ-E5R TGCATCTGTCCATGCTTTTC Antisense 5434–5453 Exon 5 B
3DL1-SEQ-E6F GCCTTTCTTTATGCCAATGT Sense 8254–8273 Exon 6 B
3DL1-SEQ-E6R CCCTTTCACTGTTGGAGTGT Antisense 8708–8727 Exon 6 B
3DL1-SEQ-E7F AGGGGTCAAACATCTCAACT Sense 12638–12657 Exon 7 B
3DL1-SEQ-E7R AGCTGTGTGCTCCCATCCT Antisense 13016–13034 Exon 7 B
3DL1-SEQ-E89F AAATGAGGACCCAGAAGTGC Sense 13372–13391 Exons 8, 9 B
3DL1-SEQ-E89R GCCTCTGAGAAGGGCGA Antisense 13676–13692 Exons 8, 9 B
3DL1/2-SEQ-E89F GGAGACAGAATCAATGGGAT Sense 15619–15638 Exons 8, 9 B
3DL1/2-SEQ-E89R GGCTGTTGTCTCCCTAGAAA Antisense 16178–16197 Exons 8, 9 B
3DL2 3DL2-SEQ-E1R CGAGATCTCCATCCCCACT Antisense 66–84 Exon 1 A2
3DL2-SEQ-E2F AGTTTACCTTCAGCCCAGCA Sense 631–650 Exon 2 A/A2
3DL2-SEQ-E2R GAGACTCCCCGACAGGACTT Antisense 848–867 Exon 2 A/A2
3DL2-SEQ-E3F AGCGGAAATGGGAGAATCTT Sense 1436–1455 Exon 3 A/A2
3DL2-SEQ-E3R CAGAAGCTCTGGGATTCAGG Antisense 1847–1866 Exon 3 A/A2
3DL2-SEQ-E4F ACCCTCACTCATTCCAGGTG Sense 3196–3215 Exon 4 A/A2
3DL2-SEQ-E4R TCTGTGTCCCAATGACAATGA Antisense 3595–3615 Exon 4 A/A2
3DL2-SEQ-E5F CTCAGGTATGAGGGGAGCTG Sense 5078–5097 Exon 5 A/A2
3DL2-SEQ-E5R TCTGCATCTGTCCATGCTTC Antisense 5515–5534 Exon 5 A
3DL2-SEQ-E6F AGGGTCCAACATTAGATAACA Sense 8492–8512 Exon 6 A/B
3DL2-SEQ-E6R CCAGGTTTCCAAAAGCAGAG Antisense 8677–8696 Exon 6 B
3DL2-SEQ-E7F GTCAATCAAGAAATGAGACAA Sense 15253–15273 Exon 7 B
3DL2-SEQ-E7R GCAATGGTCTGTGAGCTGAA Antisense 15598–15617 Exon 7 B
3DL2-SEQ-E89F TGAAATGAGGACCCAGAAGG Sense 15837–15856 Exons 8, 9 B
Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing

3DL2-SEQ-E89R AACCCCCTCAAGACCTGACT Antisense 16231–16250 Exons 8, 9 B

(continued)
449
 450
L. Hou et al.

Table 3
(continued)

KIR Sequence covers Use with

locus Primer Sequence (5¢–3¢) Strand Nucleotide positiona exonb amplicon
3DL3 3DL3-SEQ-E1R CTCGATTCCCTTCCAGGACT Antisense 38–57 Exon 1 A2
3DL3-SEQ-E2F GAGATGTTGGCTTGGAGTGC Sense 442–461 Exon 2 A2
3DL3-SEQ-E2R ATCAGTCAACCCCCTGTGTC Antisense 820–839 Exon 2 A/A2
3DL3-SEQ-E3F AGAAACGTGGAAATGGGAGA Sense 1426–1445 Exon 3 A/A2
3DL3-SEQ-E3R GAGGTGGGACAGTGAGAAGC Antisense 1823–1842 Exon 3 A/A2
3DL3-SEQ-E4F TAGACACCATGGAGGGGAAG Sense 2982–3001 Exon 4 A/A2
3DL3-SEQ-E4R AAGTCCTRGATCATTCACTC Antisense 3418–3437 Exon 4 A
3DL3-SEQ-E5F AGCTCAGGTGTGAGGAGAGC Sense 4890–4909 Exon 5 A
3DL3-SEQ-E5R TGAGCCTAAGTTCACCGGC Antisense 5083–5101 Exon 5 A
3DL3-SEQ-E5F2 ATCTATCCAGGGAGGCAGAG Sense 5063–5082 Exon 5 B
3DL3-SEQ-E5R2 TGGCTCTAGGATCACAAGACA Antisense 5277–5297 Exon 5 A/B
3DL3-SEQ-E7F CTCCTTGGGACAGCATTGAT Sense 10395–10414 Exon 7 B
3DL3-SEQ-E7R AGAAAGTCCTGCCTCTGTGG Antisense 10938–10957 Exon 7 B
3DL3-SEQ-E89F AAATGAGGACCCAGAAGTGC Sense 11231–11250 Exons 8, 9 B
3DL3-SEQ-E89R CAGCATTTGGAAGTTCCGTGTT Antisense 11562–11583 Exons 8, 9 B
 3DS1 3DS1-SEQ-E1R AGGCCCATAACTCCACCTCT Antisense 109–128 Exon 1 A
3DS1-SEQ-E2F AGTTTACCTTCAGCCCAGCA Sense 920–939 Exon 2 A
3DS1-SEQ-E2R ACAGGACTTCCCTCCCATTT Antisense 1126–1145 Exon 2 A
3DS1-SEQ-E3F1 TCTATGCAGGATGGGTCCTT Sense 1664–1683 Exon 3 A
25

3DS1-SEQ-E3F2 CAACATGAGCCCTGTGACCA Sense 1982–2001 Exon 3 B

3DS1-SEQ-E3R1 CAGAAGCTCTGGGATTCAGG Antisense 2137–2157 Exon 3 B
3DS1-SEQ-E3R2 GGTGTGAACCCCGACATG Antisense 2023–2040 Exon 3 A
3DS1-SEQ-E4F ACCCTCACTCATTCCAGGTG Sense 3509–3528 Exon 4 B
3DS1-SEQ-E4R TCCAAGTCCTGGATCATTCAC Antisense 3929–3949 Exon 4 B
3DS1-SEQ-E5F GGTCATAGAGCAGGGGAGTG Sense 5370–5389 Exon 5 B
3DS1-SEQ-E5R ATGAAGGAGGGTTTGGAGGT Antisense 5911–5930 Exon 5 B
3DS1-SEQ-E6F ACTCCCAGGGTCCAACATTA Sense 8811–8830 Exon 6 B
3DS1-SEQ-E6R TTCACAGAGCTGGGAGGTTT Antisense 9055–9074 Exon 6 B
3DS1-SEQ-E7F CATCTGGGTGCTTGTCCTAAA Sense 13138–13158 Exon 7 B
3DS1-SEQ-E7R ATCCTGCTTCCCCACATGG Antisense 13402–13420 Exon 7 B
3DS1-SEQ-E89F TCCCCCTGTTTGTTGGTATC Sense 13744–13763 Exons 8, 9 B
3DS1-SEQ-E89R CTCTGAGAAGGGCGAGTG Antisense 14051–14068 Exons 8, 9 B
a
Numbering is based on the genomic sequences in the LRC database (http://www.ncbi.nlm.nih.gov/gv/lrc/). Nucleotide 1 is the first base of exon 1
b
Exon numbering is based on nine total exons for each locus. Some of the KIR loci are missing an exon or have a pseudo exon that is not analyzed. KIR2DL1-3 and KIR2DS1-5
have a pseudo-exon 3 while KIR2DL4 and KIR2DL5 lack exon 4 (24)
Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing
451
452 L. Hou et al.

6. Ethanol: 73% solution in water.

7. Reagent grade water.
8. Thermal cycler (e.g., model 2720, Applied Biosytems, Foster
City, CA, USA).
9. 3730xl DNA Analyzer with POP7, 1× running buffer with
EDTA, and manual (Applied Biosytems).
10. Centrifuge capable of holding plates with a maximum speed of
20,000 × g (14,000 rpm) (e.g., model 5804 (for plates),
Eppendorf, Hauppauge, NY, USA).
11. Single channel and multichannel pipettors (0.5–200 mL)
and tips.
12. Semi-skirted PCR tray (Fisher Scientific, Dallas, TX, USA).
13. Tape seals (One Lambda, Canoga Park, CA, USA).
14. Agencourt SPRIPlate 96R magnet plate (Beckman Coulter).

2.9. Sequence Analysis 1. Analysis software: Assign SBT 3.2.7 (Conexio Genomics,
Including Preparation Applecross, Western Australia), HLA Librarian (Conexio
of Locus-Specific KIR Genomics), Sequencher 4.6 (Ann Arbor, MI, USA) with
Libraries manuals (see Note 9).
2. KIR nucleotide sequence databases: IPD-KIR curated coding
region sequence database at http://www.ebi.ac.uk/ipd/kir/
index.htmL; leukocyte receptor complex (LRC) database
alignment viewer for genomic sequences at http://www.ncbi.
nlm.nih.gov/gv/lrc/.

3. Methods

3.1. DNA Preparation 1. Label the appropriate number of 1.5 mL microcentrifuge tubes
and QIAamp spin columns with sample identifier (see Note 10
on laboratory).
2. Add 200 mL whole blood sample to the tube (see Note 11). If
the sample volume is less than 200 mL, add PBS to bring
sample to volume.
3. Pipet 20 mL protease into the blood sample in the tube.
4. Add 200 mL buffer AL to the sample (see Note 12). Immediately
mix by vortexing for 15 s.
5. Incubate at 56°C for 10 min.
6. Briefly centrifuge the microcentrifuge tube to remove conden-
sation drops from the inside of the lid (see Note 13).
7. Add 200 mL 96–100% ethanol to the sample and mix again by
vortexing for 15 s. Again briefly centrifuge the microcentrifuge
tube.
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 453

8. Carefully apply the sample to the QIAamp spin column in a

collection tube without wetting the rim of the spin column.
Centrifuge at 6,000 × g (8,000 rpm) for 1 min. Place the
QIAamp spin column into a clean 2-mL collection tube and
discard the tube containing the filtrate.
9. After placing the spin column into a clean collection tube,
carefully add 500 mL Buffer AW1 without wetting the rim of
the spin column. Centrifuge at 6,000 × g (8,000 rpm) for
1 min.
10. Place the spin column into a clean 2-mL collection tube and
discard tube with the filtrate. Carefully add 500 mL Buffer
AW2 without wetting the rim. Centrifuge at 20,000 × g
(14,000 rpm) for 3 min.
11. Place the QIAamp spin column in a clean 1.5 mL microcentri-
fuge tube and discard the tube with the filtrate. Add 200 mL
water and incubate at room temperature for 1–5 min.
12. Centrifuge at 6,000 × g (8,000 rpm) for 1 min. The isolated
DNA is in the liquid fraction.
13. Discard the spin column. Make sure the sample tube is labeled
correctly. Store at 4°C for short term, or −20 to −80°C for
long-term storage (see Note 14). See Note 15 for a discussion
of potential problems.

3.2. Polymerase Chain 1. See Table 1 for a listing of those KIR loci that should be
Reaction Amplification amplified following this protocol (see Note 16).
of Individual KIR Loci: 2. Thaw 10× High Fidelity PCR buffer, 50 mM MgSO4, dNTP
General mix, primer solutions, and DMSO or 5M betaine solution (see
Note 17). Mix the solutions thoroughly before use.
3. Prepare the reaction mix in a 1.5-mL tube as described in
Table 4.
4. Vortex the reaction mix and dispense 45 mL volumes into each
well of a semi-skirted PCR tray.
5. Add 5 mL of genomic DNA (50–200 ng), purified as described
in Subheading 3.1, to each well containing reaction mix (see
Note 18).
6. Set up positive and negative amplification control wells. The
positive control for each primer pair is 5 mL DNA (50–200 ng)
from a cell carrying that KIR locus. The negative control for
each primer pair is 5 mL DNA (50–200 ng) from a cell lacking
that KIR gene. For primers amplifying framework genes
(KIR2DL4, KIR3DL2, and KIR3DL3), use 5 mL water as a
negative control instead of DNA.
7. Place tape seal over entire tray and quick spin the plate in the
centrifuge to ensure all the liquid is at the bottom of the wells.
Place in the thermal cycler.
454 L. Hou et al.

Table 4
Composition of reaction master mix for Platinum Taq DNA Polymerase High Fidelity

Component Volume in each reactiona

10× High Fidelity PCR buffer 5 mL

MgSO4 (50 mM) Variable (see Table 2)
dNTP (10 mM each) 1 mL
Sense primer (10 mM) (see Table 2) 2 mL
Antisense primer (10 mM) (see Table 2) 2 mL
DMSO or 5 M betaine solution Variable (see Table 2)
Platinum Taq DNA Polymerase High Fidelity (5 U/mL) 0.5 mL
Template DNA 5 mL (added in later step in protocol)
Water Bring final volume excluding DNA to 45 mL
a
The volume for a single reaction is 50 mL including the DNA so multiply the number of amplification reactions desired
by 50 to determine how much reaction master mix to make. Always make more than you need to account for losses
during pipetting

8. Polymerase chain reaction (PCR) conditions are described in

Table 5 (see Note 19).
9. Prepare a 1.5% agarose gel in 1× TBE. Ethidium bromide
(2 mL) should be added to the gel solution.
10. After the amplification cycles are complete, confirm amplification
by electrophoresis. Mix 5 mL of each amplification reaction
with 2 mL of 5× sucrose cresol solution and load the entire
sample into one well of the polymerized agarose gel.
Electrophorese the DNA ladder as a molecular weight marker.
Electrophorese at 100 V for 20 min until the cresol red dye has
reached the bottom of the gel.
11. Visualize the bands by placing the gel on a UV translluminator.
Photograph the gel. Using the molecular weight markers,
determine the approximate molecular weight of the amplicons
by comparison. The expected sizes of the amplicons for each
locus are listed in Table 2. The presence of additional bands
indicates a potential problem (see Note 20).
12. Add the AMPure solution directly to each PCR reaction in the
PCR plate. The volume of AMPure to add is 1.8× the reaction
volume (see Note 21).
13. Mix thoroughly by pipetting and place the PCR plate onto a
magnetic plate to separate the AMPure beads from the solution.
Incubate at room temperature for approximately 5–10 min.
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 455

Table 5
Polymerase chain reaction amplification conditions

Long Template
General PCR conditions Nested PCR (see PCR (see
(see Subheading 3.2) Subheading 3.3) Subheading 3.7)

Denaturation 95°C for 5 min 92°C for 4 min 92°C for 2 min
Initial cycles 10 cycles 10 cycles: 10 cycles
95°C for 20 s 92°C for 45 s 92°C for 10 s
58–66°C for 30 s (see Table 2) 62°C for 45 s 60°C for 30 s
68°C for 3–10 min (see Table 2) 72°C for 1.5 min 68°C for 11 min
Secondary cycles 30 cycles 30 cycles 30 cycles
95°C for 20 s 92°C for 45 s 92°C for 15 s
52–64oC for 30 s (see Table 2) 57°C for 45 s 57°C for 30 s
68°C for 3–10 min (see Table 2) 72°C for 1.5 min 68°C for 11 min
Final extension 68°C for 10 min 72°C for 10 min 68°C for 10 min
Final hold 4°C 4°C 4°C

14. With the PCR plate on the magnet, aspirate the cleared solu-
tion with a pipet and discard.
15. Keeping the PCR plate on the magnet, dispense 200 mL of 70%
ethanol to each well. Allow to sit at least 30 s at room tempera-
ture. Aspirate the wash solution with a pipet, discard and repeat.
Be sure to remove as much ethanol as possible to shorten the
drying time. Dry at room temperature for 10 min.
16. To elute the purified DNA, add 30–50 mL (see Note 22) of
reagent grade water to each well and mix well by pipetting up
and down. Place the plate back on the magnet.
17. Remove the eluate containing the amplified DNA to a clean
96-well plate to begin the DNA sequencing reactions (see
Subheading 3.8).

3.3. Nested PCR for 1. See Table 1 for a listing of those KIR loci that should be
KIR2DL2 Amplicon B, amplified following this protocol.
KIR2DL3 Amplicon A, 2. Thaw Taq DNA Polymerase, 10× PCR buffer with MgCl2,
and KIR2DS4 dNTP mix, 5M betaine solution, and appropriate primer solu-
Amplicon B tions (see Table 2). Mix the solutions thoroughly before use
(see Note 23).
3. Prepare the nested PCR reaction master mix as shown in
Table 6.
4. Aliquot 45 mL of master mix into each well of a semi-skirted
PCR tray.
456 L. Hou et al.

Table 6
Composition of reaction master mix for nested polymerase
chain reaction amplification

Components Volume in each reactiona

10× PCR Buffer with MgCl2 5 mL

dNTP (10 mM each) 1 mL
Sense primer (10 mM) (see Table 2) 2 mL
Antisense primer (10 mM) (see Table 2) 2 mL
5 M betaine solution 10 mL
Taq DNA Polymerase 0.25 mL
Template DNA 5 mL (added at later step)
Water Bring final volume excluding
DNA to 45 mL
a
The volume for a single reaction is 50 mL so multiple the number of amplification
reactions desired by 50 to determine how much reaction master mix to make. Always
make more than you need to account for losses during pipetting. If you need to add
more or less volume of DNA to the reaction mix, adjust volume of water so that the
final volume for each sample is 50 mL

5. Add 5 mL of each purified PCR product (i.e., KIR2DL2 ampli-

con B, KIR2DL3 amplicon A, and KIR2DS4 amplicon B)
(50–200 ng) to each well containing reaction mix.
6. Place in the thermal cycler and perform PCR using the proto-
col in Table 5.
7. Purify the nested PCR product of KIR2DL2 and KIR2DL3
for DNA sequencing with AMPure as described in
Subheading 3.2, step 12. Purify the nested PCR product of
KIR2DS4 with AMPure as described in Subheading 3.2, step
12. If required, clone the KIR2DS4 alleles as described in
Subheading 3.6.

3.4. Isolation 1. Haplotype-specific extraction is performed using genomic

of KIR2DL2 DNA from some cells shown to carry KIR2DL2 and
and KIR2DL3 KIR2DL3 as described in Table 1.
Using HaploPrep 2. Thaw HaploPrep KIR2DL2 and KIR2DL3 locus probes and
hybridization buffer on ice (see Notes 24 and 25).
3. Prepare HaploPrep reaction mix as described in Table 7.
4. Pipet up and down to mix the reaction mix thoroughly and
dispense the volume listed in Table 7 into 1.5 mL tubes.
5. Add 5 mL genomic DNA (30–150 ng) to each tube containing
reaction mix (see Note 26).
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 457

Table 7
HaploPrep reaction master mix

Components Volume in each reactiona

Hybridization buffer H 15 mL
HaploPrep Extraction Probe 2 mL
2DL2-999T or 2DL3-1316T
Water 8 mL
Genomic DNA 5 mL (added at a later step)
a
The volume for a single reaction is 25 mL without the DNA added so multiply
the number of reactions desired by 25 to determine how much reaction
master mix to make. Always make more than you need to account for losses
during pipetting

6. Cap the tubes, mix well by vortexing, and centrifuge briefly.

Place the tubes in a heating block with a heated lid at 95°C and
incubate for 15 min to denature the DNA.
7. Insert the EZ1 HaploPrep card into the BioRobot EZ1 fol-
lowing instructions from the instrument manual.
8. Switch on the EZ1 instrument and prepare the instrument as
described in the instrument manual.
9. Allow the internal heating block of the EZ1 instrument to heat
up to 64°C. After the 15-min incubation in step 6 is complete,
remove the tubes from external heating block. Remove the
caps, and place opened sample tube containing denatured sam-
ples immediately into the EZ1 instrument heating block (see
Note 27).
10. Close the instrument door and continue to follow the instruc-
tion manual.
11. Once the HaploPrep-isolated DNA has been prepared, per-
form PCR amplification as described in Subheading 3.2 and
proceed with DNA sequencing in Subheading 3.8.

3.5. Isolation 1. This protocol is performed for some cells carrying KIR2DL3
of KIR2DL3 Locus: as described in Table 1.
Restriction Enzyme 2. Prepare the restriction enzyme reaction mix according to
Digestion Table 8.
3. Mix the reaction thoroughly and dispense indicated volume
from Table 8 into a 1.5-mL tube.
4. Add 20 mL genomic DNA (approximately 2 mg) to each tube
containing reaction mix. Incubate at 50°C for 1 h.
458 L. Hou et al.

Table 8
Restriction enzyme reaction master mix

Components Volume in each reactiona

10× NE Buffer 3 20 mL
Bc1I 3 mL
Genomic DNA 20 mL (added at a later step)
Water Bring final volume excluding
DNA to 180 mL
a
The volume for a single reaction is 200 mL so multiple the
number of digestion reactions desired by 200 to determine
how much reaction master mix to make. Always make more than
you need to account for losses during pipetting. If you need
to add more or less volume of DNA to the reaction mix,
adjust volume of water so that the final volume for each sample
is 200 mL

5. Isolate DNA by adding 200 mL phenol:chloroform:isoamyl

alcohol to each tube and vortexing (see Note 7).
6. Centrifuge briefly (1–2 min) and transfer the aqueous (top)
phase to a clean tube.
7. Add 100 mL reagent grade water to the aqueous phase and
vortex. Briefly centrifuge and transfer the aqueous phase
(approximately 300 mL) to a clean tube.
8. Add 30 mL 3M sodium acetate to the aqueous phase and place
the solution at −20°C for at least 30 min.
9. Centrifuge at 14,000 rpm for 20–30 min at room temperature.
Remove the liquid with a pipettor.
10. Wash pellet by adding 200 mL cold 70% ethanol (see Note
28).
11. Centrifuge for 10 min, remove the liquid with a pipettor, and
air dry the pellet for approximately 20 min at room
temperature.
12. Redissolve the pellet in 20 mL reagent grade water.
13. Perform PCR amplification as performed as described in
Subheading 3.2 and proceed with DNA sequencing in
Subheading 3.8.

3.6. KIR2DS4 Allele 1. Cloning is required only for PCR amplicons containing both a
Isolation by Cloning full-length allele and an allele with a deletion (see Note 29).
Prepare a nested KIR2DS4 amplicon by PCR as described in
Subheading 3.3.
2. Verify amplified products on a 1.5% agarose gel with 1Kb DNA
ladder as described in Subheading 3.2, step 9.
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 459

3. Purify the PCR products using AMPure as described in

Subheading 3.2, step 12.
4. Using the TOPO TA cloning kit, clone the PCR product into
the pCR 2.1-TOPO vector following the manufacturer’s
instructions (see Note 30).
5. Add 2ml of the TOPO cloning reaction to a vial of One Shot
Chemical E. coli and mix gently. Incubate on ice for
5–30 min.
6. Heat-shock the cells for 30 s at 42°C.
7. Add 250 mL of SOC at room temperature to the tube.
8. Incubate in a 37°C shaker (250 rpm) for 1 h before plating on
LB agar.
9. Apply 40 mL Xgal (40 mg/mL) and 40 mL 100 mM IPTG to the
surface of an LB agar plate containing ampicillin and let dry.
10. To optimize distinct colonies, plate 50 and 100 mL of each
transformation onto two separate agar plates. Incubate at 37°C
overnight.
11. Pick several isolated white colonies from the agar plate using a
sterile toothpick. Transfer each colony of bacteria into a 0.5 mL
tube containing 50 mL sterile water (see Note 31).
12. Place the tubes in a heating block at 94°C for 5 min to lyse the
bacteria and to inactivate nucleases. Centrifuge at 2,000 rpm
for 5 min.
13. Use 5 mL of the supernatant in a 50-mL PCR reaction with the
same 2DS4 nested primers and protocol as described in
Subheading 3.3.
14. Verify amplification on a 1% agarose gel as described in
Subheading 3.2, step 9.
15. Purify the PCR fragments using AMPure as described in
Subheading 3.2, step 12 and proceed with DNA sequencing in
Subheading 3.8.

3.7. Long Template 1. Amplification of long segments of DNA from KIR3DL1 and
PCR for KIR3DL1 B KIR3DL2 will require this protocol (see Table 1). Thaw 10×
and KIR3DS1 B Expand Long Template buffer 3, dNTP mix, and primer solu-
Amplicons tions for KIR3DL1 B and KIR3S1 B amplicons (see Table 2).
Vortex the solutions thoroughly before use (see Note 32).
2. Assemble the reaction mix for the Expand Long Template
PCR System as described in Table 9.
3. Vortex the reaction mix thoroughly and dispense 45 mL vol-
umes into each well of semi-skirted PCR tray.
4. Add 5 mL template DNA (100–200 ng) to each well contain-
ing reaction mix (see Note 33).
460 L. Hou et al.

Table 9
Composition of reaction master mix for Expand Long
Template PCR reaction

Components Volume in each reactiona

10× Expand Long Template buffer 3 5 mL

dNTP (10 mM) 2.5 mL
Forward primer (10 mM) 1.5 mL
(see Table 2)
Reverse primer (10 mM) 1.5 mL
(see Table 2)
Expand Long Template Enzyme mix 0.75 mL
Template DNA 5 mL (added at later step)
Water Bring final volume excluding
DNA to 45 mL
a
The volume for a single reaction is 50 mL so multiply the number of amplification
reactions desired by 50 to determine how much reaction master mix to make.
Always make more than you need to account for losses during pipetting. If you
need to add more or less volume of DNA to the reaction mix, adjust volume of
water so that the final volume for each sample is 50 mL

5. Set up positive and negative control wells as described in

Subheading 3.2, step 6.
6. Place in the thermal cycler and perform PCR using the proto-
col in Table 5.
7. Check for amplification of a band of appropriate size by elec-
trophoresis on a 1.0% agarose gel stained with ethidium bro-
mide as described in Subheading 3.2, step 9.
8. Purify and elute the PCR product with AMPure as described in
3.2, step 12 and proceed with DNA sequencing in
Subheading 3.8.

3.8. DNA Sequencing 1. Sequence the amplicons using KIR loci sequencing primers
(see Table 3). For each locus, both sense and antisense primers
are used to cover the complete sequence of the exons (Fig. 1)
(see Note 34).
2. To each well, add 2 mL of diluted Big Dye Terminator, 1 mL of
the appropriate primer (see Table 3), and 3 mL of the purified
PCR product. For exon 1 sequences for all KIR loci, add
0.3 mL DMSO to the reaction (see Note 35).
3. Place tape seal over entire tray and quick spin the plate in the
centrifuge to ensure all liquid is at the bottom of the wells.
Place in the thermal cycler.
 25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 461

Table 10
DNA sequencing reaction conditions

Conditions for all

exons except exon 1 Conditions for exon 1

30 cycles 30 cycles
96°C for 10 s 96°C for 10 s
50°C for 5 s 60°C for 1 min
60°C for 4 min Hold at 4°C
Hold at 4°C

4. Perform the DNA sequencing reaction using the protocol in

Table 10.
5. Use the Agencourt CleanSEQ kit to remove excess dye termi-
nators from the sequence reaction by adding 10 mL of
CleanSEQ magnetic beads solution to each well of the sequenc-
ing plate.
6. For a 10 mL sequencing reaction, add approximately 75 mL
73% ethanol to each well and mix thoroughly.
7. Place the sequencing plate onto the magnet to separate the
beads from the solution. Incubate approximately 3 min at
room temperature.
8. With the sequencing plate on the magnet, aspirate the cleared
solution with a pipet and discard.
9. Keeping the plate on the magnet, dispense 100 mL 73% etha-
nol to each well and allow it to sit for at least 30 s at room
temperature. Aspirate the solution and discard.
10. Add 30 mL of water to each well. The reactions are now ready
to electrophorese on the DNA analyzer.
11. Follow the instructions for operation of the DNA analyzer.
The samples are electrophoresed using ABI RunModule
“Rapidseq 36_POP7” with the default values. Longer electro-
phoresis times may be required for some sequences.
12. Sample files are analyzed as described in Subheading 3.9.

3.9. Sequence Analysis 1. Locus-specific KIR libraries must be created prior to analysis of
Including Preparation KIR sequencing data. Go to the IPD-KIR database downloads
of Locus-Specific and open up the FTP directory. Obtain the nucleotide coding
KIR Libraries region sequences of all known alleles at each KIR locus as nuc.
fasta files (e.g., KIR2DL1_nuc.fasta; one file for each locus).
Create two separate libraries for KIR2DS4, one library with
the full-length allele sequences and a second library with the
462 L. Hou et al.

sequences of the alleles exhibiting the 22 base pair (bp)

deletion.
2. Manually add the intron 8 genomic sequence from one repre-
sentative allele from each locus to the nucleotide sequence of
every allele at the locus. Use the database of the LRC to obtain
the intron sequence from the genomic DNA (see Note 36).
3. Manually add the 247 bp genomic sequence found 5¢ of exon
1 to each KIR2DL5 locus allele sequence (see Note 37).
4. Use HLA Librarian to create a sequence library and reference
file for each locus following the Library Builder user’s guide.
5. Import each nuc.fasta file containing the intron 8 sequence
into HLA Librarian assigning a name for the library and refer-
ence files (e.g., 2DL1). Enter information into the reference
file as indicated including the position of nucleotides at the 5¢
and 3¢ ends of each exon.
6. Output the files to the Assign directory following instructions
in the Assign user’s guide.
7. The library should be validated by interpreting the sequences
of multiple known KIR alleles obtained by sequencing both
homozygous and heterozygous reference cell DNA.
8. Once the library has been created, use Assign SBT 3.2.7 soft-
ware to interpret sequencing results and assign alleles (see
Notes 38 and 39).
9. The library should be updated with newer versions of the IPD-
KIR database as required (see Note 40).

4. Notes

1. Blood (8.5 mL) is collected by venipuncture into a yellow top

ACD-A tube. ACD is the preferred anticoagulant. Other anti-
coagulants (e.g., heparin) may inhibit DNA amplification dur-
ing the polymerase chain reaction. Blood can be aliquoted into
2 mL tubes and stored at -20°C until use. An alternative sam-
ple source is a buccal swab, but it is likely that the yield of DNA
will be low and insufficient for sequencing of all KIR loci.
Blood should be treated as a biohazard and handled with
caution.
2. The panel of reference cells should include cells that lack
specific KIR genes as well as cells that carry specific KIR genes.
It is helpful to know the KIR alleles carried by the cells so that
they can serve as controls for the assignment of KIR alleles.
3. Aliquot diluted primers. Repeated freezing and thawing of
diluted oligonucleotide primers should be avoided.
25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 463

4. The DNA ladder should range in size between 400 and

13,000 bp. It is helpful to have markers every 500–1,000 bp.
A high DNA mass ladder (Invitrogen) is also helpful when
judging the approximate quantity of amplicon present.
5. Handle carefully; ethidium bromide is a carcinogen.
6. It is critical to have a heated lid for the Haploprep protocol.
7. Handle phenol:chloroform:isoamyl alcohol carefully and work
in a fume hood. Alternatives to phenol:chloroform:isoamyl
alcohol extraction might be use of the Agencourt AMPure kit
(Beckman Coulter, Beverly, MA, USA) or Amicon Ultra cen-
trifugal filters (Millipore, Billerica, MA, USA), but the authors
have not tested these products in this protocol.
8. Aliquot diluted primers. Repeated freezing and thawing of
diluted oligonucleotide primers should be avoided.
9. Assign is used to obtain KIR allele assignments from the DNA
sequences obtained. HLA Librarian is used to create the locus-
specific KIR libraries. Sequencher with its library of full-length
genomic sequences and coding region sequences is used to
confirm the annealing site of PCR and sequencing primers, to
design new primers, and to aid in assigning alleles in unusual
sequences.
10. Amplicons generated in previous PCR reactions are a source of
sample contamination. By separating the source of the ampli-
cons (i.e., post-PCR activities as defined by thermal cycling
and subsequent steps) from the pre-PCR activities (as defined
by all steps up to and including assembly of the PCR reaction
just prior to placing in the thermal cycler), the potential for
contamination is greatly reduced. Ideally, the pre-PCR and
post-PCR procedures should be performed in two different
rooms, but, if not available, different areas of the laboratory
should be set aside. If all activities are to be performed in a
single room, pre-PCR activities should occur inside a laminar
flow hood, preferably equipped with a UV light. The walls of
the hood should be wiped with a freshly made 10% bleach
solution (1 part regular bleach:9 parts tap water) before pro-
cessing samples or preparing PCR samples. Dedicated equip-
ment (e.g., pipettors, test tube racks) and lab coats should be
set aside for pre-PCR procedures.
11. Typically, 200 mL of whole blood from a healthy individual will
yield 3–12 mg of DNA. Sequencing of each KIR locus requires
approximately 500 ng DNA. To sequence all the KIR loci,
5–10 mg of genomic DNA is required.
12. Never add Buffer AL directly to the protease. To obtain com-
plete lysis, the sample and the Buffer AL must be mixed imme-
diately and thoroughly.
464 L. Hou et al.

13. The speed of the quick spin should be above 1,000 rpm. Set
the speed to 8,000 rpm; press the button for 5 s and release to
achieve this speed.
14. DNA should be stored in a neutral to slightly basic buffered
solution to prevent degradation. Tris–EDTA (TE) buffer can
be used for storage. TE contains EDTA which has a high
affinity towards divalent ions like Ca+2 and Mg+2. These ions
are cofactors for many enzymes including nucleases that digest
DNA molecules. Since repeated access to a tube of genomic
DNA may introduce nucleases, TE buffer will protect DNA
from degradation during long-term storage. However, since
EDTA can bind divalent ions, it can inhibit Taq polymerase in
the PCR reaction. If DNA is stored in deionized water which
is often at an acidic pH, DNA degradation can occur by acid
hydrolysis.
15. Refer to the QIAamp® DNA Mini Kit handbook for trouble-
shooting problems.
16. It is helpful to initially assay for the presence or absence of KIR
genes using a sequence-specific priming assay as described in
Chap. 24. This will facilitate the selection of protocols to use
to isolate KIR genes for sequencing as described in Table 1.
Methods described in this chapter have been published (10–13)
(Hou, in preparation).
Some KIR haplotypes include fusion genes. For example,
KIR3DL1/KIR3DL2 hybrid alleles have been found in popu-
lations of recent African origin (13, 14). These alleles carry the
first five exons of KIR3DL1 and exons 6–9 of KIR3DL2. The
KIR3DL1 primer pairs in this protocol will amplify this chime-
ric gene. When sequencing amplicon B of KIR2DL4, be alert
for a single nucleotide deletion that removes the last nucle-
otide (811) of exon 7 in some alleles (e.g., KIR2DL4*008).
When sequencing KIR2DL5, it is possible that a cell may carry
three or four alleles, i.e., two alleles of KIR2DL5A and two
alleles of KIR2DL5B are potentially possible. An additional
two primer pairs listed in Table 1 will assist in clarifying the
allele calls in this situation. These pairs are each specific for a
subset of KIR2DL5 alleles. Sequencing primers used with
KIR2DL5 amplicon A will anneal to these two amplicons.
17. The polymerase and buffer used in the PCR reaction vary for
different loci and are described in Table 2. DMSO or 5M
betaine solution can improve and enhance the specificity of the
polymerase chain reaction. The volumes in each reaction of
MgSO4, DMSO, and 5M betaine solution are provided in
Table 2.
18. It is critical to have high-quality DNA for the PCR reaction. To
quantify the DNA and to determine its purity, read its optical
25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 465

density (OD) using a spectrophotometer. The NanoDrop spec-

trophotometer (e.g., NanoDrop ND-1000, NanoDrop
Technologies, Inc. Wilmington, DE, USA) uses very small
quantities of the solution so it or a similar instrument is recom-
mended. The DNA concentration at OD 260 nm should be
>10 ng/mL (OD260 × dilution factor × 50 = ng/mL). The purity
as measured by the ratio of the absorbance at 260 nm/absor-
bance at 280 nm (measuring protein contamination) should be
in the 1.65–1.9 range.
19. The thermal cycler should be calibrated at regular intervals
to ensure that the temperatures required for PCR are
achieved in all of the wells of the thermal cycler. This should
be done at least every 6 months or more frequently depend-
ing on the usage. The Driftcon Temperature Verification
System (CYCLERtest, Landgraaf, Netherlands) is one
instrument that might be used if this calibration is per-
formed in-house.
20. The molecular weight markers should be present as single sharp
bands. The cresol red dye runs at approximately 125 bp. Each
PCR reaction should yield a single bright band of the expected
size (see Table 2). (The deletion present in some KIR2DS4
alleles does not make a visible difference in the mobility of the
band compared to alleles without the deletion.) The presence
of additional bands suggests that the amplification conditions
were less stringent than required and the primer annealing tem-
perature should be raised until a single band is produced. The
absence of a band may indicate that the gene is absent (see Note
16) or that the amplification conditions are too stringent. To
reduce stingency, lower the annealing temperature until a single
strong band is produced. Amplification of a locus or of one of
two alleles at a locus may fail if the allele carries a nucleotide
sequence variation in a primer annealing site.
21. The AMPure kit will remove unincorporated primers, dNTPs,
and salts following the PCR.
22. Comparison of the intensity of staining of a reference mass lad-
der (see Note 4) to the staining intensity of an amplicon fol-
lowing gel electrophoresis can be used to estimate the amount
of amplified DNA in the reaction. In turn, this information can
be used to determine the amount of water used to elute purified
DNA from the AMPure beads. If the concentration of DNA is
low, elute with 30 mL instead of 50 mL of water.
23. Perform the protocol in the post-PCR laboratory since nested
PCR uses amplified DNA as a template. Use aliquots of PCR
reagents and do not return them to the pre-PCR room.
24. Probe 2DL2-999T targets nucleotide position 708 in exon 6
shared by all known KIR2DL2 alleles except KIR2DL2*004.
466 L. Hou et al.

Probe 2DL3-1316T targets nucleotide position 1024T in

exon 9 shared by all known KIR2DL3 alleles. If KIR2DL2*004
is present, the allele can be assigned based on amplicon A but
cloning or allele-specific nested PCR of the B and C amplicons
must be used to obtain the complete allele sequence. The strat-
egies used will depend on the other KIR genes found in the
sample and co-amplifying with KIR2DL2*004.
25. It is critical that the buffer be thawed on ice. HaploPrep
reagents must be always kept on ice when working with them
on the bench.
26. It is critical that the DNA is not sheared so avoid excessive
pipetting or vortexing.
27. It is critical that the solution be maintained at a high tempera-
ture to prevent renaturation of the DNA prior to exposure to
the HaploPrep reagents.
28. Be careful not to lose the pellet.
29. A known 22 bp deletion in some alleles of KIR2DS4 will make
sequencing difficult if such an allele is found together with an
allele lacking the deletion. The reading frame will be shifted
resulting in uninterpretable sequences in the region of the
deletion. In these cases, it is necessary to separate the two
alleles by cloning in order to obtain a clear sequence of each
allele in this region.
30. The amplified DNA should be obtained by PCR just prior to
cloning.
31. The efficiency at which inserts are obtained should be at least
70–80%. The white colonies contain inserted DNA (e.g.,
KIR2DS4); the blue colonies do not contain an insert.
32. It is essential to vortex buffer 3 until the salt is in solution.
33. Ensure that template DNA is of sufficiently high quality and is
not degraded. Avoid vigorous mixing or pipetting of the solu-
tion to prevent DNA from shearing.
34. The KIR sequencing primers flank each exon with the excep-
tion of exon 1 and the last two exons (exons 8 and 9). The
sequences of exon 1 for all loci except KIR2DL5 are obtained
using only an antisense primer. Since the PCR amplification
primers anneal just 5¢ of exon 1, it is not possible to obtain a
complete “read” of exon 1 sequence using either internal for-
ward primers or the forward PCR primers as sense strand
sequencing primers. The KIR2DL5 A amplicon includes 274 bp
of the 5¢ upstream region so that transcription factor binding
sites impacting gene expression (15) can be evaluated. For exons
8 and 9, one sequencing primer anneals 5¢ of exon 8 and the
second anneals 3¢ of exon 9 so that the resultant sequence
includes intron 8.
25 Killer Cell Immunoglobulin-Like Receptors (KIR) Typing by DNA Sequencing 467

35. All exon 1 sequence reactions require 5% DMSO. The thermal

cycler profile for the sequencing reaction for exon 1 is shown
in Table 10 and does not include a primer annealing step. The
sequence of exon 1 is very short and the antisense primer site
has repeated sequences so that higher denaturation and anneal-
ing temperatures are required.
36. It is recommended that locus-specific libraries be created to
facilitate the interpretation of KIR nucleotide sequences. The
intron 8 data are not analyzed so it doesn’t matter that the
intron 8 sequence in the library comes from a single allele. It is
also helpful to have the same length of nucleotides in the intron
8 library sequence so don’t insert the intron 8 sequences from
multiple alleles.
37. The 5¢ sequence for each KIR2DL5 allele can be found in the
IPD-KIR database, in GenBank and in publications.
38. The primarily heterozygous sequences are compared to a data-
base of known KIR sequences created in this section to identify
alleles. The library does not need to be created each time DNA
sequencing is performed. Manual inspection of the chromato-
graph should be performed to confirm assigned sequences and
to exclude closely related sequences. Be alert to the presence of
novel alleles.
The allele assignments for multiple loci should be consistent
with known telomere and centromere haplotype structures
(summarized in ref. (5)). For example, essentially all KIR hap-
lotypes carry the framework genes, KIR3DL3, KIR2DL4, and
KIR3DL2. Since KIR2DL2 and KIR2DL3 are alleles at a sin-
gle locus, the cell should not carry more than a total of 2 alleles
(e.g., two alleles of KIR2DL2 with KIR2DL3 absent, not two
alleles at KIR2DL2 and one allele at KIR2DL3). The same is
true for KIR3DL1 and KIR3DS1. The KIR2DL5 locus has
been duplicated; the two genes are termed KIR2DL5A and
KIR2DL5B. KIR2DL5A and KIR2DL5B should be associ-
ated with either KIR2DS3 or KIR2DS5 and specific combina-
tions of alleles at these loci have been observed (11, 16). It
should be noted that other KIR haplotypes have been described
at lower frequencies, for example, a haplotype with a duplica-
tion so that an individual carries two KIR3DL1 alleles and a
KIR3DS1 allele (14, 17, 18).
39. Poor quality sequences should not be interpreted and sequenc-
ing of those samples should be repeated.
40. The known KIR allele database, IPD-KIR, is updated at least
annually with new, modified or deleted alleles.
468 L. Hou et al.

References
1. Bashirova AA et al (2006) The killer 13. Jiang B et al (2010) The profile of KIR3DL1
immunoglobulin-like receptor gene cluster: and KIR3DS1 alleles in an African American
tuning the genome for Defense. Annu Rev population resembles that found in African
Genomics Hum Genet 7:277–300 populations. Tissue Antigens 76:64–66
2. Sanger F, Nicklen S, Coulson AR (1977) DNA 14. Norman PJ et al (2009) Meiotic recombina-
sequencing with chain-terminating inhibitors. tion generates rich diversity in NK cell receptor
Proc Natl Acad Sci USA 74:5463–5467 genes, alleles, and haplotypes. Genome Res
3. Khakoo SI, Carrington M (2006) KIR and dis- 19:757–769
ease: a model system or system of models? 15. Vilches C, Gardiner CM, Parham P (2000)
Immunol Rev 214:186–201 Gene structure and promoter variation of
4. Tomblyn M et al (2010) Decreased infections expressed and nonexpressed variants of the
in recipients of unrelated donor hematopoietic KIR2DL5 gene. J Immunol 165:6416–6421
cell transplantation from donors with an acti- 16. Ordonez D et al (2008) Duplication, mutation
vating KIR genotype. Biol Blood Marrow and recombination of the human orphan gene
Transplant 16:1155–1161 KIR2DS3 contribute to the diversity of KIR
5. Cooley S et al (2010) Donor selection for natu- haplotypes. Genes Immun 9:431–437
ral killer cell receptor genes leads to superior sur- 17. Gomez-Lozano N et al (2005) The silent
vival after unrelated transplantation for acute KIR3DP1 gene (CD158c) is transcribed and
myelogenous leukemia. Blood 116:2411–2419 might encode a secreted receptor in a minority
6. Moesta AK et al (2008) Synergistic polymor- of humans, in whom the KIR3DP1, KIR2DL4
phism at two positions distal to the ligand- and KIR3DL1/KIR3DS1 genes are dupli-
binding site makes KIR2DL2 a stronger cated. Eur J Immunol 35:16–24
receptor for HLA-C than KIR2DL3. 18. Martin MP et al (2003) Cutting edge: expan-
J Immunol 180:3969–3979 sion of the KIR locus by unequal crossing over.
7. Yawata M et al (2008) MHC class I-specific J Immunol 171:2192–2195
inhibitory receptors and their ligands structure 19. Uhrberg M et al (1997) Human diversity in
diverse human NK-cell repertoires toward a killer cell inhibitory receptor genes. Immunity
balance of missing self-response. Blood 7:753–763
112:2369–2380 20. Vilches C et al (2007) Facilitation of KIR geno-
8. Sharma D et al (2009) Dimorphic motifs in D0 typing by a PCR-SSP method that amplifies
and D1 + D2 domains of killer cell Ig-like recep- short DNA fragments. Tissue Antigens
tor 3DL1 combine to form receptors with 70:415–422
high, moderate, and no avidity for the complex 21. Gomez-Lozano N, Vilches C (2002) Geno-
of a peptide derived from HIV and HLA- typing of human killer-cell immunoglobulin-
A*2402. J Immunol 183:4569–4582 like receptor genes by polymerase chain reaction
9. Martin MP et al (2007) Innate partnership of with sequence-specific primers: an update.
HLA-B and KIR3DL1 subtypes against HIV- Tissue Antigens 59:184–193
1. Nat Genet 39:733–740 22. Murdoch S et al (2006) Detailed gene and
10. Hou L et al (2010) African Americans exhibit a allele content analysis of three homozygous
predominant allele in the midst of extensive KIR haplotypes. Tissue Antigens 68:72–77
KIR2DL1 allelic diversity. Tissue Antigens 23. Sun JY, Oki A, Senitzer D (2008) Alleles and
76:31–34 intron polymorphism of KIR3DL1 shown
11. Hou L et al (2010) Thirty allele-level haplo- by combination of allele group-specific prim-
types centered around KIR2DL5 define the ers and sequencing. Tissue Antigens 72:
diversity in an African American population. 578–580
Immunogenetics 62:491–498 24. Vilches C, Pando MJ, Parham P (2000) Genes
12. Hou L et al (2009) In contrast to other stimu- encoding human killer-cell Ig-like receptors
latory natural killer cell immunoglobulin-like with D1 and D2 extracellular domains all con-
receptor loci, several KIR2DS5 alleles predom- tain untranslated pseudoexons encoding
inate in African Americans. Hum Immunol a third Ig-like domain. Immunogenetics
70:733–737 51:639–646
 Chapter 26

An Overview of Methods Required to Evaluate Donor

NK Cell Alloreactivity for Haploidentical Haemopoietic
Stem Cell Transplantation
Andrea Velardi

Abstract
Donor-vs.-recipient NK cell alloreactivity has been established as a key therapeutic element in HLA
haplotypemismatched hematopoietic transplants in acute myeloid leukemia. NK cell allotherapy for leukemia
is deployed through stem cell transplantation and ensuing NK cell reconstitution across KIR ligand
mismatches. It is effected by functional NK cells which express inhibitory killer cell immunoglobulin-like
receptor(s) (KIRs) for self-class I ligand(s), sense missing expression of donor KIR ligand(s) in the recipient,
and mediate alloreactions. Donor-vs.-recipient NK cell alloreactivity is evaluated by integrating genetic,
phenotypic, and functional features.

Key words: HLA, KIR, Alloreactivity, Cytotoxicity, Acute leukemia, Haploidentical hematopoietic
transplantation

1. Introduction

Donor-vs.-recipient NK cell alloreactivity is a key therapeutic

element in the success of HLA haplotype-mismatched (“haploi-
dentical”) hematopoietic stem cell transplants (HSCT) for acute
leukemia. Milestones along the way towards this breakthrough
discovery were observations that: (1) extensive ex-vivo T-cell
depletion of the graft prevented graft-vs.-host disease (GvHD), in
haploidentical transplants for patients with severe combined
immunodeficiency (1); (2) a “megadose” of T-cell-depleted stem
cells ensured engraftment across MHC barriers in mice (2), and
across HLA barriers in clinical transplantation (3, 4); (3) human
NK cells exerted alloreactivity (5, 6); (4) transplantation from
haploidentical donors that were able to mount donor-vs.-recipient

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_26, © Springer Science+Business Media New York 2012

469
470 A. Velardi

NK cell alloreactions eradicated acute myeloid leukemia (AML),

favored engraftment, protected from GvHD and greatly improved
survival, as demonstrated by integrating clinical and preclinical
data ((7, 8) reviewed in refs. (9–13)).
Human NK cells possess clonally distributed inhibitory recep-
tors termed “killer cell immunoglobulin-like receptors” (KIRs)
that recognize allotypic determinants (“KIR ligands”) shared by
certain groups of HLA class I alleles. KIR2DL1 recognizes HLA-C
alleles with a Lys80 residue (HLA-Cw4 and related, “Group 2”
alleles), KIR2DL2 and KIR2DL3 recognize HLA-C with an Asn80
residue (HLA-Cw3 and related, “Group 1” alleles), KIR3DL1 is
the receptor for HLA-B alleles sharing the Bw4 specificity.
All KIR genes are randomly expressed and KIR distribution
varies on NK cells. Only NK cells which express inhibitory KIRs
for self-HLA ligands are functionally active as they become
“licensed/educated” upon interaction with self-HLA molecules
and thus enabled to exert alloreactivity against mismatched alloge-
neic targets which do not express self-HLA inhibitory KIR ligands.
NK cells which express, as their only inhibitory receptor for self, a
KIR whose ligand is a HLA class I group which is absent on allo-
geneic targets, sense the missing expression of the self-class I KIR
ligand and mediate alloreactions (“missing self” recognition).
Combined evidence from in vitro studies, murine models, and
clinical trials indicates the ability of NK cells to mediate donor-vs.-
recipient alloreactivity rests on “missing self recognition” which,
in humans, involves mismatching for each of the 3 KIR ligands
(C1, C2, Bw4) in the graft-vs.-host direction. In clinical
hematopoietic stem cell transplantation, NK cells mature in a bone
marrow microenvironment in which they are predominantly
exposed to donor HLA (on hematopoietic cells), which shapes
their repertoire to be both self (donor)-tolerant and recipient-
alloreactive. In fact, under such mismatch conditions, engrafted
stem cells give rise to an NK cell repertoire of donor origin which
includes alloreactive clones that kill recipient cryopreserved leukemic
cells (see Fig. 1). The beneficial effects of donor-vs.-recipient
NK alloreactivity in haploidentical transplants have been confirmed
in an updated analysis of 112 high-risk AML patients. The pres-
ence of NK alloreactivity was associated with significantly lower
relapse rates and better event-free survival (10). NK cell alloreac-
tivity has been shown to provide better protection from leukemia
relapse when exerted by maternal donors (14), presumably
because of the additive effect of memory T-cell immunity against
the child’s paternal HLA haplotype that results from maternal
exposure to fetal antigens during pregnancy. T-cell-depleted
megadose hematopoietic transplantation from NK alloreactive
donors has been reported to lower the risk of relapse also in pedi-
atric leukemias (15).
 26 An Overview of Methods Required to Evaluate Donor NK Cell Alloreactivity… 471

Fig. 1. Posttransplant regeneration of donor-vs.-recipient alloreactive NK cell repertoires. Top: absolute numbers of
potentially alloreactive NK cells (as evaluated by immune-fluorescence), i.e., of NK cells that express, as their only inhibi-
tory receptor for self, the KIR for which there is no ligand in the recipient (see Subheading 2.3.4). Bottom: frequencies of
alloreactive NK clones across each one of the three KIR ligand mismatches (C1, C2, Bw4), as documented by limiting
dilution cloning and cytotoxicity assays (see Subheading 2.3.5).

2. Methods
and Key Reagents
2.1. Haploidentical Conditioning of the recipient consists of 8 Gy total-body irradia-
Donor HSCT tion on day 9 in a single fraction at an instantaneous dose-rate of
Transplantation 0.16 Gy/min; lungs shielded to receive 0.04 Gy; thiotepa (5 mg/
Procedures kg daily) on days 8 and 7; fludarabine (40 mg/m2 daily) from day
7 to day 3; rabbit anti-thymocyte globulin (ATG) is given at 5 mg/
kg daily from days 5 to 2 (4, 8, 10). No immune suppression is
given after transplantation as GvHD prophylaxis; no G-CSF is
administered posttransplantation.

2.2. Graft Processing Peripheral-blood hematopoietic progenitor cells from the donor
are mobilized and collected as previously described (4, 8, 10).
CD34 positive cells are selected using the CliniMacs one-step pro-
cedure (Miltenyi Biotech, Bergisch Gladbach, Germany). Flow
cytometric analysis quantifies CD34, CD3, and CD20 positive
cells before and after selection. G-CSF stem cell mobilization and
graft processing. The stem cell “megadose” (i.e., 10 × 106 cells/kg
body weight) is routinely achieved by administering a short course
of granulocyte-colony stimulating factor (G-CSF) to the donor,
followed by leukapheresis and immunomagnetic-based CD34+
stem cell selection (see Note 1).
472 A. Velardi

2.3. Selection Full HLA haplotype-mismatched (“haplo-identical”) family donors

of Haploidentical are considered potentially able to exert donor-vs.-recipient NK cell
Donors Able to Mount alloreactivity when HLA-C and HLA-B typing shows KIR ligand
Donor-Versus- mismatches in the GvH direction, i.e., when the recipient does
Recipient NK Cell not possess one HLA-C allele group (C1 or C2) and/or the
Alloreactions HLA-Bw4 group which are present in the donor. Moreover, donors
must possess the relevant KIR gene(s) for missing self-recognition
2.3.1. Definition on recipient targets.

2.3.2. Identification Candidates are assessed for HLA compatibility by HLA typing at
of Suitable Donor-Recipient HLA-A, -B, -C, -DR, -DQ, and -DP (see elsewhere in this volume
Pairs for detailed methods). First, the recipient is HLA typed. Those
who express class I alleles belonging to all three class I groups
recognized by KIRs (HLA-C group 1, HLA-C group 2, and
HLA-Bw4 alleles) will block all NK cells from every donor and will
be resistant to alloreactive NK killing (approximately 1/3 of the
population). Patients who express only one or two of these allele
groups may find NK alloreactive donors. HLA typing on the family
will identify those haploidentical family member(s) who expresses
the class I group(s) missing in the recipient and who have, there-
fore, the potential to be a donor whose NK cells exert alloreactivity
against the recipient.

2.3.3. KIR Genotyping Donor KIR genotyping is performed to ensure the donor possesses
the relevant inhibitory KIR gene to exert alloreactivity (see Note
2). Moreover, KIR genotyping identifies donors who also possess
activating KIR genes (approximately 2/3 of Caucasian popula-
tion). Studies suggest activating KIRs may play a role in NK cell
alloreactivity and haploidentical transplantation (see Note 3).
Detailed methods for KIR typing are provided elsewhere in this
volume. In our laboratory, KIR typing is performed using a low
resolution PCR-SSP assay (KIR Genotyping Kit, Invitrogen, USA)
following the manufacturer’s instructions. This kit is designed to
identify 14 KIR genes (2DL1, 2DL2, 3DL1, 2DL4, 2DL5, 2DS1,
2DS2, 2DS3, 2DS4, 2DS5, 3DL1, 3DL2, 3DL3, 3DS1), 2 pseudo-
genes (2DP1 and 3DP1) and the common variants of KIR2DL5
(KIR2DL5A, KIR2DL5B), the KIR2DS4 allele (*001/002 and
*003), and KIR3DP1 allele (*001/002 and *003).

2.3.4. KIR Phenotyping KIR phenotyping is undertaken to identify potentially alloreactive

donor NK cells. Potentially alloreactive donor NK cells are those
cells that express, as their only inhibitory receptor for self, the KIR
for which there is no ligand in the recipient. Four-color
immunofluorescence analysis identifies KIR−/NKG2A+ vs. KIR+/
NKG2A−, CD56+/CD3− NK cells. The following mouse monoclo-
nal antibodies: allophycocyanin (APC)-conjugated anti-CD56
(IgG1; Miltenyi, Bergisch Gladbach, Germany), phycoerythrin
(PE)-Cy7-conjugated anti-CD3 (IgG1; BD Bioscience, San Diego,
 26 An Overview of Methods Required to Evaluate Donor NK Cell Alloreactivity… 473

CA), unconjugated anti-NKG2A (clone Z199, IgG2b) developed

with fluorescein isothiocyanate (FITC)-conjugated goat anti–
mouse IgG2b antibodies (Southern Biotech, Birmingham, AL),
are used in combination with the following PE-conjugated anti-
KIR antibodies (Bechman Coulter, Fullerton, CA): either anti-
KIR2DL2/3/S2 (clone GL183, IgG1), or anti-KIR2DL1/S1
(clone EB6B, IgG1), or anti-KIR3DL1/S1 (clone Z27, IgG1).
Two-color immunofluorescence analyses are used to visualize
NK cells expressing, as their only inhibitory receptor for self, a KIR
for which there was no class I ligand in the recipient. KIR2DL2/3/
S2 single-positive NK cells are identified with FITC-conjugated
anti-KIR2DL2/3/S2 antibody (clone CH-L, IgG2b, BD
Bioscience, Franklin Lakes, NJ) in combination with a cocktail
of PE-conjugated anti-KIR2DL1 (clone 143211, IgG1, R&D,
Minneapolis, MI), anti-KIR3DL1 (clone DX9, IgG1, Miltenyi),
and anti-NKG2A (clone Z199, IgG2b, Beckman Coulter, Brea,
CA) mouse antibodies. KIR2DL1 single-positive NK cells were
identified with PE-conjugated anti-KIR2DL1 (clone 143211,
IgG1, R&D, Minneapolis, MI) in combination with a cocktail of
FITC-coniugated anti-KIR3DL1 (clone DX9, IgG1, Miltenyi)
and anti-KIR2DL2/3S2 (clone CH-L, IgG2b) and anti-NKG2A
(clone Z199, IgG2b, Beckman Coulter, Brea, CA) mouse anti-
bodies developed with FITC-conjugated goat anti-mouse IgG2b
antibodies (Southern-Biotech). The same assays are used to moni-
tor posttransplant reconstitution of potentially alloreactive NK
cells (Fig. 1).

2.3.5. NK Cell Cloning This analysis aims to confirm donor NK alloreactivity against the
and Cytotoxicity Assay recipient by formally identifying donor NK cells which kill alloge-
neic recipient targets, such as PHA lymphoblasts and/or leukemia
cells (see Note 4). Large numbers of donor NK clones are gener-
ated by limiting dilution and cytotoxicity assays against recipient
target cells are used to detect the frequency of alloreactive NK
clones. Peripheral blood mononuclear cells are depleted of T cells
by negative anti-CD3 immunomagnetic selection (Miltenyi), are
plated under limiting-dilution conditions, activated with phytohe-
magglutinin (PHA; Biochrom KG, Berlin, Germany), and cultured
with interleukin-2 (Chiron BV, Amsterdam, Netherlands) and irra-
diated feeder cells. Feeder cells are obtained by pooling buffy coats
from 5 to 9 healthy donors. Such donors are not HLA typed as
PHA + IL2 activation allows efficient NK cell repertoire cloning
regardless of feeder cell HLA type. Cloning efficiencies ranges from
1 in 5 to 1 in 10 plated NK cells. Cloned NK cells are screened for
alloreactivity by standard 51-Cromium release cytotoxicity at an
effector-to-target ratio of 10:1 against recipient PHA lymphoblasts
and leukemic cells. Approximately 100 NK clones from each
person are screened. Clones exhibiting greater than 30% lysis are
scored as alloreactive. The assay is considered positive when the
474 A. Velardi

frequency of lytic clones was more than 1 in 50. The same assays
are used to monitor posttransplant reconstitution of alloreactive
NK cell clones (Fig. 1).

3. Notes

1. Median-infused cell doses per kilogram of recipient body

weight are as follows: CD34+, 13.8 × 106/kg (range, 5.1–
29.7 × 106/kg); CD3+, 1 × 104/kg (range, 0.04–3.0 × 104/kg);
and CD20+, 4.1 × 104/kg (range, 0.4–22.2 × 104/kg). The
number of infused CD34+ cells per kilogram is generally in the
range of 8–16 × 106/kg. Median recovery of CD34+ cells is
79%, and purity is 95%. Median T- and B-cell depletion is 4.5
and 3.2 log.
2. Most donors have the potential to exert NK alloreactions as
they possess a full complement of inhibitory KIR genes.
HLA-C group 1 receptor genes (KIR2DL2 and/or KIR2DL3)
are present in 100% of individuals, and the HLA-C group 2
receptor gene (KIR2DL1) in 97%. Therefore, the combination
of HLA-C group 2-positive donor/HLA-C group 2-negative
recipient which occurs in approximately 1.5% of HLA-C group
mismatched transplants, requires donor KIR2DL1 gene typing.
The HLA-Bw4 receptor gene (KIR3DL1) is found in approxi-
mately 90% of individuals.
3. Studies have suggested the specificity of alloreactive NK cells
across C1 mismatching is effected by the activating receptor
KIR2DS1 which binds C2 on targets (15–19). Such alloreac-
tive cells are relatively infrequent (approximately 1%) and are
only found in C1 homozygous individuals (who lack C2).
Patients who have all three KIR ligands (C1, C2, Bw4), who
are, therefore, not considered susceptible to donor-vs.-recipi-
ent NK cell alloreactivity, might benefit from transplantation
from haploidentical C1-homozygous, K2DS1+ donors who
possess KIR2DS1 alloreactive cells.
Additional results in haploidentical transplantation for pedi-
atric acute leukemia have emphasized the role of activating
KIR variants in NK cell alloreactivity. While donor KIR2DL1+
NK cells showed good alloreactivity against C2 negative tar-
gets, donor KIR2DL2+ and/or 2DL3+ NK cells in most cases
showed only poor allorecognition against target cells lacking
C1. Allorecognition of these C1 negative (C2 positive) targets
depended mainly on the NK cells possessing KIR2DS1 to bind
C2. These observations impact on the number of children who
might be cured of leukemia. They reduce the pool of NK
alloreactive donors because the KIR2DS1 gene is present in
 26 An Overview of Methods Required to Evaluate Donor NK Cell Alloreactivity… 475

only approximately 1/3 of Caucasians. This genetic restriction

does not apply to haploidentical transplantation for AML in
adults. In analyzing donor NK cell alloreactivity against C1
missing targets (either PHA blasts or AML cells), frequencies
of alloreactive NK cell clones did not differ significantly in
KIR2DS1 negative and positive donors, NK alloreactive reper-
toires in KIR2DS1 positive individuals were largely composed
of KIR2DS1-negative clones and, finally, negative clones killed
as efficiently as the positive (L. Ruggeri et al., unpublished).
Indeed, in an updated analysis of event-free survival of a series
of adult AML patients undergoing haploidentical transplanta-
tion, each of the 3 KIR ligand mismatches (C1, C2, Bw4)
provided a survival advantage over non-NK alloreactive (KIR
ligand-matched) transplants regardless of donor activating
KIR genetics ((10) and L. Ruggeri et al., unpublished).
4. When tested in large donor cohorts (10), functional analyses
detected high-frequency alloreactive NK clones against either
HLA-C1 or HLA-C2 group-mismatched allogeneic targets.
On the contrary, only 2/3 of HLA-Bw4-positive individuals
with the KIR3DL1 gene display alloreactive NK clones against
allogeneic HLA-Bw4-negative targets. Failure to detect allore-
active NK clones may be due to their highly variable frequen-
cies, or because certain allelic KIR3DL1 variants do not allow
receptor expression at the cell membrane (reviewed in ref.
(13)). In donor-recipient pairs that are not KIR ligand-mis-
matched in the graft-vs.-host direction, no donor alloreactive
NK clones are found, indicating that KIR ligand mismatching
is a prerequisite for NK cell alloreactivity (10).

References

1. Reisner Y, Kapoor N, Kirkpatrick D et al (1983) with one fully mismatched HLA haplotype. N
Transplantation for severe combined Engl J Med 339:186–193
immunodeficiency with HLA-A, B, D, DR 5. Ciccone E, Pende D, Viale O et al (1992)
incompatible parental marrow cells fraction- Evidence of a natural killer (NK) cell repertoire
ated by soybean agglutinin and sheep red blood for (allo) antigen recognition: definition of five
cells. Blood 61:341–348 distinct NK-determined allospecificities in
2. Bachar-Lustig E, Rachamim N, Li HW et al humans. J Exp Med 175:709–718
(1995) Nat Med 1:1268–1273 6. Colonna M, Brooks EG, Falco M et al (1993)
3. Aversa F, Tabilio A, Terenzi A et al (1994) Generation of allospecific natural killer cells by
Successful engraftment of T-cell-depleted hap- stimulation across a polymorphism of HLA-C.
loidentical “three-loci” incompatible trans- Science 260:1121–1124
plants in leukemia patients by addition of 7. Ruggeri L, Capanni M, Casucci M et al (1999)
recombinant human granulocyte colony-stim- Role of natural killer cell alloreactivity in HLA-
ulating factor-mobilized peripheral blood pro- mismatched hematopoietic stem cell transplan-
genitor cells to bone marrow inoculum. Blood tation. Blood 94:333–339
84:3948–3955 8. Ruggeri L, Capanni M, Urbani E et al (2002)
4. Aversa F, Tabilio A, Velardi A et al (1998) Effectiveness of donor natural killer cell allore-
Treatment of high-risk acute leukemia with activity in mismatched hematopoietic trans-
T-cell-depleted stem cells from related donors plants. Science 295:2097–20100
476 A. Velardi

9. Kärre K (2002) A Perfect Mismatch. Science 15. Pende D, Marcenaro S, Falco M et al (2009)
295:2029–2031 Anti-leukemia activity of alloreactive NK cells
10. Ruggeri L, Mancusi A, Capanni M et al (2007) in KIR ligand-mismatched haploidentical
Donor natural killer cell allorecognition of HSCT for pediatric patients: evaluation of the
missing self in haploidentical hematopoietic functional role of activating KIR and
transplantation for acute myeloid leukemia: redefinition of inhibitory KIR specificity. Blood
challenging its predictive value. Blood 110: 113:3119–3129
433–440 16. Witt CS, Christiansen FT (2006) The relevance
11. Velardi A (2008) Role of KIRs and KIR ligands of natural killer cell human leucocyte antigen
in hematopoietic transplantation. Curr Opin epitopes and killer cell immunoglobulin-like
Immunol 20:581–587 receptors in bone marrow transplantation. Vox
12. Christiansen FT, Velardi A (2009) Progress in Sang 90:10–20
understanding and exploiting the immune 17. Foley BA, De Santis D, Van Beelen E et al
response in solid organ and hemopoietic stem (2008) The reactivity of Bw4+ HLA-B and
cell transplantation. Curr Opin Immunol HLA-A alleles with KIR3DL1: implications for
21:522–524 patient and donor suitability for haploidentical
13. Velardi A, Ruggeri L, Mancusi A et al (2009) stem cell transplantations. Blood 112:435–443
Natural killer cell allorecognition of missing 18. Foley B, De Santis D, Lathbury L (2008)
self in allogeneic hematopoietic transplanta- KIR2DS1-mediated activation overrides
tion: a tool for immunotherapy of leukemia. NKG2A-mediated inhibition in HLA-C C2-
Curr Opin Immunol 21:525–530 negative individuals. Int Immunol 20:555–563
14. Stern M, Ruggeri L, Mancusi A (2008) Survival 19. Chewning JH, Gudme CN, Hsu KC (2007)
after T cell-depleted haploidentical stem cell KIR2DS1-positive NK cells mediate allore-
transplantation is improved using the mother sponse against the C2 HLA-KIR ligand group
as donor. Blood 112:2990–2995 in vitro. J Immunol 179:854–868
 Chapter 27

The Detection of NK Cell Alloreactivity

by Flow Cytometric CD107a Assay
Dianne De Santis, Bree Foley, Campbell S. Witt, and Frank T. Christiansen

Abstract
Natural killer (NK) cell alloreactivity can be exploited in haploidentical (one haplotype mismatched)
haematopoietic stem cell transplantation (HSCT) to prevent leukaemia relapse, rejection, and graft-vs-host
disease (GVHD) (Blood 94:333–339; Science 295:2097–2100). If NK cell alloreactivity is to be exploited
in HSCT, it is important to be able to reliably select donors who have NK alloreactivity towards the
patient. The detection of donor NK alloreactivity towards patient target cells has traditionally been evalu-
ated by NK cell cloning and 51Cr-release cytotoxicity assay. This approach is complex and time consuming
with results taking up to 6 weeks. Here, we detail a novel flow cytometric CD107a-based assay capable of
detecting NK cell alloreactivity in 14 days.

Key words: Natural killer cells, Killer immunoglobulin-like receptors, Alloreactivity, CD107a, Flow
cytometry

1. Introduction

1.1. Background NK cell cytotoxicity is regulated by a balance of inhibitory and

activating signals from receptors expressed on the cell surface. NK
cell cytotoxicity is inhibited when inhibitory receptors, such as
killer immunoglobulin-like receptors (KIR) interact with self-class
I HLA molecules on potential targets (1). The inhibitory KIR
recognise allelic epitopes on HLA-B and HLA-C molecules.
HLA-C alleles can be divided into two groups based on their
amino acid at position 80. C1-group alleles have an asparagine at
position 80 and are recognised by the inhibitory receptors
KIR2DL2 and KIR2DL3. C2-group alleles have a lysine at posi-
tion 80 and are recognised by KIR2DL1 (2–5). HLA-B alleles can
also be divided into two groups, Bw4 and Bw6, based on the

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_27, © Springer Science+Business Media New York 2012

477
478 D. De Santis et al.

amino acids 77–83. KIR3DL1 recognise HLA-B and some HLA-A

alleles with the Bw4 epitope (6–8). The range of different KIR
receptors expressed by NK clones in an individual is primarily
determined by the KIR genotype but an individual’s HLA geno-
type determines which KIR can be expressed as the only inhibi-
tory receptor on individual NK cell clones (9). Receptor selection
occurs during NK cell development to ensure only NK cells with
an inhibitory receptor for a self-ligand are permitted to become
armed for cytotoxicity (10).

1.2. Role of NK In a series of studies Ruggeri and colleagues showed that donor-
Alloreactivity vs-recipient NK cell alloreactivity resulted in less relapse, rejection,
in the Clinical Setting GVHD, and increased survival in human haploidentical transplants
and murine transplant models (11, 12). Most importantly in
haploidentical transplants, NK cells were shown to mediate GVL
effects without GVHD (11). Since the initial studies of Ruggeri
et al., a number of retrospective studies have been performed in
haploidentical and unrelated HSCT. However, many of the studies
failed to confirm the findings published by Ruggeri et al. The
extensive T-cell depletion used in Perugia compared to the other
studies has been considered as a possible explanation for the differ-
ences observed. However, other differences in transplant proto-
cols, such as conditioning regimen or stem cell source, or
heterogeneous cohorts with differences in the proportions of dis-
ease type could also be responsible for the differences observed
between the studies. If the findings of Ruggeri et al. can be
confirmed in other transplant settings using the same transplant
protocol allowing NK cell alloreactivity to be exploited in stem cell
transplants, it will be important to be able to reliably select donors
who are alloreactive towards the patient.
Donor NK alloreactivity towards a recipient can largely be
predicted if the recipient lacks one of the C1, C2, or Bw4 epitopes
that is present in the donor. However, this assumes that all HLA-C
and HLA-B alleles interact with the appropriate inhibitory KIR
receptors as predicted by the presence of the relevant amino acid
and that all alleles of the inhibitory KIR receptors interact the same
way with the HLA epitopes. Recently, these assumptions have been
shown not to be always true. The Bw4-associated HLA-B13 allele
has been shown not to interact with KIR3DL1 (13) while KIR2LD2
and KIR2DL3 may interact with both C1- and C2-group HLA-C
alleles (14–16). As there has been no comprehensive study to
determine which of the HLA alleles interact with KIR receptors, it
is not possible to reliably predict alloreactivity from the HLA and
KIR genotype. A functional assay is therefore required to confirm
NK alloreactivity.
 27 The Detection of NK Cell Alloreactivity by Flow Cytometric CD107a Assay 479

1.3. Assays NK cloning and 51chromium (Cr)-release cytotoxicity assay has

for the Detection of NK been used to detect potential NK alloreactivity. NK clones are
Cell Alloreactivity used in the cytotoxicity assay as effector cells. Briefly, enriched NK
cells are serially diluted (40, 20, 10, 5, 2 cells/100 μL) and cul-
1.3.1. NK Cloning
tured with irradiated pooled allogeneic PBMC feeder cells.
and Chromium-Release
Following 14–20 days of culture with IL-2 supplemented medium,
Cytotoxicity
each NK clone is subcultured and grown further to numbers
sufficient for use in cytotoxicity assays to determine NK cell cyto-
toxicity. As only a few percent of NK clones are expected to
demonstrate alloreactivity, 200 clones must be tested in order to
be confident of detecting the alloreactive clones. Traditionally, a
51
chromium (Cr)-release cytotoxicity assay is used in which each
NK clone is tested for the ability to kill a range of allogeneic
targets. These methods are labour intensive and time consuming
with results available in around 6 weeks.

1.3.2. The Flow Cytometric NK cells have a large number of preformed cytolytic granules in
CD107a Assay their cytoplasm. Following activation, NK cells release these gran-
ules to the site of cell–cell contact (immunological synapse). These
granules contain cytolytic proteins, such as perforin and granzyme,
that are involved in inducing target cell death (17). Lining the
membrane of the granules is the lysosomal-associated membrane
protein, LAMP-1 or CD107a (18). CD107a is a glycoprotein
representing approximately 50% of the proteins in the lysosomal
membrane. During degranulation, CD107a on the luminal mem-
brane is exposed in the immunological synapse and therefore acces-
sible for antibody binding. Betts et al. (19) showed CD107a
expression on the cell surface was a marker of cytotoxic CD8+ T
cell degranulation. Alter et al. (20) showed that CD107a can also
be detected on the surface of NK cells following stimulation with
target cells lacking HLA ligands. By modifying the method
described by Alter et al. we have developed a rapid flow cytometric
assay to detect KIR-dependent alloreactive NK cells. Here, we pro-
vide a detailed protocol for the rapid, efficient detection of donor
NK alloreactivity by a flow cytometric, CD107a-based assay.

2. Materials

1. Target EBV-transformed BLCL (Table 1) (see Note 1).

2. Peripheral blood lymphocyte (PBMC) feeder cells
(Subheading 3.1).
3. 1× PBS—17 g NaCl, 2.68 g Na2HPO42H2O (disodium hydro-
gen orthophosphate) and 0.78 g NaH2PO42H2O (sodium
dihydrogen orthophosphate) were dissolved in 2 L Milli-Q
water and the pH adjusted to 7.2.
480 D. De Santis et al.

Table 1
HLA types and KIR epitopes of BLCL used as target cells
used in the CD107a assay

KIR epitopes

Identifier C1 C2 BW4 HLA-A HLA-B HLA-C

“721.221” − − − − − −
“All” + + + 01, 03 27, 44 01, 04
“C1” − + + 02 57 06
“C2” + − + 33 58 03
“Bw4” + + − 03, 11 07, 35 04, 07

4. Heat inactivated foetal calf serum (HIFCS) (see Note 2).

Foetal calf serum (Thermo Trace, Melbourne, Australia) was
incubated at 56°C with shaking for 30 min, aliquotted and
stored at −20°C.
5. Ficoll-Paque Plus (Pharmacia, Uppsala, Sweden).
6. RPMI 1640 medium (Invitrogen, Carlsbad, USA) supplemented
with 500 IU/mL penicillin, 500 μg/mL streptomycin,
2 mM L-glutamine, and 10% sterile FCS.
7. DMSO (BDH, Poole, UK).
8. GammaCell 3000 ELAN gamma irradiator (MDS Nordion,
Canada).
9. Rosette-Sep Human NK enrichment cocktail (StemCell
Technologies, Canada).
10. 2-ME (ICN Biochemicals, OH, USA)—dilute 1:250 in RPMI
and store at −20°C. A 1:1,000 dilution of this stock is then
added to NK cell culture medium.
11. NK culture medium—RPMI 1640 medium supplemented
with penicillin (100 IU/mL), streptomycin (100 μg/mL),
L-glutamine (2 mM), 10% HIFCS, 1% non-essential amino
acids, 1% pyruvate, and 2-mercaptoethanol.
12. Centrifuge (Multifuge 3, Heraeus).
13. IL-2 (Chiron, Emeryville. USA).
14. 96-well round bottom culture plates (BD, Franklin Lakes,
USA).
15. Antibodies (see Table 2).
16. Monensin (6 μg/mL), (BD GolgiStop BD Biosciences).
17. Paraformaldehyde (Sigma P6148)—Dissolve 2 g paraform-
aldehyde in 100 mL PBS and heat solution to 56°C with
 27 The Detection of NK Cell Alloreactivity by Flow Cytometric CD107a Assay 481

Table 2
Antibodies used in the CD107a experiments

Antibody Fluorochrome Isotype Clone Dilution Supplier

CD56 PeCy7 IgG1 B159 1/40 BD Biosciences

CD158a PE IgG1 EB6B 1/10 Beckman Coulter
CD158b PE IgG1 GL183 1/10 Beckman Coulter
CD158a APC IgG1 EB6B 1/25 Beckman Coulter
CD158b APC IgG1 GL183 1/25 Beckman Coulter
CD158e PE IgG1 z27.3.7 1/10 Beckman Coulter
NKB1(CD158e) PE IgG1 DX9 1/10 BD Biosciences
CD107a FITC IgG1 H4A3 1/20 BD Biosciences

gently shaking. Cool solution and adjust pH to 7.4. Store

solution at 4°C wrapped in foil.
18. Trypan blue (ICN Biochemicals, Ohio USA)—Dissolve 0.5 g
trypan blue in 100 mL 1× PBS. Sterilise solution using a 0.22-
μM filter and store at room temperature.
19. Freezing medium—RPMI 1640, 20% HIFCS, and 10% DMSO.
20. Flow buffer—add 2% HIFCS and 0.1% sodium azide to 1×
PBS and store at 4°C.
21. BD FACSCanto Flow Cytometer with BD FACSDiva Software.
22. Adjustable pipettes (2, 10, 20, 100, and 200 μL).

3. Methods

3.1. Preparation 1. PBMC are isolated from ten random buffy coats. From the ten
of Peripheral Blood random donors, centrifuge 10 mL of acid citrate dextrose
Lymphocytes Feeder (ACD) blood at 300 × g for 10 min and harvest 1 mL of buffy
Cells coat from each donor (see Note 3).
2. Dilute each buffy coat 1:3 in PBS containing 2% HIFCS
(see Note 2), overlay diluted buffy coat onto Ficoll-paque™
PLUS and centrifuge at 1,200 × g for 20 min with no brake
(see Note 4).
3. Harvest the cells from the interface and wash 3 times with
PBS/2% HIFCS at 300 × g for 7 min.
4. Resuspend the cell pellet in 1 mL RPMI/10% HIFCS, count
and adjust to 106/mL.
482 D. De Santis et al.

5. Each of the donor’s cells at 106/mL should be pooled in equal

volumes. The total number of cells from each donor must be
the same. Any remaining donor cells can be stored in liquid
nitrogen for later use.
6. Irradiate pooled allogeneic PBMC at 30 Gy (see Note 5). After
irradiation, wash cells twice with RPMI/10% HIFCS at 300 × g
for 7 min to remove any free radicals in cell suspension gener-
ated by the irradiation process.
7. Following final wash, resuspend cell pellet in 1-mL NK cell
culture medium, count and adjust to 3 × 106/mL ready for use.
Irradiated feeder cells can also be stored in liquid nitrogen for
later use (see Note 6).

3.2. Enrichment of NK 1. To 4 mL of donor ACD blood add 200 μL of Rosette-Sep

Cells and Preparation Human NK cell enrichment cocktail (50-μL cocktail mix/mL
of 12-Day Polyclonal ACD blood) and incubate for 20 min at room temperature
NK Cultures with gentle mixing (see Note 7).
2. Following incubation, dilute with equal volume of PBS/2%
HIFCS and overlay onto 4 mL of Ficoll-paque™ PLUS.
Centrifuge at 1,200 × g for 20 min with no brake.
3. Following centrifugation, harvest the cell interface and wash
cell pellet 3 times with PBS/2% HIFCS at 300 × g for 7 min
(see Note 8).
4. During washing steps, prepare NK medium supplemented
with IL-2 to a final concentration of 200 IU/mL.
5. Resuspend the cell pellet in 1 mL NK cell medium supplemented
with IL-2 (as prepared in step 4), count and adjust to a final
concentration of 3 × 105/mL with supplemented NK medium.
6. To prepare polyclonal NK cultures, to each well of a 96-well
plate, add 100 μL of NK cell suspension from step 5 and
100 μL of the irradiated PBMC feeder cells from
Subheading 3.1, step 7. The ratio of NK cells:PBMC feeder
cells should be 1:10.
7. On day 3 and every third day, remove 100 μL of supernatant
from each well without disturbing the cell pellet and replace
with 100 μL of fresh NK culture medium supplemented with
200 IU IL-2/mL. Resuspend the cell pellet, cover the plate,
and incubate at 37°C 5% CO2.
8. After 12 days, pool the polyclonal NK cells into a tube, count
and adjust to a working cultured polyclonal NK cells can
also be frozen and stored in liquid nitrogen for later use (see
Note 6).

3.3. Preparation EBV-transformed BLCL expressing the HLA alleles that encode
of EBV-Transformed the presence or absence of the C1, C2, and Bw4 epitopes are used
BLCL Target Cells as target cells in the CD107a assay. The 721.221 cell line, a class I
 27 The Detection of NK Cell Alloreactivity by Flow Cytometric CD107a Assay 483

negative human BLCL, is also used as a positive control (see

Table 1, Note 1).
1. Maintain cell lines in NK cell culture medium and subculture
every 2–3 days to 1 × 105 cells/mL.
2. Subculture and feed target BLCL the day before use. On the
day of the assay, count the cells and adjust to 1 × 106/mL ready
for use.

3.4. CD107a 1. Before use in the CD107a assay, 12-day cultured polyclonal
Cytotoxicity Assay NK cells are activated by further culture for 48 h with NK
medium supplemented with 400 IU IL-2/mL (see Note 10).
To each well of a 96-well plate, add 100 μL of the cultured
polyclonal NK cells from Subheading 3.2, step 8 and 100 μL
of IL-2 supplemented NK medium. Incubate for 48 h at 37°C
5% CO2.
2. After the 48-h incubation, pool cells from each well, count and
adjust to 1 × 106/mL in fresh NK cell culture medium.
3. Activated cultured polyclonal NK cells from each donor should
be incubated alone (background control), with the HLA class
I negative 721.221 target (positive control), a target express-
ing all epitopes (negative control), target cells expressing C2
and Bw4 and lacking only C1 (C1−), target cells expressing
C1 and Bw4 and lacking only C2 (C2−), and target cells
expressing both C1 and C2 but not Bw4 (Bw4−) and if avail-
able transformed EBV-BLCL from the recipient of HSCT. In
duplicate wells, add 100 μL of 1 × 106/mL EBV-BLCL target
cells to wells of a 96-well culture plate followed by 100 μL of
1 × 106/mL activated cultured polyclonal NK cells (1:1 ratio).
4. To the effector/target mix, add 5 μL of anti-CD107a-FITC
(see Table 2) antibody and incubate at 37°C in 5% CO2 for 1 h.
5. After the hour incubation, add 20 μL of monensin (6 μg/mL)
at 1/15 dilution to the cells and further incubate for 5 h at
37°C in 5% CO2 (see Note 11).
6. Following the incubation, wash the cells by adding 100 μL of
flow buffer to each well followed by centrifugation at 200 × g
for 5 min and then incubate with an antibody cocktail of a
1/40 dilution of anti-CD56-PECy7 and 1/20 dilution of the
appropriate KIR antibodies (see Tables 2 and 3) to a total
volume of 50 μL in the dark for 15 min at 4°C (see Note 12).
7. Wash cells twice in 100 μL flow buffer by centrifugation at
200 × g for 5 min.
8. Finally, resuspend the cells in 180-μL flow buffer containing
1% paraformaldehyde and analyse.
484 D. De Santis et al.

Table 3
KIR antibody combinations and BLCL target cells to be used to detect NK
alloreactivity by the flow cytometric CD107a assay

NK cell alloreactivity KIR antibody

(relevant KIR receptor) BLCL target cell combinations NK cell gating

C1 (2DL2) C1− CD158b−APC, CD158a−/CD158b+/CD158e−

CD158a−PE,
CD158e−PE
C2 (2DL1) C2− CD158a−APC, CD158a+/CD158b−/CD158e−
CD158b−PE,
CD158e−PE
Bw4 (3DL1) Bw4− CD158e−PE, CD158a−/CD158b−/CD158e+
CD158a−APC,
CD158b−APC

3.5. Data Analysis The analysis method is critical in the detection of NK alloreactivity.
To minimise the effect of irrelevant receptors, it is important to
restrict the analysis to only NK cells expressing the inhibitory KIR
receptor relevant to the missing epitope against the relevant target
cell. An example of the analysis method used for the detection of
potential C1 alloreactive NK cells in a C1+, C2+ NK donor tested
against a C1− target is shown in Fig. 1. The analysis method is
described below with reference to Fig. 1.
1. Analyse the duplicate tubes on a BD FACSCanto™ flow cyto-
meter using BD FACSDiva software (or similar). The final
percentage of cells is an average of the duplicate tubes.
2. Gate events initially on forward (FSC) and side scatter (SSC)
to identify lymphocytes. In the example shown in Fig. 1a,
16,217 out of 96,301 events were identified as viable lympho-
cytes (P1).
3. Then, gate CD56 vs. SSC to acquire at least 10,000 CD56+
cells. Fig. 1b identifies 11,084 events corresponding to CD56
cells (P2).
4. Further gate these CD56+ cells and examine for KIR expres-
sion to identify cells expressing the KIR phenotype of interest,
e.g. CD158a+/158b− or CD158b+/158a−. In the example
shown in Fig. 1c gate P4 (upper left quadrant) identifies cells
expressing CD158b but not CD158a (2,129 events, 19.2%)
while gate P5 identifies cells expressing CD158a but not
CD158b (1,410 events, 12.7%).
5. To determine the percentage of cells with the KIR phenotype
of interest that also express CD107a+, use the gates generated
in step 4 (P4 and P5) and analyse in separate bivariate plots the
 27 The Detection of NK Cell Alloreactivity by Flow Cytometric CD107a Assay 485

a All Events b P1

200 250
(⫻ 1.000)

105
CD56 PE-Cy7-A
104
P2

FSC-A
150

103
100
50

102
50 100 150 200 250 50 100 150 200 250
SSC-A (⫻ 1.000) SSC-A (⫻ 1.000)

c d e
P2 P5 P4

105
105
105

Q1-2 Q2-2

CD158BA PC-A
CD158BA PC-A

P4
CD158A PE-A

5% 27.4%
104

104
104

Q2-1
103
103

103
Q3-2 Q4-2
102

102
102

P5 Q3-1 Q4-1

102 103 104 105 102 103 104 105 102 103 104 105
CD158A PE-A CD107A FITC-A CD107A FITC-A

Fig. 1. Flow cytometric histograms showing the detection of potential C1 alloreactive NK cells in a C1+, C2+ NK donor
tested against a C1− target. (a) Data gated on forward (FSC) and side scatter (SSC) to identify lymphocytes (16,217 events).
(b) Data gated on CD56 vs. SSC to acquire at least 10,000 CD56+ cells (11,084 events). (c) CD56+ cells expressing the KIR
phenotype of interest, e.g. CD158a+/158b− or CD158b+/158a−. Gate P4 (upper left quadrant ) identifies cells expressing
CD158b but not CD158a (2,129 events, 19.2%) while gate P5 identifies cells expressing CD158a but not CD158b (1,410
events, 12.7%). (d, e) The number of cells in gates P4 and P5, respectively, that express CD107a. The number of CD158a+/
CD158b−/CD107a+ cells is 71 (5%) (d, Q2-1) while the number of CD158b+/CD158a−/CD107a+ cells is 583 (27.4%)
(e, Q2-2). The frequency of C1 alloreactive NK cells is calculated by dividing the number of CD158b+/CD158a−/CD107a+
cells (583) by the total number of CD56+ cells (11,084), giving a frequency of 0.0526 (5.26%), followed by the subtraction
of the percentage of such cells present in the tube that lacks a target, i.e. the background control (0.16%, data not shown).
The frequency of C1 alloreactive NK cells in the example shown in figure is therefore 5.1%. The frequency of C2 alloreac-
tive NK cells against this target (C1−) is calculated in the same way, as expected the frequency of C2 alloreactive NK cells
in this example is low, 0.0056 (0.56%).

cells expressing the receptor of interest (e.g. CD158a) on one

axis and CD107a on the other axis. In the example shown in
Fig. 1, the number of CD158a+/CD158b−/CD107a+ cells is
71 (5%) (Fig. 1d, Q2-1) while the number of CD158b+/
CD158a−/CD107a+ cells is 583 (27.4%) (Fig. 1e, Q2-2).
6. To predict the potential for NK alloreactivity, determine the
number of CD56+ cells that express the relevant inhibitory
receptor to the missing epitope on target cells and lack other
inhibitory KIR and that are CD107a+ as a percentage of all NK
cells. For example in Fig. 1, the frequency of C1 alloreactive
NK cells is calculated by dividing the number of CD158b+/
CD158a−/CD107a+ cells (583) by the total number of CD56+
486 D. De Santis et al.

cells (11,084, see step 3), giving a frequency of 0.0526 (5.26%),

followed by the subtraction of the percentage of such cells
present in the tube that lacks a target, i.e. the background con-
trol (0.16%, data not shown). The frequency of C1 alloreactive
NK cells in the example shown in Fig. 1 is therefore 5.1%. The
frequency of C2 alloreactive NK cells against this target (C1−)
is calculated in the same way. As expected the frequency of C2
alloreactive NK cells in this example is only 0.0056 (0.56%).
In NK donors whose HLA type would predict inability to
generate alloreactive NK cells for a particular epitope, CD107a
expression is present at a frequency of <1%, while in most individu-
als whom alloreactivity is predicted, CD107a+ cells are present at a
frequency of >1% (see Fig. 3 in (21)).

4. Notes

1. Donor NK alloreactivity is identified when a target cell lacking

a single epitope is killed by NK cells expressing the relevant
KIR. The “C1−”, “C2−”, and “Bw4−” EBV-BLCL target cells
used in the cytotoxicity assay to demonstrate potential NK
alloreactivity were selected based on the absence of a single
HLA epitope as defined by the HLA type of each cell. The
particular cells in Table 1 are suggestions only which we find
useful and the use of these particular cells are not essential;
however, the absence of a single epitope is crucial. In addition
to the target cells used here, transformed EBV-BLCL cells
from a recipient of HSCT could also be used in the cytotox-
icity assay.
2. HIFCS was prepared by heating foetal calf serum (FCS) for
30 min at 56°C. Heating serum inactivates protease and com-
plement proteins which historically were thought to inhibit or
destroy cells under certain culture conditions. The use of FCS
(without heat inactivation) in the protocols described here was
not evaluated and therefore testing is recommended before
FCS is substituted for HIFCS.
3. In our initial experiments, ten random buffy coats were selected
to prepare the pooled PBMC feeders; however, in subsequent
experiments it was shown that alloreactive polyclonal NK
cultures could be successfully grown using feeder cells pre-
pared from at least five buffy coats.
4. The ficoll gradient should be centrifuged with no brake to
ensure that the cell interface layer is not disturbed while the
centrifuge slows down.
27 The Detection of NK Cell Alloreactivity by Flow Cytometric CD107a Assay 487

5. To ensure irradiation of PBMC feeder cells is successful, that

is, cells are sufficiently irradiated to inhibit cell growth but the
production of important growth factors is not, PBMC prepa-
rations should be adjusted to no more than 106/mL. The
amount of irradiation or the time of irradiation may need to be
adjusted in individual laboratories depending on instrumenta-
tion used. Initial experiments using 20 Gy was insufficient to
prevent PBMC growth and therefore the amount of irradia-
tion was increased to 30 Gy.
6. Cultured polyclonal NK cells can be stored in liquid nitrogen
for later use. To prepare cells for freezing, count and adjust the
cells to a final concentration of 10 × 106/mL in freezing
medium. Freezing medium should be made fresh and added to
the cells drop-wise on ice. To prepare frozen cells for use,
remove cells from liquid nitrogen and thaw quickly in a 37°C
waterbath and immediately transfer to RPMI/10%HIFCS.
Pellet cells by centrifugation at 300 × g for 7 min and resus-
pended in 1 mL RPMI/10% HIFCS or NK cell medium.
Dilute cells either in 1/2 or 1/5 in trypan blue, count and
check cell morphology using a Neubauer haemocytometer and
adjust to 3 × 106/mL for use.
7. Rosette-Sep has been optimised for use with whole blood;
however, cells can be enriched from buffy coat or isolated
PBMC. The concentration of nucleated cells in the sample
should not exceed 5 × 107 cells/mL, and red blood cells should
be present at a ratio of at least 30–50 RBCs per nucleated cell.
8. The NK cell interface following centrifugation during the NK
cell enrichment process is often not clearly visible. To ensure all
NK cells are harvested take up interface layer along with the
ficoll from just beneath the NK cell layer.
9. Polyclonal NK cultures should be harvested just before the
cultures reach the end of their growth cycle. The culture time
may vary between NK cell donors, some NK donors reach
the end of their growth cycle around day 11 while others at
day 14.
10. It is important that the NK polyclonal cultures are activated by
incubating for 48 h with 400 IU IL-2/mL before use in the
CD107a assay. We have shown that very few freshly isolated
NK cells which have not been cultured for 12 days expressed
CD107a in response to BLCL targets.
11. BD GolgiStop™ which contains monensin, blocks the intracel-
lular protein transport processes resulting in the accumulation
of CD107a in the cell allowing the glycoprotein to be detected
by flow cytometric analysis.
12. We used an antibody cocktail that included the antibodies
CD107a, CD56, CD158a, CD158b, and NKB1 which detect
488 D. De Santis et al.

the receptors of interest in our experiments, however, other or

additional antibodies can be used. We have noted in our exper-
iments that when using CD158-PE antibody obtained from
BD Biosciences at least a 1/5 dilution of antibody is required.

References
1. Valiante NM, Uhrberg M, Shilling HG, Lanier LL, Parham P (1997) Human diversity
Lienert-Weidenbach K, Arnett KL, D’Andrea in killer cell inhibitory receptor genes.
A, Phillips JH, Lanier LL, Parham P (1998) Immunity 7:753–763
Functionally and structurally distinct NK cell 10. Kim S, Poursine-Laurent J, Truscott SM,
receptor repertoires in the peripheral blood of Lybarger L, Song YJ, Yang L, French AR,
two human donors. Immunity 7:739–751 Sunwoo JB, Lemieux S, Hansen TH, Yokoyama
2. Colonna M, Spies T, Strominger JL, Ciccone WM (2005) Licensing of natural killer cells by
E, Moretta A, Moretta L, Pende D, Viale O host major histocompatibility complex class I
(1992) Alloantigen recognition by two human molecules. Nature 436:709–713
natural killer cell clones is associated with 11. Ruggeri L, Capanni M, Casucci M, Volpi I,
HLA-C or a closely linked gene. Proc Natl Tosti A, Perruccio K, Urbani E, Negrin RS,
Acad Sci USA 89:7983–7985 Martelli MF, Velardi A (1999) Role of natural
3. Colonna M, Brooks EG, Falco M, Ferrara GB, killer cell alloreactivity in HLA-mismatched
Strominger JL (1993) Generation of allospecific hematopoietic stem cell transplantation. Blood
natural killer cells by stimulation across a poly- 94:333–339
morphism of HLA-C. Science 260: 12. Ruggeri L, Capanni M, Urbani E, Perruccio K,
1121–1124 Shlomchik WD, Tosti A, Posati S, Rogaia D,
4. Moretta A, Vitale M, Bottino C, Orengo AM, Frassoni F, Aversa F, Martelli MF, Velardi A
Morelli L, Augugliaro R, Barbaresi M, Ciccone (2002) Effectiveness of donor natural killer cell
E, Moretta L (1993) P58 molecules as putative alloreactivity in mismatched hematopoietic
receptors for major histocompatibility complex transplants. Science 295:2097–2100
(MHC) class I molecules in human natural 13. Foley BA, De Santis D, Van Beelen E, Lathbury
killer (NK) cells. Anti-p58 antibodies reconsti- LJ, Christiansen FT, Witt CS (2008) The reac-
tute lysis of MHC class I-protected cells in NK tivity of Bw4+ HLA-B and HLA-A alleles with
clones displaying different specificities. J Exp KIR3DL1: implications for patient and donor
Med 178:597–604 suitability for haploidentical stem cell trans-
5. Vales-Gomez M, Reyburn HT, Erskine RA, plantations. Blood 112:435–443
Strominger J (1998) Differential binding to 14. Winter CC, Gumperz JE, Parham P, Long EO,
HLA-C of p50-activating and p58-inhibitory Wagtmann N (1998) Direct binding and func-
natural killer cell receptors. Proc Natl Acad Sci tional transfer of NK cell inhibitory receptors
USA 95:14326–14331 reveal novel patterns of HLA-C allotype recog-
6. Cella M, Longo A, Ferrara GB, Strominger JL, nition. J Immunol 161:571–577
Colonna M (1994) NK3-specific natural killer 15. Moesta AK, Norman PJ, Yawata M, Yawata N,
cells are selectively inhibited by Bw4-positive Gleimer M, Parham P (2008) Synergistic poly-
HLA alleles with isoleucine 80. J Exp Med morphism at 2 positions distal to the ligand-
180:1235–1242 binding site makes KIR2DL2 a stronger
7. Litwin V, Gumperz JE, Parham P, Phillips JH, receptor for HLA-C than KIR2DL3.
Lanier LL (1994) NKB1: a natural killer cell J Immunol 180:3969–3979
receptor involved in the recognition of poly- 16. Pende D, Marcenaro S, Falco M et al (2008)
morphic HLA-B molecules. J Exp Med Anti-leukemia acitivity of alloreactive NK cells
180:537–543 in KIR ligand-mismatched haploidentical
8. Gumperz JE, Litwin V, Phillips JH, Lanier LL, HSCT for pediatric patients: evaluation of the
Parham P (1995) The Bw4 public epitope of functional role of activating KIR and
HLA-B molecules confers reactivity with natural re-definition of inhibitory KIR specificity.
killer cell clones that express NKB1, a putative Blood 113:3119–3129
HLA receptor. J Exp Med 181:1133–1144 17. Djeu JY, Jiang K, Wei S (2002) A view to a kill:
9. Uhrberg M, Valiante NM, Shum BP, Shilling signals triggering cytotoxicity. Clin Cancer Res
HG, Lienert-Weidenbach K, Corliss B, Tyan D, 8:636–640
 27 The Detection of NK Cell Alloreactivity by Flow Cytometric CD107a Assay 489

18. Fukuda M (1991) Lysosomal membrane glyco- 20. Alter G, Malenfant JM, Altfeld M (2004)
proteins. Structure, biosynthesis, and intracellu- CD107a as a functional marker for the
lar trafficking. J Biol Chem 266:21327–21330 identification of natural killer cell activity. J
19. Betts MR, Brenchley JM, Price DA, De Rosa Immunol Methods 294:15–22
SC, Douek DC, Roederer M, Koup RA (2003) 21. De Santis D, Foley BA, John E, Senitzer D,
Sensitive and viable identification of antigen- Christiansen FT, Witt CS (2010) Rapid, flow
specific CD8+ T cells by a flow cytometric assay cytometric assay for NK alloreactivity reveals
for degranulation. J Immunol Methods 281: exceptions to rules governing alloreactivity. Biol
65–78 Blood Marrow Transplant 16(2):179–191
 Chapter 28

Clinical Production and Therapeutic Applications

of Alloreactive Natural Killer Cells
David H. McKenna, Diane M. Kadidlo, Sarah Cooley, and Jeffrey S. Miller

Abstract
Recent advances have improved our understanding of natural killer (NK) cell-mediated alloreactivity after
hematopoietic cell transplantation (HCT) or with adoptive transfer. NK cells contribute to a graft-versus-
leukemia effect and may play a role in preventing graft-versus-host disease or controlling infectious diseases
after allogeneic HCT. New discoveries in NK cell biology, including characterization of NK cell receptors
and their interactions with self-HLA molecules and a better understanding of the mechanism of NK cell
education have led to the development of novel strategies to exploit NK cell alloreactivity against tumors.
While early studies using autologous NK cells lacked efficacy, the use of adoptively transferred NK cells to
treat hematopoietic malignancies has been expanding. The production of allogeneic donor NK cells
requires efficient removal of T- and B cells from clinical-scale leukapheresis collections. The goal of this
chapter is to review NK cell biology, NK cell receptors, the use of NK cells as therapy and then to discuss
the clinical decisions resulting in our current good manufacturing practices processing and activation of
human NK cells for therapeutic use.

Key words: NK cells, Immunotherapy, Cell processing, Current good manufacturing practices

1. Introduction

1.1. NK Cell Biology NK cells are large granular lymphocytes that were first described in
the 1970s for their ability to lyse virally infected and tumor target
cells without MHC-restriction or prior sensitization (1, 2). Human
NK cells are found in the bone marrow, spleen, lymph nodes, and
peripheral blood (PB), where they comprise ~10–15% of the lym-
phocyte pool. NK cells play an important role in immune surveil-
lance and response to infections and link the innate and adaptive
immune systems (3, 4). Activated NK cells produce cytokines and
chemokines, including granulocyte colony stimulatory factor
(G-CSF), TNF, interferon gamma (IFN-γ), and transforming

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_28, © Springer Science+Business Media New York 2012

491
492 D.H. McKenna et al.

growth factor beta (TGF-β). The approximately 10% of NK cells

that are CD56bright proliferate better and produce more cytokines
(especially IFN-γ) than the more cytotoxic CD56dim subset, which
expresses Fc receptors that mediate antibody-dependent cellular
cytotoxicity (ADCC) (5–7). Cytokine-activated cells show more
expansion potential, increased cytokine production, and higher
cytotoxicity against targets than do resting NK cells (8). NK cells
respond to IL-2, IL-15, and IL-21, all of which signal via the IL-2
receptor γ chain (9–11), and to the combination of IL-12 and
IL-18 which is an especially strong inducer of IFN-γ production
(12). IL-2 is the only US Food and Drug Administration (FDA)-
approved common γ chain cytokine available to activate NK cells
in vivo or ex vivo.

1.2. NK Cell Receptors In 1985, Ljunggren and Karre described the phenomenon of “miss-
Control Function, ing self recognition,” whereby the loss of MHC class I expression
Self-Tolerance, renders autologous targets more sensitive to NK-mediated killing.
and Alloreactivity The mechanism by which these innate killer cells recognize class I
loss in tumors or virally infected cells (13) was explained with the
discovery of killer-cell immunoglobulin-like receptors (KIR). Human
KIR gene content can be simplified into two haplotypes, each of
which contain the three framework genes (KIR3DL3, KIR2DL4,
and KIR3DL2), and a variable number of activating and inhibitory
genes from the centromeric (Cen) or telomeric (Tel) ends. KIR A
haplotypes contain five inhibitory receptors (2DL1, 2DL3, 3DL1,
3DL2, 3DL3), and just one activating receptor (2DS4). KIR B hap-
lotypes have variable gene content and contain more activating KIR.
Binding of higher affinity inhibitory KIR to their cognate ligands,
self-class I HLA, suppresses NK cell effector responses, including
cell-mediated lysis and cytokine release (14). KIR2DL1, KIR2DL2/
KIR2DL3, and KIR3DL1 bind HLA class I C2, C1 and Bw4 alleles,
respectively. In contrast, the natural ligands for activating KIR remain
unknown.
Several other families of activating and inhibitory receptors affect
NK cell function. The NKG2 family of C-type lectin receptors can be
inhibitory (NKG2A) or activating (NKG2C/E), and recognize non-
classical HLA-E (15). NKG2D recognizes stress-induced molecules,
such as MHC class I polypeptide-related sequence A/B (MICA and
MICB) or viral particles (16). Other receptors include the natural
cytotoxicity receptors (NCR) NKp30, NKp46 and NKp44, DNAM-
1, and Nectin-2 (CD122), 2B4, leukocyte-associated immunoglobu-
lin-like receptor-1 (LAIR-1) (17, 18), and Ig-like (ILT) receptors
(19–23). NK cell function is determined by the net sum of signals
delivered through inhibitory and activating NK cell receptors.
The mechanism by which NK cells acquire self-tolerance and
alloreactivity has been referred to as “NK cell education.” Several
models have been proposed to explain the integration of inhibitory
receptor expression with the acquisition of effector functions.
 28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 493

These models show that human NK cells lacking inhibitory recep-

tors are hyporesponsive (24, 25) and therefore self-tolerant. As
strategies to use NK cells therapeutically are considered, it is criti-
cally important to understand effector function in vivo. For exam-
ple, although reconstituting NK cells are abundant after transplant,
their KIR expression is decreased and dysregulated KIR expression
correlates with clinical outcomes (26, 27). As delayed NK cell edu-
cation may result in NK cells with abnormal function, adoptive
transfer of mature NK cells, educated in the donor, may be a more
attractive strategy than using NK cells derived from hematopoietic
stem cells which must develop in the patient.

1.3. Clinical NK cells mediate lysis of hematologic malignancies and variety of

Applications solid tumors, including breast, ovarian, hepatocellular and colon
of NK Cells cancers, neuroblastoma, and melanoma (28–33). The therapeutic
efficacy of NK cells is primarily controlled by inhibitory receptor
interactions and the relative resistance of some tumors may be due
to a higher expression of class I HLA. Coengagement of activating
NK cell receptors such as NKG2D also plays an important role in
immune surveillance and tumor lysis (34, 35).
The two main strategies to harness the therapeutic power of
alloreactive NK cells are: (1) hematopoietic cell transplantation
(HCT) (36) and (2) adoptive transfer of NK cells (37). The Perugia
group first proposed the KIR-ligand incompatibility model, which
predicts that donor-derived NK cells will be alloreactive when recip-
ients lack C2, C1, or Bw4 alleles that are present in the donor.
Many groups, including our own (38–40), have tested the clinical
efficacy of selecting donors for NK cell therapy or transplantation
based on their predicted alloreactivity against the host. The poten-
tial benefits include: (1) decreased rates of GVHD (41, 42), (2)
decreased rates of graft rejection mediated by NK cell lysis of host
T cells, (3) better anti-tumor activity via direct cytotoxicity (43),
(4) improved engraftment mediated by NK cell release of hematopoi-
etic cytokines (44, 45), and (5) enhanced immune reconstitution
and decreased infections (46–48). The main benefit of KIR ligand
mismatch is seen in myeloid malignancies (49), but less so after
T-cell replete transplants (50, 51). The inconsistent effects of KIR-
ligand mismatch and KIR-ligand absence strategies may be due to
differences in NK cell development and education based on the
stem cell source, conditioning regimen, extent of T-cell depletion,
and the use of post-HCT immunosuppression (52–55). Another
approach to exploit the beneficial effects of NK cells after HCT is to
consider the full KIR genotype of the donors. Donors with certain
KIR B haplotypes (defined on http://www.ebi.ac.uk/ipd/kir/
donor_b_content.html) are associated with less relapse and improved
survival after unrelated donor HCT for AML (56, 57).
The other therapeutic approach is to use NK cells for adoptive
transfer. Several studies have shown that activation of autologous
494 D.H. McKenna et al.

NK cells lacked efficacy (58). Presumably, inhibition of NK cells by

“self” HLA molecules expressed by the tumor resulted in poor
tumor lysis. Additionally, NK cells from cancer patients exhibit
functional defects, which may prevent them from killing targets or
producing cytokines capable of mediating an immune response. To
overcome limitations of autologous NK cell therapy, in 2005 we
pioneered the use of haploidentical NK cells in patients with
advanced cancer (37). The safety and success of adoptive NK cell
infusions was established in a trial using haploidentical, related-
donor NK cell products with subcutaneous IL-2 to induce in vivo
expansion of alloreactive NK cells which persisted in vivo for up to
1 month (37). In patients with refractory leukemia, this persistence
and expansion correlated with achieving a clinical remission in
those treated with refractory disease having failed standard therapy.
NK cell expansion is dependent on a lymphodepleting chemother-
apy regimen given to create space and on the enhancement of
endogenous cytokines to support the expansion of donor NK cells
(59). This preparative therapy also has the potential to decrease
immune suppressive factors released by tumor cells, which may
interfere with adoptive transfer (60, 61). Addition of a nonmy-
eloablative dose of 400 cGy of TBI followed by a haploidentical
CD34-selected stem cell product from the same donor, resulted in
improved rates of successful in vivo NK cell expansion and clinical
remission, but with late complications including deaths due to
infection (62). A similar approach has been used in a pediatric
cohort, where haploidentical NK cell infusions were given with
T-cell and B-cell deplete haploidentical grafts to treat neuroblas-
toma (63). Other groups have used NK cell donor lymphocyte
infusions (DLI) after haploidentical HCT to consolidate engraft-
ment in adults with AML (64) or children with leukemia and solid
tumors (65, 66). Clinically effective strategies exploiting NK cell
alloreactivity to treat cancer require robust NK cell development
and education. Ex vivo expansion of NK cell products can increase
the cell dose and allow for additional activation of the products.
Further improvements may be made by selecting donors based on
their NK receptor profile, by using concurrent exogenous cytokine
therapy with IL-15, which is in phase I testing at the NCI (67), by
manipulation of tumor ligand expression or interactions with other
immune cells. For example, eliminating host regulatory T cells,
which can suppress NK cell proliferation and killing, may also
improve the immune effector functions of the expanding NK cells
(68). Ultimately, combination therapy using several strategies at
once will likely prove most successful.

1.4. Clinical Factors Studies using autologous NK cells did not require extensive pro-
Influencing cessing because the non-NK cells (T cells, B cells, and monocytes),
Development of an also autologous to the patient, were not felt to pose a significant
NK Cell Product risk even after ex vivo IL-2 activation. However, when crossing
28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 495

allogeneic barriers, several issues need to be considered. The goal

was to collect an adequate NK cell dose from a single leukapheresis
product. The primary concern is the known risk of graft-versus-
host disease mediated by donor T cells. Even small T-cell numbers
in a blood transfusion can mediate potentially lethal graft-versus-host
disease if the recipient is immunocompromised by cancer or other
pathology (69). Therefore, initial NK cell enrichment strategies
used T-cell depletion alone accomplished by removal of CD3+
cells. Using this approach, T cells which comprised 63% cells in the
starting product, decreased to <1% after processing, a 2.7 log T-cell
depletion (70). Successful adoptive transfer is defined by persis-
tence and in vivo expansion of donor NK cells. Noting variability
in the magnitude of NK cell expansion, we hypothesized that con-
taminating B cells or monocytes might inhibit NK cells after adop-
tive transfer. We tested a more rigorous processing approach,
positively selecting CD56+ NK cells after CD3 depletion. While
this processing yielded a highly pure NK cell product, with mini-
mal contaminating B cells, T cells or monocytes, unfortunately the
NK cells failed to persist and expand in vivo when tested in patients
with refractory AML. This failure may be explained by (1) the
importance of monocyte accessory cells for in vivo NK cell expan-
sion, as has been noted in vitro (71) or (2) the approximate 40–50%
decrease in NK cell dose in the final product after both CD3 deple-
tion and CD56 selection. Irrespective of the exact reason, the poor
clinical results and lack of NK cell expansion prompted us to revert
back to T-cell depletion alone for subsequent trials where we have
verified the anti-leukemia activity of haploidentical NK cells (72).
We have infused over 100 haploidentical NK cell products and
have noted some unusual complications due to allogeneic donor B
cells persisting in the recipient. We observed several cases of post-
transplant lymphoproliferative disease (PTLD), a potentially lethal
complication (73). This Epstein-Barr virus-driven lymphoprolifera-
tion was in B cells of donor origin infused with the NK cell product,
and patients receiving NK cells as part of a transplant or who had failed
a prior transplant were most susceptible to this complication. We have
also seen two patients with immune hemolysis (one clinically significant
and one clinically silent) as a result of passenger B lymphocyte syn-
drome (PBLS) (74). Both of these patients received adoptive NK cell
therapy for solid tumors and had not had prior transplants. We dem-
onstrated that contaminating B cells in the NK cell product were
responsible for PBLS. Therefore, our current GMP processing utilizes
immunomagnetic selection to achieve both CD3 and CD19 deple-
tion to minimize these risks while preventing GVHD (75).
There are several ways to manufacture NK cell products using
immunomagnetic selection-based methods. The method described
below is now our standard practice for processing NK cells
for clinical trials in both nontransplant and transplant settings.
This method involves CD3 (T cell) and CD19 (B cell)-depletion of
496 D.H. McKenna et al.

a nonmobilized peripheral blood mononuclear cell apheresis collec-

tion followed by an overnight culture and activation with IL-2. This
process takes approximately 9 h in total [4 h (preculture processing);
3 h (postculture processing); 2 h (lot release testing)].

2. Materials

2.1. Equipment 1. Miltenyi Biotec CliniMACS Cell Selection System (Miltenyi

Biotec, Bergisch Gladbach, Germany).
2. COBE 2991 Cell Processor (CaridianBCT, Lakewood,
Colorado, USA).
3. Incubator, 5% CO2/37°C.

2.2. Main Supplies 1. CliniMACS PBS/EDTA Buffer (Miltenyi Biotec).

and Reagents 2. CliniMACS CD3 and CD19 Microbeads (Miltenyi Biotec).
2.2.1. Day 1 (Preculture) 3. CliniMACS Tubing Set (LS) (Miltenyi Biotec).
Processing 4. COBE 2991 Processing Set (CaridianBCT).
5. Human serum albumin (HSA) 25%.
6. Pall SQ40S Filter (Pall Corp., Port Washington, New York,
USA).
7. Transfer Bags, 150–2,000 mL.
8. Interleukin-2.
9. Fluorinated ethylene propylene (FEP) Bags: 290, 500, 730 mL
(AFC, Gaithersburg, MD, USA).
10. Human AB serum, heat-inactivated.
11. X-VIVO 15 Medium (Lonza Walkersville, Inc., Walkersville,
MD, USA).

2.2.2. Day 2 (Postculture) 1. COBE 2991 Processing Set (CaridianBCT).

Processing 2. HSA (5%).
3. HSA (25%).
4. Double Coupler Adapter.
5. Transfer Bag (150, 300 mL).

3. Methods

3.1. cGMP Production In general, clinical production and administration of novel cell
of NK Cells therapies in the USA requires submission to and approval by the
FDA for an investigational new drug (IND) application. Further,
 28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 497

the manufacturing of such products typically requires the infra-

structure of a “cGMP facility” which includes not only the physical
structure (e.g., clean rooms, air handlers, etc.), but the appropriate
expertise in manufacturing, quality assurance (QA), and materials
management (76).

3.2. Procurement 1. Prior to leukapheresis, allogeneic donors must undergo a clini-

of the Leukapheresis cal evaluation for donor eligibility, including a medical history
screening and testing for transmissible diseases comparable to
current standard practice for whole blood donations (see Note 1).
We have focused on related donor haploidentical NK cells but
it is not known whether matching in the adoptive transfer
setting is of benefit. Donors must be in good general health by
established criteria (see Note 2).
2. It is not yet known whether donor choice based on KIR-ligand
mismatch or KIR genotyping provides any clinical benefit and
these strategies are being tested.
3. A nonmobilized mononuclear cell collection (i.e., leukaphere-
sis) is collected with an apheresis instrument using established,
standard collection procedures.
4. Aseptic technique is followed to prevent contamination of the
collection and infection at the site of the intravenous catheter(s).
5. The leukapheresis is then transported immediately in a validated
shipping container to the laboratory. Shipping containers should
be durable, leak-proof, resistant to breakage and able to with-
stand temperature extremes. Container selection will depend
upon distance traveled and mode of transportation.

3.3. Cell Processing The apheresis product is T cell (CD3)-depleted and B cell (CD19)-
depleted using the Miltenyi Biotec CliniMACS® Cell Selection
System and CD3 and CD19 MicroBeads and reagent (Miltenyi
Biotec, Bergisch Gladbach, Germany). Up to 2.50 × 1010 total
nucleated cells (TNC) may be labeled with the MicroBeads and
separated on an LS column (Depletion Program 2.1) which is
placed in the magnetic field of the CliniMACS® device. Cells are
resuspended at 2 × 106 cells/mL in X-VIVO 15, without gentami-
cin and phenol red (Cambrex BioScience, Walkersville, Maryland),
supplemented with 1,000 U/mL IL-2 (Chiron Corporation,
Emeryville, CA) and 10% human AB serum, heat-inactivated
(Valley Biomedical Products and Services, Inc., Winchester, VA) in
VueLife™ Teflon® (FEP) bags (American Fluoroseal Corporation,
Gaithersburg, MD). Cells are then incubated overnight for 8–16 h
(37°C/5% CO2). Cells are washed twice with a 5% HSA, and the
final formulation consists of nucleated cells in approximately
80–150 mL of 5% HSA solution (<100 × 106 NC/mL). The target
dose for infusion is £ 8 × 107 NC/kg.
498 D.H. McKenna et al.

3.4. Day 1 (Preculture) 1. General cleanroom preparation

Processing (a) Review and documentation of current cleanroom sanitiza-
3.4.1. Pre-CD3 Depletion tion and biological safety cabinet calibration records.
2. Leukapheresis receipt and inspection
(a) Documentation of arrival time and collection center.
(b) Review label information, including: product name, vol-
ume, product identifier, and additives used.
(c) Confirm product identity by
● Performing ABO/Rh blood typing from sample taken
from the product.
● Verifying product label information corresponds with
physician cell processing order.
(d) Conduct a gross inspection of product/bag ensuring
product bag is intact and product is free of clots/
discoloration.
(e) Review of donor infectious disease testing and donor eligi-
bility determination.
3. Sampling for QC testing
(a) Nucleated cell (NC) count; differential; flow cytometry
(CD3, CD20, CD56, 7-AAD viability).
4. Determination of volume and initial NC count
(a) Calculate the TNC in the product bag.
(b) If TNC number exceeds the limit, calculate the volume of
cells to remove from the bag (see Note 3). The remaining
cells may be cryopreserved for future use or discarded.

3.4.2. CD3/CD19 Depletion 1. Culture medium preparation

(a) Inspect bottles of X-VIVO 15 and heat-inactivated type
AB human serum for part number, expiry date, container
integrity, and appearance of reagent
(b) Add 110 mL of heat-inactivated type AB human serum to
each 1,000-mL bottle of X-VIVO 15
(c) Filter prepared medium with a 0.22-μm filter unit, label
accordingly, and place into 37°C incubator to warm
2. Supplement CliniMACS buffer with 0.5% HSA
(a) Add 20 mL of 25% HSA to each buffer bag and label
accordingly
3. Wash cells
(a) Inspect a COBE 2991 processing set for expiry date and
integrity
(b) Load processing set onto COBE 2991 for wash procedure
and attach hemostats to all lines
28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 499

(c) Attach bag of cells and one bag of CliniMACS buffer/0.5%

HSA to the processing set
(d) In manual mode, set the COBE 2991 settings as follows:
centrifuge speed—1,200 rpm (198 × g); super out rate
450 mL/min; minimum agitate time—30 s; super out
volume—600 mL
(e) Open hemostat on cell suspension line and transfer the
cells to the COBE processing bag
(f) Fill COBE processing bag with CliniMACS buffer/0.5%
HSA. Perform one wash (15 min at 1,200 rpm, 198 × g),
and expel as much supernatant as possible into waste bag
4. Label CD3+ and CD19+ cells
(a) Remove one vial each of CliniMACS CD3 and CD19
Microbeads from refrigerator; inspect vial and 150-mL
transfer bag for expiry date and integrity
(b) Label bag appropriately and insert dispensing pin into one
port
(c) Add 52 mL CliniMACS buffer/0.5% HSA and entire vials
of CD3 and CD19 Microbeads to transfer bag; inject
approximately 90 mL of air from biosafety cabinet into
bag, and mix well
(d) Attach transfer bag to blue line of COBE processing set;
remove hemostat attached to blue line tubing; add
Microbeads/buffer solution to processing bag; clear the
blue line of as many Microbeads as possible using air from
transfer bag. Place hemostat on the blue line
(e) Lift seal weight and open sliding covers to remove the
COBE Processing Bag from the instrument
(f) Resuspend cells in bag by gently massaging
(g) Return bag to centrifuge bowl, close sliding covers, and
replace seal weight
(h) Push in rapid succession Start/Spin, Super out, and
Agitate/Wash In; incubate product with constant agita-
tion for approximately 30 min
(i) Fill COBE Processing bag with CliniMACS buffer/0.5%
HSA; remove air from processing bag and hemostat all lines
(j) Press Start/Spin; wash product once at 1,500 rpm (300 × g)
for 15 min; at end of wash press Super Out; express as
much supernatant as possible into waste bag; press Stop
(k) Heat seal all tubing lines except the line attached to the
buffer bag
(l) Remove COBE Processing bag and attached buffer bag
from COBE 2991 and place into biosafety cabinet; detach
and remove remainder of processing set and discard
500 D.H. McKenna et al.

(m) Transfer cell suspension from COBE Processing bag into

transfer bag; rinse COBE Processing bag with small amount
of CliniMACS buffer/0.5% HSA and add diluted cells to
transfer bag; dilute cell suspension with CliniMACS
buffer/0.5% HSA to final volume of approximately 100 mL
(n) Remove sample for NC count
5. CD3/CD19 depletion
(a) Turn on CliniMACS and choose Depletion Program 2.1
(b) When prompted indicate LS tubing set, NC concentra-
tion, percentage of labeled cells (70%), and product
volume
(c) The CliniMACS program will calculate volume of buffer
needed for the depletion process and volumes expected
for collection in the negative fraction bag, the buffer waste
bag, and cell fraction bag; adjust bags accordingly (if
>400 mL)
(d) Inspect CliniMACS LS tubing set and transfer bag for
expiry date and integrity
(e) Label bags accordingly; the transfer bag will be the cell
collection bag and should be labeled “CD3/CD19 nega-
tive”; label the bag on the tubing set designated “negative
fraction” as “CD3 positive”
(f) Load the tubing set onto the CliniMACS instrument and
attach the required amount of buffer solution
(g) Perform prime and integrity testing following prompts
(h) Connect the bag containing washed cells to the Pall filter
and proceed with the cell separation
(i) Once the separation is completed, heat seal and detach
both fractions (i.e., CD3−/CD19− and CD3+/CD19+)
and set aside for further processing; discard remainder of
tubing set
6. Volume reduction of CD3−/CD19− fraction
(a) Weigh, count, and calculate volume and perform NC
count (see Note 4)
(b) Inspect COBE processing set for expiry date and
integrity
(c) Load set onto COBE 2991; place hemostats on all lines
(d) In manual mode, set the COBE 2991 settings as follows:
centrifuge speed—1,500 rpm (300 × g); super out rate
450 mL/min; minimum agitate time—N/A; super out
volume—600 mL
(e) Open the hemostat on the cell suspension line and transfer
cells to the COBE processing bag
 28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 501

(f) Volume reduce product at 1,500 rpm (300 × g) for 3 min

per spin; at the end of super out, expel as much superna-
tant as possible into the waste bag
(g) Remove the COBE processing bag from the COBE 2991
and place in the biosafety cabinet; discard the remainder of
the processing set
(h) Transfer the cell suspension into a transfer bag; add
warmed cell culture medium (X-VIVO 15) that has been
heated in a 37°C incubator to the drained COBE process-
ing bag and add rinse to the transfer bag
7. Culture preparation and QC testing of CD3−/CD19− and
CD3+/CD19+ fractions
(a) Weigh and calculate volume. Remove a small aliquot of
cells from product bag and perform NC count
(b) Inspect vial of IL-2 for expiry date, integrity, gross appear-
ance. Do not use reagent if it appears cloudy, contains pre-
cipitates or clumps.
(c) Add warmed cell culture medium (X-VIVO 15) to obtain
a target cell concentration of 2 × 106 nucleated cells/mL
(d) Add IL-2 to obtain a concentration of 1,000 U/mL of
culture medium
(e) Remove samples for QC testing of CD3−/CD19− fraction
(e.g., flow cytometry, cytotoxicity, sterility) and CD3+/
CD19+ fraction (e.g., NC count, flow cytometry)

3.4.3. Overnight Culture 1. Inspect appropriately sized culture bags (FEP) for expiry date
in X-VIVO 15/IL-2 and integrity
2. Transfer cells into culture bags and incubate in 37°C/5% CO2
incubator for 8–16 h

3.5. Day 2 (Postculture) 1. General cleanroom preparation

Processing (a) Review and documentation of current room sanitization
3.5.1. Wash and and biological safety cabinet calibration records
Resuspension 2. Volume reduction/wash (2×)
(a) Remove bags of cells from incubator
(b) Inspect an appropriate number of bottles of 5% HSA for
expiry date, integrity, and gross appearance
(c) Inspect COBE processing set for expiry date and integrity
(d) Load set onto COBE 2991; place hemostats on all lines
(e) In manual mode, set the COBE 2991 settings as follows:
centrifuge speed—1,500 rpm (300 × g); super out rate
450 mL/min; minimum agitate time—30 s; super out
volume—600 mL
502 D.H. McKenna et al.

(f) Open the hemostat on the cell product line and transfer
cells to the COBE processing bag
(g) Volume reduce product at 1,500 rpm (300 × g) for 3 min
per spin; at the end of super out, expel as much super-
natant as possible into the waste bag
(h) Fill the processing bag with 5% HSA; remove air from pro-
cessing bag; wash product twice at 1,500 rpm (300 × g)
for 3 min per spin; expel as much supernatant as possible
into the waste bag
(i) Heat seal and remove the processing bag from the COBE
2991; detach and discard the remainder of the processing
set
(j) Transfer the cells into an appropriately sized transfer bag;
rinse the processing bag with 5% HSA; and add to the
transfer bag
(k) Dilute the final product to approximately 100 mL total
volume with 5% HSA
3. Dose adjust (if necessary)
(a) Calculate available dose and determine if a portion of cells
needs to be removed

3.6. Sampling for QC/ 1. All lot release testing is performed on the final product prior to
Lot Release Testing release of the product for infusion. Lot release criteria are
shown in Table 1.

Table 1
Natural killer cells: lot release criteria

Assay Test method Specification

Viability 7-AAD, flow cytometry ³ 70.0%

(see Note 5)
NK cell (CD56+/CD3−) enumeration Flow cytometry (see Note 5) ³ 30.0% (see Note 7)
(see Notes 6 and 7)
T cell (CD3+) enumeration Flow cytometry (see Note 5) <3.00 × 105 T cells/kg
(see Note 8)
B cell (CD20+) enumeration Flow cytometry (see Note 5) <3.0%
(see Note 9)
Endotoxin LAL method per FDA guidance <5.0 EU/kg
Gram stain Clinical microbiology lab No organisms
NC dose (see Note 10) Hematology analyzer £ 8.00 × 107/kg
 28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 503

2. Sterility testing is performed on the final product; however, it

is not completed prior to release of the NK cell product. In the
event that sterility testing becomes positive, lab staff immedi-
ately notifies QA and the medical director who contacts the
patient physician for appropriate clinical action.

3.7. Final Labeling and 1. Label appropriately and include the following statement:
Release “Caution: New Drug—Limited by Federal Law to Investigational
Use” per 21 CFR 312.6
2. Release NK cell product once lot release testing has been
completed, passed, and reviewed by Quality Assurance unit.

4. Notes

1. Donors are typically related (sibling, parent, offspring, parent

or offspring of an HLA identical sibling) HLA-haploidentical
(3–5 of 6 HLA-A, B, and C) match to recipient, between 14
and 75 years of age, and not pregnant.
2. Infectious disease testing typically includes assays for the detec-
tion of HIV, HCV, and WNV (by nucleic acid testing), anti-
HIV I/II, anti-HTLV I/II, anti-HBc Ab, HBsAg, anti-HCV,
anti-CMV, Chagas disease, and Treponema pallidum (by
serology).
3. The maximum number of NCs to be loaded on the CliniMACS
is 2.5 × 1010. Loading beyond this limit may require a second
column.
4. Remove small aliquot of cells from the bag (1 mL) and per-
form either a manual or automated nucleated cell count.
5. At our institution, all flow cytometry readouts for lot release
are performed by a CLIA certified clinical laboratory. Adherence
to these lot release criteria depend on the quality of these read-
outs. Since these lot criteria are for clinical safety, we recom-
mend that these measures be performed in a certified clinical
laboratory (where routine quality control is required) rather
than a research laboratory.
6. Failure to meet NK cell content lot release in the final product
occurs occasionally and is most often a result of a low NK cell
content in the starting apheresis product (products containing
<5% NK cells) rather than a failure of the CD3 and CD19
depletion. Products with low NK cell content have been infused
safely (with FDA communication by PI) but it is still unknown
what NK cell content threshold, if any, will correlate with clini-
cal efficacy.
504 D.H. McKenna et al.

7. Lot release for NK cell purity has been used as 20% in non-
transplant settings of NK cell infusion.
8. The average T-cell content of the final product in a recent non-
transplant trial contained a median-infused T-cell dose of
5.6 × 104 CD3+ T cells/kg. Caution should be used in setting
this parameter depending on the clinical setting relative to the
risk of transfusion-associated graft-versus-host disease. For
example, the threshold for T-cell dose may need to be lower
when NK cell products are used as a part of haploidentical
transplant. Safe T-cell doses may vary based on the clinical set-
ting where the NK cell product is administered.
9. Absolute safe thresholds for B cell content relative to uncom-
mon complications associated with product B cells have not
been firmly established. These thresholds may vary depending
on the immune competence of the recipient.
10. An upper dose limit has been set but is rarely reached based on
the processing of a single apheresis product (15 L processing
volume; roughly 5 h collection). Larger or multiple collection
above this threshold have not been tested for safety.

References
1. Herberman RB, Ortaldo JR (1981) Natural 9. Carson WE, Giri JG, Lindemann MJ, Linett
killer cells: their roles in defenses against dis- ML, Ahdieh M, Paxton R et al (1994)
ease. Science 214(4516):24–30 Interleukin (IL) 15 is a novel cytokine that acti-
2. Kiessling R, Petranyi G, Klein G, Wigzel H vates human natural killer cells via components
(1975) Genetic variation of in vitro cytolytic of the IL-2 receptor. J Exp Med 180(4):
activity and in vivo rejection potential of non- 1395–1403
immunized semi-syngeneic mice against a mouse 10. Trinchieri G (2003) Interleukin-12 and the
lymphoma line. Int J Cancer 15(6):933–940 regulation of innate resistance and adaptive
3. Moretta L, Ferlazzo G, Bottino C, Vitale M, immunity. Nat Rev Immunol 3(2):133–146
Pende D, Mingari MC et al (2006) Effector and 11. Young HA, Ortaldo J (2006) Cytokines as criti-
regulatory events during natural killer-dendritic cal co-stimulatory molecules in modulating the
cell interactions. Immunol Rev 214:219–228 immune response of natural killer cells. Cell Res
4. Walzer T, Dalod M, Robbins SH, Zitvogel L, 16(1):20–24
Vivier E (2005) Natural-killer cells and den- 12. Singh SM, Yanagawa H, Hanibuchi M, Miki T,
dritic cells: “l’union fait la force”. Blood Okamura H, Sone S (2000) Augmentation by
106(7):2252–2258 interleukin-18 of MHC-nonrestricted killer
5. Cooper MA, Fehniger TA, Caligiuri MA (2001) activity of human peripheral blood mononu-
The biology of human natural killer-cell subsets. clear cells in response to interleukin-12. Int J
Trends Immunol 22(11):633–640 Immunopharmacol 22(1):35–43
6. Fehniger TA, Caligiuri MA (2001) Interleukin 13. Ljunggren HG, Karre K (1990) In search of the
15: biology and relevance to human disease. ‘missing self’: MHC molecules and NK cell rec-
Blood 97(1):14–32 ognition. Immunol Today 11(7):237–244
7. Caligiuri MA (2008) Human natural killer cells. 14. Vales-Gomez M, Reyburn HT, Mandelboim
Blood 112(3):461–469 M, Strominger JL (1998) Kinetics of interac-
8. Rayner AA, Grimm EA, Lotze MT, Chu EW, tion of HLA-C ligands with natural killer cell
Rosenberg SA (1985) Lymphokine-activated inhibitory receptors. Immunity 9(3):337–344
killer (LAK) cells. Analysis of factors relevant to 15. Houchins JP, Yabe T, McSherry C, Bach FH
the immunotherapy of human cancer. Cancer (1991) DNA sequence analysis of NKG2, a
55(6):1327–1333 family of related cDNA clones encoding type II
 28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 505

integral membrane proteins on human natural reconstitution is altered by T cells in the graft
killer cells. J Exp Med 173(4):1017–1020 and correlates with clinical outcomes after unre-
16. Lopez-Botet M, Angulo A, Guma M (2004) lated donor transplantation. Blood
Natural killer cell receptors for major histocom- 106(13):4370–4376
patibility complex class I and related molecules 28. Bottino C, Castriconi R, Moretta L, Moretta A
in cytomegalovirus infection. Tissue Antigens (2005) Cellular ligands of activating NK recep-
63(3):195–203 tors. Trends Immunol 26(4):221–226
17. Moretta L, Bottino C, Pende D, Castriconi R, 29. Soriani A, Zingoni A, Cerboni C, Iannitto ML,
Mingari MC, Moretta A (2006) Surface NK Ricciardi MR, Di Gialleonardo V et al (2009)
receptors and their ligands on tumor cells. ATM-ATR dependent up-regulation of
Semin Immunol 18(3):151–158 DNAM-1 and NKG2D ligands on multiple
18. Meyaard L (2008) The inhibitory collagen myeloma cells by therapeutic agents results in
receptor LAIR-1 (CD305). J Leukoc Biol enhanced NK cell susceptibility and is associ-
83(4):799–803 ated with a senescent phenotype. Blood
19. Riteau B, Menier C, Khalil-Daher I, Martinozzi 113(15):3503–3511
S, Pla M, Dausset J et al (2001) HLA-G1 co- 30. Stein MN, Shin J, Gudzowaty O, Bernstein
expression boosts the HLA class I-mediated AM, Liu JM (2006) Antibody-dependent cell
NK lysis inhibition. Int Immunol 13(2): cytotoxicity to breast cancer targets despite
193–201 inhibitory KIR signaling. Anticancer Res
20. Riteau B, Rouas-Freiss N, Menier C, Paul P, 26(3A):1759–1763
Dausset J, Carosella ED (2001) HLA-G2, -G3, 31. Pende D, Spaggiari GM, Marcenaro S, Martini
and -G4 isoforms expressed as nonmature cell S, Rivera P, Capobianco A et al (2005) Analysis
surface glycoproteins inhibit NK and antigen- of the receptor-ligand interactions in the natu-
specific CTL cytolysis. J Immunol 166(8): ral killer-mediated lysis of freshly isolated myel-
5018–5026 oid or lymphoblastic leukemias: evidence for
the involvement of the Poliovirus receptor
21. Shiroishi M, Tsumoto K, Amano K, Shirakihara
(CD155) and Nectin-2 (CD112). Blood
Y, Colonna M, Braud VM et al (2003) Human
105(5):2066–2073
inhibitory receptors Ig-like transcript 2 (ILT2)
and ILT4 compete with CD8 for MHC class I 32. Re F, Staudacher C, Zamai L, Vecchio V, Bregni
binding and bind preferentially to HLA-G. M (2006) Killer cell Ig-like receptors ligand-
Proc Natl Acad Sci USA 100):8856–8861 mismatched, alloreactive natural killer cells lyse
primary solid tumors. Cancer 107(3):640–648
22. Cerboni C, Achour A, Warnmark A, Mousavi-
Jazi M, Sandalova T, Hsu ML et al (2006) 33. Armeanu S, Bitzer M, Lauer UM, Venturelli S,
Spontaneous mutations in the human CMV Pathil A, Krusch M et al (2005) Natural killer
HLA class I homologue UL18 affect its binding cell-mediated lysis of hepatoma cells via specific
to the inhibitory receptor LIR-1/ILT2/CD85j. induction of NKG2D ligands by the histone
Eur J Immunol 36(3):732–741 deacetylase inhibitor sodium valproate. Cancer
Res 65(14):6321–6329
23. Kirwan SE, Burshtyn DN (2005) Killer cell
Ig-like receptor-dependent signaling by Ig-like 34. Guerra N, Tan YX, Joncker NT, Choy A,
transcript 2 (ILT2/CD85j/LILRB1/LIR-1). J Gallardo F, Xiong N et al (2008) NKG2D-
Immunol 175(8):5006–5015 deficient mice are defective in tumor surveil-
lance in models of spontaneous malignancy.
24. Anfossi N, Andre P, Guia S, Falk CS, Roetynck Immunity 28(4):571–580
S, Stewart CA et al (2006) Human NK cell edu-
cation by inhibitory receptors for MHC class I. 35. Brenner CD, King S, Przewoznik M, Wolters I,
Immunity 25(2):331–342 Adam C, Bornkamm GW et al (2010)
Requirements for control of B-cell lymphoma
25. Cooley S, Xiao F, Pitt M, Gleason M, McCullar by NK cells. Eur J Immunol 40(2):494–504
V, Bergmann TL et al (2007) A subpopulation
36. Ruggeri L, Capanni M, Urbani E, Perruccio K,
of human peripheral blood NK cells that lacks
Shlomchik WD, Tosti A et al (2002)
inhibitory receptors for self MHC is develop-
Effectiveness of donor natural killer cell allore-
mentally immature. Blood 110:578–586
activity in mismatched hematopoietic trans-
26. Shilling HG, McQueen KL, Cheng NW, plants. Science 295(5562):2097–2100
Shizuru JA, Negrin RS, Parham P (2003)
37. Miller JS, Soignier Y, Panoskaltsis-Mortari A,
Reconstitution of NK cell receptor repertoire
McNearney SA, Yun GH, Fautsch SK et al
following HLA-matched hematopoietic cell
(2005) Successful adoptive transfer and in vivo
transplantation. Blood 101(9):3730–3740
expansion of human haploidentical NK cells in
27. Cooley S, McCullar V, Wangen R, Bergemann patients with cancer. Blood 105(8):
TL, Spellman S, Weisdorf DJ et al (2005) KIR 3051–3057
506 D.H. McKenna et al.

38. Davies SM, Ruggieri L, DeFor T, Wagner JE, 48. Farrell HE, Vally H, Lynch DM, Fleming P,
Weisdorf DJ, Miller JS et al (2002) Evaluation Shellam GR, Scalzo AA et al (1997) Inhibition
of KIR ligand incompatibility in mismatched of natural killer cells by a cytomegalovirus MHC
unrelated donor hematopoietic transplants. class I homologue in vivo. Nature 386(6624):
Killer immunoglobulin-like receptor. Blood 510–514
100(10):3825–3827 49. Giebel S, Locatelli F, Lamparelli T, Velardi
39. Farag SS, Bacigalupo A, Eapen M, Hurley C, A, Davies S, Frumento G et al (2003) Survival
Dupont B, Caligiuri MA et al (2006) The effect advantage with KIR ligand incompatibility in
of KIR ligand incompatibility on the outcome hematopoietic stem cell transplantation from
of unrelated donor transplantation: a report unrelated donors. Blood 102(3):814–819
from the center for international blood and 50. Beelen DW, Ottinger HD, Ferencik S,
marrow transplant research, the European Elmaagacli AH, Peceny R, Trenschel R et al
blood and marrow transplant registry, and the (2005) Genotypic inhibitory killer immuno-
Dutch registry. Biol Blood Marrow Transplant globulin-like receptor ligand incompatibility
12(8):876–884 enhances the long-term antileukemic effect of
40. Miller JS, Cooley S, Parham P, Farag SS, unmodified allogeneic hematopoietic stem cell
Verneris MR, McQueen KL et al (2007) Missing transplantation in patients with myeloid leuke-
KIR ligands are associated with less relapse and mias. Blood 105(6):2594–2600
increased graft-versus-host disease (GVHD) 51. Lowe EJ, Turner V, Handgretinger R, Horwitz
following unrelated donor allogeneic HCT. EM, Benaim E, Hale GA et al (2003) T-cell
Blood 109(11):5058–5061 alloreactivity dominates natural killer cell allore-
41. Shlomchik WD, Couzens MS, Tang CB, McNiff activity in minimally T-cell-depleted HLA-non-
J, Robert ME, Liu J et al (1999) Prevention of identical paediatric bone marrow transplantation.
graft versus host disease by inactivation of host Br J Haematol 123(2):323–326
antigen-presenting cells. Science 52. Clausen J, Kircher B, Auberger J, Schumacher
285(5426):412–415 P, Ulmer H, Hetzenauer G et al (2010) The
42. Lundqvist A, McCoy JP, Samsel L, Childs R (2007) role of missing killer cell immunoglobulin-like
Reduction of GVHD and enhanced antitumor receptor ligands in T cell replete peripheral
effects after adoptive infusion of alloreactive Ly49- blood stem cell transplantation from HLA-
mismatched NK cells from MHC-matched donors. identical siblings. Biol Blood Marrow Transplant
Blood 109(8):3603–3606 16(2):273–280
43. Russell JH, Ley TJ (2002) Lymphocyte- 53. Linn YC, Phang CY, Lim TJ, Chong SF, Heng
mediated cytotoxicity. Annu Rev Immunol KK, Lee JJ et al (2010) Effect of missing killer-
20:323–370 immunoglobulin-like receptor ligand in recipients
44. Murphy WJ, Keller JR, Harrison CL, Young undergoing HLA full matched, non-T-depleted
HA, Longo DL (1992) Interleukin-2-activated sibling donor transplantation: a single institution
natural killer cells can support hematopoiesis experience of 151 Asian patients. Bone Marrow
in vitro and promote marrow engraftment Transplant 45(6):1031–1037
in vivo. Blood 80(3):670–677 54. Gagne K, Busson M, Bignon JD, Balere-Appert
45. Siefer AK, Longo DL, Harrison CL, Reynolds ML, Loiseau P, Dormoy A et al (2009) Donor
CW, Murphy WJ (1993) Activated natural killer KIR3DL1/3DS1 gene and recipient Bw4 KIR
cells and interleukin-2 promote granulocytic ligand as prognostic markers for outcome in
and megakaryocytic reconstitution after synge- unrelated hematopoietic stem cell transplanta-
neic bone marrow transplantation in mice. tion. Biol Blood Marrow Transplant 15(11):
Blood 82(8):2577–2584 1366–1375
46. Chen C, Busson M, Rocha V, Appert ML, 55. Triplett BM, Horwitz EM, Iyengar R, Turner
Lepage V, Dulphy N et al (2006) Activating V, Holladay MS, Gan K et al (2009) Effects of
KIR genes are associated with CMV reactiva- activating NK cell receptor expression and NK
tion and survival after non-T-cell depleted cell reconstitution on the outcomes of unre-
HLA-identical sibling bone marrow transplan- lated donor hematopoietic cell transplantation
tation for malignant disorders. Bone Marrow for hematologic malignancies. Leukemia
Transplant 38(6):437–444 23(7):1278–1287
47. Stern M, Elsasser H, Honger G, Steiger J, 56. Cooley S, Trachtenberg E, Bergemann TL,
Schaub S, Hess C (2008) The number of acti- Saeteurn K, Klein J, Le CT et al (2009) Donors
vating KIR genes inversely correlates with the with group B KIR haplotypes improve relapse-
rate of CMV infection/reactivation in kidney free survival after unrelated hematopoietic cell
transplant recipients. Am J Transplant 8(6): transplantation for acute myelogenous leuke-
1312–1317 mia. Blood 113(3):726–732
 28 Clinical Production and Therapeutic Applications of Alloreactive Natural Killer Cells 507

57. Cooley S, Weisdorf DJ, Guethlein LA, Klein JP, in childhood acute myeloid leukemia. J Clin
Wang T, Le CT et al (2010) Donor selection for Oncol 28(6):955–959
natural killer cell receptor genes leads to supe- 67. Lugli E, Goldman CK, Perera LP, Smedley J,
rior survival after unrelated transplantation for Pung R, Yovandich JL et al (2010) Transient
acute myelogenous leukemia. Blood 116(14): and persistent effects of IL15 on lymphocyte
2411–2419 homeostasis in nonhuman primates. Blood
58. Burns LJ, Weisdorf DJ, DeFor TE, Vesole DH, 116(17):3238–3248
Repka TL, Blazar BR et al (2003) IL-2-based 68. Barao I, Hanash AM, Hallett W, Welniak LA,
immunotherapy after autologous transplanta- Sun K, Redelman D et al (2006) Suppression
tion for lymphoma and breast cancer induces of natural killer cell-mediated bone marrow
immune activation and cytokine release: a cell rejection by CD4+ CD25+ regulatory T
phase I/II trial. Bone Marrow Transplant cells. Proc Natl Acad Sci USA 103(14):
32(2):177–186 5460–5465
59. Dudley ME, Wunderlich JR, Robbins PF, Yang 69. Bordin JO, Heddle NM, Blajchman MA (1994)
JC, Hwu P, Schwartzentruber DJ et al (2002) Biologic effects of leukocytes present in trans-
Cancer regression and autoimmunity in patients fused cellular blood products. Blood
after clonal repopulation with antitumor lym- 84(6):1703–1721
phocytes. Science 298(5594):850–854 70. McKenna DH Jr, Sumstad D, Bostrom N,
60. Bachanova V, Burns LJ, McKenna DH, Kadidlo DM, Fautsch S, McNearney S et al
Curtsinger J, Panoskaltsis-Mortari A, Lindgren (2007) Good manufacturing practices produc-
BR et al (2010) Allogeneic natural killer cells tion of natural killer cells for immunotherapy: a
for refractory lymphoma. Cancer Immunol six-year single-institution experience. Transfusion
Immunother 59(11):1739–1744 47(3):520–528
61. Geller MA, Cooley S, Judson PL, Ghebre R, 71. Miller JS, Oelkers S, Verfaillie C, McGlave P
Carson LF, Argenta PA et al (2010) A phase II (1992) Role of monocytes in the expansion of
study of allogeneic natural killer cell therapy to human activated natural killer cells. Blood
treat patients with recurrent ovarian and breast 80(9):2221–2229
cancer. Cytotherapy 13(1):98–107 72. Godal R, Bachanova V, Gleason M, McCullar
62. Cooley S, Gada P, McKenna D, McCullar V, V, Yun GH, Cooley S, Verneris MR, McGlave
Fautsch S, Verneris MR et al (2008) Successful PB, Miller JS (2010) Natural killer cell kill-
haploidentical hematopoietic cell engraftment ing of AML and ALL blasts by killer cell
using a non-myeloablative preparative regimen immunoglobulin-like receptor negative nat-
including natural killer (NK) cells. In: American ural killer cells after NKG2A and LIR-1
Society of Hematology 50th annual meeting, blockade. Biol Blood Marrow Transplant
San Francisco, CA: Blood, p 827 16(5): 612–621
63. Pfeiffer MM, Feuchtinger T, Teltschik HM, 73. Curtis RE, Travis LB, Rowlings PA, Socie G,
Schumm M, Muller I, Handgretinger R et al Kingma DW, Banks PM et al (1999) Risk of
(2010) Reconstitution of NK cell receptors lymphoproliferative disorders after bone mar-
influences NK activity and relapse rate after row transplantation: a multi-institutional study.
haploidentical transplantation of T and B cell Blood 94(7):2208–2216
depleted grafts in children. Haematologica 74. Reed M, Yearsley M, Krugh D, Kennedy MS
95(8):1381–1388 (2003) Severe hemolysis due to passenger lym-
64. Passweg JR, Tichelli A, Meyer-Monard S, Heim phocyte syndrome after hematopoietic stem cell
D, Stern M, Kuhne T et al (2004) Purified transplantation from an HLA-matched related
donor NK-lymphocyte infusion to consolidate donor. Arch Pathol Lab Med 127(10):
engraftment after haploidentical stem cell trans- 1366–1368
plantation. Leukemia 18(11):1835–1838 75. Hummon D, Zantek ND, Sumstad D, Miller
65. Koehl U, Esser R, Zimmermann S, Tonn T, JS, McKenna DH (2009) Transfusion-associated
Kotchetkov R, Bartling T et al (2005) Ex vivo graft-versus-host disease: a perspective from a
expansion of highly purified NK cells for immu- cell therapy laboratory. Transfusion 49(5):
notherapy after haploidentical stem cell trans- 1018–1019
plantation in children. Klin Padiatr 217(6): 76. McKenna DH, Kadidlo DM, Miller JS, Orchard
345–350 PJ, Wagner JE, McCullough J (2005) The
66. Rubnitz JE, Inaba H, Ribeiro RC, Pounds S, Minnesota molecular and cellular therapeutics
Rooney B, Bell T et al (2010) NKAML: a pilot facility: a state-of-the-art biotherapeutics engi-
study to determine the safety and feasibility of neering laboratory. Transfus Med Rev 19(3):
haploidentical natural killer cell transplantation 217–228
 Chapter 29

Minor Histocompatibility Antigen Typing by DNA

Sequencing for Clinical Practice in Hematopoietic
Stem-Cell Transplantation
Eric Spierings and Els Goulmy

Abstract
In HLA-matched stem-cell transplantation (SCT), minor H antigens are key molecules driving allo-
immune responses in both graft-versus-host disease (GvHD) and in graft-versus-leukemia (GvL) reactivity.
Dissection of the dual function of minor H antigens became evident through their different modes of
tissue and cell expression, i.e., hematopoietic system restricted or broad. Broadly expressed minor H
antigens are the targets of immune responses in both arms of graft-versus-host (GvH) responses, i.e.,
both GvHD and GvL, whereas the immune responses against the hematopoietic system-specific minor H
antigens are restricted to the GvL arm of SCT. Evidently, it is this latter group of minor H antigens that
can function as curative tools for stem-cell (SC)-based immunotherapy of hematological malignancies and
disorders. The HLA-matched patient/donor combinations, incompatible for one of the hematopoietic-
specific minor H antigens, are suitable for minor H antigen immunotherapy (Goulmy, Immunol Rev
157:125–140, 1997). Information on the minor H antigen phenotype is therefore needed. Hereto,
genomic typing for minor H antigens has been implemented in many HLA laboratories. Here, we firstly
summarize the relevance of minor H antigens particularly in hematopoietic SCT. Secondly, we describe
a method for typing the various polymorphic minor H antigens molecularly identified to date by
DNA sequencing.

Key words: Minor histocompatibility antigens, Hematopoietic stem-cell transplantation, Graft-versus-host

disease, Graft-versus-leukemia, Genotyping, polymorphism, immunogenetics, immunotherapy

1. Introduction

1.1. Background Minor H antigens are peptides derived from polymorphic self-proteins
that can differ both between unrelated individuals and between
related individuals. Minor H peptides are presented by the various
MHC class I and class II molecules. Their immune recognition is a

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_29, © Springer Science+Business Media New York 2012

509
510 E. Spierings and E. Goulmy

typical example of HLA-restricted recognition by T cells in man

(2) The MHC-minor H peptide-complexes can function as barri-
ers in allogeneic HLA-matched stem-cell transplantation (SCT)
and in solid organ transplantation (2, 3). In man, the impact of the
minor H antigenic systems became apparent in 1976 through a
serious clinical problems after HLA identical sibling SCT (2).
Although in vitro techniques unequivocally showed cellular immu-
nological reactivity between the SC recipient and the HLA identi-
cal SC donor (4, 5), it was not until 1995 that the molecular
characteristics of human minor H antigens were clarified (6–8).
Their chemical identification was instrumental in novel research
aimed at understanding the immunobiology of minor H antigenic
systems and for exploring their clinical usefulness in SCT.
Interestingly, minor H antigenic immune responses are known to
occur in the physiological situation of HLA haplo-mismatched
pregnancy as well (9). Namely, multiparous women demonstrate
that pregnancy can lead to allo-immune responses against the
infant’s paternal minor H antigens (9). Similarly, minor H antigen-
specific CTLs directed at the noninherited maternal minor H anti-
gens have been demonstrated in cord blood (10). Both minor H
antigen sensitization and tolerization occur in mothers and in their
adult offspring (11). Evidently, this natural allogeneic situation is a
most interesting source for studying minor H antigen immuniza-
tion patterns in all its aspects providing the basic information for
transplantation-related reactions and for the development of minor
H antigen-based immunotherapeutic strategies.

1.2. Minor H Antigen Autosomally encoded minor H antigens show different phenotype
Immunogenetics frequencies; the latter also differ in the various ethnic populations.
Evidently, the number of patients eligible for minor H antigen-
based immunotherapy depends on their phenotype frequencies in
the relevant ethnic population (12). The high phenotype frequency
of HLA-A2 in combination with a relatively high disparity rate for
HA-1 in all ethnic populations studied, currently mark HA-1 as
the most favorable minor H antigen for clinical application in all
ethnic populations. Similar results, however limited to only one or
two ethnic groups, were obtained for the minor H antigens ACC-
1, ACC-2, and HB-1Y. Based upon frequencies, the ACC-1 minor
H antigen is applicable in the Asian/Pacific population, with 6.5%
disparity in sibling pairs and 12.1% in MUD pairs. Similarly, ACC-2
and HB-1Y can only be applied in a significant proportion of the
pairs in the Caucasian and Mulatto population. Finally, the chance
of having at least one mismatch for one of the eight hematopoietic
system-restricted minor H antigen systems studied herein was cal-
culated as 21.2 and 33.6% for Caucasian sibling and MUD pairs,
respectively. In the black population, these numbers were 7.4 and
11.9%, respectively. In all these estimations, the HLA allele
frequencies of the respective populations were taken into account.
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 511

Table 1
Y-chromosome-encoded minor H antigens order by HLA-restriction molecule

Minor H antigen HLA restriction Gene Peptide References

A1/HY HLA-A1 USP9Y IVDCLTEMY (59, 60)

A2/HY HLA-A2 KDM5D FIDSYICQV (61)
A33/HY HLA-A33 TMSB4Y EVLLRPGLHFR (62)
B27/HY HLA-B27 DDX3Y RDSRGKPGY (63)
B52/HY HLA-B52 RPS4Y1 TIRYPDPVI (64)
B60/HY HLA-B60 UTY RESEEESVSL (65)
B7/HY HLA-B7 KDM5D SPSVDKARAEL (6)
B8/HY HLA-B8 UTY LPHNHTDL (66)
DQ5/HY HLA-DQ5 DDX3Y HIENFSDIDMGE (67)
DR15/HY HLA-DR15 DDX3Y SKGRYIPPHLR (68)
DRB3*0301/HY HLA-DRB3*0301 RPS4Y1 VIKVNDTVQI (69)

Since this study on ten minor H antigens was reported, 27 novel

autosomally encoded minor H antigens have been identified. It is
now estimated that the current set of 37 autosomally encoded
minor H antigens leads to at least one identifiable disparity in more
than 60% of SCT MUD pairs in the Caucasian population.

1.3. Minor H Antigen The Y-chromosome encoded HY antigens (Table 1), most of which
Tissue Distribution are broadly expressed, are reported to clinically contribute to graft-
in Relation to Their versus-leukemia (GvL) activity (13). Among others, CTLs specific
Clinical Relevance for the broadly expressed HY antigen can be detected in associa-
tion with leukemia remission (14).
The involvement of broadly expressed minor H antigens in the
GvHD arm of SCT was first shown by in vitro functional assays.
Namely, broadly expressed minor H antigens, such as the HY anti-
gens, are, among others, expressed on cell types of organs affected
by GvHD, such as fibroblasts, melanocytes, and keratinocytes (15).
This indicates that CTLs directed to broadly expressed minor H
antigens are particularly relevant for the development of GvHD.
To further address this issue, skin sections were incubated with
CTLs specific for the broadly expressed minor H antigen, HY, or
for the hematopoietic system-restricted minor H antigens HA-1
and HA-2. CTLs specific for the HY minor H antigen induced
severe GvH reactions of grades III–IV, while CTLs specific for
HA-1 and HA-2 induced no or weak GvH reactions (16). We
recently developed specific T-cell staining reagents to directly study
512 E. Spierings and E. Goulmy

infiltrating HY-specific T cells in the skin of sex-mismatched SCT

recipients. Indeed, these in situ results fortify the presence of
HY-specific T cells in fresh/frozen skin samples taken from male
patients suffering from acute GvHD after sex-mismatched SCT
(17).
Most autosomally encoded minor H antigens demonstrate
expression limited to but expressed on all cells of the hematopoietic
system, including leukemic cells and leukemic progenitor cells. In
vitro experiments have demonstrated that CTLs specific for HA-1
and HA-2 are capable of lysing leukemic cells (18). Clinically, CTLs
specific for the hematopoiesis-restricted minor H antigens HA-1
and HA-2 coincide with remission of hematological malignancies
after donor lymphocyte infusion (DLI) (19). Since T cells directed
to hematopoiesis-restricted minor H antigens do not target the
nonhematopoietic skin components, they may not be involved in
local GvHD. However, there exist controversial clinical reports on
the role of HA-1 in the development of GVHD as well. The skin
explant assay, as described and applied above, does not provide a
complete picture of GvHD because antigen presenting cells (APCs)
migrate out of the skin samples during its in vitro processing.
Therefore, we analyzed the behavior of HA-1-specific CTLs in an
ex vivo in situ skin explant model wherein HA-1-expressing den-
dritic cells (DCs) were added as APCs. Infiltration and activation of
HA-1 CTLs occurred only in those cases where both HLA-A2 and
HA-1 were expressed, either by the skin or by the DCs, or by the
combination of HLA-A2+ skin and HA-1+ DCs. The infiltrated and
activated CTLs did not cause skin tissue destruction. These results
point towards the role of recipient’s HA-1+ DCs in the chimeric
patient suffering from GvHD after HA-1-mismatched SCT and
provide a first step in understanding the reported association of
HA-1 mismatching with clinical GvHD (20).

1.4. Minor H Antigens Minor H antigens with expression limited to cells of the hematopoi-
as Treatment Modality etic system are especially relevant for the GvL activity (21–23). The
latter type of minor H antigens can be exploited as curative tools
for SC-based immunotherapy of hematological malignancies and
disorders (Goulmy, Immunol Rev 157:125–140, 1997). Hereto,
HLA-matched minor H antigen-mismatched recipient/donor
combinations are selected. Two treatment modalities currently
exploit minor H antigen disparity between SC donor and recipient.
One modality is the adoptive transfer of in vitro generated SC
donor-derived minor H antigen-specific CTLs. The other, a more
practical and far less laborious strategy, is post-SCT “vaccination.”
Both strategies, which are briefly discussed below, are currently
ongoing in clinical phase I/II studies.
Adoptive transfer of human minor H antigen-specific CTLs
has been shown to be successful in the treatment of human leuke-
mia and solid tumors in an experimental nonobese diabetic/severe
combined immunodeficiency mouse model (24, 25). Translation
29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 513

of this approach to the human situation is feasible. Namely, minor

H antigen-specific CTLs can be generated in large numbers in vitro
by co-culture of donor lymphocytes with donor-derived dendritic
cells which are minor H peptide-loaded or retrovirally transduced
with the minor H gene (26, 27). Artificial APCs coated with HLA-
A2/minor H antigen complexes, CD80 and CD54 might help to
selectively enrich minor H antigen-specific CTLs for adoptive
immunotherapy (28).
Another form of adaptive cellular immunotherapy is the transfer
of minor H antigen-specific T-cell receptor (TCR) genes into periph-
eral blood mononuclear cells of the SCT donor. In vitro, successful
transfer for the minor H antigens HA-2 (29) and HA-1 (30) has
been demonstrated. Implementation of the TCR gene-transduced
cells into clinical trials is currently not opportune. Two crucial barri-
ers still have to be overcome. Firstly, the relatively low avidity of the
transferred TCRs is associated with low lysis capacity. Secondly,
autoreactivities caused by pairing of the endogenous α-chain with
the transduced β-chain and vice versa cannot be excluded.
Interestingly, minor H antigens could potentially be applied in
a partly HLA-mismatched setting of SCT. This supposition is based
on in vitro results demonstrating the generation of CTLs specific
for the self-antigens HLA-A2/cyclin-D1 (31), and the minor H
antigen HA-1 (32) in an HLA-mismatched SCT setting. Although
really challenging and with a potential broad applicability, unde-
sired alloreactivity is a major problem of this approach. So far no
protocols exist that reliably and reproducibly result in clinically
graded non-self HLA/minor H antigen-specific CTLs.
In the treatment modality of minor H antigen peptide vacci-
nation, the relevant recipient-specific minor H peptides are admin-
istered to the patient after SCT and/or DLI to boost and thus
to further enhance the donor-derived minor H antigen-specific
T cells in vivo (21). The observation that HA-1 and HA-2 CTLs
emerge after DLI following allogeneic SCT underscores the feasi-
bility of this approach (33, 34). Minor H antigen vaccination will
boost minor H antigen-specific CTLs in patients with partial or
full donor chimerism, i.e., when host APCs are progressively
replaced by minor H antigen-negative donor APCs. The optimal
protocol of boosting minor H antigen-specific CTLs by vaccina-
tion is still unclear with regard to the extent of the immune
response and of the avoidance of tolerance. Although variables,
such as the delivery system (naked DNA, recombinant vectors,
recombinant protein or, dose, route of administration, frequency
of vaccination, and immunological adjuvants (35, 36)), still need
to be investigated, the preferable peptide length for vaccination
has been determined (37).
Recently, a phase I/II clinical minor H peptide vaccination
trial has been initiated for multiple myeloma patients who do not
respond on the regular allogeneic SCT treatment. Herein, myeloma
514 E. Spierings and E. Goulmy

is targeted after SCT and DLI by patient vaccination with patient-

dendritic cells loaded with graft-versus-myeloma-associated minor
H antigens (T. Mutis, H. Lokhorst, L. Hambach, E. Goulmy,
2012 still in progress).
A number of methods have been described for allelic typing of
minor H antigens, including allele-specific PCR based on sequence
specific primers (PCR-SSP; gene-specific PCR; real-time PCR;
PCR with melting curve analyses; reference strand conformation
analysis (RSCA); sequencing-based typing (SBT) and multiplex
PCT-SSO). PCR-SSP for minor H antigens has already been
described in great detail and is not described here (38). However,
an updated list of primer sets that can be used in this protocol is
presented in Tables 2 and 3. Here we focus on an SBT protocol, as
this technique allows typing of the largest number of minor H
antigens and is the most universally applicable protocol.

2. Materials

2.1. Materials 1. PCR primers as listed in Table 4.

2. Standard thin-wall PCR tubes.
3. Amplitaq DNA polymerase (Applied Biosystems).
4. GeneAmp 10× PCR Buffer II (Applied Biosystems).
5. MgCl2 (25 mM; Applied Biosystems).
6. dNTPs (10 mM; Invitrogen).
7. High-performance liquid chromatography (HPLC) water.
8. Agarose, molecular biology grade.
9. Gel Red (Gentaur).
10. Loading buffer (0.5 g Orange G and 40 g sucrose in 100 mL
H2O).
11. Tris-acetate EDTA (TAE) buffer.
12. Exonuclease I (Westburg).
13. SAP (GE Healthcare).
14. Sequence primers as listed in Table 4.
15. Big Dye Terminator (BDT) Cycle sequence kit v1.1 (Applied
Biosystems).
16. Sephadex G-50 Superfine DNA Grade (GE Healthcare).
17. Sephadex 96 wells matrix plate, (Millipore).
18. 96-Wells MultiScreen filter plate (Millipore).
19. POP7 (Applied Biosystems).
Table 2
Oligonucleotide primers for the uniform allele-specific PCR-SSP

Minor H antigen Allele Forward primer Reverse primer Amplicon size (bp)

HA-1 H CTTAAGGAGTGTGTGCTGCA ACTCCTACACATCCCTCAGAG 190

R CTTAAGGAGTGTGTGTTGCG ACTCCTACACATCCCTCAGAG 190
29

HA-2 V ACAGTCTCTGAGTGGCTCAG GCAGCTCCTGGTAGGGGTTAAC 271

M ACAGTCTCTGAGTGGCTCAG GCAGCTCCTGGTAGGGGTTAAT 271
HA-3 M GTCCTTCAGAGAGACTTGGGCAT AGACTCAGCAGGTTTGTTAC 129
T GTCCTTCAGAGAGACTTGGGCAC AGACTCAGCAGGTTTGTTAC 318
HA-8 R TGCAGTCAGCAGATCACCG CTTCTGGGCAACAGTTATGGA 186
P TGCAGTCAGCAGATCACCC CTTCTGGGCAACAGTTATGGA 186
HB-1 Y GCCATTCTTTTCTATAGGTTCTTTGT AGGGCATATGTTCCACTTGCTT 213
H ATTCTTTTCTATAGGTTCTCTGC AGGGCATATGTTCCACTTGCTT 210
ACC-1 Y CATTGCCTCAACAGCTTCAAG GGTTGTGGTATCTGTAGGGCGT 136
C CATTGCCTCAACAGCTTCAAG GGTTGTGGTATCTGTAGGGCGC 136
ACC-2 D GATGGAAAAGGAGTTTGAAGGCGA CAGCCTCCGTTTTGCCTTATC 197
G GATGGAAAAGGAGTTTGAAGGCGG CAGCCTCCGTTTTGCCTTATC 197
SP110 R AATGTGGTTTGAAGACCAAAAGT CTTGTACTCTCATCTTACCTCC 185
G AATGTGGTTTGAAGACCAAAAGT CTTGTACTCTCATCTTACCTCT 185
PANE1 R AGGCAAGTCCCACACTCG AATGGGGTAAATGACGTGCTG 207
stop CAGGCAAGTCCCACACTCA AATGGGGTAAATGACGTGCTG1 207
Minor Histocompatibility Antigen Typing by DNA Sequencing…

(continued)
515
 Table 2
516

(continued)

Minor H antigen Allele Forward primer Reverse primer Amplicon size (bp)

LB-ECGF H CCTGGCTCCGCGTGCACCG GAGGGGGCGGCGAATGTCGT 105

R CCTGGCTCCGCGTGCACCG AGGGGGCGGCGAATGGCGC 105
CTSH R TGCCGCTGCTCTGCGCCA CAGAGTTCCTGGCCTCTCAC 210
G TGCCGCTGCTCTGCGCCG CAGAGTTCCTGGCCTCTCAC 210
LRH1 5C GCCCACACACACTCTGAGAA ACCTGATTGTGACCCACAAC 193
E. Spierings and E. Goulmy

4C GCCCACACACACTCTGAGAA AACCTGATTGTGACCACAAC 193

LB-ADIR F GCTGAAGAGCGTCCAAGTACC GCGCCAGCTTTGGCTCTTTT 183
S GCTGAAGAGCGTCCAAGTACC CGCCAGCTTTGGCTCTTTC 183
Nucleotides in bold indicate introduced mismatches with the genomic sequence for improved performance
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 517

Table 3
Oligonucleotide primers for gene-specific PCR

Minor H Size
antigen Forward primer Reverse primer (bp)
UGT2B17
5¢ region GGCAGTATCTTGCCAATGT AGACTCCAAGTGCCAGTT 341
Exon 1a TGTTGGGAATATTCTGACTATAA CCCACTTCTTCAGATCATATGC 352
Exon 1b AAATGACAGAAAGAAACAA GCATCTTCACAGAGCTTATAT 443
Exon 6 GAATTCATCATGATCAACCG ACAGGCAACATTTTGTGATC 201
HY (SRY) TGGCGATTAAGTCAAATTCGC CCCCCTAGTACCCTGACAAT 136
GTATT

Table 4
Oligonucleotide primers for sequencing-based typing

Minor H Amplicon
antigen Type of primer Primer sequence size (bp)
HA-1 Forward PCR AAGGACCGTTGTTTGTGAAAGC
Reverse PCR CAGGAAACAGCTATGACCCGTT 523
GTCATGCCTTCATCCTAC
Forward sequence AAGGACCGTTGTTTGTGAAAGC
Reverse sequence CAGGAAACAGCTATGACC
HA-2 Forward PCR CCCCTTCTTCCTTCTCCACT
Reverse PCR CAGGAAACAGCTATGACCAGCC 526
AGGTTCGCTCTCTTA
Forward sequence CACAGGAGATGCTCTAATGG
Reverse sequence TGTGGGTGCCTGTAAG
HA-3 Forward PCR GTGACTGACCCACAGGGAGT
Reverse PCR CCCAGGTGGGAAAGTACTCA 498
Forward sequence TACCTCTTGATTGGGAGAAA
Reverse sequence CTTGACTTGTTCGATGACAG
HA-8 Forward PCR GCTACCTCTTCTGGGCAACA
Reverse PCR CAGGAAACAGCTATGACGCCTG 402
GCTCTTTAAATGCAC
Forward sequence GCCATTGGAGTTAGAATCTG
Reverse sequence TCAGCTAATCAGGTTACTCC
(continued)
518 E. Spierings and E. Goulmy

Table 4
(continued)

Minor H Amplicon
antigen Type of primer Primer sequence size (bp)
HB-1 Forward PCR CCTGAGCCTTCTGACCTCAC
Reverse PCR AAACTCTGGGCTTCGAGTCA 215
Forward sequence CACAGTGGAGAAACCAGAAC
Reverse sequence AAGAGCTGTGCCTCAGATAC
ACC-1 Forward PCR CATTGCCTCAACAGCTTCAA
and-2
Reverse PCR CAGGAAACAGCTATGACAGCCT 478
CCGTTTTGCCTTATC
Forward sequence CTCAAGACTTTGCTCTCCAC
Reverse sequence CATTATGAACTCCGCAAC
PANE1 Forward PCR GCAGGCCACGGAGAT
Reverse PCR GTCAAGCCCTGACTGGACAT 308
Forward sequence GGAAGGAGGAACAGC
Reverse sequence ACCCTCACAGGTCCTCCA
SP110 Forward PCR CCATCACCTAAGGTCACCAA
Reverse PCR TGAAGAGTGGGCCTTGATTT 573
Forward sequence CAAGATTGGGAGATCAATGT
Reverse sequence TAAAATTCAACCCTCCACAG
CTSH Forward PCR GAGAACCCCCGCCTATAATG
Reverse PCR GCCGACAGGTGAACTAGAGC 507
Forward sequence TCTGCGTACCTAAGGAGTTC
Reverse sequence TCCCAGTTGACGCTC
LRH-1 Forward PCR CCTTCATTCTGCCAGGAGAG
Reverse PCR CAGGAAACAGCTATGACCCTCA 378
GTGCCTCTCTGGTTC
Forward sequence AGAATAGGGGTTCTTCTTCG
Reverse sequence GGGTTCTTCTTCGCTT
(continued)
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 519

Table 4
(continued)

Minor H Amplicon
antigen Type of primer Primer sequence size (bp)
LB-ADIR-1 Forward PCR CTCCTCGTCCTCGTCACAGT
Reverse PCR CAGGAAACAGCTATGACAGG 447
ACCCCAGACAGGGTTT
Forward sequence GCTGAAGAGCGTCCAGTA
Reverse sequence CCACAGCAACACCAAT
CD19 Forward PCR CTCTCTCTGCGGGTCTTTGT
Reverse PCR CAGGAAACAGCTATGACTTCA 557
GCTCTAGGCTCAGCAA
Forward sequence GTGTCTCTGCATTTGGTTCT
Reverse sequence CACAGGACAGCCAGAG
Sequences in bold represent the M13-reverse primer sequence

20. 10× Genetic Analyzer Buffer with EDTA (Applied Biosystems).

21. DNA size marker, ranging from 100 bp to at least 1,000 bp
with a size interval of approximately 100 bp, e.g., the 1 kb + ladder
from Invitrogen.

2.2. Special Equipment 1. Standard micropipets ranging from 0.2 μL to 1.0 mL.
2.2.1. Pre-PCR Area 2. Filter tips (0.5–10; 10–100; 100–1,000 μL).
3. Thin-walled PCR tubes, strips, or individual tubes (Eppendorf).

2.2.2. Thermal Cycler Area 1. Thermal cycler with a 96-well block. We have successfully used
the Dyad Disciple (MJ-Research) and the models 2720 and
9700 from Applied Biosystems.

2.2.3. Post-PCR Area 1. Electrophoresis system. A horizontal electrophoresis system

for submerged gel electrophoresis with a tray that accommo-
dates six 26-well combs enables genotyping of 150 DNAs in a
single run and allows inclusion of one size marker per row.
2. Combs compatible with multichannel pipets facilitate fast load-
ing of a full gel.
3. DNA size marker 1 kb + ladder (Invitrogen).
4. A multichannel pipettor in the 0.5–10.0 μL range for gel
loading.
5. Geldoc 2000 (Biorad) or alternative DNA visualization system.

2.2.4. Sequencing Area 1. Plate centrifuge (Beckman Coulter).

2. Automated DNA sequencer and consumables. We use the ABI
3730, 48 capillary sequencer (Applied Biosystems).
520 E. Spierings and E. Goulmy

2.2.5. Allele Identification 1. SBTEngine® Software (Genome Diagnostics) for automated

Software allele assignment.
2. If SBTEngine® is not available, general sequence analysis pro-
grams as BioEdit can be used for manual analysis.

3. Methods

3.1. DNA Preparation 1. Different methods for isolation of genomic DNA are available.
For optimal results, high-quality DNA is required. The
described protocols assume a DNA concentration of 30 ng/μL.
If required, dilute the DNA stock solution in HPLC-grade
water. Higher or lower concentrations of DNA may lead to
unsuccessful amplification.
2. Control DNA can be obtained from the IHW cell panel.
A large number of the cell lines in this panel has been typed for
minor H antigens (38). See Note 1 for preventing carryover
contamination.

3.2. PCR Amplification 1. Dilute PCR primers to a concentration of 100 pmol/μL.

2. For each Minor-H antigen to be tested, prepare a master mix
(21.75 μL/reaction) containing 2.5 μL GeneAmp 10× PCR
Buffer II, 1.5 μL MgCl2, 1 μL dNTPs, 16.55 μL H2O, 0.1 μL
of the appropriate 5¢ PCR primer, and 0.1 μL of the appropri-
ate 3¢ PCR primer.
3. Mix master mix well.
4. Master mixes can be stored at −20°C for later use.
5. Take the appropriate volume of master mix, depending on the
number of amplifications (see Note 2).
6. Add Amplitaq (0.25 μL/reaction) to the master mix (see
Note 2).
7. Add 22 μL master mix to a thin-wall PCR tube (see Note 3).
8. Add 3 μL (90 ng) of the test DNA to each aliquot of the mas-
ter mixes. Only samples with the appropriate HLA restriction
molecule for each minor H antigen should be tested (see
Note 4).
9. Incubate the samples in a thermal cycler according to the fol-
lowing thermal cycler profile (final volume 25 μL): 1 cycle:
95°C, 10 min; 35 cycle: 95°C, 30 s/54°C, 30 s/72°C, 60 s; 1
cycle: 72°C, 10 min; hold: 15°C. This protocol can be applied
for all minor H antigens.
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 521

10. Analysis of amplification products: apply 5 μL of each

amplification reaction with 3 μL loading buffer to a 1.5% aga-
rose gel. Run gel for 30 min at 200 V. Visualize the amplification
products with a Geldoc.
11. The expected sizes of the PCR products have been listed for
each minor H antigen in Table 4.
12. The DNA size marker can be diluted for semi-quantitative
comparison. In our experience, sequencing of PCR products
with a lower intensity than the 1/8 dilution of the 1 kb + DNA
marker (Invitrogen), in general appears to be unsuccessful. It
is therefore advisable to only proceed with the products show-
ing a stronger intensity than the 1/8 diluted marker. See Note
5 for troubleshooting.

3.3. Purification Digestion with Exonuclease I and SAP is performed to remove

of the Amplification residual PCR primers and dNTPs.
Products
1. Add 1 μL SAP and 0.5 μL Exonuclease I to each amplification
product.
2. Incubate the samples in a thermal cycler: 37°C, 15 min/80°C,
20 min/15°C, hold until ready for use in sequencing reaction.

3.4. DNA Sequencing 1. Dilute sequence primers to a concentration of 100 pmol/μL

Reaction (see Table 4 for sequences).
2. For each sequencing primer, prepare the appropriate amounts
of each of the sequencing reaction mixes containing 1 μL BDT
Ready Reaction v1.1 mix, 1.5 μL BDT buffer 5×, 0.1 μL of the
appropriate sequence primer, and 6.4 μL H2O per sequencing
reaction.
3. Add 9 μL sequencing mix to a thin-wall PCR tube.
4. Add 1 μL of the purified PCR amplification product—see
Subheading 3.3, step 2.
5. Incubate the samples in a thermal cycler: 1 cycle: 96°C, 10 s;
25 cycles: 96°C, 10 s/50°C 10 s/60°C, 120 s; 15°C—hold
until ready for purification.

3.5. Purification 1. Add Sephadex to the matrix plate (Millipore), fill each matrix
of Sequencing well to the rim.
Reaction Products 2. Place the Millipore Multiscreen plate upside down on top of
the filled matrix plate.
3. Transfer the Sephadex from the matrix plate to the Millipore
Multiscreen plate by turning the matrix plate to the top
position.
522 E. Spierings and E. Goulmy

4. Add 300 μL H2O per well of the Millipore Multiscreen

plate.
5. Keep at room temperature for 3 h or overnight at 4°C.
6. Put a dummy 96-well flat bottom plate under the Multiscreen
plate.
7. Spin down each pair of the stacked plates for 5 min at 670 × g,
discard the dummy plate.
8. Overlay the sequencing product on top of the packed Sephadex
in the Multiscreen plate, one product per well.
9. Put a sequencing product collection plate under the Multiscreen
plate.
10. Spin down the plates for 5 min at 670 × g.
11. Add a sequencing cover lid to the collection plates or use a
plastic seal.
12. Spin down 5 s at 225 × g to remove the air bubbles from the
sequencing product.

3.6. Settings 1. Install your sequencer with a 50-cm capillary.

for Capillary 2. Use liquid polymer POP-7.
Electrophoresis
3. If your sequencer is not able to perforate the seal or the
sequencing cover lid, remove the seal or the cover.
4. Place the sequencing plate in a cassette and load into the auto-
mated sequencer.
5. Start the sequencing procedure using the settings given
below:
Run Module: BDx FastSeq50 POP7
(a) Prerun time: 180 s
(b) Prerun voltage: 15.0 kV
(c) Injection time: 3 s
(d) Injection voltage: 1.0 kV
(e) Electrophoresis voltage: 13.4 kV
(f) Temperature: 60°C

3.7. Sequence Analysis Analyses can be performed manually with general sequencing anal-
ysis software like BioEdit, or automated with SBTEngine®, which
includes data files for a number of minor H antigens. An example
of automated allele assignments is depicted in Fig. 1. Homozygosity
can be confirmed with PCR-SSP (see Note 6).
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 523

Fig. 1. Sequencing-based typing results for minor H antigen HA-1. The cell-lines used in this validation protocol were
previous typed by PCR-SSP. Analyses were performed with SBTEngine© (GenDx BV). Left panel shows a reference sample
that is homozygous for the HA-1H allele. Numbers at top indicate the codon position. The sequence underneath the codon
position displays the potential nucleic acids and their position, based upon the known HA-1 alleles. The position of the
nonsynymous SNP at position 416 has been marked with a circle. The right panel displays the result for the HA-1R refer-
ence sample. In the middle, a heterozygous reference sample has been depicted. Both the heterozygous synonymous SNP
at position 412 and the nonsynonymous SNP at position 416 have been underlined.

4. Notes

1. Carryover contamination from previous PCR can be a

significant problem due to the abundance of PCR products and
to the ideal structure of the contaminant materials for
re-amplification. Therefore, controls to detect contamination
should always be included. Moreover, a strict physical separa-
tion of pre- and post amplification areas is strongly advised. An
alternative option to control carryover contamination with
PCR products is using uracil DNA glycosylase (UDG) in all
PCR samples (39). The procedure requires the following two
steps: (1) incorporating dUTP in all PCR products (by substi-
tuting dUPT for dTTP), and (2) treating all subsequent fully
preassembled starting reactions with UDG, followed by ther-
mal inactivation of UDG. UDG cleaves the uracil base from
uracil-containing DNA, but has no effect on thymidine-con-
taining DNA. This cleavage blocks replication by DNA poly-
merases. UDG does not react with dUTP, and is inactivated
during denaturation at 95°C.
 524

Table 5
Autosomally encoded minor H antigens, ordered alphabetically

Original name HLA restriction Gene Chrom. Peptide dbSNP References

ACC-1 HLA-A24 BCL2A1 15 DYLQYVLQI rs1138357 (40)

ACC-2 HLA-B44 BCL2A1 15 KEFEDDIINW rs3826007 (40)
ACC-6 HLA-B44 HMSD 18 MEIFIEVFSHF rs9945924 (41)
E. Spierings and E. Goulmy

C19orf48 HLA-B7 C19orf48 19 TPNQRQNVC rs5818907 (42)

CD19 HLA-DQA1*05/ CD19 16 RTLDKVLEV rs2173904 (43)
B1*02
CTSH/A31 HLA-A31 CTSH 15 ATLPLLCAR rs2289702 (44)
CTSH/A33 HLA-A33 CTSH 15 WATLPLLCAR rs2289702 (44)
HA-1/A2 HLA-A2 HMHA1 19 VLHDDLLEA rs1801284 (45)
HA-1/B60 HLA-B60 HMHA1 19 KECVLHDDLL rs1801284 (46)
HA-2 HLA-A2 MYO1G 7 YIGEVLVSV rs61739531 (8)
HA-3 HLA-A1 AKAP13 15 VTEPGTAQY rs7162168 (47)
HA-8 HLA-A2 KIAA0020 9 RTLDKVLEV rs2173904 (48)
HB-1 HLA-B44 HMHB1 5 EEKRGSLHVW rs161557 (49)
HEATR1 HLA-B8 HEATR1 1 ISKERAEAL rs2275687 (50)
LB-ADIR-1 HLA-A2 TOR3A 1 SVAPALALFPA rs2296377 (51)
LB-APOBEC3B-1K HLA-B7 APOBEC3B 22 KPQYHAEMCF rs2076109 (52)
LB-ARHGDIB-1R HLA-B7 ARHGDIB 12 LPRACWREA rs4703 (53)
LB-BCAT2-1R HLA-B7 BCAT2 19 QPRRALLFVIL rs11548193 (52)
Original name HLA restriction Gene Chrom. Peptide dbSNP References
LB-EBI3-1I HLA-B7 EBI3 19 RPRARYYIQV rs4740 (52)
LB-ECGF-1 HLA-B7 TYMP 22 RPHAIRRPLAL n.a. (52)
LB-ERAP1-1R HLA-B7 ERAP1 5 HPRQEQIALLA rs26653 (52)
LB-GEMIN4-1V HLA-B7 GEMIN4 17 FPALRFVEV rs4968104 (52)
LB-LY75-1K HLA-DRB1*1301 LY75 2 GITYRNKSLM rs12692566 (53)
LB-MR1-1R HLA-DRB3*0202 MR1 1 YFRLGVSDPIRG rs2236410 (53)
29

LB-MTHFD1-1Q HLA-DRB1*0301 MTHFD1 14 SSIIADQIALKL rs2236225 (53)

LB-PDCD11-1F HLA-B7 PDCD11 10 GPDSSKTFLCL rs2986014 (52)
LB-PI4K2B-1 HLA-DQ6 PI4K2B 4 SRSSSAELDRSR rs313549 (54)
LB-PRCP-1D HLA-A2 PRCP 11 FMWDVAEDLKA rs2298668 (52)
LB-PTK2B-1T HLA-DRB3*0101 PTK2B 8 VYMNDTSPL rs751019 (53)
LB-SSR1-1S HLA-A2 SSR1 6 SLAVAQDLT rs10004 (52)
LB-WNK1-1I HLA-A2 WNK1 12 RTLSPEIITV rs12828016 (52)
LRH-1 HLA-B7 P2RX5 17 TPNQRQNVC rs5818907 (55)
PANE1 HLA-A3 CENPM 22 RVWDLPGVLK n.a. (56)
SLC1A5 HLA-B61 SLC1A5 19 AEATANGGLAL rs3027956 (57)
SP110 HLA-A3 SP110 2 SLPRGTSTPK rs1365776 (58)
UGT2B17 HLA-A29/B44 UGT2B17 4 AELLNIPVLY n.a. (59)
UGT2B17 HLA-A2 UGT2B17 4 CVATMIFMI n.a. (57)
Minor Histocompatibility Antigen Typing by DNA Sequencing…
525
526 E. Spierings and E. Goulmy

2. Mixes for a maximum of ten samples are optimal in our hands.

Master mixes for PCR and sequencing can be stored at −20°C
for at least 1 year. Storing of PCR and sequencing reaction
mixes after the addition of AmpliTaq or BDT is not advised.
Storing mixes after addition of the enzymes results in a rapidly
reduction of the PCR yields and sequencing qualities.
3. Preparation of samples for PCR preferentially occurs on ice.
4. Minor H antigens are only relevant when presented in the cor-
rect HLA molecule. Check the presence of the HLA-restriction
molecule before typing certain minor H antigens. For this pur-
pose, Table 5 can be used.
5. If PCR reactions show a consistently poor result for various
samples, adaptations to the protocol should be made. These
adaptations could include changes to the PCR temperature
profile. For example, PCR yields for HA-1 may improve when
increasing the annealing temperature to 56°C. Alternatively,
PCR primer concentrations can be doubled.
6. Due to the use of nonpolymorphic amplification primers, SBT
has a low risk to miss an allele due to suboptimal amplification
conditions. If required, homozygosity may be confirmed using
PCR-SSP.

Acknowledgments

We thank Talitha de Hoop en Walter van Ginkel, University

Medical Center Utrecht, for developing and validating the SBT
protocols, Jos Pool and Jos Drabbels, Leiden University Medical
Center for developing and validating the uniform PCR-SSP proto-
cols, and Erik Rozemuller, Genome Diagnostics BV, Utrecht, for
the implementation of automated allele assignment for minor H
antigens after SBT in SBTEngine®.

References
1. Goulmy E (1997) Human minor histocompat- 4. Goulmy E, Gratama JW, Blokland E, Zwaan
ibility antigens: new concepts for marrow trans- FE, van Rood JJ (1983) A minor transplanta-
plantation and adoptive immunotherapy. tion antigen detected by MHC-restricted cyto-
Immunol Rev 157:125–140 toxic T lymphocytes during graft-versus-host
2. Goulmy E, Termijtelen A, Bradley BA, van disease. Nature 302:159–161
Rood JJ (1976) Alloimmunity to human H-Y. 5. van Els CA, D’Amaro J, Pool J, Blokland E,
Lancet 2:1206 Bakker A, van Elsen PJ et al (1992)
3. Goulmy E, Bradley BA, Lansbergen Q, van Immunogenetics of human minor histocom-
Rood JJ (1978) The importance of H-Y incom- patibility antigens: their polymorphism and
patibility in human organ transplantation. immunodominance. Immunogenetics 35:
Transplantation 25:315–319 161–165
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 527

6. Wang W, Meadows LR, den Haan JM, Sherman 16. Dickinson AM, Wang XN, Sviland L, Vyth-
NE, Chen Y, Blokland E et al (1995) Human Dreese FA, Jackson GH, Schumacher TN et al
H-Y: a male-specific histocompatibility antigen (2002) In situ dissection of the graft-versus-host
derived from the SMCY protein. Science activities of cytotoxic T cells specific for minor his-
269:1588–1590 tocompatibility antigens. Nat Med 8:410–414
7. Goulmy E, Pool J, van den Elsen PJ (1995) 17. Kim YH, Faaij CM, van Halteren AG, Schrama E,
Interindividual conservation of T-cell receptor de Jong TA, Schøller J, Egeler RM, Pavel S, Vyth-
beta chain variable regions by minor histocom- Dreese FA, van Tol MJ, Goulmy E, Spierings E.
patibility antigen-specific HLA-A*0201- In situ detection of HY-specific T cells in acute
restricted cytotoxic T-cell clones. Blood 85: graft-versus-host disease-affected male skin after
2478–2481 sex-mismatched stem cell transplantation. Biol
8. den Haan JM, Sherman NE, Blokland E, Blood Marrow Transplant. 2012 Mar;18(3):
Huczko E, Koning F, Drijfhout JW et al (1995) 381–7. Epub 2011 Nov 4.
Identification of a graft versus host disease- 18. van der Harst D, Goulmy E, Falkenburg J,
associated human minor histocompatibility Kooij-Winkelaar Y, van Luxemburg-Heijs S,
antigen. Science 268:1476–1480 Goselink H et al (1994) Recognition of minor
9. Verdijk RM, Kloosterman A, Pool J, Van De histocompatibility antigens on lymphocytic and
KM, Naipal AM, Van Halteren AG et al (2004) myeloid leukemic cells by cytotoxic T-cell
Pregnancy induces minor histocompatibility clones. Blood 83:1060–1066
antigen-specific cytotoxic T cells: implications 19. Marijt W, Heemskerk M, Kloosterboer F,
for stem cell transplantation and immunother- Goulmy E, Kester M, van der Horn M et al
apy. Blood 103:1961–1964 (2003) Hematopoiesis-restricted minor histo-
10. Mommaas B, Stegehuis-Kamp JA, Van Halteren compatibilty antigens HA-1- or HA-2-specific
AG, Kester M, Enczmann J, Wernet P et al T cells can induce complete remissions of
(2004) Cord blood comprises antigen- relapsed leukemia. Proc Natl Acad Sci USA
experienced T cells specific for maternal minor 100:2742–2747
histocompatibility antigen HA-1. Blood 105: 20. Kim YH, Vyth-Dreese FA, Schrama E, Pavel S,
1823–1827 Bajema I, Goulmy E et al (2010) Exogenous
11. Van Halteren AG, Jankowska-Gan E, Joosten addition of minor H antigen HA-1+ dendritic
A, Blokland E, Pool J, Brand A et al (2009) cells to skin tissues ex vivo causes infiltration
Naturally acquired tolerance and sensitization and activation of HA-1 specific cytotoxic T
to minor histocompatibility antigens in cells. Biol Blood Marrow Transplant 17(1):
healthy family members. Blood 114: 69–77
2263–2272 21. Mutis T, Goulmy E (2002) Hematopoietic
12. Spierings E, Hendriks M, Absi L, Canossi A, system-specific antigens as targets for cellular
Chhaya S, Crowley J et al (2007) Phenotype immunotherapy of hematological malignancies.
frequencies of autosomal minor histocompati- Semin Hematol 39:23–31
bility antigens display significant differences 22. Klein CA, Wilke M, Pool J, Vermeulen C,
among populations. PLoS Genet 3:e103 Blokland E, Burghart E et al (2002) The
13. Gratwohl A, Hermans J, Niederwieser D, van hematopoietic system-specific minor histocom-
Biezen A, van Houwelingen HC, Apperley patibility antigen HA-1 shows aberrant expres-
J (2001) Female donors influence transplant- sion in epithelial cancer cells. J Exp Med
related mortality and relapse incidence in male 196:359–368
recipients of sibling blood and marrow trans- 23. Fujii N, Hiraki A, Ikeda K, Ohmura Y, Nozaki
plants. Hematol J 2:363–370 I, Shinagawa K et al (2002) Expression of
14. Takami A, Sugimori C, Feng X, Yachie A, minor histocompatibility antigen, HA-1, in
Kondo Y, Nishimura R et al (2004) Expansion solid tumor cells. Transplantation 73:
and activation of minor histocompatibility 1137–1141
antigen HY-specific T cells associated with 24. Hambach L, Nijmeijer BA, Aghai Z, Schie ML,
graft-versus-leukemia response. Bone Marrow Wauben MH, Falkenburg JH et al (2006)
Transplant 34:703–709 Human cytotoxic T lymphocytes specific for a
15. de Bueger MM, Bakker A, Van Rood JJ, Van single minor histocompatibility antigen HA-1
der Woude F, Goulmy E (1992) Tissue distri- are effective against human lymphoblastic leu-
bution of human minor histocompatibility anti- kaemia in NOD/scid mice. Leukemia
gens ubiquitous versus restricted tissue 20:371–374
distribution indicates heterogeneity among 25. Hambach L, Vermeij M, Buser A, Aghai Z, Van
human cytotoxic T lymphocyte-defined non- der KT, Goulmy E (2008) Targeting a single
MHC antigens. J Immunol 149:1788–1794 mismatched minor histocompatibility antigen
528 E. Spierings and E. Goulmy

with tumor-restricted expression eradicates 35. Scheibenbogen C, Letsch A, Schmittel A,

human solid tumors. Blood 112:1844–1852 Asemissen AM, Thiel E, Keilholz U (2003)
26. Mutis T, Verdijk R, Schrama E, Esendam B, Rational peptide-based tumour vaccine devel-
Brand A, Goulmy E (1999) Feasibility of immu- opment and T cell monitoring. Semin Cancer
notherapy of relapsed leukemia with ex vivo- Biol 13:423–429
generated cytotoxic T lymphocytes specific for 36. Figdor CG, de Vries I, Lesterhuis WJ, Melief
hematopoietic system- restricted minor histo- CJ (2004) Dendritic cell immunotherapy: map-
compatibility antigens. Blood 93:2336–2341 ping the way. Nat Med 10:475–480
27. Mutis T, Ghoreschi K, Schrama E, Kamp J, 37. Hambach L, Aghai Z, Pool J, Kroger N,
Heemskerk M, Falkenburg JH et al (2002) Goulmy E (2010) Peptide length extension
Efficient induction of minor histocompatibility skews the minor HA-1 antigen presentation
antigen HA-1-specific cytotoxic T-cells using toward activated dendritic cells but reduces
dendritic cells retrovirally transduced with its presentation efficiency. J Immunol 185:
HA-1-coding cDNA. Biol Blood Marrow 4582–4589
Transplant 8:412–419 38. Spierings E, Drabbels J, Hendriks M, Pool J,
28. Oosten LE, Blokland E, Van Halteren AG, Spruyt-Gerritse M, Claas F et al (2006) A uni-
Curtsinger J, Mescher MF, Falkenburg JH et al form genomic minor histocompatibility anti-
(2004) Artificial antigen-presenting constructs gen typing methodology and database designed
efficiently stimulate minor histocompatibility to facilitate clinical applications. PLoS One
antigen-specific cytotoxic T lymphocytes. Blood 1:e42
104:224–226 39. Longo MC, Berninger MS, Hartley JL (1990)
29. Heemskerk MH, Hoogeboom M, de Paus RA, Use of uracil DNA glycosylase to control carry-
Kester MG, van der Hoorn MA, Goulmy E over contamination in polymerase chain reac-
et al (2003) Redirection of antileukemic reac- tions. Gene 93:125–128
tivity of peripheral T lymphocytes using gene 40. Akatsuka Y, Nishida T, Kondo E, Miyazaki M,
transfer of minor histocompatibility antigen Taji H, Iida H et al (2003) Identification of a
HA-2-specific T-cell receptor complexes polymorphic gene, BCL2A1, encoding two
expressing a conserved alpha joining region. novel hematopoietic lineage-specific minor his-
Blood 102:3530–3540 tocompatibility antigens. J Exp Med 197:
30. Mommaas B, Van Halteren AG, Pool J, Van V, 1489–1500
Wieles B, Heemskerk MH et al (2005) Adult 41. Kawase T, Akatsuka Y, Torikai H, Morishima S,
and cord blood T cells can acquire HA-1 Oka A, Tsujimura A et al (2007) Alternative
specificity through HA-1 T-cell receptor gene splicing due to an intronic SNP in HMSD gen-
transfer. Haematologica 90:1415–1421 erates a novel minor histocompatibility antigen.
31. Sadovnikova E, Jopling LA, Soo KS, Stauss HJ Blood 110:1055–1063
(1998) Generation of human tumor-reactive 42. Tykodi SS, Fujii N, Vigneron N, Lu SM, Mito
cytotoxic T cells against peptides presented by JK, Miranda MX et al (2008) C19orf48 encodes
non-self HLA class I molecules. Eur J Immunol a minor histocompatibility antigen recognized
28:193–200 by CD8+ cytotoxic T cells from renal cell
32. Mutis T, Blokland E, Kester M, Schrama E, carcinoma patients. Clin Cancer Res 14:
Goulmy E (2002) Generation of minor histo- 5260–5269
compatibility antigen HA-1-specific cytotoxic 43. Spaapen RM, Lokhorst HM, Van de OK,
T cells restricted by nonself HLA molecules: a Otterud BE, Dolstra H, Leppert MF et al
potential strategy to treat relapsed leukemia (2008) Toward targeting B cell cancers with
after HLA-mismatched stem cell transplanta- CD4+ CTLs: identification of a CD19-encoded
tion. Blood 100:547–552 minor histocompatibility antigen using a novel
33. Marijt WA, Heemskerk MH, Kloosterboer FM, genome-wide analysis. J Exp Med 205:
Goulmy E, Kester MG, van der Hoorn MA 2863–2872
et al (2003) Hematopoiesis-restricted minor 44. Torikai H, Akatsuka Y, Miyazaki M, Tsujimura
histocompatibility antigens HA-1- or HA-2- A, Yatabe Y, Kawase T et al (2006) The human
specific T cells can induce complete remissions cathepsin H gene encodes two novel minor his-
of relapsed leukemia. Proc Natl Acad Sci USA tocompatibility antigen epitopes restricted by
100:2742–2747 HLA-A*3101 and -A*3303. Br J Haematol
34. Kircher B, Stevanovic S, Urbanek M, 134:406–410
Mitterschiffthaler A, Rammensee HG, 45. den Haan JM, Meadows LM, Wang W, Pool J,
Grunewald K et al (2002) Induction of HA-1- Blokland E, Bishop TL et al (1998) The minor
specific cytotoxic T-cell clones parallels the histocompatibility antigen HA-1: a diallelic
therapeutic effect of donor lymphocyte infu- gene with a single amino acid polymorphism.
sion. Br J Haematol 117:935–939 Science 279:1054–1057
 29 Minor Histocompatibility Antigen Typing by DNA Sequencing… 529

46. Mommaas B, Kamp J, Drijfhout JW, Beekman R et al (2005) A frame-shift polymorphism in

N, Ossendorp F, Van Veelen P et al (2002) P2X5 elicits an allogeneic cytotoxic T lympho-
Identification of a novel HLA-B60-restricted cyte response associated with remission of
T cell epitope of the minor histocompatibility chronic myeloid leukemia. J Clin Invest 115:
antigen HA-1 locus. J Immunol 169: 3506–3516
3131–3136 56. Brickner AG, Evans AM, Mito JK, Xuereb SM,
47. Spierings E, Brickner AG, Caldwell JA, Zegveld Feng X, Nishida T et al (2006) The PANE1
S, Tatsis N, Blokland E et al (2003) The minor gene encodes a novel human minor histocom-
histocompatibility antigen HA-3 arises from patibility antigen that is selectively expressed in
differential proteasome-mediated cleavage of B-lymphoid cells and B-CLL. Blood 107:
the lymphoid blast crisis (Lbc) oncoprotein. 3779–3786
Blood 102:621–629 57. Kamei M, Nannya Y, Torikai H, Kawase T,
48. Brickner AG, Warren EH, Caldwell JA, Taura K, Inamoto Y et al (2009) HapMap scan-
Akatsuka Y, Golovina TN, Zarling AL et al ning of novel human minor histocompatibility
(2001) The immunogenicity of a new human antigens. Blood 113:5041–5048
minor histocompatibility antigen results from 58. Warren EH, Gavin MA, Xuereb SM, Koo KK,
differential antigen processing. J Exp Med Mito JK, Brown ML et al (2002) A single
193:195–206 nucleotide polymorphism in the SP110
49. Dolstra H, Fredrix H, Maas F, Coulie PG, nuclear phosphoprotein gene creates a minor
Brasseur F, Mensink E et al (1999) A human histocompatibility antigen whose expression
minor histocompatibility antigen specific for B is regulated by interferon-gamma. Blood
cell acute lymphoblastic leukemia. J Exp Med 100:73A
189:301–308 59. Murata M, Warren EH, Riddell SR (2003) A
50. Bleakley M, Otterud BE, Richardt JL, Mollerup human minor histocompatibility antigen result-
AD, Hudecek M, Nishida T et al (2010) ing from differential expression due to a gene
Leukemia-associated minor histocompatibility deletion. J Exp Med 197:1279–1289
antigen discovery using T-cell clones isolated 60. Vogt MH, de Paus RA, Voogt PJ, Willemze R,
by in vitro stimulation of naive CD8+ T cells. Falkenburg JH (2000) DFFRY codes for a new
Blood 115:4923–4933 human male-specific minor transplantation
51. van Bergen CA, Kester MG, Jedema I, antigen involved in bone marrow graft rejec-
Heemskerk MH, van Luxemburg-Heijs SA, tion. Blood 95:1100–1105
Kloosterboer FM et al (2007) Multiple 61. Pierce RA, Field ED, den Haan JM, Caldwell
myeloma-reactive T cells recognize an activa- JA, White FM, Marto JA et al (1999) Cutting
tion-induced minor histocompatibility anti- edge: the HLA-A*0101-restricted HY minor
gen encoded by the ATP-dependent histocompatibility antigen originates from
interferon-responsive (ADIR) gene. Blood DFFRY and contains a cysteinylated cysteine
109:4089–4096 residue as identified by a novel mass spectro-
52. Van Bergen CA, Rutten CE, Van Der Meijden metric technique. J Immunol 163:6360–6364
ED, Van Luxemburg-Heijs SA, Lurvink EG, 62. Meadows L, Wang W, den Haan JM, Blokland
Houwing-Duistermaat JJ et al (2010) High- E, Reinhardus C, Drijfhout JW et al (1997)
throughput characterization of 10 new minor The HLA-A*0201-restricted H-Y antigen con-
histocompatibility antigens by whole genome tains a posttranslationally modified cysteine that
association scanning. Cancer Res 70: significantly affects T cell recognition. Immunity
9073–9083 6:273–281
53. Stumpf AN, van der Meijden ED, van Bergen 63. Torikai H, Akatsuka Y, Miyazaki M, Warren
CA, Willemze R, Falkenburg JH, Griffioen M EH, Oba T, Tsujimura K et al (2004) A novel
(2009) Identification of 4 new HLA-DR- HLA-A*3303-restricted minor histocompati-
restricted minor histocompatibility antigens as bility antigen encoded by an unconventional
hematopoietic targets in antitumor immunity. open reading frame of human TMSB4Y gene. J
Blood 114:3684–3692 Immunol 173:7046–7054
54. Griffioen M, van der Meijden ED, Slager EH, 64. Rosinski KV, Fujii N, Mito JK, Koo KK, Xuereb
Honders MW, Rutten CE, van Luxemburg- SM, Sala-Torra O et al (2008) DDX3Y encodes
Heijs SA et al (2008) Identification of phos- a class I MHC-restricted H-Y antigen that is
phatidylinositol 4-kinase type II beta as HLA expressed in leukemic stem cells. Blood 111:
class II-restricted target in graft versus leukemia 4817–4826
reactivity. Proc Natl Acad Sci USA 105: 65. Ivanov R, Aarts T, Hol S, Doornenbal A,
3837–3842 Hagenbeek A, Petersen E et al (2005)
55. de Rijke B, van Horssens-Zoetbrood A, Identification of a 40S ribosomal protein
Beekman JM, Otterrud B, Maas F, Woestenenk S4-derived H-Y epitope able to elicit a
530 E. Spierings and E. Goulmy

lymphoblast-specific cytotoxic T lymphocyte DBY gene codes for an HLA-DQ5 restricted

response. Clin Cancer Res 11:1694–1703 human male specific minor histocompatibility
66. Vogt MH, Goulmy E, Kloosterboer FM, antigen involved in GvHD. Blood 99:
Blokland E, de Paus RA, Willemze R et al 3027–3032
(2000) UTY gene codes for an HLA-B60- 69. Zorn E, Miklos DB, Floyd BH, Mattes-Ritz A,
restricted human male-specific minor histo- Guo LX, Soiffer RJ et al (2004) Minor histo-
compatibility antigen involved in stem cell graft compatibility antigen DBY elicits a coordi-
rejection: characterization of the critical poly- nated B and T cell response after allogeneic
morphic amino acid residues for T-cell recogni- stem cell transplantation. J Exp Med 199:
tion. Blood 96:3126–3132 1133–1142
67. Warren EH, Gavin MA, Simpson E, Chandler P, 70. Spierings E, Vermeulen C, Vogt MH, Doerner
Page DC, Disteche C et al (2000) The human LEE, Falkenburg JHF, Mutis T et al (2003)
UTY gene encodes a novel HLA-B8-restricted Identification of HLA class II-restricted H-Y-
H-Y antigen. J Immunol 164:2807–2814 specific T-helper epitope evoking CD4+
68. Vogt MH, van den Muijsenberg J, Goulmy E, T-helper cells in H-Y-mismatched transplanta-
Spierings E, Kluck P, Kester M et al (2002) The tion. Lancet 362:610–615
 Chapter 30

Donor Registries and Search Strategies

Carolyn K. Hurley, Machteld Oudshoorn, and Michelle Setterholm

Abstract
The optimal donor of hematopoietic progenitor cells shares alleles of the major histocompatibility genes
with the recipient. This chapter describes the strategies aimed at identifying such a matched donor from
registries of volunteers or from umbilical cord blood banks.

Key words: Hematopoietic progenitor cell transplantation, Histocompatibility, Human leukocyte

antigen, Registries, Cord blood banks, Unrelated donor search

1. Introduction

Transplantation of hematopoietic progenitor cells (HPCs) is one

therapy for a variety of fatal blood diseases, such as acute myeloid
leukemia and aplastic anemia (1). To provide the best chance for an
optimal outcome to a transplant, the patient (i.e., the recipient) and
the cell donor must express the same histocompatibility molecules
on their cell surfaces (i.e., be “matched”). In humans, the major
histocompatibility molecules are the human leukocyte antigens
(HLA), a highly polymorphic family of six proteins termed HLA-A,
HLA-B, HLA-C, HLA-DR, HLA-DQ, and HLA-DP. The genes
encoding each HLA protein and their alleles are inherited in tradi-
tional Mendelian fashion. Because of the extensive allelic diversity
in the human population, a patient requiring an HPC transplant is
most likely to find a match among their siblings. However, unre-
lated donors are the source of HPCs for the approximately 70% of
patients who do not have a suitable family donor. To facilitate the
search for a matched unrelated donor, many countries have esta-
blished registries of volunteers who offer to provide HPCs if matched

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_30, © Springer Science+Business Media New York 2012

531
532 C.K. Hurley et al.

to a searching patient (2, 3). An alternative source of matched

HPCs is umbilical cord blood which has been banked for public use
(4). This chapter focuses on the procedures for an unrelated donor/
umbilical cord blood search and discusses strategies that might be
employed to find the optimal HLA matched volunteer donor or
umbilical cord blood unit for a specific patient.

2. Materials

2.1. Web Sites Useful 1. Bone Marrow Donors Worldwide (BMDW): This Web site
in Search (http://www.bmdw.org/) provides a comprehensive listing of
almost all registries and cord blood banks in the world and
their contact information. It also provides the tools to perform
an initial search of a single worldwide database of donors to
assess the likelihood of identifying a matched donor (5). This
database does not substitute, however, for a search of a specific
registry or cord blood bank because it is only a snapshot of
each registry/bank at one recent time and does not directly
allow access to specific donors/cord blood units for further
testing.
2. HLA nomenclature and allele designations: The Web site,
http://hla.alleles.org/nomenclature/, is the World Health
Organization-designated HLA nomenclature Web site and it
includes both current and previous HLA designations (6, 7).
The IMmunoGeneTics/HLA Web site (http://www.ebi.ac.
uk/imgt/hla/) is a source of information on each HLA allele
including whether the sequence has been reported more than
once (i.e., has been confirmed) and the race/ethnicity of the
individuals in whom the allele was first identified (8). It pro-
vides tools to compare HLA sequences and provides a listing
of alleles and genotypes that cannot be distinguished from one
another (i.e., are ambiguous) based on the nucleotide sequence
of their most variable exons. It also provides a tool to link sero-
logic and DNA-based HLA assignments based on an HLA
“dictionary” (9).
3. Allele and haplotype frequencies: The frequencies of alleles and
haplotypes (see Note 1) within the US National Marrow
Donor Program (NMDP) registry can be found at http://
bioinformatics.nmdp.org/ (see Note 2) (10). A list of rare
alleles identifies alleles that have never been observed or have
seldom been observed within the NMDP registry. This site
also includes a tool to identify the alleles that are contained
within the letter codes used to report HLA types (e.g.,
DRB1*04:AB where AB indicates 01 or 02 defining the assign-
ment DRB1*04:01 or DRB1*04:02) (7). Two sources of
 30 Donor Registries and Search Strategies 533

worldwide allele frequencies are http://www.allelefrequencies.

net (11) and http://www.pypop.org/popdata/ (12). Tools to
predict the frequencies of specific haplotypes can be found at
http://www.hlaexplorer.net and http://bioinformatics.nmdp.
org/ under HaploStats.

3. Methods

3.1. HLA Typing 1. Obtain a biological sample from the patient and from nuclear
of Patient and Family family members if transplantation is being considered as a
potential treatment option (see Note 3) and perform HLA
typing (see Note 4). If available, include parents of the patient,
siblings, and biological children. The patient should be typed
for HLA-A,-B, -C, -DR, -DQ (see Note 5) at allele or high
resolution (see Note 6). The resolution of the typing of the
family members should be sufficient to clearly distinguish the
HLA types segregating in the family (see Note 7).
2. If sufficient family members are available, confirm the HLA
assignments of the patient through observing the same assign-
ments in other family members, identify the HLA types carried
by each haplotype segregating in the family (see Note 8), and
determine if there is a family member who is HLA matched.
The latter should be further evaluated to determine if they
might serve as an HPC donor (see Note 9).
3. Perform an unrelated donor and umbilical cord blood search
as described below so that these options for treatment of the
patient can be considered.

3.2. Designing 1. Begin by evaluating whether the patient’s HLA assignments

an Unrelated Donor are complete. For potential 8/8 matches, the typing must
Search Strategy include HLA loci HLA-A, -B, -C, and -DRB1 (see Note 10).
In case 10/10 matches are preferred, the typing must include
HLA-A, -B, -C, -DRB1, and -DQB1 assignments. Also deter-
mine if the level of typing (i.e., the resolution of the typing) of
the patient is adequate (see Note 11).
2. Examine the HLA type of the patient to identify any uncommon
or unexpected assignments (see Note 12). This step evaluates
the accuracy of the HLA assignments.
3. Establish the patient’s HLA haplotypes (see Note 13).
4. Determine the frequency of the patient’s HLA alleles and hap-
lotypes in various ethnic populations based on Web sites listing
the frequencies of alleles and haplotypes. Table 1 provides an
example of the evaluation of the patient’s HLA assignments in
preparation for a search.
534 C.K. Hurley et al.

Table 1
Example of how a patient typing is evaluated

Web site providing

Issue noted information Action/resolution
Patient typing: A*02:05, 68:04; B*35:DNXN, 53:01; DRB1*03:01, 13:03; C*04:01; DQB1*02:01, 02:02
Is information included on all If the transplant center would like to
the loci required for search for a 10/10 match, assign-
matching? ments must be included for 5 loci;
for a 8/8 match, 4 loci. The typing
includes all the required loci for
either level of match
Evaluate resolution of testing: http://bioinformatics.nmdp. Expand to full allele string. Depending
allele B*35:DNXN has an org DNA type lookup tool on the assignments, request
allele code additional typing to increase the
resolution. In this case, the assign-
ment includes only alleles that share
the sequence of exons 2 and 3:
B*35:01, B*35:40N, B*35:42,
B*35:57, B*35:94
Evaluate resolution of testing: http://hla.alleles.org/alleles/ Ask the laboratory if it specifically
C*04:01 allele is at high index.html G groups excluded the other alleles that
resolution but this level of carry the same nucleotide sequence
resolution required special as C*04:01 in the HLA-C exons
testing by the HLA specifying the peptide binding site.
laboratory For example, did the laboratory
specifically exclude the non-ex-
pressed allele C*04:09N?
(see Note 46)
Evaluate frequency of alleles: http://bioinformatics.nmdp. In family? If yes, acceptable. If no,
allele A*68:04 is org rare allele list repeat HLA typing of HLA-A
uncommon preferably with different reagents/
method
Investigate B/C and DR/DQ http://bioinformatics.nmdp. Concordant with family typing?
haplotypes: B *35:DNXN org/ under HaploStats Common or uncommon? In this
with C *04:01, B*53:01 case, the B-C associations are
with C*04:01, common. Both DRB1 alleles are
DRB1*03:01 with frequently observed with
DQB1*02:01 or *02:02, DQB1*02:01 but not with
DRB1*13:03 with DQB1*02:02 so this is unusual.
DQB1*02:01 or *02:02 The laboratory should be asked
to confirm the DQB1*02:02
assignment if it is not observed
in the family
Investigate multilocus haplo- http://bioinformatics.nmdp. Concordant with family typing?
types (A/B/C/DRB1/ org under HaploStats Common or uncommon?
DQB1)
 30 Donor Registries and Search Strategies 535

3.3. Submitting 1. An unrelated donor or umbilical cord blood search should

a Search Request be initiated to identify donors and cord blood units whose
to a National or HLA typing is potentially identical to the patient’s (see Notes
International Registry 14 and 15).
2. Collect the following information about the patient: HLA
typing data, preferably including the patient’s HLA haplo-
types; date of birth; diagnosis date; gender; weight; cytomega-
lovirus (CMV) status; ethnicity (see Note 16).
3. Determine the parameters of the search, including the type of
search to be performed, i.e., adult progenitor cell donor, cord
blood unit or both (see Note 17); urgency of the search; the
specific HLA loci to be considered during matching; and
whether or not mismatches are acceptable. These criteria are
found in the transplant center protocol (see Note 18).
4. Perform a search of BMDW which provides you with a com-
prehensive overview of almost all HPC sources worldwide (see
Notes 19 and 20).
5. Complete a search request form and submit the search request
form to all registries of interest or electronically enter the
patient’s information directly using software or Web access
where available. A generic form accepted by many registries
and cord blood banks (e.g., WMDA Preliminary Search
Request Form) can be found under http://www.worldmarrow.
org. Alternatively, a search can be started through a registry to
registry network, such as the European Marrow Donor
Information System (see Note 21).

3.4. Reviewing 1. Obtain a search report from the registries and banks contacted
a Search Report (see Note 22). Most registries will display adult donor matches
to Select Potential based on HLA-A,-B,-DRB1 typing and will list both 6/6 and
Adult Volunteer 5/6 matches. The selection of umbilical cord blood units is
Donors for Additional described in Subheading 3.6.
HLA Testing 2. Ensure that the correct match criteria are known for the patient
(whether an 8/8 or 10/10 match is preferred; which and how
many mismatches are allowed; and any other donor-specific
characteristics that need to be considered, such as gender, age,
or CMV status). This will guide the selection of the best poten-
tial donors and help to increase the likelihood of a successful
transplantation (see Note 23).
3. Expand any donor typing results (see Note 24) that may have
letter codes (e.g., A*01:CRY) as necessary to understand the
full set of alleles in the assignments. See Note 25 for a tool to
assist with this task.
4. Evaluate the donor typing results for completeness according to
the transplant center match criteria. For example, if an 8/8 match
536 C.K. Hurley et al.

is preferred, do the donor typing results include HLA-A, -B, -C,

and -DRB1? If incomplete, utilize haplotype frequency data
relative to the ethnicity of each donor to help predict the most
likely HLA type of the missing loci (see Note 26).
5. Evaluate the donor typing results to determine if the level of
resolution at each locus is sufficient in order to make a deter-
mination of the likelihood to match the patient’s alleles. Utilize
allele and haplotype frequency data and the donor’s ethnicity
to predict the most likely higher resolution types for each locus
for the donors of interest on the search report (see Note 27).
Table 2 provides an example of the evaluation of a donor’s
HLA assignments.
6. Prioritize the donors deemed most likely to match the patient’s
HLA type using all of the information gathered in the analyses
above. If multiple, suitably matched donors are available on
the search list, apply the additional transplant protocol criteria
to the non-HLA information given for each donor to perform
a secondary sort. For example, prioritize younger donors over
older.
7. Request a biological sample from the selected donors for further
typing from the Registry (see Note 28).
8. Perform HLA typing of the samples to obtain more information
as to the HLA alleles carried at key loci and to verify the HLA
typing assignment (see Note 29).
9. Select the best donor based on the transplant center criteria
(see Notes 30 and 31).
10. Contact the registry to request that a specific volunteer be
medically evaluated for donation.

3.5. Refining the 1. When the search report or the analyses performed above does
Search Strategy not yield donors who potentially match the patient, a strategy
to Identify a to locate the best potential mismatch may need to be imple-
Mismatched or mented. In these instances, a donor with only one mismatch
Mistyped Donor (7/8 or 9/10) may be considered as an alternative depending
on the transplant center protocol (see Note 32). The search
reports from most registries include 5/6 matched donors (i.e.,
with one mismatch at one of the three loci, HLA-A, -B,
-DRB1). A longer list of potential donors may be needed if the
search was initiated through EMDIS since some registries limit
the search results to either 50 or 100 donors.
2. Evaluate the patient for the presence of IgG antibodies directed
to specific HLA antigens (see Note 33).
3. Evaluate the methodology that may have been utilized to gen-
erate the HLA data of the potential donors. Assignments made
using older serologic typing techniques have a higher likeli-
hood of appearing falsely disparate than typing techniques that
Table 2
Examples of how potential donor typing is evaluated for patient in Table 1

Issue noted Web site providing information Action/resolution

Patient typing: A*02:05, 68:04; B*35:DNXN, 53:01; DRB1*03:01, 13:03; C*04:01
Donor 1 typing: A2, 68; B35, 53; DRB1* 03:DHJB, 13:BGTY
Allele codes need expansion http://bioinformatics.nmdp.org DHJB: 03:01/03:06/03:18/03:19/03:22/03:28/03:37
DNA type lookup tool
HaploStats BGTY: 13:03/13:33/13:66
The patient’s DR alleles are included in the donor assignments and, based on the
30

frequency of the alleles in the code, are likely to be matched (i.e., DRB1*13:03
is more common than DRB1*13:33 or DRB1*13:66)
Determine likelihood for HaploStats Predict higher resolution and likelihood of allele match with patient based on
allele match: serologic ethnicity. A*02:01 is more frequent than A*02:05; it is unlikely that the donor
HLA explorer
typing at HLA-A and -B carries A*68:04. B*35:01 is common and may be present in B35 typed indi-
DNA dictionary vidual although other B*35 alleles (B*35:02, B*35:03) are possible; B*53:01 is
likely to be present
Evaluate typing for http://bioinformatics.nmdp.org/ Predict HLA-C alleles that might be present. If B*35:01 and B*53:01 are present,
completeness: HLA-C index.html, haplotype table both B alleles are frequently found with C*04:01
not typed
Is resolution sufficient The donor is likely an HLA-A mismatch (i.e., not A*68:04). A donor matched at
to make a determination 7/8 or 9/10 may be acceptable to the transplant center. Depending on other
of match? donors available on the search report, determine whether this donor should be
typed at higher resolution
Donor Registries and Search Strategies
537
538 C.K. Hurley et al.

utilize DNA-based methodologies. If the patient carries an

antigen that is known to be difficult to distinguish by serologic
methods, then donors typed serologically who have a high
likelihood of matching the patient may appear as mismatched
on the search report (see Note 34). Utilize the HLA diction-
ary to identify any of these antigens (see Note 35). Perform a
new search if needed by substituting the patient’s difficult allele
or antigen with an alternate choice as indicated by the HLA
dictionary “assigned type” field (see Note 36).
4. Evaluate the possibility that allele level differences between a
donor and the patient may be permissible mismatches due to the
location of the disparity. Some alleles that have different assign-
ments are actually identical, either genotypically or phenotypi-
cally, in the region that encodes the peptide binding site and
may not impact the transplant outcome (see Notes 37 and 38).
5. Review the patient’s HLA assignment to determine if an
uncommon haplotype or allele may be present (see
Subheading 3.2, step 4). In these cases, it may be possible to
substitute a donor with a single mismatched allele at the locus
in question and meet the 7/8 or 9/10 transplant center crite-
ria (see Note 39).
6. Prioritize the loci involved and the selection of potentially
mismatched donors according to the transplant center crite-
ria and using recommendations given in recent publications
(see Note 40). Avoid choosing donors carrying HLA anti-
gens to which the patient is sensitized.

3.6. Searching 1. Review the cord blood units that have been listed on the
for a Cord Blood Unit patient’s search report (see Note 41).
2. Identify units with greater than a minimal required cell dose
based on the transplant center protocol (see Note 42).
3. Identify units that may meet the transplant center’s HLA match-
ing criteria for a cord blood unit (see Note 43). The units will
likely be HLA typed at a low resolution and higher resolution
testing may be required for optimal donor selection.
4. Select one or more units that have both an acceptable cell dose
and an acceptable HLA match. Request that the Registry/
Cord Blood Bank initiate further HLA testing of the selected
unit or request a sample for typing at your local HLA labora-
tory (see Note 44). Also request a Unit Report that gives more
information about the unit plus the maternal health history.
5. Consider whether the unit has attached segments for verification
typing (see Note 45) in making your decision about whether
to request the unit for transplant.
6. Request that the Registry/Cord Blood Bank reserve the unit
for your patient until you have completed your evaluation.
 30 Donor Registries and Search Strategies 539

4. Notes

1. A haplotype is the combinations of HLA alleles that are found

on a single copy of chromosome 6. A haplotype may refer to a
combination of two or more loci. Some allele combinations
such as the combination of DR and DQ alleles found on a
haplotype or the combination of HLA-B and HLA-C alleles
found on a haplotype can be predicted based on previous pop-
ulation studies. For example, haplotypes with DRB1*01:01
almost always carry DQB1*05:01. Tables of these associations
can be found at http://bioinformatics.nmdp.org/ and at
http://hlaexplorer.net.
2. The excel files on this Web site can be filtered and sorted.
3. It is important to evaluate the likelihood of finding a suitable
HPC donor early during the initial evaluation of the patient
when various treatment options are being considered. The
outcome of a transplant is more favorable when the patient is
transplanted early in the course of their disease (13).
4. See chapters on HLA typing for more information on the meth-
odology. Blood may not be the best sample source for some
patients because of the low number of circulating white cells,
because of the possibility of mutations in the HLA genes due to
the unstable nature of tumors, or because of the presence of
transfused cells from another individual. In these cases, a buccal
swab or hair follicles might be the preferred sample source.
5. DRB1 and DQB1 genes specify one polypeptide of the DR
and DQ heterodimers. These genes are routinely typed during
donor selection (14). The DRA and DQA1 genes are not usu-
ally typed as they are less polymorphic and provide only limited
additional information. The secondary DRB loci, DRB3,
DRB4, and DRB5, are often typed. One purpose for typing
the HLA loci suggested is to provide further support for the
HLA assignments. Specific associations are found between: (1)
DRB1 and DQB1 alleles (e.g., DRB1*01:01 is found predom-
inantly associated with DQB1*05:01); (2) HLA-B and -C
alleles (e.g., B*07:02 with C*07:02) and (3) DRB1 and the
secondary DRB loci (DRB1*15:01 with DRB5 alleles). These
associations should be consistent with the HLA assignments.
6. Allelic resolution is defined as a DNA-based typing result that
is consistent with a single allele. An allele is defined as a unique
nucleotide sequence for a gene as defined by the use of all of
the digits in a current allele name. High resolution typing is
defined as a typing which includes alternative alleles that spec-
ify the same protein sequence for the region of the HLA mol-
ecule that forms the peptide binding site. The peptide binding
540 C.K. Hurley et al.

site is composed of the amino-terminal domains of the HLA

proteins (i.e., alpha1 and alpha2 domains for HLA class I mol-
ecules and alpha1 and beta1 domains for HLA class II mole-
cules). This is the region that binds self or foreign peptides and
that is recognized by T-cell receptors. High resolution also
excludes alleles that are not expressed as cell-surface proteins,
especially if these null alleles are thought to be common.
7. In general, low resolution HLA typing may be used for typing
the patient’s family members except in the cases where the par-
ents are not available or where similar alleles are segregating in
the family. A low resolution DNA-based typing result is at the
level of the digits comprising the first field in the DNA-based
nomenclature (e.g., A*11; B*07).
8. Haplotypes are defined by segregation in the family. For exam-
ple, father is typed as A*01, A*02, B*07, B*08; mother as
A*11, A*24, B*18, B*35; sib 1 as A*01, A*24, B*08, B*18;
and sib 2 as A*01, A*11, B*08, B*35. Based on segregation,
the haplotypes are A*01-B*08, A*02-B*07, A*11-B*35, and
A*24-B*18.
9. If there is not a suitable match in the nuclear family, it may be
possible to find a match in the extended family (e.g., aunts,
uncles, cousins) (15). This is possible if the patient carries a
common haplotype which might have been inherited by chance
in the extended family. It is also possible if there is a history of
consanguinity in the family (e.g., cousin–cousin marriages) or
if there are other close family relationships (e.g., two siblings
from one family married to two siblings from another family).
10. 8/8 means 2 alleles at 4 loci, usually defined as HLA-A, -B, -C,
and -DRB1.
11. Low or intermediate resolution typing will only provide a
rough indication of possible donors. Patient typing at this res-
olution should not be used for an unrelated donor search.
Time and money may be wasted if donors with the highest
chance of being identical with the patient are not rapidly
selected based on the patient’s allele/high resolution typing
assignment.
12. It can be useful to know whether the patient has an allele that
is rare or an unexpected allele combination as this may be a
typing error. Rare alleles in patients should always be confirmed
either in the family (i.e., by finding another family member car-
rying the same rare allele) or by using a different typing tech-
nique or reagent set. To know whether the patient has a rare
allele is also important as it may affect the likelihood of finding
a perfectly matched donor. If no potentially identical donor is
available, it is best to look for a donor who is identical for all
alleles except the one rare allele. It is also important to realize
 30 Donor Registries and Search Strategies 541

that there are alleles that may be rare or uncommon in all but
one ethnic group. In this case, finding a matched donor can be
greatly facilitated by selecting potential donors from the ethnic
group in which your patient’s allele is less rare or even
common.
Furthermore, it may be very difficult to find donors if the
patient has a rare HLA association (e.g., HLA-B + HLA-C
allele association) and this needs to be identified by using HLA
haplotype frequency tables. For example, if the patient has a
rare HLA-B + HLA-C association and no potentially identical
donors are present, one can substitute the HLA-B allele for
another allele that is commonly associated with the patient’s
HLA-C allele. See Subheading 3.5 for more information on
search strategies to find a mismatched donor.
13. Preferentially, family typing is available and the patient’s haplo-
types can be established through family segregation analyses.
If family data are not available, there are haplotype frequency
tables that can be used to predict which haplotypes the patient
might carry and also to provide information whether or not
the patient has common haplotypes. Understanding common
haplotype or allele associations can be useful for predicting
which alleles are most likely to be present in donors who have
only low resolution typing information and no family data to
define haplotypes. Two useful Web sites for haplotype analysis
are http://bioinformatics.nmdp.org and http://www.hlaex-
plorer.net.
14. Registries and banks have guidelines for who can submit a
search request. Usually, search requests must come from a
physician or from a transplant center.
15. Typically, searches are performed in two steps. In the prelimi-
nary search, a physician will receive a summary indicating
whether or not there are potentially matching donors in a
registry or cord blood bank. If the registry appears to include
suitable donors, a transplant center will formalize the search.
In a formal search, the HLA types of each potentially matching
donor/unit will be listed individually and the transplant center
can request biologic samples from these donors/units for
further HLA and infectious disease testing.
16. It is important to confirm the patient’s typing using a second
biological sample to ensure that no errors are made in the HLA
assignments to be used in the unrelated donor/umbilical cord
blood search.
17. It is recommended that the search identify the best unrelated
donor and the best umbilical cord blood unit to provide alter-
natives for the physician. An urgent search might focus only on
umbilical cord blood units because they are more rapidly
available.
542 C.K. Hurley et al.

18. The level of matching required for optimal outcome has been
discussed in the literature, e.g., (13, 14, 16, 17).
19. Access to the search tool in BMDW requires that a transplant
center register with BMDW. The search is submitted online.
Matches identified in the BMDW worldwide registry will iden-
tify one or more countries where potential donors or cord
blood units can be found for the patient. The transplant center
can then focus the search on registries and cord blood banks
from those countries. Contact information for these registries
and banks can be found on the BMDW Web site.
20. The national registry in the country of the transplant center
caring for the patient may also assist in contacting multiple
registries and banks. For example, a search of the NMDP in
the USA will also interrogate the inventory of the BMDW.
Any physician may request a search of the NMDP on behalf of
a patient through the Office of Patient Advocacy (http://www.
bethematch.org). There are also several publicly available Web
sites that provide summarized search results (e.g., MatchView
at http://www.marrow.org). When using the NMDP, staff will
perform all of the contacts and assist with HLA expertise to
guide the process of selecting the best potential donor or cord
blood unit.
21. Search requests often go from registry to registry using the
European Marrow Donor Information System (EMDIS).
Transplant centers may have the ability to directly enter the
search information into a software application.
22. Searches performed electronically using BMDW or NMDP
generate donor lists immediately. Most registries and cord
blood banks will provide a search report within 24 h when
using the EMDIS system. The matches displayed usually
include 6/6 (HLA-A,-B,-DRB1) and 5/6 matched donors
but the criteria for the donors displayed on a search should be
confirmed with the registry.
23. Non-HLA factors that are often considered in donor selection
have been reviewed by Confer et al. (18).
24. The donors/units listed on the report are sorted by the extent
of matching with the patient. It is possible, however, that the
optimal donor may not appear at the top of the list so it is criti-
cal to examine all of the donors on the report.
25. An HLA typing result may include a two or more letter code
following the first colon in the allele name (e.g., A*02:AF
where AF is a code standing for 01 or 09 indicating that an
individual carries A*02:01 or A*02:09). The Web site, http://
bioinformatics.nmdp.org, has a tool to define alleles included
in any letter codes. Further information on HLA assignments
can be found at http://hla.alleles.org and http://www.ebi.ac.
uk/imgt/hla/.
 30 Donor Registries and Search Strategies 543

26. There are several approaches that can be used to predict the
most likely missing data. The HaploStats tool and the HLA
Explorer will generate a table of the most frequently seen
HLA-C types associated with a known HLA-B and the most
frequently seen HLA-DQB1 types associated with a known
DRB1 antigen sorted by likelihood and ethnicity. The complete
list of HLA-A, -B, -C, -DRB1, and -DQB1 haplotypes found on
http://bioinformatics.nmdp.org can be used by searching on
the known HLA-A, -B, and DRB1 haplotype and then viewing
the possibilities of full haplotype combinations considering the
frequencies of each and the racial/ethnic group of the donor
(if known).
27. The HaploStats and HLA Explorer tools will each display the
possible high resolution types given any low resolution input.
The difference in the two tools is that HaploStats requires all
known loci to be entered and HLA Explorer will provide data
on any single low resolution type. They both sort by frequency
within racial/ethnic populations.
28. High or allele resolution typing should be performed to identify
those donors who carry the same HLA alleles as the patient.
This extended typing should include all loci to be considered for
matching. Verification typing using a fresh biological sample
from the donor should be performed in order to ensure that this
individual was the one with the HLA assignment appearing on
the search report. The resolution of verification typing should
be sufficient to confirm the original HLA assignments.
29. It is recommended that more than one donor be selected for
additional HLA testing. This is because some donors may not
carry the same alleles as the patient following higher resolution
testing or because some volunteers will be unwilling or unable
to donate or cannot be located. The number of donors to
select will depend on the likelihood of the patient’s alleles and
haplotypes being present in the donors. For common alleles
and haplotypes, 3–5 donors might be selected for further HLA
testing. Uncommon alleles and haplotypes may require 10 or
more donors to be selected for extended HLA typing. The
registry may have information on the rate of donor availability
to help with this decision.
30. The donor’s registry or the cord blood bank will request that
your extended HLA typing results be provided to them to
update the donor/unit file. Follow the policy of the registry
for the loci and level of resolution to be tested when reporting
the results for each donor.
31. If there are multiple suitable matched donors, HLA assignments
at other HLA loci, not routinely considered by the transplant
center in determining the match, can be used to select among
the donors. For example, if the transplant center requests an
8/8 match and there are several 8/8 matched donors, matching
544 C.K. Hurley et al.

for HLA-DQB1, HLA-DRB3/DRB4/DRB5, or HLA-DPB1

alleles might be used to prioritize donors.
32. A single mismatched donor is defined as a 7/8 or 9/10
matched donor (14).
33. Prior sensitization to specific HLA antigens may result in rejec-
tion of grafts expressing those antigens (19, 20). Antibody
assays are described in Chapter 17.
34. For example, if the patient carries A*01:01, A*74:01, it may
be possible that those donors typed only as HLA-A1 using
serology might actually carry HLA-A74 since HLA-A74 was
frequently missed by serology. Lists of frequently missed or
misassigned serologic antigens have been reported (21, 22).
35. The HLA dictionary can also be found online at http://www.
ebi.ac.uk/imgt/hla/ (9).
36. An example from the HLA dictionary is B*39:20 that has an
assigned type of “blank” (67%) or B38 (33%). If the patient
carries B*39:20, B*07:02, an alternative phenotype search
might be submitted in which the patient’s HLA-B assignment
is listed as B*38:XX, B*07:02. In some cases, acceptable
donors may not appear on the search report because their HLA
typing appears to be more mismatched than the registry’s min-
imum criteria. For example, a search report usually lists 6/6
and 5/6 matched donors. A donor who appears as a 4/6 match
will not be shown on the report. An example for a patient
typed as A*02:01, A*03:01, B*39:20, B*07:02 is a donor
typed as A11, A3, B38, B7. Submission of an alternative phe-
notype search replacing B*39:20 with B38 may yield addi-
tional 5/6 donors who did not appear on the initial search.
37. For example, a patient may carry DRB1*14:01 and a potential
donor might carry DRB1*14:54. These two alleles specify
proteins that are identical in amino acid sequence in the por-
tion of the HLA molecule called the peptide binding site.
Note 6 further defines this region of the HLA molecule. Since
it is thought that allorecognition focuses on this region of the
HLA molecule, the single amino acid difference outside of the
peptide binding site that distinguishes these two HLA mole-
cules is thought to be unlikely to trigger allorecognition (23).
Thus, differences in HLA molecules that fall outside of the
peptide binding site might be considered permissive. A list of
alleles that share the sequence of their peptide binding site can
be found at http://hla.alleles.org/nomenclature/.
38. Several studies have discussed permissive mismatches (13, 23–26)
or have evaluated tools to select these mismatches (27, 28).
39. For example, if the patient carries an uncommon HLA-B locus
allele, the search may yield a donor carrying a differing HLA-B
locus allele that still matches the patient’s HLA-C alleles.
 30 Donor Registries and Search Strategies 545

For example, if the patient carries B*35:15-C*07:02, the rare

B*35:15 may be substituted by B*07:02, as B*07:02 is strongly
associated with C*07:02. Likewise, this strategy can be utilized
to identify an HLA-DRB1 mismatched donor who may mis-
match the patient’s uncommon HLA-DRB1 yet has high
potential to match the patient’s HLA-DQB1 allele. For exam-
ple, if the patient carries DRB1*04:07-DQB1*03:01, the rare
DRB1*04:07 may be substituted by DRB1*11:01 as
DRB1*11:01 is strongly associated with DQB1*03:01.
40. The selection of which locus to mismatch has been discussed in
the literature (13, 14, 16). Evidence guiding the selection of a
donor with the most permissive mismatch at a locus is limited;
most studies have evaluated strategies that ultimately did not
predict permissive mismatches (13, 26–28).
41. The units shown on the search report likely meet some mini-
mum HLA match and/or total nucleated cell (TNC) criteria
defined by the registry or cord blood bank. The units may be
sorted into categories based on the degree of match or size of
the unit. It should not be assumed that the optimal unit is
found at the top of the list; all units on the report should be
evaluated.
42. Cell dose is one of two critical factors in the selection of a unit;
HLA match is the second critical factor. The unit selected must
contain a minimum of 2.5–3 × 107 precryopreserved TNC per
kilogram of recipient body weight (29). If the possibility of
using two cord blood units is being considered, the minimum
TNC/kg for a unit as part of a double cord transplantation
may be lower, but together the 2 units should have at least
2.5–3 × 107 TNC/kg.
43. The criteria that are often used for matching is a minimum of
a four of six match, two antigen level matches at HLA-A and
HLA-B and an allele match at HLA-DRB1. An antigen match
is defined as a match at the first two digits of the allele name.
For example, a patient typed as A*02:05 and a unit typed as
A*02:01 would be considered an antigen match. Matching
among B*15 and B*40 alleles is the exception. In these cases,
the serologic assignment of the allelic products determines
whether the patient and donor are matched at the antigen
level. For example, a patient typed as B*15:01 (serologic B62)
would not be an antigen match with a donor typed as B*15:03
(serologic B70). Use the HLA dictionary to obtain the sero-
logic types associated with B*15 and B*40 alleles (9).
44. Recent reports suggest that better matching of the patient and
the umbilical cord unit may result in better outcome (29).
Therefore, it is recommended that the unit be tested at allele
to high level resolution at HLA-A, -B, -C, -DRB1 so that the
best matched unit can be selected for the patient.
546 C.K. Hurley et al.

45. It is important to insure that the cord blood unit is the one
with the HLA typing found on the search report. This
verification typing should be performed on a biological sample
found in an aliquot physically attached to the unit.
46. The specifications for the resolution of HLA typing should be
established by the transplant center in discussion with the his-
tocompatibility laboratory. The resolution of the patient’s typ-
ing assignment will determine which donors appear on a search
report and may vary among registries. For example, donors
listed on a search report for a patient typed as C*04:01 may
differ from the donors listed for a patient typed as C*04:01:01G
(which includes alternative alleles C*04:01:01:01,
C*04:01:01:02, C*04:01:01:03, C*04:01:01:04, C*04:09N,
C*04:28, C*04:30, C*04:41, C*04:79, C*04:82, C*04:84)
or typed as C*04:FEAS (which includes alternative alleles
C*04:01, C*04:09N, C*04:28, C*04:30, C*04:41).

References

1. Copelan EA (2006) Hematopoietic stem- 10. Maiers M, Gragert L, Klitz W (2007) High-
cell transplantation. N Engl J Med 354: resolution HLA alleles and haplotypes in the
1813–1826 United States population. Hum Immunol
2. Hurley CK et al (2007) Overview of registries, 68:779–788
HLA typing and diversity, and search algo- 11. Middleton D et al (2003) New allele frequency
rithms. Tissue Antigens 69(Suppl 1):3–5 database. Tissue Antigens 61:403–407. http://
3. Foeken LM et al (2010) Monitoring the inter- www.allelefrequencies.net
national use of unrelated donors for transplan- 12. Solberg OD et al (2008) Balancing selection
tation: the WMDA annual reports. Bone and heterogeneity across the classical human
Marrow Transplant 45:811–818 leukocyte antigen loci: a meta-analytic review
4. Welte K et al (2010) International exchange of of 497 population studies. Hum Immunol
cord blood units: the registry aspects. Bone 69:443–464
Marrow Transplant 45:825–831 13. Lee SJ et al (2007) High-resolution donor-
5. Oudshoorn M et al (1994) Bone Marrow recipient HLA matching contributes to the suc-
Donors Worldwide: a successful exercise in cess of unrelated donor marrow transplantation.
international cooperation. Bone Marrow Blood 110:4576–4583
Transplant 14:3–8 14. Bray RA et al (2008) National marrow donor
6. Marsh SG et al (2010) Nomenclature for fac- program HLA matching guidelines for unrelated
tors of the HLA system, 2010. Tissue Antigens adult donor hematopoietic cell transplants. Biol
75:291–455 Blood Marrow Transplant 14:45–53
7. Bochtler W et al (2007) World Marrow Donor 15. Schipper RF, D’Amaro J, Oudshoorn M (1996)
Association guidelines for use of HLA nomen- The probability of finding a suitable related
clature and its validation in the data exchange donor for bone marrow transplantation in
among hematopoietic stem cell donor registries extended families. Blood 87:800–804
and cord blood banks. Bone Marrow Transplant 16. Petersdorf EW (2008) Optimal HLA matching
39:737–741 in hematopoietic cell transplantation. Curr
8. Robinson J et al (2009) The IMGT/HLA Opin Immunol 20:588–593
database. Nucleic Acids Res 37:D1013–D1017 17. Shaw BE et al (2010) Diverging effects of
9. Holdsworth R et al (2009) The HLA diction- HLA-DPB1 matching status on outcome fol-
ary 2008: a summary of HLA-A, -B, -C, lowing unrelated donor transplantation depend-
-DRB1/3/4/5, and -DQB1 alleles and their ing on disease stage and the degree of matching
association with serologically defined HLA-A, for other HLA alleles. Leukemia 24:58–65
-B, -C, -DR, and -DQ antigens. Tissue Antigens 18. Confer DL et al (2010) Selection of adult
73:95–170 unrelated hematopoietic stem cell donors:
 30 Donor Registries and Search Strategies 547

beyond HLA. Biol Blood Marrow Transplant of HLA mismatches on bone marrow transplant
16:S8–S11 outcomes in the United States. Biol Blood
19. Spellman S et al (2010) The detection of donor- Marrow Transplant 15:971–981
directed, HLA-specific alloantibodies in recipi- 25. Kawase T et al (2007) High-risk HLA allele
ents of unrelated hematopoietic cell mismatch combinations responsible for severe
transplantation is predictive of graft failure. acute graft-versus-host disease and implication
Blood 115:2704–2708 for its molecular mechanism. Blood
20. Ottinger HD et al (2002) Positive serum cross- 110:2235–2241
match as predictor for graft failure in HLA- 26. Wade JA et al (2007) HLA mismatching within
mismatched allogeneic blood stem cell or outside of cross-reactive groups (CREGs) is
transplantation. Transplantation 73:1280–1285 associated with similar outcomes after unrelated
21. Lau M, Park MS, Terasaki PI (1997) hematopoietic stem cell transplantation. Blood
International cell exchange 1974–1996, a 109:4064–4070
23-year documentation. (eds. Terasaki PI and 27. Duquesnoy R et al (2008) HLAMatchmaker-
Gjertson DW. It was published by the UCLA defined triplet matching is not associated with
Tissue Typing Laboratory in Los Angeles better survival rates of patients with class I HLA
California) 85–124 allele mismatched hematopoietic cell trans-
22. Noreen HJ et al (2001) Validation of DNA- plants from unrelated donors. Biol Blood
based HLA-A and HLA-B testing of volunteers Marrow Transplant 14:1064–1071
for a bone marrow registry through parallel 28. Spellman S, Klein J, Haagenson M, Askar M,
testing with serology. Tissue Antigens Baxter-Lowe LA, He J, et al (2011) Scoring
57:221–229 HLA Class I Mismatches by HistoCheck Does
23. Xiao Y et al (2009) Evaluating the potential Not Predict Clinical Outcome in Unrelated
impact of mismatches outside the antigen rec- Hematopoietic Stem Cell Transplantation. Biol
ognition site in unrelated hematopoietic stem Blood Marrow Transplant 2011 Sep 29
cell transplantation: HLA-DRB1*1454 and 29. Kamani N et al (2008) State of the art review:
DRB1*140101. Tissue Antigens 73:595–598 HLA matching and outcome of unrelated
24. Baxter-Lowe LA et al (2009) HLA-A dispari- donor umbilical cord blood transplants. Biol
ties illustrate challenges for ranking the impact Blood Marrow Transplant 14:1–6
 Chapter 31

Cytokine Gene Polymorphisms: Methods of Detection

and Biological Significance
Gurvinder Kaur and Narinder Mehra

Abstract
In recent years, several functional polymorphisms, particularly, SNPs have been identified in cytokines and
their receptor genes that regulate levels of cytokine expression. These have been implicated as immune
prognostic markers in diseases, including differential response to therapy and as biomarkers of graft out-
come following organ and stem cell transplantation. Population distribution of cytokine gene polymor-
phisms (CGPs) reveals significant variations in allele frequencies in different ethnic groups and this might
explain, to some extent, the observed differences in SNP associations with various diseases and immune-
pathologies. A number of molecular methods are available for defining CGPs. These include PCR-SSP,
AFLP, Taqman probe assays as well as sequencing based typing. Of these, the PCR-based sequence-specific
primer based test is the most widely accepted technique. This chapter describes steps involved in this pro-
cedure along with sources for procuring essential reagents. An important aspect of CGP analyses is the
correct interpretation of results particularly determination of their multilocus haplotypes.

Key words: Cytokinesm, Single nucleotide polymorphisms, PCR, Population diversity, SSP primers

1. Introduction

1.1. What Are Cytokines (Greek cyto-, cell; and -kinos, movement) are pleiotro-
Cytokines? pic immuomodulatory proteins produced by a wide range of cell
types through antigen-specific and nonspecific stimuli. The T-cell
subset paradigm and its immunological cytokine networking have
attracted much attention over the last decade. With the recent
knowledge on cytokine gene polymorphisms (CGPs), differences
between individuals have been discovered that influence not only
cytokine gene expression, but also susceptibility to diseases, their
progression, severity, and clinical outcomes. Ethnic differences

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_31, © Springer Science+Business Media New York 2012

549
550 G. Kaur and N. Mehra

observed in patterns of CGPs at the population level are mainly a

result of natural selection (1, 2) imposed by invading microbiota,
environmental factors, and complex host–pathogen interactions (3).
These correlate with population-based variations in the ability to
mount an immune response and therefore act as important biomark-
ers of risk assessment in various diseases as well as for the develop-
ment of graft- versus host disease (GvHD) or graft rejection following
organ and hematopoietic stem cell transplantation (1).
Several studies have examined a possible relationship between
CGPs and a variety of diseases, including inflammatory, infectious,
autoimmune, malignant, and transplant-related conditions. Online
databases of established associations between cytokine gene
diversity and disease have been developed (1). The CGP-based
studies received a practical boost during the 13th International
Histocompatibility Workshop (13IHWC) when specific cytokine
genotyping CTS-SSP tray kits were developed for the first time by
Joannis Mytilineos and his group for testing by individual partici-
pating laboratories. These kits are now made available by the
Collaborative Transplant Study (CTS) group led by Gerhard
Opelz at the Univ of Heidelberg, Germany and are being used
extensively for investigating the role of CGP SNPs in organ and
stem cell transplantation (4) (http://www.ctstransplant.org). The
main rationale behind this collaborative effort is to encourage
investigations on immunogenetic markers for a given disease,
ascertain genotype–phenotype correlates and predict clinical
responses to therapeutics, vaccines, and transplant outcomes. Data
generated over the years has identified several important SNPs in
various cytokines that are important markers not only for better
understanding of the etiology and pathology of a disease, but also
as potential biomarkers of disease susceptibility and severity.

1.2. Cytokine Genes Genetic polymorphisms are frequently encountered in the 5¢ and
Polymorphisms 3¢ regulatory regions or introns of a cytokine and/ or its receptors
(Fig. 1) and these have been shown to play an important role in
regulating the extent of gene expression through one or more of
the following possible mechanisms: (1) by altering the structure of
transcription factor binding sites within promoters, (2) manipulat-
ing the structure of enhancers or silencers within introns or at dis-
tant regions, (3) altering the binding sites within nuclear matrix for
architectural transcription factors that modulate promoter activity
(1, 5, 6).
Cytokines and their receptor genes are generally conserved in
their exonic sequences with a few exceptions like IL4 receptor,
TNFb, TGFb, GM-CSF, and IL2 receptor-g gene (6). Most of the
observed mutations are synonymous, i.e., do not affect the protein
sequences, although they might influence cellular levels of the
cytokine expression. Broadly, three different approaches have been
employed for the study of CGPs. These include case-control
 31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 551

Fig. 1. Commonly studied polymorphisms in human type 1, type II, and other cytokines and their receptors. The chromo-
somal localization, db SNP names, positions of functional polymorphisms in relation to gene start site and the nucleotide
substitutions are given.

comparisons, transmission disequilibrium tests, and genome-wide

studies. Population-based studies are important for recording the
frequency distribution of different cytokine gene variants and for
defining differences among various population groups (7–13).
Such information is extremely helpful in designing translational
studies.

1.3. Genotype– Recent reports have established that some CGPs correlate with the
Phenotype levels of cytokine expression and their phenotypic balance helps to
Correlations determine and/or predict the type of immune response. Table 1
summarizes the established cytokine genotype–phenotype correla-
tions determined largely on the basis of their in vitro levels of
expression (14–24).

1.4. Relevance Some genetic polymorphisms in cytokine genes affect their tran-
of Cytokine Gene scription levels. Therefore, these polymorphisms may collectively
Polymorphisms modulate the net magnitude of cytokine-mediated immune
responses following transplantation or as a result of infection. This
underlies the importance of studying SNPs in relation to CGPs in
a variety of situations. Although CGPs play an important role in
552 G. Kaur and N. Mehra

Table 1
Established cytokine alleles/genotypes/haplotypes that
correlate with their in vitro expression levelsa

Cytokine Genotype Expression

IL1b +3962 (nt5887)T Increased

TGFb1 +896T, +915G Increased
TNFa -308A Increased
TNFa -863A Decreased
IFNg +874T Increased
IL2 -330T Increased
IL 6 -174G Increased
IL6b -597G,-572G,-373A(9)/T(11),-174G Increased
b
IL6 -597A,-572G,-373A(8)/T(12),-174G Decreased
IL10 -1082G,-819C,-590C Increased
IL10 -1082A,-819C,-592C Decreased
IL10 -1082A,-819T,-592A Decreased
IL13 -1055T Increased
Fas -670A Increased
Fas L -843C Increased
a
Important references (14–24)
b
In case of IL6, the A(9)/T(11) and A(8)/T(12) represent repeats of A and T nucle-
otides that constitute a part of the IL6 haplotypes associated with increased and
decreased expression, respectively

allograft survival and for evaluating incidence of acute or chronic

rejection, the relevance of ethnic differences in their frequencies
have not been fully explored yet. Observations based on racial vari-
ations in allograft loss, such as a relatively higher rate of renal
allograft failure among Blacks could be correlated with ethnic dif-
ferences in cytokine-mediated immune reactivity through an SNP
mapping approach. It has been shown that renal transplant recipi-
ents with genotypes predicting high TNFa, IL10, and gIFN are
more likely to develop acute chronic rejection (15). Ethnicities of
Asian descent as compared to Whites have been documented to
possess high frequencies of genotypes that result in low expression
of gIFN and IL 10 but high IL6 (12). Similarly, the role of CGPs
in infectious or autoimmune diseases has been widely studied and
found to correlate with Th1/Th2 dominated immune responses
(2, 3, 6, 25).
 31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 553

1.5. Methods of Several assays are available for defining CGPs or their effects on
Detecting Cytokine cytokine expression and these can be prioritized depending on the
Gene Polymorphisms study design and requirements. Their absolute or relative expres-
sion correlates can be estimated using ELISA, CELISA, ELISPOT,
real-time RT-PCR, Luminex monitoring, western blotting and
immunohistochemistry in conjunction with densitometry, flow
cytometry, and microarrays.
Similarly, several techniques are available for studying the bial-
lelic CGPs. These include PCR-SSP (or ARMS PCR), AFLP,
Taqman probe assays, microarrays, and sequencing-based typing.
Of these, the PCR-SSP is the most widely used and well-established
procedure that has been adopted by numerous studies, including
the 13IHWC as well as by the CTS, Heidelberg. Cytokine geno-
typing SSP kits can be procured on request from the CTS,
University of Heidelberg or from commercial vendors like One
lambda, USA. Primers can also be designed individually using soft-
ware like the Primer express and other online packages and in-house
SSP assays can be developed.
Figure 2 describes the basic principle of PCR-SSP procedure.
The PCR primers are designed deliberately to include a perfect
match with a single allele (i.e., a corresponding mismatch with the
other allele). As a result of stringent PCR conditions, allele-specific
amplification is initiated with only the perfectly matched primers
and continued exponentially during the subsequent PCR cycles.
The amplicons are resolved in agarose gels and interpreted. It
is pertinent to include an internal control of reference for
amplification in order to rule out amplification failures, if any. An
internal control is also needed to check amplification related errors
or variations, e.g., quality of template, carryover of salts and other

Fig. 2. Principle underlying the PCR-SSP test: (a) a perfectly matched primer leads to amplification while in (b) a mismatch
at the 3’end of the primer results in no amplification.
554 G. Kaur and N. Mehra

inhibitors, pippeting errors, etc., and more importantly to avoid

interpreting false negative results. The PCR primer mixes are pre-
pared for not only scoring alleles, but also for the identification of
genotypes and multilocus haplotypes involving multiple cytokines,
e.g., TGFb, TNFa, IL 2, IL 4, IL 6, and IL 10.
The PCR-SSP technique for cytokine genotyping using kits
procured from the CTS (Heidelberg) is described in detail in the
following sections. This kit enables detection of SNPs in the pro-
moter regions of IL1a, IL1b, IL1R, IL1RA, IL2, IL4, IL6, IL10,
IL12, TNFa, and gIFN genes, as well as in the untranslated regions
of TGFb and IL4Ra genes.

2. Materials

2.1. Extraction of DNA Salting out procedure is one of the most commonly employed
(Salting-Out Procedure methods of extraction of DNA and is described in this chapter
Using Ammonium although a number of commercial kits are also available for DNA
Acetate) extraction from peripheral blood and other tissues.
1. Source of DNA—(see Note 1).
2. Table top Centrifuge.
3. ND-1000 Spectrophotometer (NanoDrop).
4. Incubator/water bath.
5. Vortex mixer.
6. Microfuge.
7. Autopippettes and tips.
8. 1.5-mL polypropylene microfuge tubes.
9. pH meter.
10. Autoclave.
11. Speed-Vac (optional).
Reagents
1. Preparation of red cell lysis buffer (RCLB).
(a) Prepare 1 M ammonium chloride (NH4Cl) Stock Solution:
Dissolve 53.49 g of NH4Cl (M.W. 53.49) in 900 mL of
sterile molecular grade H2O (Milli Q) and make up the
final volume to 1 L.
(b) Prepare 1 M sodium bicarbonate (NaHCO3) Stock Solution:
Dissolve 84.01 g of NaHCO3 (M.W. 84.01) in 900 mL of
sterile H2O and bring the final volume to 1 L.
(c) Mix 144 mL of 1 M NH4Cl and 1 mL of 1 M NaHCO3
and 855 mL of MilliQ H2O to obtain a final volume of 1 L
and label as RCLB.
 31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 555

2. Nucleus lysis buffer (NLB).

(a) Prepare 5 M NaCl Stock solution: Dissolve 292.20 g of
NaCl (M.W. 58.44) in 700 mL of Milli Q grade water and
make the final volume to 1 L.
(b) Prepare 0.5 M Na2EDTA Solution (pH 8.0): Dissolve
186.12 g of Na2EDTA (M.W. 372.24) in 750 mL of MilliQ
H2O. Adjust pH to 8 by adding approximately 12 g NaOH
pellets and make final volume to 1 L.
(c) Prepare 2 M Tris–HCl (pH 8.0) Solution: Weigh and dis-
solve 242.28 g of Tris (M.W. 121.14) in 700 mL of Milli
Q grade H2O, adjust the pH to 8 with concentrated HCl,
and bring the final volume to 1 L.
(d) Mix 4 mL of 5 M NaCl + 200 mL of 0.5 M Na2EDTA (pH
8) + 250 mL of 2 M Tris–HCl (pH 8). Bring the final vol-
ume to 50 mL with Milli Q grade H2O and label as NLB.
Both RCLB and NLB should be autoclaved and stored at room
temperature.
3. Proteinase K solution.
Dissolve 10 mg proteinase K in 1 mL of sterile water, aliquot
into 200 mL fractions in microfuge tubes and store in –20°C.
4. TE solution (1 mM Tris 0.1 mM EDTA pH7.5).
Dissolve 121.16 mg Tris (MW 121.16) and 37.224 mg EDTA
(M.W. 372.24) in 950 mL dH2O, adjust the pH to 7.5 and
make the final volume to 1 L with MilliQ H2O and autoclave.
5. 4 M ammonium acetate.
Dissolve 154.16 g ammonium acetate (MW 77.08 g) in
500 mL final volume Milli Q water and autoclave.
6. 70% ethanol and isopropyl alcohol.

2.2. PCR-SSP 1. Thermal cycler (ABI 9700 or equivalent).

Using CTS Kits 2. Sterile molecular grade water.
3. Cytokine CTS-PCR SSP Tray kit (Product no.124) (www.
ctstransplant.org). The kit is supplied with 3 mL aliquots of
master mix (CYT 5%). It is recommended to prepare further
276 mL aliquots of this mix sufficient for each PCR tray and
store frozen at −20°C to avoid repeat thawing.
4. Taq DNA polymerase (5 U/mL) (Fermentas or Inno-Train or
Applied Biosystems). Store at −20°C.

2.3. Gel Electrophoresis 1. Apparatus for horizontal agarose gel electrophoresis.

2. Power pack for gel electrophoresis apparatus.
3. UV transilluminator (or gel documentation system).
4. Agarose high quality electrophoresis grade and with low elec-
troendosmosis (EEO) (see Note 2).
556 G. Kaur and N. Mehra

5. TBE buffer (0.5×):

(a) Prepare a 10× TBE stock by mixing 108 g Tris base + 55 g
boric acid and 40 mL 0.5 M EDTA (pH 8.0) and make up
the volume to 1 L with distilled water.
(b) Autoclave and store at room temperature.
(c) At the time of use, dilute the stock to 0.5× concentration.
6. Ethidium bromide 10 mg/mL (Caution: Ethidium bromide
is a potent carcinogen and must be handled carefully with
gloves).

3. Methods

3.1. Extraction of DNA 1. Approximately 5 mL of blood is collected by venipuncture in

K2-EDTA vacutainer tubes and mixed properly to avoid coag-
3.1.1. Collection
ulation. Heparin is not used as an anticoagulant since it has
of Peripheral Blood
been reported to have a potent inhibitory effect on PCR.

3.1.2. Lysis of RBC 1. Add 40 mL of RCLB to 5 mL of blood sample and mix

properly.
2. Incubate the contents at room temperature for 30 min and
centrifuge at 1,800 × g force for 20 min at room temperature.
3. Drain off the supernatant gently and resuspend the remaining
pellet in 40 mL RCLB.
4. Repeat step 2.
5. Drain the supernatant completely and proceed with nucleus
lysis steps as described below.

3.1.3. Nucleus Lysis 1. Resuspend the pellet obtained in Subheading 3.2, steps 1 and
and Protein Digestion 5 in 3 mL NLB (see Note 3).
2. Add 200 mL of 10% SDS followed by 20 mL (10 mg/mL) of
Proteinase-K solution. Mix thoroughly (see Note 4).
3. The mixture is kept at 37–42°C for 6 h to overnight. This is
the optimal temperature for maximum digestion. After
proper digestion, the solution becomes visibly clear with no
debris.

3.1.4. Desalting and A volume of 4 mL of 4 M ammonium acetate is added to the con-

Protein Precipitation tents obtained in Subheading 3.1.3, step 3 for deproteinization
and vortexed extensively. Add 3 mL of chloroform. The high
concentration of 4 M ammonium acetate dehydrates proteins and
 31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 557

separates them as a precipitate while chloroform helps in removal

of lipid contamination.
1. The mixture is vortexed and centrifuged at 1,800 × g force for
30 min. Protein settles in between the lower most layer of
chloroform and upper aqueous layer.
2. The aqueous layer containing DNA is transferred carefully into
another new 50-mL costar centrifuge tube.
3. Add 7 mL of isopropyl alcohol in order to precipitate DNA.
Swirl the tube gently until DNA begins to precipitate and
appears to float as jelly or whitish mass of thread-like
structures.
4. Carefully transfer the visible DNA thread like structure to a
1.5-mL sterile microfuge tube containing 70% ethanol to wash
off any salt trapped with DNA.
5. Microfuge the tube for 15 s, drain off the supernatant. Repeat
the 70% ethanol wash once more.
6. Add 100 mL TE solution to the DNA pellet and let it dis-
solve. DNA dissolves better in low concentration salt solution
(i.e., TE solution) than in distilled water. Do not use force to
dissolve DNA. It should dissolve spontaneously over a few
hours.

3.1.5. Quantification of DNA 1. Use 1 mL of DNA to measure optical density (OD) at a

wave length of 260l and 280l using an ND-1000 spectro-
photometer (NanoDrop). Pure DNA preparation gives
260l/280l ratio of around 1.8 (between 1.5 and 2.0) (see
Note 5). The OD value at 260l is used to calculate the
concentration of DNA. NOTE: 1 OD at 260l is equivalent
to 50 mg DNA.

3.1.6. Storage of DNA 1. DNA is usually stored at 4°C for daily use and at −80°C for
several months to years.

3.2. PCR-SSP Using Each 96-well tray supplied by the CTS for cytokine genotyping
CTS Tray Kit contains prepipetted lyophilized primer mixes for two DNA , i.e.,
48 wells per individual. The PCR mix positions corresponding to
every locus specificities in the tray are shown in Fig. 3 and Table 2.
A separate master mix (CYT) is also supplied with the kit and con-
tains the reaction buffer but no Taq polymerase. The reader is
advised to refer to the procedure manual that is supplied along
with the kit and take a note of lot specific differences if any.
1. Program the thermal cycler (ABI 9700) as per following
conditions and save the program (see Note 6).
558 G. Kaur and N. Mehra

Fig. 3. An example of a 2% agarose gel electrophoregram showing PCR-SSP genotyping for cytokine and receptor genes
using the CTS trays obtained from University of Heidelberg, Germany.
Table 2
List of various CGPs, their gene location, allele specificities, and amplicon sizes obtained using CTS PCR-SSP kits

Corresponding Size of Size of the

SNP-dba Position Position Nucleotide allele/genotype/ allele-specific amplification
Cytokine Location name on tray on tray Mix No. specificity haplotype amplicon (bp) control
IL 1a Promoter rs 1800587 H1 H7 1 T at position -889 T 220 440
G1 G7 2 C at position -889 C 220 440
IL 1b Promoter rs 16944 F1 F7 1 C at position -511 C 215 440
E1 E7 2 T at position -511 T 215 440
rs 1143634 D1 D7 3 T at position +3962 T 340 440
C1 D1 4 C at position +3962 C 340 440
IL 1R Promoter rs 2234650 B1 B7 1 C at position C 290 440
pstI 1970
A1 A7 2 T at position T 290 440
pstI 1970
IL 1RA Promoter rs 315952 H2 H8 1 T at position mspa T 330 440
11100
G2 G8 2 C at position mspa C 330 440
11100
IL 4Ra 1st Exon rs 1801275 F2 F8 1 G at position +1902 G 140 440
E2 E8 2 A at position +1902 A 140 440
IL 12B Promoter rs 3212227 D2 D8 1 C at position -1188 C 800 440
C2 C8 2 A at position -1188 A 800 440
gIFN Intron 1 rs 2430561 B2 B8 1 A at position +874 A 180 440
A2 A8 2 T at position +874 T 180 440
(continued)
Table 2
(continued)

Corresponding Size of Size of the

SNP-dba Position Position Nucleotide allele/genotype/ allele-specific amplification
Cytokine Location name on tray on tray Mix No. specificity haplotype amplicon (bp) control
TGFb 1st Exon rs 1800470 H3 H9 1 C at codon 10 CG or CC 195 440
G3 G9 2 T at codon 10 TG or TC 195 440
rs 1800471 F3 F9 3 C at codon 10; CG 80 440
G at codon 25
E3 E9 4 C at codon 10; CC 80 440
C at codon 25
D3 D9 5 T at codon 10; TG 80 440
G at codon 25
C3 C9 6 T at codon 10; TC 80 440
C at codon 25
TNFa Promoter rs 1800629 B3 B9 1 G at position -308; GG 110 440
G at position -238
rs 361525 A3 A9 2 A at position -308; AG 110 440
G at position -238
H4 H9 3 G at position -308; GA 110 440
A at position -238
G4 G9 4 A at position -308; AA 110 440
A at position -238
IL 2 Promoter rs 2069762 F4 F10 1 T at position -330; TG 560 90
G at position +166
rs 2069763 E4 E10 2 G at position -330; GG 560 90
G at position +166
 D4 D10 3 G at position -330; GT 560 90
T at position +166
C4 C10 4 T at position -330; TT 560 90
T at position +166
IL 4b Promoter rs 2243248 B4 B10 1 T at position -1098; TT* 560 90
T at position -590
rs 2243250 A4 A10 2 T at position -1098; TC* 560 90
C at position -590
H5 H11 3 G at position -1098; GT* 560 90
T at position -590
G5 G11 4 G at position -1098; GC* 560 90
C at position -590
rs 2070874 F5 F11 5 T at position -590; *TT 610 90
T at position -33
E5 E11 6 T at position -590; *TC 610 90
C at position -33
D5 D11 7 C at position -590; *CT 610 90
T at position -33
C5 C11 8 C at position -590; *CC 610 90
C at position -33
IL 6 Promoter rs 1800795 B5 B11 1 G at position -174; GG 430 90
G at position +565
rs 1800797 A5 A11 2 C at position -174; CG 430 90
G at position +565
H6 H12 3 G at position -174; GA 430 90
A at position +565
G6 G12 4 C at position -174; CA 430 90
A at position +565
(continued)
 Table 2
(continued)

Corresponding Size of Size of the

SNP-dba Position Position Nucleotide allele/genotype/ allele-specific amplification
Cytokine Location name on tray on tray Mix No. specificity haplotype amplicon (bp) control
IL 10b Promoter rs 1800896 F6 F12 1 G at position -1082; GC* (GCC or 305 90
C at position -819 GCA)
rs 1800971 E6 E12 2 G at position -1082; G*C (GCC or 530 90
C at position -592 GAC)
rs 1800872 D6 D12 3 A at position -1082; AC* (ACC or 305 90
C at position -819 ATC)
C6 C12 4 A at position -1082; AT* (ATC or 305 90
T at position -819 ATA)
B6 B12 5 A at position -1082; A*C (ACC or 530 90
C at position -592 ATC)
A6 A12 6 A at position -1082; A*A (ACA or 530 90
A at position -592 ATA)
a
The reader is advised to refer to Fig. 1 for specific SNP-db names mentioned in this table
b
In case of IL4 and IL10, it is possible to deduce 3 locus haplotypes where * in the haplotype means any nucleotide at one of the three positions. In case of IL4, these haplotypes refer
to -1098,-590, and -33 positions. Similarly, the haplotypes in IL10 encompass the positions -1082,-819, and -592
31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 563

Initial denaturation 94°C 2 min

Denaturation 94°C 15 s
10 cycles
Annealing + extension 65°C 1 min
Denaturation 94°C 15 s
20 cycles
Annealing 61°C 50 s
Extension 72°C 30 s
Hold 4°C 15 min

2. Prepare the PCR reaction mix for a single DNA sample as

shown below:

Component Volume (mL)

Master mix CYT 138

Taq Pol (5 U/mL) 3
Water 329
DNA (40 ng/mL) 50
Final volume 520

3. Each CTS tray contains primers for genotyping two DNA sam-
ples. Use scissors to cut the tray into two halves. The PCR
reaction mix can be prepared in volumes depending on the
number of samples to be tested (see Notes 7–9).
4. Remove the CTS tray from the 4°C refrigerator. The lower
edge of the tray is marked with a black line. The starting posi-
tion in the tray is H1 and the series is oriented from H1 (bot-
tom) followed by G1, F1, etc., to A1 (top) until the last well
A6. The second set of the kit begins from H7 going on to A7
and so on until A12. See Table 2 for all the corresponding
positions on the tray.
5. Transfer 10 mL of reaction mix from Subheading 3.2, step 2
into each of the 48 wells (see Note 10).
6. Seal the PCR tray with Strip caps provided with the kit and
place in a thermal cycler that preprogrammed as in step 1
above. Ensure the caps completely seal the wells to prevent
evaporation.
7. Start thermal cycling, after completion, remove the tray and
proceed with gel electrophoresis.
8. A 2% agarose gel can be prepared and kept ready while the
PCR is running. To avoid contamination with PCR amplification
products, it is necessary to perform the test in two separate
working areas: the preamplification and the postamplification
areas.
564 G. Kaur and N. Mehra

3.3. Gel Electrophoresis 1. While the PCR is running (Subheading 3.2, step 7), a 2% agarose
gel can be poured and kept ready. For this, boil 25 agarose in
0.5× TBE, cool to ~60°C and add 5 mL ethidium bromide
(0.5 mg/mL final concentration). Mix and pour into gel tray
prearranged with combs.
2. Allow the gel to set at room temperature for 1 h.
3. After the PCR is over, remove the strip caps carefully from the
tray.
4. Remove the combs and ensure that wells are not broken. Place
the gel inside and just under the surface of electrophoresis run-
ning buffer (0.5× TBE). Load 10 mL amplicons starting from
H1 towards A1 as described in Fig. 3.
5. Connect the electrodes in the power-pack and run the electro-
phoresis at 100 V until just before the tracking dye exits the
gel.
6. Visualize the gel under UV illumination (312 nm) and save the
image or take a picture for interpretation and documentation.

3.4. Interpretation Each of the primer mixes contain internal control primer pairs
of Results corresponding to either b-globin (90 bp) or CRP gene (440 bp).
Note the presence of these internal control bands in all the lanes
3.4.1. Positive Internal
as per their expected correct band sizes (Table 2 and Fig. 3). The
Control Amplification
440 bp band should be visible in lanes corresponding to IL1a,
IL1b, IL1R, IL1RA, IL4Ra, IL12, gIFN, TGFb, and TNFa.
Similarly, the 90 bp band should be observed in lanes correspond-
ing to IL 2, IL 4, IL 6, and IL 10.

3.4.2. Allele-Specific 1. Record the amplification patterns obtained for each of the
Amplification primer pairs in every well after confirmation of the correct
band sizes as shown in Table 2. The wells where allele-specific
amplicon is present may show weak or sometimes no internal
control band (see Note 11).
2. Some of the cytokine genotyping reagents in the kit allow
detection of certain haplotypes depending on characteristic
amplification patterns. Refer Table 2 for deducing correct
haplotypes.
3. Disregard any products of incorrect size whether strong or
weak (see Note 12).
4. A primer dimer diffused band of <80 bp may be observed
occasionally but can be distinguished from the allele-specific
amplifications which are relatively sharper and usually 80 bp or
more in size.
5. False negatives, if any, may be caused by inefficient amplification
or poor quality of DNA or variations in thermal cycler and can
be ruled out individually.
 31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 565

6. In case of gIFN, mix 1 may yield relatively weaker amplification

than mix 2.
7. For TGFb, a nonspecific smear may be observed but should be
discernible from the small sized (80 bp) amplicons.
8. When IL4 amplifications are being interpreted, note that mix
6 may produce faint nonspecific amplification. An allele-specific
PCR product is usually strongly positive.
9. The test should be repeated if there is no amplification of con-
trol bands as well as allele-specific bands or if reaction pattern
is inconclusive.

3.4.3. An Example The first observation to look for is an internal control amplification
of Interpretation of 90 bp (Fig. 3, lanes 43–48). Now, refer to Table 2 for expected
of Allele-Specific band sizes for allele-specific amplifications relating to IL10 mixes
Amplification 1–6 corresponding to the same lanes 43–48. As can be seen in
for IL 10 Fig. 3, there is a positive band of 305 bp (mix 1) in lane number
43 and a band of size 530 bp (mix 2) in lane 44 while all the
remaining lanes 45–48 do not show any allele-specific
amplification(s) for IL 10. A combined interpretation of GC*
and G*C as suggested against lanes 43 and 44 (Table 2) hence
suggests that the individual tested is homozygous for the GCC
haplotype.

4. Notes

1. EDTA blood is the most widely used material to isolate genomic

DNA. A great care needs to be taken to avoid contamination
of DNA from another source because PCR technique is a pow-
erful method that would amplify contaminated DNA up to the
detection level. White blood cells (WBCs) are used as the major
source of DNA and can be easily obtained by removing the red
blood cells (RBCs) after lysis from the whole blood.
2. It is advisable to use high quality electrophoresis grade agarose
with low EEO (preferably < 0.13). EEO refers to the flow of
liquid under the influence of an electric field due to immobi-
lized charge groups on the matrix. The anionic groups of aga-
rose are affixed to the matrix; however, counter cations tend to
migrate towards cathode within the matrix, giving rise to EEO.
This movement of water counteracts the otherwise electro-
phoretic movement of DNA towards anode, hence leads to
convection and smearing of bands.
3. The NLB contains Tris, NaCl, and EDTA. Tris and NaCl are
used for buffering the reaction (pH-8). Intrinsic DNAse activity
is inhibited by EDTA as it chelates the divalent ions (Mg++,
Ca++, Zn++) into their respective salts as precipitates. These divalent
566 G. Kaur and N. Mehra

ions are cofactors for the DNAse enzyme and by removing

them the intrinsic DNAse activity is inhibited and DNA frag-
mentation avoided. SDS is an ionic detergent which disturbs
membranal structure, denatures proteins and can digest the
DNAse enzyme.
4. Proteinase-K is a proteolytic enzyme that breaks the peptide
bonds and helps to get protein-free DNA.
5. A ratio of 280/260 between 1.5 and 2.0 is acceptable. In case,
the DNA isolated does not give this ratio, it is advisable to
reextract DNA from a fresh sample.
6. Ensure that thermal cycler is calibrated and verified.
7. Use of excess DNA quantity (>200 ng/mL) may result in
nonspecific amplifications, false positives, and smearing in
the gel.
8. Use of less DNA (<10 ng/mL) may result in weak amplification
bands or false negatives.
9. A poor or no amplification indicates poor quality or degraded
DNA. It is advisable to take fresh blood sample and reextract
the DNA.
10. When dispensing 10 mL of reaction mix to each well, care
should be taken to release the drop onto well side walls near the
top and allow it to slide down (you may gently tap the tray).
Do not touch the bottom of the well. Make sure that each of
the well receives the drop (look for pink color) before proceed-
ing to thermal cycler. Ensure that the caps are sealed completely
on top of the wells to avoid evaporation. The time required
between addition of samples to wells and initiation of PCR
must be minimized and performed as efficiently as possible.
11. In the presence of a positive typing band (specific amplification
of a cytokine allele), the product of the internal control primer
may be weak or absent due to the difference in concentration
and melting temperatures between the specific primer pairs
and the internal control primer pair.
12. In a few instances, e.g., if DNA quality is not very good or if
there was a delay in starting the thermal cycler etc., non-specific
amplifications may occur. Hence, any band (whether strong or
weak) that does not correspond to its expected correct size
should be ignored.

Acknowledgements

The authors acknowledge Joannis Mytilineos and his team for

developing the Cytokine genotyping reagents and the CTS group
at Univ Heidelberg for kindly providing these on regular basis.
 31 Cytokine Gene Polymorphisms: Methods of Detection and Biological Significance 567

The financial support provided by the Indian Council of Medical

Research (ICMR) and Department of Biotechnology (DBT),
Govt. of India are gratefully acknowledged. The laboratory mem-
bers, particularly Neeraj Kumar, Chowphi Rapthap, Gaurav
Sharma, and Shekhar Neolia are acknowledged for optimizing the
above procedures.

References
1. Hollegaard MV, Bidwell JL (2006) Cytokine Greek Cypriot population. Eur J Immmunogenet
gene polymorphism in human disease: on-line 30:341–343
databases, supplement 3. Genes Immun 13. Hoffmann SC, Stanley EM, Cox ED et al
7:269–276 (2002) Ethnicity greatly influences cytokine
2. Kaur G, Mehra N (2009) Genetic determinants gene polymorphism distribution. Am J
of HIV-1 infection and progression to AIDS: Transplant 2(6):560–567
immune response genes. Tissue Antigens 14. Dudler J, Li J, Pagnotta M et al (2006) Gene
74:373–385 transfer of programmed death ligand-1.Ig pro-
3. Ryan AW, Mapp J, Myna S et al (2006) Levels longs cardiac allograft survival. Transplantation
of interpopulation differentiation among dif- 82(12):1733–1737
ferent functional classes of immunologically 15. Hahn AB, Kasten-Jolly JC, Conctantino DM
important genes. Genes and Immunity et al (2001) TNF-alpha, IL-6, IFN-gamma,
7:179–183 and IL-10 gene expression polymorphisms and
4. Mytilineos J, Laux G, Opelz G (2004) Relevance the IL-4 receptor alpha-chain variant Q576R:
of IL10, TGFb1, TNFa and IL4Ra gene poly- effects on renal allograft outcome.
morphisms in kidney transplantation: a collab- Transplantation 72(4):660–665
orative transplant study report. Am J Transplant 16. Pociot F, Mølvig J, Wogensen L et al (1992) A
4:1684–1690 TaqI polymorphism in the human interleukin-1
5. Bidwell J, Keen L, Gallagher G et al (1999) beta (IL-1 beta) gene correlates with IL-1 beta
Cytokine gene polymorphism in human dis- secretion in vitro. Eur J Clin Invest 22(6):
ease: on-line databases. Genes Immun 1:3–19 396–402
6. Bidwell J, Keen L, Gallagher G et al (2001) 17. Awad MR, El-Gamel A, Hasleton P et al (1998)
Cytokine gene polymorphism in human dis- Genotypic variation in the transforming growth
ease: on-line databases, supplement 1. Genes factor-beta1 gene: association with transform-
Immun 2:61–70 ing growth factor-beta1 production, fibrotic
7. Kaur G, Rapthap CC, Kumar N et al (2007) lung disease, and graft fibrosis after lung trans-
Frequency distribution of cytokine gene poly- plantation. Transplantation 66(8):1014–1020
morphisms in the healthy North Indian popula- 18. Pociot F, Wilson AG, Nerup J et al (1993) No
tion. Tissue Antigens 69:113–120 independent association between a tumor
8. Pyo CW, Hur SS, Kim YK et al (2003) necrosis factor-alpha promotor region poly-
Polymorphisms of IL1B, IL1RN, IL2, IL4, morphism and insulin-dependent diabetes mel-
IL6, IL10 and IFNg genes in the Korean popu- litus. Eur J Immunol 23(11):3050–3053
lation. Hum Immunol 64:979–989 19. Fishman D, Faulds G, Jeffery R (1998) The
9. Kurzawski M, Pawlik A, Czerny B et al (2005) effect of novel polymorphisms in the interleu-
Frequencies of the common promoter poly- kin-6 (IL-6) gene on IL-6 transcription and
morphisms in cytokine genes in a polish popu- plasma IL-6 levels, and an association with
lation. Int J Immunogenet 32:285–291 systemic-onset juvenile chronic arthritis. J Clin
10. Uboldi de Capei MU, Dametto E, Fasano ME Invest 102(7):1369–1376
et al (2003) Genotyping for cytokine 20. Terry CF, Loukaci V, Green FR (2000)
polymorphisms: allele frequencies in the Italian Cooperative influence of genetic polymor-
population. Eur J Immunogenet 30:5–10 phisms on interleukin 6 transcriptional regula-
11. Trajkov D, Arsov T, Petlichkovski A et al (2005) tion. J Biol Chem 275(24):18138–18144
Cytokine gene polymorphisms in population of 21. Turner DM, Williams DM, Sankaran D et al
ethnic Macedonians. Croat Med J 46(4):685–692 (1997) An investigation of polymorphism in
12. Costeas PA, Koumas L, Koumouli A et al (2003) the interleukin-10 gene promoter. Eur J
Cytokine polymorphism frequencies in the Immunogenet 24(1):1–8
568 G. Kaur and N. Mehra

22. Crawley E, Woo P, Isenberg DA et al (1999) 25. Wang K, Li M, Hakonarson H (2010) Analysing
Single nucleotide polymorphic haplotypes of biological pathways in genome-wide associa-
the interleukin-10 5¢ flanking region are not tion studies. Nat Rev Genet 11:843–854
associated with renal disease or serology in
Caucasian patients with systemic lupus erythe-
matosus. Arthritis Rheum 42(9):2017–2018
Useful Website Links
23. van der Pouw Kraan TC, van Veen A, Boeije
LC et al (1999) An IL-13 promoter polymor- h t t p : // w w w 3 . n c b i . n l m . n i h . g o v / O m i m /
phism associated with increased risk of allergic searchomim.html
asthma. Genes Immun 1(1):61–65
24. Mullighan CG, Marshall SE, Bunce M et al http://www.immport.org/immportWeb/que-
(1999) Variation in immunoregulatory genes ryref/geneListSummary.do
determines the clinical phenotype of common http://snpper.chip.org/
variable immunodeficiency. Genes Immun http://genome.cse.ucsc.edu/
1:137–148 http://cmbi.bjmu.edu.cn/cmbidata/cgf/
 Chapter 32

IMGT® Tools for the Nucleotide Analysis of Immunoglobulin

(IG) and T Cell Receptor (TR) V-(D)-J Repertoires,
Polymorphisms, and IG Mutations: IMGT/V-QUEST
and IMGT/HighV-QUEST for NGS
Eltaf Alamyar, Patrice Duroux, Marie-Paule Lefranc,
and Véronique Giudicelli

Abstract
IMGT/V-QUEST is the highly customized and integrated online IMGT® tool for the standardized analysis
of the immunoglobulin (IG) or antibody and T cell receptor (TR) rearranged nucleotide sequences. The
analysis of these antigen receptors represents a crucial challenge for the study of the adaptive immune response
in normal and disease-related situations. The expressed IG and TR repertoires represent a potential of 1012
IG and 1012 TR per individual. This huge diversity results from mechanisms that occur at the DNA level dur-
ing the IG and TR molecular synthesis. These mechanisms include the combinatorial rearrangements of the
variable (V), diversity (D) and joining (J) genes, the N-diversity (deletion and addition at random of nucle-
otides during the V-(D)-J rearrangement) and, for IG, somatic hypermutations. IMGT/V-QUEST identifies
the V, D, J genes and alleles by alignment with the germline IG and TR gene and allele sequences of the
IMGT reference directory. The tool describes the V-REGION mutations and identifies the hot spot posi-
tions in the closest germline V gene. IMGT/V-QUEST integrates IMGT/JunctionAnalysis for a detailed
analysis of the V-J and V-D-J junctions and IMGT/Automat for a complete annotation of the sequences
and also provides IMGT Collier de Perles. IMGT/HighV-QUEST, the high-throughput version of IMGT/
V-QUEST, implemented to answer the needs of deep sequencing data analysis from Next Generation
Sequencing (NGS), allows the analysis of thousands of IG and TR sequences in a single run. IMGT/V-
QUEST and IMGT/HighV-QUEST are available at the IMGT® Home page, http://www.imgt.org.

Key words: IMGT, Immunogenetics, Immunoinformatics, Immunoglobulin, Antibody, T cell receptor,

Somatic hypermutation, Allelic polymorphism, High-throughput sequencing, IMGT-ONTOLOGY

1. Introduction

The efficiency of the adaptive immune response and its capability

of recognizing a large number of different antigens depend on
the huge diversity of the antigen receptors, antibodies or

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_32, © Springer Science+Business Media New York 2012

569
570 E. Alamyar et al.

immunoglobulins (IG) of the B lymphocytes, and T cell receptors

(TR) of the T lymphocytes. The genes which code the IG and TR
are highly polymorphic and are organized in clusters in several loci
(three loci for IG and four for TR in human) located on different
chromosomes (four in human) in the genome (1, 2). The molecular
synthesis of the IG and TR chains is particularly complex and
unique. It includes several mechanisms that occur at the DNA
level: combinatorial rearrangements of the variable (V), diversity (D)
and joining (J) genes that code the variable domain or V-DOMAIN,
exonuclease trimming at the ends of the V, D, and J genes and the
random addition of nucleotides by the terminal deoxynucleotidyl
transferase (TdT) that create the junction N-diversity, and somatic
hypermutations (for IG only) (1, 2). The IG and TR repertoires
show an extraordinary diversity with a potential of 1012 IG and 1012
TR per individual and the only limiting factor is the number of
genetically programmed B and T cells for an organism. Therefore,
the analysis of the IG and TR genes and of their expression repre-
sents a crucial challenge for the understanding of the immune
response in normal and pathological situations.
IMGT/V-QUEST (V-QUEry and STandardization) (3–7) is
the IG and TR nucleotide sequence analysis tool of IMGT®, the
international ImMunoGeneTics information system® (8) (http://
www.imgt.org), created in 1989 at Montpellier, France (CNRS
and Université Montpellier 2). IMGT/V-QUEST (Subheading 2)
identifies the V, D, and J genes in rearranged IG and TR sequences
and, for the IG, the nucleotide (nt) mutations and amino acid (AA)
changes resulting from somatic hypermutations. The tool inte-
grates IMGT/JunctionAnalysis (9), for the detailed characteriza-
tion of the V-D-J or V-J junctions. The standardized analysis is
based on the concepts of IMGT-ONTOLOGY, the first ontology
in immunogenetics and immunoinformatics, which are generated
from the axioms of the Formal IMGT-ONTOLOGY or IMGT-
Kaleidoscope (10–13), and more particularly the axioms of
CLASSIFICATION (gene and allele nomenclature),
DESCRIPTION (standardized labels), and NUMEROTATION
(IMGT unique numbering and IMGT Colliers de Perles). That
standardization has allowed a considerable upgrade, from IMGT/
V-QUEST online analyzing 50 sequences per run, to IMGT/
HighV-QUEST (14), the high-throughput version of IMGT/V-
QUEST, analyzing up to 150,000 sequences per run, with the
same degree of resolution and high quality results. IMGT/HighV-
QUEST (Subheading 3) was implemented to answer the needs of
data analysis from the Next Generation Sequencing (NGS). The
standardized results provided by IMGT/V-QUEST and IMGT/
HighV-QUEST are crucial for the analysis of repertoire of IG anti-
body sites, and TR recognition sites and IG and TR polymor-
phisms, in normal and pathological situations. IMGT/V-QUEST
and IMGT/HighV-QUEST are constantly improved following
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 571

the results of fundamental and clinical research. New alleles are

annotated in IMGT® and integrated in the IMGT/V-QUEST and
IMGT/HighV-QUEST reference directories as soon as they are
confirmed experimentally and publicly available in the generalist
databases and entered in IMGT/GENE-DB (15). Rules for
sequence analysis have been improved through scientific collabora-
tion and with the constant feedback of clinicians and scientists. By
their wide range of customizable and integrated functionalities and
by their high level of standardization, IMGT/V-QUEST and
IMGT/HighV-QUEST provide results of much value for the
comparative analysis of the IG and TR sequences (including those
of the newly sequenced vertebrate genomes), for statistical analysis
of the junctions and of the repertoires, for antibody engineering
and therapeutic antibodies.

2. IMGT/V-QUEST

The link to the IMGT/V-QUEST tool is available from the IMGT®

Home page, http://www.imgt.org in the “IMGT tools” section.
Clicking on that link gives access to the IMGT/V-QUEST
Welcome page (Fig. 1).

2.1. IMGT/V-QUEST 1. In the IMGT/V-QUEST Welcome page (Fig. 1), locate the
Welcome Page section that corresponds to the receptor type (IG or TR) of
your sequences (see Note 1).
2. Locate the species or taxon that corresponds to the IMGT ref-
erence directory against which you want your sequences to be
analyzed (see Note 2).
3. Click on the species (for instance “Human”) or taxon. This
will give access to the IMGT/V-QUEST Search page (Fig. 2).

2.2. IMGT/V-QUEST IMGT/V-QUEST can analyze IG and TR rearranged nucleotide

Search Page sequences from genomic DNA or cDNA, indifferently. Sequences
must be submitted in FASTA format (up to 50 sequences per run).
2.2.1. Entering Nucleotide
The sequences must be copied and pasted in the text area of
Sequences for the Analysis
“Nucleotide sequences,” or uploaded as a text file by selecting the
option “Or give the path access to a local file containing your
sequence(s) in FASTA format” and using the “Browse” button to
select the file.

2.2.2. Selection The user can choose between three types of display for the results:
of the Result Display “Detailed view,” “Synthesis view,” or “Excel file” and the para-
meters of the analysis can be customized with a large choice of
options in “Advanced parameters” (Fig. 2).
572 E. Alamyar et al.

Fig. 1. IMGT/V-QUEST Welcome page. The two upper sections correspond to the immunoglobulin (IG) or antibody and T cell
receptor (TR) analyses, respectively. The lowest section is for download of the IMGT reference directory sets.

1. Detailed view.
“Detailed view” provides the results for each sequence,
individually. They may be displayed in HTML (by default) or
Text format, with a user-defined number of nucleotides per
line in alignments (60 by default). There is a choice of 14 dif-
ferent result displays for the analysis of the V-DOMAIN. The
results the most commonly requested are selected by default.
You can check or uncheck the result displays as needed:
(a) Alignment for V-GENE.
(b) Alignment for D-GENE.
(c) Alignment for J-GENE.
(d) “Results of IMGT/JunctionAnalysis” (9) (with or without
full list of eligible D-GENEs).
(e) Sequence of the JUNCTION (“nt” and “AA”).
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 573

Fig. 2. The IMGT/V-QUEST Search page. The Search page is that for IG: “Analyse your immunoglobulin (IG) or antibody
sequences.” Your species (or taxon) selection from the IMGT/V-QUEST Welcome page is recalled at the top of the page
(here “Human”).
574 E. Alamyar et al.

(f) V-REGION alignment.

(g) V-REGION translation.
(h) V-REGION protein display.
(i) V-REGION mutation and AA change table.
(j) V-REGION mutation and AA change statistics.
(k) V-REGION mutation hot spots.
(l) Sequences of V-, V-J-, or V-D-J-REGION (“nt” and
“AA”) with gaps in FASTA and access to IMGT/
PhyloGene for V-REGION (“nt”).
(m) Annotation by IMGT/Automat.
(n) IMGT Colliers de Perles:
– Link to IMGT/Collier-de-Perles tool.
– IMGT Colliers de Perles (for a nb of sequences <5).
– No IMGT Collier de Perles.
2. IMGT/V-QUEST “Synthesis view”
The “Synthesis view” consists in an interesting alternative
to the “Detailed view” if several sequences of the same run use
the same V gene and allele. Indeed, sequences expressing the
same V gene and allele will be aligned together. The results
may be displayed in HTML (by default) or Text format, with a
user-defined number of nucleotides per line in alignments (60
by default). There are seven different result displays for the
alignments of sequences that express the same V gene and
allele and the results of IMGT/JunctionAnalysis (9):
(a) Alignment for V-GENE.
(b) V-REGION alignment.
(c) V-REGION translation.
(d) V-REGION protein display.
(e) V-REGION protein display (with AA class codons).
(f) V-REGION protein display (only AA changes displayed).
(g) V-REGION most frequently occurring AA.
(h) Results of IMGT/JunctionAnalysis.
Three of these options (e), (f), and (g) are specific to
IMGT/V-QUEST “Synthesis View” (see Note 3).
3. Excel file.
“Excel file” allows the users to get the results of the analysis
in a spreadsheet with the choice of eleven sheets. The “Summary”
and “Parameters” sheets are always provided.
(a) Summary.
(b) IMGT-gapped-nt-sequences.
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 575

(c) nt-sequences.
(d) IMGT-gapped-AA-sequences.
(e) AA-sequences.
(f) Junction.
(g) V-REGION-mutation-and-AA-change-table.
(h) V-REGION-nt-mutation-statistics.
(i) V-REGION-AA-change-statistics.
(j) V-REGION-mutation-hot-spots.
(k) Parameters.

2.2.3. Advanced Default values of “Advanced parameters” have been set for classical
Parameters analysis. They should be modified for more sophisticated queries
or for unusual sequences (e.g., with insertions or deletions). The
options include:
1. “Selection of IMGT reference directory set” with a choice of
four sets:
(a) F + ORF.
(b) F + ORF + in frame P (by default).
(c) F + ORF including orphons.
(d) F + ORF + in frame P including orphons.
Where F is functional, ORF is open reading frame, P is
pseudogene. This allows sequences to be compared with only
relevant gene sequences (e.g., orphon sequences are relevant
for genomic but not for expressed repertoire studies). The
selected set can also be chosen either “With all alleles” or
“With allele *01 only” (see Note 4).
2. “Search for insertions and deletions in V-REGION” (“Yes” or
“No”): somatic hypermutations by nucleotide insertions and
deletions in the V-REGION are rare events that are known to
occur in normal and malignant cells (16). By default, IMGT/
V-QUEST does not search for insertions and deletions. If “Yes”
is selected, the analysis is slower and the number of submitted
sequences in a single run is limited to ten (see Note 5).
3. Parameters for IMGT/JunctionAnalysis:
(a) “Nb of accepted D-GENE”: this selection is provided for
the IGH junctions (for the IG) and for the TRB and TRD
junctions (for the TR).
(b) “Nb of accepted mutations” in 3¢V-REGION,
D-REGION, and 5¢J-REGION (by default, no mutation
is accepted for the TR junctions. For IGH, 2, 4, and 2
mutations are accepted in the 3¢V-REGION, D-REGION,
and 5¢J-REGION, respectively. For the IGK and IGL loci,
576 E. Alamyar et al.

the maximum number of accepted mutations is 7 in the

3¢V-REGION and 7 in the 5¢J-REGION).
4. Parameters for “Detailed view”:
(a) “Nb of nucleotides to exclude in 5¢ of the V-REGION for
the evaluation of the nb of mutations”: in case of primer
specific nucleotides.
(b) “Nb of nucleotides to add (or exclude) in 3¢ of the
V-REGION for the evaluation of the alignment score”: in
case of low (or high) exonuclease activity.

2.3. IMGT/V-QUEST The top of “Detailed view” results are displayed in “A. Detailed
Output for Detailed results for the IMGT/V-QUEST analyzed sequences.” The top of
View this page indicates the number of analyzed sequences with links to
individual results. Use the link associated with a sequence name to
go directly to its individual result that comprises the sequence and
“Result summary” followed by the different displayed results as
selected in the Search page.

2.3.1. Sequence The sequence and “Result summary” are shown at the top of each
and Result Summary individual result (Fig. 3). The IMGT reference directory set against
which the sequence was analyzed (e.g., human IG set) is indicated

Fig. 3. IMGT/V-QUEST “Detailed view” result page: Sequence and “Summary.” The upper part of an individual result shows
the sequence and “Result summary.” The seq1 is the human IGH rearranged sequence with the accession number
AB012909 in the IMGT/LIGM-DB database (34).
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 577

(Fig. 3). A sequence submitted in antisense orientation will be

shown as complementary reverse sequence, which is in V gene
sense orientation. The “Result summary” provides a crucial feature
that is the evaluation of the user sequence functionality performed
automatically by IMGT/V-QUEST: productive (if no stop codon
and in-frame junction) or unproductive (if stop codons and/or
out-of-frame junction) (see Note 6). It also summarizes the main
characteristics of the analyzed sequence which include (1) the
names of the closest “V-GENE and allele” (e.g., IGHV3-30*04)
and “J-GENE and allele” (e.g., IGHJ4*02) (see Note 7) with the
alignment score (see Note 8), the percentage of identity and the
ratio of the number of identical nucleotides (nt)/number of aligned
nt, (2) the name of the closest “D-GENE and allele” (IGHD5-
5*01) determined by IMGT/JunctionAnalysis (9) with the
D-REGION reading frame, (3) the FR-IMGT lengths (e.g.,
[25.17.38.11]), the CDR-IMGT lengths (e.g., [8.8.11]), and the
amino acid (AA) JUNCTION sequence which characterizes a V
domain. The information shown in Fig. 3 (online in orange)
“Productive IGH rearranged sequence” and “93.06%” is used by
clinicians analyzing the level of mutations in chronic lymphocytic
leukemia (CLL) IGHV rearranged sequences as a prognostic crite-
rion (17) (see Note 9). IMGT/V-QUEST provides warnings (not
shown) that appear as notes in red to alert the user, if potential
insertions or deletions are suspected in the V-REGION (see
Note 10), or if other possibilities for the J-GENE and allele names
are identified (see Note 11).

2.3.2. Sequence If insertions and/or deletions are suspected, as mentioned above

and Result Summary with (Subheading 2.3.1), the user can go back to the IMGT/V-QUEST
the Option “Search for Search page and select the option “Search for insertions and dele-
Insertions and Deletions” tions” in “Advanced parameters.” If indeed insertions and/or
deletions are detected, they will be described in the “Result
summary” row (Fig. 4) with their localization in FR-IMGT or
CDR-IMGT, the number of inserted or deleted nucleotides and,
for insertions, the inserted nucleotides, the presence or absence of
frameshift, the V-REGION codon from which the insertion or
deletion starts and the nt position in the user sequence. The inser-
tions are highlighted in capital letters in the user sequence (Fig. 4)
and the tool runs a classical IMGT/V-QUEST search after having
removed the insertion(s) from the user sequence. In case of dele-
tions, the tool adds gaps to replace the identified deletions before
running a classical IMGT/V-QUEST search. Users should be
aware that an insertion or a deletion at the beginning of FR1-
IMGT or at the end of the FR3-IMGT may not be detected.

2.3.3. Alignments for V, D, The alignments for V, D, J genes and alleles, if selected in the
and J Genes and Alleles Search page, display the alignments with the five closest V, D, and
J gene alleles, respectively, with their alignment score and their
578 E. Alamyar et al.

Fig. 4. IMGT/V-QUEST “Detailed view” result page: Sequence and “Result summary” with the option “Search for insertions
and deletions.” An insertion of 9 nt was detected by IMGT/V-QUEST. The analysis is then performed after having removed
the insertion.

identity percentage (Fig. 5). The alignment for D-GENE and allele
should be considered with caution since it may show discrepancies
with the results obtained by IMGT/JunctionAnalysis (9) (see
Note 12).

2.3.4. “Results of IMGT/ The results of IMGT/JunctionAnalysis comprise (Fig. 6):

JunctionAnalysis” and
1. “Analysis of the JUNCTION” (top section of Fig. 6): it
“Sequence of the
provides the name of the D gene and allele (for IGH, TRB,
JUNCTION (‘nt’ and ‘AA’)”
and TRD sequences) and shows the details of the junction
at the nucleotide level (nucleotides trimmed at the V, D, and J
gene ends and N nucleotides added by the TdT) with an accu-
rate delimitation of the 3¢V-REGION (from the 2nd-CYS 104
to the 3¢ end of the V-REGION), D-REGION, and
5¢J-REGION (from the 5¢ end of the J-REGION to J-TRP 118
or J-PHE 118) (see Note 13). Dots represent the nucleotides
that have been trimmed by the exonuclease. The number of
 32
IMGT® Tools for the Nucleotide Analysis of Immunoglobulin…

Fig. 5. IMGT/V-QUEST “Detailed view” result page: “Alignment for V-GENE and allele identification” and “Alignment for J-GENE and allele identification.” Dashes indicate identical
nucleotides. Dots indicate gaps. FR-IMGT and CDR-IMGT are delimited according to the IMGT unique numbering (19–21). The seq1 is the human IGH rearranged sequence with the
579

accession number AB012909 in the IMGT/LIGM-DB database (34).

580 E. Alamyar et al.

Fig. 6. IMGT/V-QUEST “Detailed view” result page: “Results of IMGT/JunctionAnalysis” (with “Analysis of the JUNCTION,”
“Eligible D genes” and “Translation of the JUNCTION”) and “Sequence of the JUNCTION (‘nt’ and ‘AA’).” The seq1 is the
human IGH rearranged sequence with the accession number AB012909 in the IMGT/LIGM-DB database (34).
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 581

mutations in the 3¢V-REGION, D-REGION, and 5¢J-REGION

is indicated, for example Vmut 0, Dmut 3, Jmut 2 (Fig. 6) and
the corresponding mutated nucleotides are underlined.
Nucleotides of the N-REGION are displayed in N1 and N2.
The ratio of the number of g + c nucleotides to the total num-
ber of N nucleotides is indicated under “Ngc,” for example in
Fig. 6 “Ngc” is 4/5 as there are 4 “c” or “g” in N1 + N2 and
the total of nt is 5 (g (N1), acgc (N2)). By default, the number
of allowed mutations in the IG 3¢V-REGION, (D-REGION
for IGH) and 5¢J-REGION is limited (parameters ¢2,(4),2¢ for
IGH, parameters ¢7,7¢ for IGK and IGL). Moreover, in case of
unmutated IGHV genes (no mutation from FR1-IMGT to
FR3-IMGT), the tool does not allow any mutation in the
3¢V-REGION and 5¢J-REGION and automatically limits the
number of allowed mutations in the D-REGION to 2 (para-
meters “0,(2),0”) in order to reflect the low probability of
somatic hypermutations. By default, mutations are not allowed
for the TR (see Note 14).
2. “Eligible D genes” (middle section of Fig. 6): this option
allows to compare the IMGT/JunctionAnalysis D gene
identification with all D genes which match the junction with
their corresponding score. The D gene allele with the best
score, selected by IMGT/JunctionAnalysis, is shown in the
rectangle in Fig. 6. The location d[7–19]s[10–22] means that
the D-REGION nucleotides (d) from positions 7–19 of the
germline IGHD5-5*01 match the sequence (s) of the junction
from positions 10–22.
3. “Translation of the JUNCTION” (lower section of Fig. 6): it
displays the junction with amino acids colored according to
the 11 IMGT physicochemical classes (18) (see Note 15), the
frame (“+” and “−” indicate in-frame and out-of-frame, respec-
tively), the CDR3-IMGT length, the molecular mass, and the
isoelectric point (pI). In case of frameshifts, gaps (represented
by one or two dots) are inserted to maintain the J-REGION
frame. The corresponding codon which cannot be translated is
represented by “#.”
4. “Sequence of the JUNCTION (‘nt’ and ‘AA’)”: this provides
the JUNCTION in nucleotides (nt) and amino acids (AA)
according to the IMGT unique numbering, and in the FASTA
format with the correctly formatted header required as input
by IMGT/JunctionAnalysis online (9).

2.3.5. “V-REGION The results provide three displays of the V-REGION (Fig. 7):
Alignment,” “V-REGION
Translation,” and 1. The “V-REGION alignment according to the IMGT unique
“V-REGION Protein numbering” displays the nt sequences with the FR-IMGT and
Display” CDR-IMGT delimitations according to the IMGT unique
numbering (19–21).
582 E. Alamyar et al.

Fig. 7. IMGT/V-QUEST “Detailed view” result page: Three displays of the V-REGION. The seq1 is the human IGH rearranged
sequence with the accession number AB012909 in the IMGT/LIGM-DB database (34). Only a part of each of the three
displays is shown.
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 583

2. The “V-REGION translation” displays the nt sequence and

deduced AA translation of the input sequence, aligned with
the closest germline V-REGION, and with the FR-IMGT and
CDR-IMGT delimitations.
3. The “V-REGION protein display” displays the deduced AA
translation of the input sequence, aligned with the V-REGION
translation of the closest germline V-GENE and with the
FR-IMGT and CDR-IMGT delimitations, and with the third
line of the alignment and shown in bold, the AA of the input
sequence which are different from the closest germline
V-REGION.

2.3.6. Analysis of the The analysis of the mutations in the V-REGION is performed by
Mutations: “V-REGION comparison of the analyzed sequence with the closest germline V
Mutation and AA Change gene and allele.
Table,” “V-REGION
1. The “V-REGION mutation and AA change table” (Fig. 8)
Mutation and AA Change
lists the nt mutations and the corresponding AA changes if the
Statistics,” and “V-REGION
mutations are not silent. They are described for each FR-IMGT
Mutation Hot Spots”
and CDR-IMGT, with their nt and codon position according
to the IMGT unique numbering (19–21) and for the AA
changes according to the IMGT AA classes (18). For example,
c1 > g, Q1 > E (++−) means that the nt mutation (c > g) at nt 1
leads to an AA change (Q > E) at codon 1 with the same
hydropathy (+) and volume (+), but with different physico-
chemical properties (−) classes. It is the first time that such
qualification of amino acid replacement is provided. This has
led to identify four types of AA changes: very similar (+++),
similar (++−, +−+), dissimilar (−−+, −+−, +−−), and very dis-
similar (−−−).
2. The “V-REGION mutation and AA change statistics” (Fig. 8)
comprises two tables, the first one for “Nucleotide (nt) muta-
tions” described as silent/nonsilent, transitions/transversions,
and the second one for “Amino acid (AA) changes.” For both
tables, results are given for the V-REGION and for FR-IMGT
and CDR-IMGT. Statistics are calculated up to the 3¢ end of
the V-REGION identified in the input sequence (this includes
the 3¢ last two identical nucleotides with the closest germline
V-REGION). The numbers in parentheses, in the V-REGION
and CDR3-IMGT columns, correspond to the statistics calcu-
lated up to the 3¢ end of the closest germline V-REGION and
therefore may include nt and AA differences due to the junc-
tion diversity.
In order to avoid to count sequence differences due to the 5¢
primer in “V-REGION mutation and AA change table” and in
“V-REGION mutation and AA change statistics,” before launching
the analysis it is possible to exclude in 5¢ of the V-REGION, a
given number of nucleotides, defined by the user, in “Advanced
584 E. Alamyar et al.

Fig. 8. IMGT/V-QUEST “Detailed view” result page: “V-REGION mutation and AA change table” and “V-REGION mutation and
AA change statistics.” Results are for the human IGH rearranged sequence AB012909 of the IMGT/LIGM-DB database (34)
by comparison with the germline IGHV3-30*04.

parameters” (Parameters for “Detailed view” in the IMGT/V-QUEST

Search page).
The positions of the mutations have been correlated with
the presence of specific patterns in the germline V gene and allele
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 585

Fig. 9. IMGT/V-QUEST “Detailed view” result page: “V-REGION mutation hot spots (motifs and localization of hot spots in
germline V-REGION).” Mutation hot spots are those of the human IGHV3-30*04 allele.

(see Somatic hypermutations, in IMGT Education: http://www.

imgt.org/IMGTeducation/Tutorials/IGandBcells/_UK/
SomaticHypermutations/). The “V-REGION mutation hot spots”
table (Fig. 9) shows the localization of the hot spot patterns (a/t)a
(or wa) and (a/g)g(c/t)(a/t) (or rgyw) and their complementary
reverse motifs t(a/t) (or tw) and (a/t)(a/g)c(c/t) (or wrcy) in the
closest germline V gene and allele.

2.3.7. Sequences of V-, This result provides nt and AA sequences with gaps according to
V-J- or V-D-J-REGION the IMGT unique numbering (19–21) and includes (Fig. 10) the
(“nt” and “AA”) with Gaps “V-REGION nucleotide sequence in FASTA format” with access
in FASTA and Access to the IMGT/PhyloGene tool (22), “V-REGION amino acid in
to IMGT/PhyloGene FASTA format,” “V-REGION amino acid on one line,” V-J or
for V-REGION (“nt”) V-D-J-REGION nt and AA sequences in FASTA format, and V-J or
V-D-J-REGION AA sequence on one line. In that option, the J read-
ing frame of sequences with an out-of-frame junction is not restored
(the sequence does not include “#”) and a note alerts the user.

2.3.8. Annotation by IMGT/ The results of the analysis of the input sequence by IMGT/
Automat Automat (23, 24) provide a full automatic annotation for the V-J-
REGION or V-D-J-REGION using IMGT® standardized labels
(Fig. 11).
586 E. Alamyar et al.

Fig. 10. IMGT/V-QUEST “Detailed view” result page: V-REGION and V-(D)-J-REGION sequences with IMGT gaps. Sequences
(“nt” and “AA”) are for the human IGH rearranged sequence AB012909 of the IMGT/LIGM-DB database (34). The 3¢ end of
the “on one line” sequences are not shown owing to the figure format.

2.3.9. IMGT Colliers The “IMGT Collier de Perles” (25–27) can be displayed either as
de Perles a “link to IMGT/Collier-de-Perles tool” or as a direct “IMGT
Collier de Perles (for a number of sequences <5)” representation
integrated in IMGT/V-QUEST results (Fig. 12), depending on the
user selection in the IMGT/V-QUEST Search page (see Note 16).

2.4. IMGT/V-QUEST The “Synthesis view” results are displayed in “B. Synthesis for the
Output for “Synthesis IMGT/V-QUEST analyzed sequences.” At the top of the page,
View” the number of analyzed sequences is indicated.

2.4.1. Summary Table The “Summary table” (Fig. 13) displays one row for each input
sequence with the corresponding results, including: (1) the name
of the sequence (Sequence ID), (2) the name of the closest
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 587

Fig. 11. IMGT/V-QUEST “Detailed view” result page: “Annotation by IMGT/Automat.” Results are for the human IGH
rearranged sequence AB012909 of the IMGT/LIGM-DB database (34). Only a part of the annotation is shown.

V-GENE and allele (see Note 7), (3) the functionality of the
sequence (when found, the presence of stop codons is indicated),
(4) the V-REGION score (see Note 8), (5) the V-REGION per-
centage of identity with, between parentheses, the ratio of number
of identical nucleotides (nt)/number of aligned nt (see Note 17),
(6) the name of the closest J-GENE and allele (see Note 18), (7)
and, provided according to the IMGT/JunctionAnalysis results,
the D-GENE and allele name, the D reading frame, the CDR-
IMGT lengths, the AA JUNCTION, and the JUNCTION frame.
In the absence of results of IMGT/JunctionAnalysis, only the AA
JUNCTION defined by IMGT/V-QUEST is displayed.

2.4.2. Results of IMGT/ Below the “Summary table,” use the links associated to the locus
JunctionAnalysis name(s) (e.g., IGH) to display “Results of IMGT/JunctionAnalysis”
for sequences identified by the tool as belonging to the same locus
588 E. Alamyar et al.

Fig. 12. IMGT/V-QUEST “Detailed view” result page: IMGT Collier de Perles. The IMGT Collier de Perles was obtained with
the option “IMGT Collier de Perles (for a nb of sequences <5)” of the human IGH rearranged sequence AB012909 of the
IMGT/LIGM-DB database (34). Anchor positions are shown as squares. In FR-IMGT, the hydrophobic amino acids (hydropa-
thy index with positive value) and tryptophan (W) found at a given position in more than 50% of sequences are displayed
(online with a blue background color).

(Fig. 14). Results are displayed for the IGH, IGK, or IGL locus for
IG sequences, and for the TRA, TRB, TRG, or TRD locus for TR
sequences. They are similar to those obtained for individual
sequences detailed in Subheading 2.3.4.

2.4.3. Alignment with the The synthesis results provide six different displays (if all were
Closest Alleles selected): “Alignment for V-GENE,” “V-REGION alignment
according to the IMGT unique numbering” and “V-REGION
translation” (Fig. 15), and “V-REGION protein display” in three
different formats (Fig. 16).

2.4.4. V-REGION Most This section shows, for each FR-IMGT and CDR-IMGT, and for
Frequently Occurring AA each position according to the IMGT unique numbering (19–21),
Per Position and Per the most frequently occurring amino acid (AA) (Fig. 17).
FR-IMGT and CDR-IMGT
 32

Fig. 13. IMGT/V-QUEST “Synthesis view” result page: Top of the page: The seq1, seq2, seq3, seq4, seq5, seq6, and seq7 correspond to the DQ100777, AB021524, AB012909,
AB021511, AB021516, AB021514, and AB021539 accession numbers, respectively, from the IMGT/LIGM-DB database (34).
IMGT® Tools for the Nucleotide Analysis of Immunoglobulin…
589
 590
E. Alamyar et al.

Fig. 14. IMGT/V-QUEST “Synthesis view” result page: Results of IMGT/JunctionAnalysis. The results show the “Analysis of the JUNCTIONs” and the “Translation of the JUNCTIONs”
for the IGH sequences of the submitted analysis. Owing to the figure format, Vmut, Dmut, Jmut, Ngc, molecular mass, and pI are not shown.
Fig. 15. IMGT/V-QUEST “Synthesis view” result page: “Alignment for V-GENE,” “V-REGION alignment according to the IMGT
unique numbering,” and “V-REGION translation.” Displayed sequences are those of the input set, which have been identified
as using the same closest germline V gene and allele (here IGHV3-30*04). Dashes indicate identical nucleotides. Dots
indicate gaps. FR-IMGT and CDR-IMGT are delimited according to the IMGT unique numbering (19–21). The sequence of
the V-REGION of the closest germline V is shown on the first line and the input sequences are aligned, with nt mutations
indicated by comparison. Underlined nucleotides in the closest germline V correspond to mutation hot spot motif localizations.
Only a part of each of the three displays is shown.
 592
E. Alamyar et al.

Fig. 16. IMGT/V-QUEST “Synthesis view” result page. V-REGION protein displays: Three different displays are shown. “V-REGION protein display,” “V-REGION protein display with
colored AA according to the AA IMGT classes,” (online colored AA) and “V-REGION protein display (only AA changed displayed).” The first line shows the AA sequence of the closest
V-REGION allele against which are aligned the translations of the input sequences identified by IMGT/V-QUEST as using the same V gene and allele (here IGHV3-30*04).
 32
IMGT® Tools for the Nucleotide Analysis of Immunoglobulin…

Fig. 17. IMGT/V-QUEST “Synthesis view” result page: “V-REGION most frequently occurring AA per position and per FR-IMGT and CDR-IMGT.” Results are for a set of four sequences
expressing IGHV3-30*04 (accession numbers AB012909, AB021511, AB021516, AB021514 from IMGT/LIGM-DB database (34)). For each set of user sequences that express the
same closest germline V gene and allele, the most frequently occurring amino acid is shown per amino acid (or codon) position and per FR-IMGT and CDR-IMGT. The display highlights
the conserved as well as the polymorphic positions.
593
594 E. Alamyar et al.

2.5. IMGT/V-QUEST “Excel file” allows the users to open and save a spreadsheet includ-
Output for Excel File ing the results of the IMGT/V-QUEST analysis. The file contains
11 sheets (if all were selected in the IMGT/V-QUEST Search
page, see Subheading 2.2.2). The “Summary” and “Parameters”
are always selected.
The content of the eleven sheets is detailed in the IMGT/
HighV-QUEST “Download” (Subheading 3.4).

3. IMGT/
HighV-QUEST
IMGT/HighV-QUEST, the high-throughput version of IMGT/
V-QUEST, can analyze large numbers of IG and TR-rearranged
sequences (up to 150,000) in a single run.

3.1. IMGT/HighV- As for the other IMGT® databases and tools, IMGT/HighV-
QUEST Welcome Page QUEST is freely available for academics. However, the IMGT/
HighV-QUEST Welcome page requires user identification and
provides, for new users, a link to register. User identification has
been set to avoid nonrelevant use and overload of the server and to
contact the user if needed.

3.2. IMGT/HighV- 1. After user identification, a unique IMGT/HighV-QUEST

QUEST Search Page Search page is displayed (Fig. 18). It is very similar to the
classical IMGT/V-QUEST Search page with however, at the
top, four additional fields:
(a) The analysis title (which must be provided by the user).
(b) The selection of the species.
(c) The selection of the receptor type or locus. IMGT/HighV-
QUEST allows the user to restrict the analysis to a particu-
lar locus, for example, “IGH” or “TRA.” Values for
receptor type or locus are displayed only if the correspond-
ing IMGT reference directory is available.
(d) The choice of e-mail notifications. Because of their large
number of sequences, the analyses are first queued on the
IMGT® server and they are performed depending on the
available resources. Therefore, the results are not displayed
immediately and users may choose to be notified by e-mail:
when analysis is queued, when analysis is submitted, when
analysis is completed (selected by default), and/or before
the results are removed (always selected). These choices
are not exclusive.
2. The sequences must be uploaded using the simple text format-
ted file (the copy and paste option is not available). Then, the
customization for the “Detailed View” (except for IMGT
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 595

Fig. 18. IMGT/HighV-QUEST Search page.

596 E. Alamyar et al.

Collier de Perles results and IMGT/Phylogene which are not

available), “Files in CSV” (equivalent of the “Excel file” of
IMGT/V-QUEST) and “Advanced parameters” is identical to
those of IMGT/V-QUEST.

3.3. IMGT/HighV- From their account, the user can check the status of his/her analyses
QUEST “Analysis at any time by displaying the “Analysis history” (Fig. 19) (available
History Page” from a link from the top of the Search page). The table displays
each analysis with its title, the user name, the status of the analysis,
the submission date with a predicted completion time, the number
of submitted sequences, the species and the receptor type or locus
(as selected by the user), and the actions that can be performed by
the user. A user may cancel a queued or a running analysis at any
time. When the analysis is completed, the user can download the
results as a single file in ZIP format. The size of the ZIP file and the
number of included files are indicated in the table (Fig. 19). If
the user removes a completed analysis, all related files (sequences
and results) are definitively deleted from the server.

3.4. IMGT/HighV- The downloaded ZIP file (Fig. 20) contains a main folder with
QUEST “Download” eleven files (equivalent to the eleven sheets of the excel file pro-
vided by the classical IMGT/V-QUEST) in CSV format, and one
subfolder with individual files, in Text, for each sequence (provid-
ing “Detailed view” results). Text and CSV formats have been cho-
sen in order to facilitate statistical studies for further interpretation
and knowledge extraction. The eleven files (the “Summary” and
“Parameters” files are always provided) comprise:
1. The “Summary” file (25–29 columns) provides the synthesis
of the analysis (the sequence names, the sequence functional-
ity, the names of the closest V, D, and J genes and alleles with
identity percentage, FR-IMGT and CDR-IMGT lengths,
amino acid JUNCTION, the description of insertions and
deletions if any …).
2. The “IMGT-gapped-nt-sequences” file (18 columns) includes
the sequence identification (name, functionality, names of the
closest V, D, and J genes and alleles), the nucleotide sequences
that have been gapped according to the IMGT unique num-
bering (19–21) for V-D-J-REGION, V-J-REGION,
V-REGION, FR1-IMGT, CDR1-IMGT, FR2-IMGT, CDR2-
IMGT, FR3-IMGT, and the nucleotide sequences of CDR3-
IMGT, JUNCTION, J-REGION, and FR4-IMGT.
3. The “nt-sequences” file (63–78 columns) includes the sequence
functionality, the names of the closest V, D, and J genes and
alleles, and the nucleotide sequences of all labels that can be
automatically annotated by IMGT/Automat (23, 24) in the
analyzed sequences.
32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 597

Fig. 19. IMGT/HighV-QUEST “Analysis history” page.

598 E. Alamyar et al.

Fig. 20. IMGT/HighV-QUEST content of a results ZIP file.

4. The “IMGT-gapped-AA-sequences” file (18 columns) includes

the sequence identification, the amino acid sequences that have
been gapped according to the IMGT unique numbering for
V-D-J-REGION, V-J-REGION, V-REGION, FR1-IMGT,
CDR1-IMGT, FR2-IMGT, CDR2-IMGT, FR3-IMGT, and
the amino acid sequences of CDR3-IMGT, JUNCTION,
J-REGION and FR4-IMGT.
5. The “AA-sequences” file (18 columns) includes the same col-
umns as the “IMGT-gapped-AA-sequences” file, but sequences
are without IMGT gaps.
6. The “Junction” file includes the sequence identification and
the results of IMGT/JunctionAnalysis (9) (33 columns for
IGL, IGK, TRA and TRG sequences, 46 (if one D), 66 (if two D)
or 77 (if 3 D) columns for IGH, TRB and TRD sequences).
7. The “V-REGION-mutation-and-AA-change table” file (11
columns) includes the sequence name, the functionality, the
name of the closest V gene and allele, and list of mutations
(nucleotide mutation, AA change, AA class identity or change
for V-REGION, FR1-IMGT, CDR1-IMGT, FR2-IMGT,
CDR2-IMGT, FR3-IMGT and CDR3-IMGT).
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 599

8. The “V-REGION-nt-mutation-statistics” file (130 columns)

includes the sequence name, the sequence functionality, the
name of the closest V gene and allele, then the number of posi-
tions including IMGT gaps, the number of nucleotides, the
number of identical nucleotides, the total number of muta-
tions, the number of silent mutations, the number of nonsilent
mutations, the number of transitions (a > g, g > a, c > t, t > c),
and the number of transversions (a > c, c > a, a > t, t > a, g > c,
c > g, g > t, t > g) for V-REGION, FR1-IMGT, CDR1-IMGT,
FR2-IMGT, CDR2-IMGT, FR3-IMGT, and CDR3-IMGT.
9. The “V-REGION-AA-change-statistics” file (189 columns)
includes the sequence name, the sequence functionality, the
name of the closest V gene and allele, then the number of AA
positions including IMGT gaps, the number of AA, the num-
ber of identical AA, the total number of AA changes, the num-
ber of AA changes according to AAclassChangeType (+++,
++−, +−+, +−−, −+−, −−+, −−−), and the number of AA class
changes according to AAclassSimilarityDegree (number of Very
similar, number of Similar, number of Dissimilar, number of
Very dissimilar) for V-REGION, FR1-IMGT, CDR1-IMGT,
FR2-IMGT, CDR2-IMGT, FR3-IMGT, and CDR3-IMGT.
10. The “V-REGION-mutation-hot-spots” file (eight columns)
includes the sequence functionality, the name of the closest V
gene and allele, and the hot spots motifs (a/t)a, t(a/t), (a/g)
g(c/t)(a/t), (a/t)(a/g)c(c/t) identified in the closest
germline V-REGION and with CDR-IMGT and FR-IMGT
localizations.
11. The “Parameters” file includes the date of the analysis, the
IMGT/V-QUEST programme version, the IMGT/V-QUEST
reference directory release, and the parameters used for the
analysis: the species, the receptor type or locus, the IMGT ref-
erence directory set, “with allele *01” (if selected), “Search for
insertions and deletions,” the number of nucleotides to add
(or exclude) in 3¢ of the V-REGION for the evaluation of the
alignment score, and the number of nucleotides to exclude in
5¢ of the V-REGION for the evaluation of the number of
mutations, if these 2 numbers are not 0 (default value).

4. Notes

1. The tool will automatically identify the locus (IGH, IGK, or

IGL for IG sequences, or TRA, TRB, TRG, or TRD for TR
sequences).
2. The IMGT/V-QUEST reference directory sets include IMGT
reference sequences from all functional genes and alleles, all
600 E. Alamyar et al.

open reading frame (ORF) and all in-frame pseudogenes (P)

alleles. By definition, these sets contain one sequence for each
allele. IMGT reference directories have been set up for species
which have been extensively studied, such as human and
mouse. This also holds for the other species or taxons with
incomplete IMGT reference directory sets. In those cases,
results should be interpreted considering the status of the
IMGT reference directory (information on the updates on the
IMGT® Web site). Links to the IMGT/V-QUEST reference
directory sets are available from the IMGT/V-QUEST
Welcome page.
3. The options (e) and (f) provide different displays of the results
shown in option (d), whereas the option (g) allows to evaluate
the occurrence of amino acids changes at each position.
4. By default, the user sequences are compared with all genes and
alleles. However, the option “With allele *01 only” is useful if
the user sequences need to be compared in “Detailed view”
with different genes. In “Synthesis view,” this option allows to
align all user sequences that express the same gene indepen-
dently of the allelic polymorphism.
5. This option is usually selected in a second step after a first anal-
ysis by IMGT/V-QUEST that provided warnings for potential
insertions or deletions (Subheading 2.3.1).
6. The “Functionality” is an important concept for IG and TR
nucleotide sequence identification, as it allows to distinguish
on the one hand the functionality of germline and conven-
tional genes: functional, ORF (open reading frame), or pseudo-
gene, and on the other hand the functionality of IG and TR
rearranged sequences: productive or unproductive. An IG or a
TR rearranged sequence is productive if no stop codon has
been detected in the V-D-J-REGION or in the V-J-REGION,
and if the junction is in-frame. An IG or a TR rearranged
sequence is unproductive if stop codons are detected in the
V-D-J-REGION or in the V-J-REGION and/or if the junction
is out-of-frame.
7. The IMGT® IG and TR gene names (1, 2, 28, 29) were
approved by the Human Genome Organization (HUGO)
Nomenclature Committee (HGNC) in 1999 (30) and are the
official references for the World Health Organization-
International Union of Immunological Societies (WHO-IUIS)
Nomenclature Subcommittee for IG and TR (31, 32). The
complete list of the human IG and TR gene names has been
entered in IMGT/GENE-DB (15), the IMGT® gene data-
base, and in Entrez Gene at the National Center for
Biotechnology Information (NCBI) (33). IMGT® IG and TR
gene and allele names are based on the concepts of classification
32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 601

of IMGT-ONTOLOGY (CLASSIFICATION axiom) (10–13).

IMGT® reference sequences have been defined for each allele
of each gene based on one or, whenever possible, several of the
following criteria: germline sequence, first sequence published,
longest sequence, mapped sequence (34, 35). IMGT® IG
gene and allele names are used for the antibody description
in the WHO International Nonproprietary Name (INN) pro-
gramme (36).
8. The score of the alignment for 2 sequences is calculated by
counting +5 for each position where nucleotides are identical
(match) and −4 for each position with different nucleotides
(mismatch).
9. In addition to the analysis of the somatic mutations, results of
IMGT/V-QUEST, frequently used by clinicians, also include
the sequences of the V-(D)-J junctions which are used in the
synthesis of specific probes for the follow-up of residual dis-
eases in leukemias and lymphomas.
10. Potential insertions or deletions are suspected by IMGT/V-
QUEST when the V-REGION score is very low (less than
200), and/or the percentage of identity is less than 85%, and/
or when the input sequence has different CDR1-IMGT and/
or CDR2-IMGT lengths, compared to those of the closest
germline V.
11. The note (in red online) indicates the additional J-GENE and
allele names and the criterion used for their identification (usu-
ally highest number of consecutive identical nucleotides).
12. The way to identify the closest germline D is different between
IMGT/V-QUEST and IMGT/JunctionAnalysis since the
evaluation of the alignment score is different. In case of dis-
crepancy, the results of IMGT/JunctionAnalysis are the most
accurate. However, the alignment provided by IMGT/V-
QUEST that is less stringent and displays several D genes and
alleles may be helpful to solve ambiguous cases when IMGT/
JunctionAnalysis does not provide results for the D gene
and allele.
13. Amino acids of the V-DOMAIN are numbered according to
the IMGT unique numbering (19–21). Position 104 (second-
CYS) and position 118 (J-PHE or J-TRP) that belong to the
FR3-IMGT and FR4-IMGT, respectively, are the anchors of
the CDR3-IMGT. They correspond to the 5¢ and 3¢ ends of
the JUNCTION.
14. The default can be modified in “Advanced parameters,” but
only for TR with mutations in the V-REGION (between posi-
tions 1 and 104).
15. The 20 amino acids have been classified in eleven IMGT
“Physicochemical” classes which are based on “Hydrophathy,”
602 E. Alamyar et al.

“Volume,” and “Chemical” characteristics (http://www.imgt.

org/, section “Amino acids” in IMGT Education > Aide-
mémoire).
16. The IMGT Colliers de Perles (25–27) allows the user to bridge
the gap between sequences and structures (37, 38) and is
widely used in the protocol for antibody humanization or for
the evaluation of therapeutic monoclonal antibodies (39, 40):
the provided information is useful to localize the amino acids
of the loops CDR1-IMGT, CDR2-IMGT, CDR3-IMGT that
are involved in the contacts with the antigen. It allows study of
the physicochemical properties of the amino acids at a given
position in a pool of sequences. IMGT/Collier-de-Perles tool
can be customized to display the amino acids colored accord-
ing to their hydropathy, volume, or IMGT physicochemical
classes (18). By default, the IMGT Colliers de Perles are dis-
played on one layer. They can also be displayed on two layers
in order to get a graphical representation closer to the 3D
structure. In the case of sequences with an out-of-frame junc-
tion, the J reading frame is not restored (the sequence does not
include “#”).
17. A note (in red online) may appear with the V-GENE and allele
name when potential insertions or deletions are suspected (cri-
teria detailed in Note 10). In those cases, the alignment for
this sequence has to be checked in “A. Detailed view,” using
the advanced parameters “Search for insertions and deletions
in V-REGION” (Subheading 2.2.3).
18. A note (in red online) may appear to inform on other possibili-
ties for the J-GENE and allele name as described in Note 11.

Acknowledgements

We are grateful to Gérard Lefranc for helpful discussion and to the

IMGT® team for its expertise and constant motivation. IMGT®
received funding from Centre National de la Recherche Scientifique
CNRS, Ministère de l’Enseignement Supérieur et de la Recherche
MESR (University Montpellier 2), Région Languedoc-Roussillon
(GPTR, GEPETOS), Agence Nationale de la Recherche ANR
(ANR-06-BYOS-0005-01), and European Community
(ImmunoGrid, FP6-2004-IST-4). This work was granted access to
the HPC resources of CINES under the allocations 2010-036029
and 2011-036029 made by GENCI (Grand Equipement National
de Calcul Intensif).
 32 IMGT® Tools for the Nucleotide Analysis of Immunoglobulin… 603

References
1. Lefranc M-P, Lefranc G (2001) The immuno- 12. Lefranc M-P, Giudicelli V, Regnier L, Duroux
globulin FactsBook. Academic Press, London, P (2008) IMGT, a system and an ontology that
UK, pp 1–458 bridge biological and computational spheres in
2. Lefranc M-P, Lefranc G (2001) The T cell bioinformatics. Brief Bioinform 9:263–275
receptor FactsBook. Academic Press, London, 13. Duroux P, Kaas Q, Brochet X, Lane J, Ginestoux
UK, pp 1–398 C, Lefranc M-P, Giudicelli V (2008) IMGT-
3. Giudicelli V, Chaume D, Lefranc M-P (2004) Kaleidoscope, the formal IMGT-ONTOLOGY
IMGT/V-QUEST, an integrated software pro- paradigm. Biochimie 90:570–583
gram for immunoglobulin and T cell receptor 14. Alamyar E, Giudicelli V, Duroux P, Lefranc
V-J and V-D-J rearrangement analysis. Nucleic M-P (2010) IMGT/HighV-QUEST: a high-
Acids Res 32:W435–W440 throughput system and web portal for the anal-
4. Lefranc M-P (2004) IMGT, the international ysis of rearranged nucleotide sequences of
ImMunoGeneTics information system. antigen receptors-high-throughput version of
Methods Mol Biol 248:27–49 IMGT/V-QUEST. Journées Ouvertes en
5. Giudicelli V, Lefranc M-P (2005) Interactive Biologie, Informatique et Mathématiques,
IMGT on-line tools for the analysis of immu- September 7-9, 2010, Montpellier, France
noglobulin and T cell receptor repertoires. In: (JOBIM 2010), Poster #60, http://www.
Veskler BA (ed) New research on immunology. jobim2010.fr/?q=fr/node/55. Accessed 31
Nova Science Publishers, Inc., New York, pp Jan 2011
77–105 15. Giudicelli V, Chaume D, Lefranc M-P (2005)
6. Brochet X, Lefranc M-P, Giudicelli V (2008) IMGT/GENE-DB: a comprehensive database
IMGT/V-QUEST: the highly customized and for human and mouse immunoglobulin and T
integrated system for IG and TR standardized cell receptor genes. Nucleic Acids Res
V-J and V-D-J sequence analysis. Nucleic Acids 33:D256–D261
Res 36:W503–W508 16. Belessi C, Davi F, Stamatopoulos K, Degano
7. Giudicelli V, Lefranc M-P (2008) IMGT® M, Andreou TM, Moreno C, Merle-Béral H,
standardized analysis of immunoglobulin rear- Crespo M, Laoutaris NP, Montserrat E,
ranged sequences. In: Ghia P, Rosenquist R, Caligaris-Cappio F, Anagnostopoulos AZ, Ghia
Davi F (eds) Immunoglobulin gene analysis in P (2006) IGHV gene insertions and deletions
chronic lymphocytic leukemia (chapter 2). in chronic lymphocytic leukemia: “CLL-biased”
Wolters Kluwer Health, Italy, pp 33–52 deletions in a subset of cases with stereotyped
receptors. Eur J Immunol 36:1963–1974
8. Lefranc M-P, Giudicelli V, Ginestoux C,
Jabado-Michaloud J, Folch G, Bellahcene F, 17. Ghia P, Stamatopoulos K, Belessi C, Moreno
Wu Y, Gemrot E, Brochet X, Lane J, Regnier C, Stilgenbauer S, Stevenson F, Davi F,
L, Ehrenmann F, Lefranc G, Duroux P (2009) Rosenquist R (2007) ERIC recommendations
IMGT®, the international ImMunoGeneTics on IGHV gene mutational status analysis in
information system®. Nucleic Acids Res chronic lymphocytic leukemia. Leukemia 21:1–3
37:D1006–D1012 18. Pommié C, Levadoux S, Sabatier R, Lefranc G,
9. Yousfi Monod M, Giudicelli V, Chaume D, Lefranc M-P (2004) IMGT standardized crite-
Lefranc M-P (2004) IMGT/JunctionAnalysis: ria for statistical analysis of immunoglobulin
the first tool for the analysis of the immuno- V-REGION amino acid properties. J Mol
globulin and T cell receptor complex V-J and Recognit 17:17–32
V-D-J JUNCTIONs. Bioinformatics 19. Lefranc M-P (1997) Unique database number-
20:i379–i385 ing system for immunogenetic analysis.
10. Giudicelli V, Lefranc M-P (1999) Ontology for Immunol Today 18:509
immunogenetics: the IMGT-ONTOLOGY. 20. Lefranc M-P (1999) The IMGT unique num-
Bioinformatics 15:1047–1054 bering for Immunoglobulins, T cell receptors
11. Lefranc M-P, Giudicelli V, Ginestoux C, Bosc and Ig-like domains. Immunologist 7:
N, Folch G, Guiraudou D, Jabado-Michaloud 132–136
J, Magris S, Scaviner D, Thouvenin V, Combres 21. Lefranc M-P, Pommié C, Ruiz M, Giudicelli V,
K, Girod D, Jeanjean S, Protat C, Yousfi- Foulquier E, Truong L, Thouvenin-Contet V,
Monod M, Duprat E, Kaas Q, Pommié C, Lefranc G (2003) IMGT unique numbering
Chaume D, Lefranc G (2004) IMGT- for immunoglobulin and T cell receptor vari-
ONTOLOGY for immunogenetics and immu- able domains and Ig superfamily V-like
noinformatics. In Silico Biol 4:17–29 domains. Dev Comp Immunol 27:55–77
604 E. Alamyar et al.

22. Elemento O, Lefranc M-P (2003) IMGT/ T cell receptors report. Immunogenetics
PhyloGene: an on-line tool for comparative 59:899–902
analysis of immunoglobulin and T cell receptor 32. Lefranc M-P (2008) WHO-IUIS nomenclature
genes. Dev Comp Immunol 27:763–779 subcommittee for immunoglobulins and T cell
23. Giudicelli V, Chaume D, Jabado-Michaloud J, receptors report August 2007, 13th International
Lefranc M-P (2005) Immunogenetics sequence Congress of Immunology, Rio de Janeiro,
annotation: the Strategy of IMGT based on Brazil. Dev Comp Immunol 32:461–463
IMGT-ONTOLOGY. Stud Health Technol 33. Maglott D, Ostell J, Pruitt KD, Tatusova T
Inform 116:3–8 (2011) Entrez gene: gene-centered information
24. Giudicelli V, Protat C, Lefranc M-P (2003) at NCBI. Nucleic Acids Res 39:D52–D57
The IMGT strategy for the automatic annota- 34. Giudicelli V, Duroux P, Ginestoux C, Folch G,
tion of IG and TR cDNA sequences: IMGT/ Jabado-Michaloud J, Chaume D, Lefranc M-P
Automat. In: INRIA (DISC/Spid), Paris, (2006) IMGT/LIGM-DB, the IMGT com-
DKB-31. Proceedings of the European prehensive database of immunoglobulin and T
Conference on Computational Biology (ECCB cell receptor nucleotide sequences. Nucleic
2003), pp 103–104 Acids Res 34:D781–D784
25. Ruiz M, Lefranc M-P (2002) IMGT gene 35. Lefranc M-P (1998) IMGT (ImMunoGeneTics)
identification and Colliers de Perles of human locus on focus. A new section of experimental
immunoglobulins with known 3D structures. and clinical immunogenetics. Exp Clin
Immunogenetics 53:857–883 Immunogenet 15:1–7
26. Kaas Q, Lefranc M-P (2007) IMGT Colliers de 36. Lefranc M-P (2011) Antibody nomenclature:
Perles: standardized sequence-structure repre- from IMGT-ONTOLOGY to INN definition.
sentations of the IgSF and MhcSF superfamily MAbs 3:1
domains. Curr Bioinform 2:21–30 37. Kaas Q, Ruiz M, Lefranc M-P (2004)
27. Kaas Q, Ehrenmann F, Lefranc M-P (2007) IMGT/3Dstructure-DB and IMGT/Structural
IG, TR and IgSF, MHC and MhcSF: what do Query, a database and a tool for immunoglobu-
we learn from the IMGT Colliers de Perles? lin. T cell receptor and MHC structural data.
Brief Funct Genomic Proteomic 6:253–264 Nucleic Acids Res 32:D208–D210
28. Lefranc M-P (2001) Nomenclature of the 38. Ehrenmann F, Kaas Q, Lefranc M-P (2010)
human immunoglobulin genes. In: Coligan JE, IMGT/3Dstructure-DB and IMGT/
Bierer BE, Margulies DH, Shevach EM, DomainGapAlign: a database and a tool for
Strober W (eds) Current protocols in immu- immunoglobulins or antibodies, T cell recep-
nology. Wiley, Hoboken, NJ, USA, pp tors, MHC, IgSF and MhcSF. Nucleic Acids
A.1P.1–A.1P.37 Res 38:D301–D307
29. Lefranc M-P (2001) Nomenclature of the human 39. Lefranc M-P (2009) Antibody databases and
T cell receptor genes. In: Coligan JE, Bierer BE, tools: the IMGT® experience. In: Zhiqiang An
Margulies DH, Shevach EM, Strober W (eds) (ed) Therapeutic monoclonal antibodies: from
Current protocols in immunology. Wiley, bench to clinic (chapter 4). Wiley, Hoboken,
Hoboken, NJ, USA, pp A.1O.1–A.1O.23 NJ, USA, pp 91–114
30. Wain HM, Bruford EA, Lovering RC, Lush 40. Ehrenmann F, Duroux P, Giudicelli V, Lefranc
MJ, Wright MW, Povey S (2002) Guidelines M-P (2010) Standardized sequence and struc-
for human gene nomenclature. Genomics ture analysis of antibody using IMGT®. In:
79:464–470 Kontermann R, Dübel S (eds) Antibody engi-
31. Lefranc M-P (2007) WHO-IUIS nomencla- neering (chapter 2), vol 2, Springer. Berlin,
ture subcommittee for immunoglobulins and Heidelberg, pp 11–31
 Chapter 33

IMGT/DomainGapAlign: The IMGT® Tool for the Analysis of IG,

TR, MH, IgSF, and MhSF Domain Amino Acid Polymorphism
François Ehrenmann and Marie-Paule Lefranc

Abstract
IMGT/DomainGapAlign is the online tool of IMGT®, the international ImMunoGeneTics information
system®, for the analysis of amino acid sequences and two-dimensional (2D) structures of domains. IMGT/
DomainGapAlign allows the analysis of the closest variable (V) and constant (C) domains of immuno-
globulins (IG) or antibodies, T cell receptors (TR), and immunoglobulin superfamily (IgSF) proteins, and
of the groove (G) domains of major histocompatibility (MH; in humans, HLA for human leukocyte anti-
gen) and MH superfamily proteins. IMGT/DomainGapAlign aligns the user own sequences against the
IMGT domain reference directory, displays amino acid changes, creates IMGT gaps, and delimits the
domain strands and loops (and helix for G domain) according to the IMGT unique numbering. IMGT/
DomainGapAlign is coupled to the IMGT/Collier-de-Perles tool that draws standardized IMGT Colliers
de Perles. The analysis is based on the IMGT-ONTOLOGY concepts of identification, classification,
description, and numerotation generated from the axioms of the Formal IMGT-ONTOLOGY or IMGT-
Kaleidoscope. IMGT/DomainGapAlign provides an invaluable help for antibody engineering and anti-
body humanization as it precisely defines the standardized framework regions (FR-IMGT) and
complementarity determining regions (CDR-IMGT) to be grafted. IMGT/DomainGapAlign is freely
available at http://www.imgt.org.

Key words: IMGT, Immunoglobulin, Antibody, T cell receptor, IMGT-ONTOLOGY, Immuno-

globulin superfamily, Major histocompatibility, HLA, MH superfamily, Antibody humanization

1. Introduction

1.1. Overview IMGT®, the international ImMunoGeneTics information system®

(1), http://www.imgt.org, is specialized in the sequences, struc-
tures, and genetic data of the immunoglobulins (IG) or antibodies,
T cell receptors (TR), major histocompatibility (MH; in humans,
HLA for human leukocyte antigen), proteins of the immunoglob-
ulin superfamily (IgSF) and MH superfamily (MhSF), and related

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_33, © Springer Science+Business Media New York 2012

605
606 F. Ehrenmann and M.-P. Lefranc

proteins of the immune system (RPI) (1). IMGT® comprises

databases, online tools, and more than 15,000 pages of Web
resources (1). IMGT/DomainGapAlign (2) is the IMGT® online
tool for the analysis of amino acid sequences and two-dimensional
(2D) structures of domains. IMGT/DomainGapAlign standard-
ization is based on the IMGT-ONTOLOGY concepts of
identification, classification, description, and numerotation gener-
ated from the axioms of the Formal IMGT-ONTOLOGY or
IMGT-Kaleidoscope (3–6). IMGT/DomainGapAlign provides
the standardized IMGT gene and allele names (CLASSIFICATION
axiom), the standardized IMGT labels (DESCRIPTION axiom),
and the IMGT unique numbering (NUMEROTATION axiom).
Three types of domains can be analyzed by IMGT/
DomainGapAlign: the variable (V) (7–9) and constant (C) (10)
domains of the IG and TR and other IgSF proteins, and the
groove (G) (11) domain of MH and other MhSF proteins.
IMGT/DomainGapAlign identifies domains in terms of sequence
identity and aligns the user amino acid (AA) sequences with the
closest V, C, or G domain from the IMGT domain reference direc-
tory. For the IG and TR V domains (V-DOMAIN) which result
from V-(D)-J rearrangements (12, 13), the tool displays align-
ments with the translation of the germline V and joining (J) genes
from IMGT/GENE-DB (14). IMGT/DomainGapAlign creates
gaps in the user amino acid sequences, delimits domains, strands,
turns and loops, helix for G, framework regions (FR-IMGT) and
complementarity determining regions (CDR-IMGT) for
V-DOMAIN, according to the IMGT unique numbering (7–11).
IMGT/DomainGapAlign also displays the amino acid changes,
highlights them in IMGT Colliers de Perles (15–17) generated
using the IMGT/Collier-de-Perles tool, and provides tables with
their detailed description, according to the eleven IMGT amino
acid physicochemical classes (18). The IMGT/DomainGapAlign
tool allows analysis of several sequences simultaneously (up to 50)
and users can choose how many alignments to display for each
sequence. IMGT/DomainGapAlign is freely available at http://
www.imgt.org.

1.2. Basic Information Basic information related to domain types, IMGT domain refer-
for IMGT/Domain- ence directory, IMGT unique numbering, and IMGT Collier de
GapAlign Perles, necessary for an effective use of IMGT/DomainGapAlign,
is briefly reviewed.

1.2.1. Domain Types 1. V domain.

A V domain (8, 9) comprises about 100 amino acids and is
made of nine antiparallel beta strands (A, B, C, C¢, C″, D, E, F,
and G) linked by beta turns (AB, CC¢, C″D, DE, and EF) or
loops (BC, C¢C″ and FG) and forming a sandwich of two sheets
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 607

Table 1
V domain strands and loops, IMGT positions and lengths, based on the IMGT unique
numbering for V domain (V-DOMAIN and V-LIKE-DOMAIN) (7–9)

V domain strands IMGT Characteristic V-DOMAIN FR-IMGT

and loopsa positions Lengthsb Residue@Positionc and CDR-IMGT

A-STRAND 1–15 15 (14 if gap at 10) FR1-IMGT

B-STRAND 16–26 11 1st-CYS 23
BC-LOOP 27–38 12 (or less) CDR1-IMGT
C-STRAND 39–46 8 CONSERVED- FR2-IMGT
C¢-STRAND 47–55 9 TRP 41
C¢C″-LOOP 56–65 10 (or less) CDR2-IMGT
C″-STRAND 66–74 9 (or 8 if gap at 73) FR3-IMGT
D-STRAND 75–84 10 (or 8 if gaps at 81, 82)
E-STRAND 85–96 12 Hydrophobic 89
F-STRAND 97–104 8 2nd-CYS 104
FG-LOOP 105–117 13 (or less, or more) CDR3-IMGT
G-STRAND 118–128 11 (or 10) V-DOMAIN FR4-IMGT
J-PHE 118 or
J-TRP 118d
a
IMGT® labels (concepts of description) are written in capital letters
b
In number of amino acids (or codons)
c
Residue@Position is an IMGT® concept of numerotation that numbers the position of a given residue (or that of a
conserved property amino acid class), based on the IMGT unique numbering
d
In the IG and TR V-DOMAIN, the G-STRAND (or FR4-IMGT) is the C-terminal part of the J-REGION, with
J-PHE or J-TRP 118 and the canonical motif F/W-G-X-G at positions 118–121. The JUNCTION refers to the
CDR3-IMGT plus the two anchors 2nd-CYS 104 and J-PHE or J-TRP 118

(see Table 1). The sheets are closely packed against each other
through hydrophobic interactions giving a hydrophobic core
and joined together by a disulfide bridge between 1st-CYS at
position 23 in the B-STRAND in the first sheet and the 2nd-
CYS 104 in the F-STRAND in the second sheet. The V domain
type (8, 9) includes the V-DOMAIN of the IG and TR, which
corresponds to the V-J-REGION or V-D-J-REGION encoded
by V-(D)-J rearrangements (12, 13), and the V-LIKE-
DOMAIN of the IgSF other than IG and TR (19–21).
2. C domain.
A C domain (10) comprises about 100 amino acids and is
made of seven antiparallel beta strands (A, B, C, D, E, F, and
G) linked by beta turns, a transversal strand (CD) and loops
(BC and FG), on two sheets (see Table 2). A C domain has a
topology and a three-dimensional structure similar to that of a V
608 F. Ehrenmann and M.-P. Lefranc

Table 2
C domain strands, turns, and loops, IMGT positions and lengths, based on the IMGT
unique numbering for C domain (C-DOMAIN and C-LIKE-DOMAIN) (10)

C domain strands, Characteristic Residue@

turns, and loopsa IMGT positions Lengthsb Positionc

A-STRAND 1–15 15 (14 if gap at 10)

AB-TURN 15.1–15.3 0–3
B-STRAND 16–26 11 1st-CYS 23
BC-LOOP 27–31, 34–38 10 (or less)
C-STRAND 39–45 7 CONSERVED-TRP 41
CD-STRAND 45.1–45.9 0–9
D-STRAND 77–84 8 (or 7 if gap at 82)
DE-TURN 84.1–84.7, 85.1–85.7 0–14
E-STRAND 85–96 12 Hydrophobic 89
EF-TURN 96.1–96.2 0–2
F-STRAND 97–104 8 2nd-CYS 104
FG-LOOP 105–117 13 (or less, or more)
G-STRAND 118–128 11 (or less)
a ®
IMGT labels (concepts of description) are written in capital letters
b
In number of amino acids (or codons)
c
Residue@Position is an IMGT® concept of numerotation that numbers the position of a given residue (or that of a
conserved property amino acid class), based on the IMGT unique numbering

domain, but without the C¢ and C″ strands and the C¢C″ loop.
The C domain type (10) includes the C-DOMAIN of the IG
and TR (12, 13) and the C-LIKE-DOMAIN of the IgSF other
than IG and TR (19–23).
3. G domain.
A G domain (11) comprises about 90 amino acids and is
made of four antiparallel beta strands linked by turns and a
helix that sits on the beta strands, its axis forming an angle of
about 40° with the strands (see Table 3). Two G domains are
needed to form the MhSF groove made of a “floor” and two
“walls.” Each G domain contributes groove structure by its
four strands and turns to half of the groove floor and by its
helix to one wall of the groove (11). The G domain type
includes the G-DOMAIN of the MH (11) and the G-LIKE-
DOMAIN of the MhSF other than MH (22, 24).

1.2.2. IMGT Domain The IMGT domain reference directory is the IMGT reference
Reference Directory directory for V, C, and G domains (2). It is manually curated and
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 609

Table 3
G domain strands, turns, and helix, IMGT positions and lengths, based on the IMGT
unique numbering for G domain (G-DOMAIN and G-LIKE-DOMAIN) (11)

G domain strands, Characteristic Residue@Positionc

turns, and helixa IMGT positions Lengths b
and additional positionsd

A-STRAND 1–14 14 7A, CYS-11

AB-TURN 15–17 3 (or 2 or 0)
B-STRAND 18–28 11 (or 10e)
BC-TURN 29–30 2
C-STRAND 31–38 8
CD-TURN 39–41 3 (or 1f)
D-STRAND 42–49 8 49.1–49.5
HELIX 50–92 43 (or less or more) 54A, 61A, 61B, 72A, CYS-74, 92A
a
IMGT® labels (concepts of description) are written in capital letters
b
In number of amino acids (or codons)
c
Residue@Position is an IMGT® concept of numerotation that numbers the position of a given residue (or that of a
conserved property amino acid class), based on the IMGT unique numbering
d
For details on the characteristic Residue@Position and additional positions, see ref. (11)
e
Or 9 in some G-BETA (11)
f
Or 0 in some G-ALPHA2-LIKE (11)

contains the amino acid sequences of the domains delimited accord-

ing to the IMGT rules (based on the exon delimitations) (10).
Sequences are from the IMGT Repertoire (1) or from IMGT/
GENE-DB (14) (see Note 1). Owing to the particularities of the
V-DOMAIN synthesis (12, 13), there is no V-DOMAIN in
the IMGT reference directory. Instead, the directory comprises the
translation of the IG and TR germline V and J genes (V-REGION
and J-REGION, respectively). The IMGT domain reference
directory provides the IMGT “gene” and “allele” names
(“CLASSIFICATION” axiom) (see Note 2). Data are comprehen-
sive for human and mouse IG and TR and human MH, whereas for
other species and IgSF and MhSF they are added progressively.

1.2.3. IMGT Unique 1. IMGT unique numbering for V domain.

Numbering The V domain strands and loops and their delimitations
and lengths are based on the IMGT unique numbering for V
domain (V-DOMAIN and V-LIKE-DOMAIN) (7–9)
(Table 1). In the IG and TR V-DOMAIN, the G-STRAND is
the C-terminal part of the J-REGION, with J-PHE or J-TRP
118 and the canonical motif F/W-G-X-G at positions 118–
121 (8, 9). The loop length (number of amino acids (or
codons), that is number of occupied positions) is a crucial and
610 F. Ehrenmann and M.-P. Lefranc

original concept of IMGT-ONTOLOGY. The lengths of the

loops BC (or CDR1-IMGT for V-DOMAIN), C¢C″ (or
CDR2-IMGT for V-DOMAIN), and FG (or CDR3-IMGT
for V-DOMAIN) characterize the V domain. The lengths of
the three loops BC, C¢C″, and FG are shown in number of
amino acids (or codons), into brackets and separated by dots.
For example, [9.6.9] means that the BC, C¢C″, and FG loops
(or CDR1-IMGT, CDR2-IMGT, and CDR3-IMGT for a
V-DOMAIN) have a length of 9, 6, and 9 amino acids (or
codons), respectively.
2. IMGT unique numbering for C domain.
The C domain strands, turns, and loops, and their delimita-
tions and lengths are based on the IMGT unique numbering for C
domain (C-DOMAIN and C-LIKE-DOMAIN) (10) (Table 2).
3. IMGT unique numbering for G domain.
The G domain strands, turns, and helix and their delimita-
tions and lengths are based on the IMGT unique numbering
for G domain (G-DOMAIN and G-LIKE-DOMAIN) (11)
(Table 3).

1.2.4. IMGT Colliers IMGT Colliers de Perles are standardized IMGT 2D graphical
de Perles representations of protein domains (15–17) that can be obtained
using the IMGT/Collier-de-Perles tool available from the IMGT®
Home Page at http://www.imgt.org (for sequences already
gapped according to the IMGT unique numbering) or using the
IMGT/Collier-de-Perles tool integrated in IMGT/V-QUEST
(25) (starting from V-(D)-J nucleotide sequences) or integrated in
IMGT/DomainGapAlign (2) (starting from V, C, or G domain
amino acid sequences).

2. IMGT/
DomainGapAlign
2.1. IMGT/Domain- 1. The IMGT/DomainGapAlign Welcome page is accessed by
GapAlign Welcome clicking the link IMGT/DomainGapAlign (“IMGT tools”
Page section) in the IMGT® Home page at http://www.imgt.org.
2. In the IMGT/DomainGapAlign Welcome page (Fig. 1), locate
the text area and paste your amino acid sequences in FASTA
format. Alternatively, you can upload a file. A precise delimita-
tion of the domain sequences is not required; however, if the
sequence contains several domains, the sequence should be
split between the different domains. Several domain amino
acid sequences can be analyzed simultaneously (up to 50), but
each sequence must have a distinct name and belong to the
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 611

Fig. 1. IMGT/DomainGapAlign Welcome page.

same domain type. If not, the query needs to be launched for

each domain type, successively. If the limits and the numbers of
domains of an amino acid sequence are unknown, you can pro-
gressively analyze the protein, shortening the sequence once a
domain has been identified by the tool (it should be reminded
that the first domain identified by the tool is not necessarily the
first one in the protein). Sequences to test IMGT/
DomainGapAlign are available by clicking on the link “here.”
3. In the “Select a Domain type” drop-down list, select a domain
type V, C, or G (detailed in Subheading 1.2.1).
4. In the “Select a Species” drop-down list, select a Species. If the
selection is “All species,” the IMGT/DomainGapAlign tool
will compare your sequence with all sequences available for the
612 F. Ehrenmann and M.-P. Lefranc

selected domain type in the IMGT reference directory (detailed

in Subheading 1.2.2).
5. In the “Smith-Waterman score above” drop-down list, you can
modify the Smith-Waterman score for the alignments to dis-
play. Selecting a higher score corresponds to a higher selection
of the results to display. For example, choosing a “Smith-
Waterman score above 200” will only provide and display
alignments for which the Smith-Waterman score is superior to
200 (see Note 3).
6. In the “Displayed alignments” drop-down list, select the num-
ber of alignments to display (by default three) and tick off the
checkbox if you want to display IMGT Colliers de Perles.
7. Check the “Advanced parameters” section if you would like to
modify parameters. For alignment, IMGT/DomainGapAlign
use parameters by default: The E-value (see Note 4) is set to
200, the “Gap penalty for query” (relative to the user sequence)
is −5, and the “Gap penalty for reference” (relative to the
IMGT reference sequence) is −20 (see Note 5). These param-
eters can be modified for special queries.
8. Press the “Align and IMGT-gap my sequence(s)” button
to launch the analysis. This will return the IMGT/
DomainGapAlign results page.

2.2. IMGT/Domain- The IMGT/DomainGapAlign comprises three parts: the top of

GapAlign Results Page the results page, the Results summary and AA changes, and the
IMGT Colliers de Perles.

2.2.1. Top of the Results The top of the results page for V domain (Fig. 2), C domain
Page (Fig. 3), and G domain (Fig. 4) displays:
1. The “Sequence name” (as provided by the user).
2. The “Closest reference gene and allele(s) from the IMGT domain
directory” section. The domain type (V, C or G) and the species
as selected by the user are indicated in the section title (online in
orange). The following results are displayed: “Gene and allele
name” (see Note 2), “Species,” “Domain number,” “Smith-
Waterman Score” (online in orange), label of the domain (online
in color) as identified in the closest reference gene and allele
domain with its “percentage of identity” and “Overlap” score. If
several closest gene and alleles are displayed, the user can select
“Align your sequence with” to display the corresponding align-
ment. For V-DOMAIN (Fig. 2a), the closest reference gene and
alleles section shows the results for the V-REGION and
J-REGION of the V and J genes and alleles, respectively.
3. The Alignment(s) with the domain of the closest gene and
allele from the IMGT domain directory. The domain type (V,
C or G) and the species as selected by the user as indicated in
 33

Fig. 2. IMGT/DomainGapAlign top of the results page for a V domain. (a) V-DOMAIN. The user amino acid (AA) sequence is aligned with the closest germline V-REGION and J-REGION,
with IMGT gaps and delimitations of the FR-IMGT and CDR-IMGT according to the IMGT unique numbering (8, 9). In this example, the user sequence is the V-DOMAIN of the heavy
chain (VH) of the monoclonal antibody (mAb) alemtuzumab (2). The V-REGION and J-REGION of the alemtuzumab (2) VH is identified as having 73 and 92.9% identity with the Homo
IMGT/DomainGapAlign: The IMGT® Tool for the Analysis…

sapiens IGHV4-59*01 and IGHJ4*01, respectively. The alemtuzumab domain, an antibody humanized before the IMGT standardized delimitations of the FR-IMGT and CDR-IMGT,
shows 14 AA changes in the FR-IMGT, which most probably explains the immunogenicity of that antibody. (b) V-LIKE-DOMAIN. The CD58 domain is a V-LIKE-DOMAIN. The domain AA
sequence of the 3D structure 1ci5 from IMGT/3Dstructure-DB (2) shows four AA changes compared to the CD58*01 allele.
613
614 F. Ehrenmann and M.-P. Lefranc

Fig. 2. (continued)
 33

Fig. 3. IMGT/DomainGapAlign top of the results page for a C domain. (a) C-DOMAIN. The C domain of an IG lambda chain (or C-LAMBDA) is a C-DOMAIN. The domain AA sequence of
the 3D structure 1aqk from IMGT/3Dstructure-DB (2) shows 3 AA changes with the closest allele IGLC3*03. (b) C-LIKE-DOMAIN. The beta2 microglobulin domain is a C-LIKE-DOMAIN.
The AA sequence of the 3D structure 2d4d from IMGT/3Dstructure-DB (2) shows 3 AA changes compared to the B2M*01 allele.
IMGT/DomainGapAlign: The IMGT® Tool for the Analysis…
615
616 F. Ehrenmann and M.-P. Lefranc

Fig. 3. (continued)
 33

Fig. 4. IMGT/DomainGapAlign top of the results page for a G domain. (a) G-DOMAIN. The first G domain of a MH1 chain is a G-DOMAIN (more precisely G-ALPHA1) (11). In this example,
the user sequence MamuA is the G-ALPHA1 of a MH1 protein of the rhesus monkey Macaca mulatta (2) aligned with the G-ALPHA1 of the closest human HLA-A allele. The query was
with a selection of “Homo sapiens” as a species. (b) G-LIKE-DOMAIN. MICA is a member of the MhSF (11). The alignment identifies the closest domain as being [D2] (G-ALPHA2-LIKE)
of MICA*65 (24).
IMGT/DomainGapAlign: The IMGT® Tool for the Analysis…
617
618 F. Ehrenmann and M.-P. Lefranc

Fig. 4. (continued)
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 619

the section title (online in orange). The alignments are shown,

based on the IMGT unique numberings for V domain (7–9)
(Fig. 2), C domain (10) (Fig. 3), and G domain (11) (Fig. 4).
Below the alignment, the label of the domain (online in color),
as identified by the tool for the closest reference gene and
allele, is indicated with a horizontal line. For the V-DOMAIN
(Fig. 2a), the V-REGION, N (for (N-D)-REGION), and
J-REGION are delimited by IMGT/DomainGapAlign.
4. The Region(s) and domain(s) identified in your sequence (by
comparison with the closest genes and alleles). The species and
gene and allele name identified by IMGT/DomainGapAlign
are recalled (online in dark red). The domains (or regions for
the V-DOMAIN) identified by the tool are colored according
to the IMGT color menu.
5. The links to the sequence, with or without gaps, in FASTA
format (HTML page or downloadable).

2.2.2. Results Summary The Results summary (by comparison with the closest gene and
and AA Changes allele) and AA changes for V domain (Fig. 5), C domain (Fig. 6),
and G domain (Fig. 7) are shown as tables:
1. The Results summary table has three columns that are com-
mon to the V, C, and G domains; and four additional columns
for the V domain. The three common columns are:
– Sequence name.
– Domain (V-REGION for V-DOMAIN, V-LIKE-
DOMAIN, C-DOMAIN, C-LIKE-DOMAIN,
G-DOMAIN, G-LIKE-DOMAIN) identity percentage.
– Total number of AA changes in the domain.

The four additional columns for the V domain (Fig. 5)

are: “CDR-IMGT lengths,” “Number of different AA in
CDR1-IMGT and CDR2-IMGT,” “FR-IMGT lengths,” and
“Number of different AA in FR-IMGT” (V-DOMAIN)
(Fig. 5a). “Loop lengths,” “Number of different AA in loops,”
“Strand lengths,” and “Number of different AA in strands”
(V-LIKE-DOMAIN) (Fig. 5b).
2. The tables for AA changes are located below the Results sum-
mary and provide AA changes in:
– Strands and Loops (for V domain) (Fig. 5).
– Strands, Turns, and Loops (for C domain) (Fig. 6).
– Strands, Turns, and Helix (for G domain) (Fig. 7).
In these tables, IMGT amino acid (AA) changes are
described according to the IMGT AA hydropathy, volume,
and physicochemical classes (18) (see Note 6).
620 F. Ehrenmann and M.-P. Lefranc
 33

Fig. 5. IMGT/DomainGapAlign Results summary for a V domain. (a) V-DOMAIN. The results summary provides sequence name, V-REGION identity percentage, CDR-IMGT lengths,
number of different AA in CDR1- and CDR2-IMGT, FR-IMGT lengths, number of different AA in FR-IMGT, and total number of AA changes. AA changes are shown for strands and loops,
and for FR-IMGT and CDR-IMGT (8, 9). (b) V-LIKE-DOMAIN. The results summary provides sequence name, V-LIKE-DOMAIN, identity percentage, loop lengths, number of different AA
in loops, strand lengths, number of different AA in strands, and the total number of AA changes in V-LIKE-DOMAIN. Loop lengths are those of [BC.C¢C″.FG] that correspond to the
CDR-IMGT of the V-DOMAIN. Strand lengths are those of [(A + B).(C + C').(C″ + D + E + F).G] that correspond to the FR-IMGT of the V-DOMAIN (8, 9). AA changes are shown for strands
IMGT/DomainGapAlign: The IMGT® Tool for the Analysis…

and loops (9).

621
 622
F. Ehrenmann and M.-P. Lefranc

Fig. 6. IMGT/DomainGapAlign Results summary for a C domain. Results summary are similar for a C-DOMAIN or a C-LIKE-DOMAIN. A C-DOMAIN (1aqk_L_CL, same sequence as in
Fig. 3a) is shown as an example. The Results summary provides sequence name, C-DOMAIN (or C-LIKE-DOMAIN), identity percentage, and total number of AA changes. AA changes
are shown for strands (A, B, C, CD, D, E, F, and G), turns (AB, DE, EF), and loops (BC and FG) (10).
 33

Fig. 7. IMGT/DomainGapAlign Results summary for a G domain. Results summary are similar for a G-DOMAIN or a G-LIKE-DOMAIN. A G-DOMAIN (G-ALPHA, same sequence as in
Fig. 4a) is shown as an example. The results summary provides sequence name, G-DOMAIN (or G-LIKE-DOMAIN), identity percentage, and total number of AA changes. AA changes
are shown for strands (A, B, C and D), turns (AB, BC and CD), and helix (11).
IMGT/DomainGapAlign: The IMGT® Tool for the Analysis…
623
624 F. Ehrenmann and M.-P. Lefranc

Fig. 8. IMGT Collier de Perles for V domain. IMGT Collier de Perles on one and two layers are on the left hand side and right
hand side of the figure, respectively. IMGT Colliers de Perles on one and two layers with AA changes (online pink border)
are shown in the lower section of the figure. (a) V-DOMAIN. In an IG or TR V-DOMAIN, the G-STRAND (or FR4-IMGT) is the
C-terminal part of the J-REGION. The FR4-IMGT is at least composed of 9 or 10 amino acids beyond the phenylalanine F
(J-PHE 118) or tryptophan W (J-TRP 118) of the motif F/W-G-X-G that characterizes the J-REGION. The example is the VH
of alemtuzumab (as in Figs. 2a and 5a). The CDR-IMGT lengths are [8.10.12]. (b) V-LIKE-DOMAIN. The example is the
V-LIKE-DOMAIN of Homo sapiens CD58 (same sequence as in Figs. 2b and 5b). The loop lengths are [5.4.7].

2.2.3. IMGT Colliers de If selected in the IMGT/DomainGapAlign Welcome page (step 6

Perles in Subheading 2.1), IMGT Colliers de Perles on one and two
layers, and with or without amino acid (AA) changes are displayed.
IMGT Colliers de Perles for the V domain, C domain, and G
domain are shown in Figs. 8–10, respectively. In IMGT Colliers de
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 625

Fig. 8. (continued)

Perles for V and C domains (15–17), highlighted positions (online

in blue) mean that the AA of the user sequence at these positions
is hydrophobic (hydropathy index with positive value) or is a tryp-
tophan (W), like in 50% or more of analyzed V domains. Anchor
positions are in square (see Note 7). Hatched positions correspond
to gaps according to the IMGT unique numbering (7–10). Proline
are shown in yellow. Positions with bold (online red) letters indi-
cate the four conserved positions that are common to a V domain
and to a C domain: 1st-CYS 23, CONSERVED-TRP 41, hydro-
phobic 89, 2nd-CYS 104, and a fifth conserved position that is
specific to the V-DOMAIN: J-TRP or J-PHE 118 (Table 1).
626 F. Ehrenmann and M.-P. Lefranc

Fig. 9. IMGT Collier de Perles for C domain. IMGT Collier de Perles on one and two layers are on the left hand side and right
hand side of the figure, respectively. IMGT Colliers de Perles on one and two layers with AA changes (online pink border)
are shown in the lower section of the figure. (a) C-DOMAIN. The example is the Homo sapiens C-LAMBDA of 1aqk (1aqk_L_
CL) as in Figs. 3a and 6. (b) C-LIKE-DOMAIN. The example is the B2M C-LIKE-DOMAIN of 2d4d (same sequence as in
Fig. 3b).
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 627

Fig. 9. (continued)

The MH groove (16, 17) which binds peptide consists of

two G domains, belonging to the same chain (MH1-ALPHA) for
MH class I (MH1) or to two different chains (MH2-ALPHA and
MH2-BETA) for MH class II (MH2) (11). For the MhSF other
than MH (or RPI-MH1Like) (see Note 8), the two G domains
628 F. Ehrenmann and M.-P. Lefranc
 33

Fig. 10. IMGT Collier de Perles for G domain. IMGT/DomainGapAlign works per domain. IMGT Collier de Perles of the MH groove are shown separately (2). (a) G-DOMAIN. The example
is a MH1 groove. The domain G-ALPHA1 ([D1] of the MH1-ALPHA chain) is automatically located by the tool in the upper part of the IMGT Collier de Perles groove representation (the
example is G-ALPHA1 of MamuA, 3jtt_A_MamuA, left hand side of the figure), whereas the domain G-ALPHA2 ([D2] of the MH1-ALPHA chain) is automatically located in the lower
part of the groove (the example is G-ALPHA2 of MamuA, 3jtt_A_MamuA, right hand side of the figure). (b) G-LIKE-DOMAIN. The domain G-ALPHA1-LIKE ([D1] of the MH1LIKE-ALPHA
chain) is automatically located in the upper part (G-ALPHA1-LIKE, 1b3j_A_MICA, left hand side of the figure) where the domain G-ALPHA2-LIKE ([D2] of the MH1LIKE-ALPHA chain)
is automatically located in the lower part (G-ALPHA2-LIKE, 1b3j_A_MICA, right hand side of the figure) (same sequence as in Fig. 4b). Note that in IMGT/3Dstructure-DB (2), the two
IMGT/DomainGapAlign: The IMGT® Tool for the Analysis…

G domains are displayed together.

629
630 F. Ehrenmann and M.-P. Lefranc

also belong to the same chain (MH1LIKE-ALPHA) (11, 24). The

domain shown in the upper part of the IMGT Colliers de Perles
groove representation is, for MH1, G-ALPHA1 ([D1] of the
MH1-ALPHA chain), for MH2, G-ALPHA ([D1] of the MH2-
ALPHA chain) and for RPI-MH1Like, G-ALPHA1-LIKE ([D1]
of the MH1LIKE-ALPHA chain). The domain shown in the lower
part is for MH1, G-ALPHA2 ([D2] of the MH1-ALPHA chain),
for MH2, G-BETA ([D1] of the MH2-BETA chain), and for RPI-
MH1Like, G-ALPHA2-LIKE ([D2] of the MH1LIKE-ALPHA
chain) (Fig. 10).

3. Conclusion

IMGT/DomainGapAlign provides an invaluable help for amino

acid sequence analysis of V domain, C domain, and G domain
and bridges the gap between amino acid (and nucleotide)
sequences and 3D structures (26). IMGT/DomainGapAlign is
the first tool available for the analysis of IG and TR V-(D)-J
rearrangements starting from amino acid sequences. Indeed,
IMGT/DomainGapAlign is able to identify and to delimit pre-
cisely the V-REGION, (N-D)-REGION, and J-REGION and to
characterize the AA differences compared to the closest germline
V and J genes and alleles. IMGT/DomainGapAlign is essential
for standardized domain amino acid comparison and polymor-
phisms in molecular immunogenetics (19–21, 23, 24) and in
structural immunology for RPI.IgSF (19–21, 23) and MhSF
(22, 24) protein 2D structures, for IG/antigen (27, 28) and
TR/pMH (29–31) interaction analysis. IMGT/DomainGapAlign
is widely used in biotechnology related to antibody engineering
and humanization design (27, 28, 32, 33) based on CDR graft-
ing as it precisely defines the standardized FR-IMGT and CDR-
IMGT. IMGT/DomainGapAlign facilitates the identification of
potential immunogenic residues and amino acid polymorphism
at given positions in chimeric or humanized antibodies, including
amino acid polymorphisms of the constant domains (34, 35).

4. Notes

1. The IMGT reference directory comprises domain sequences of

functional (F), ORF (open reading frame), and in-frame
pseudogene (P) (Functionality is according to the IMGT
Scientific chart rules (http://www.imgt.org/IMGTScientific
Chart/ SequenceDescription/IMGTfunctionality.html), based
on the concepts of identification, generated from the
IDENTIFICATION axiom) (3–6).
33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 631

2. In IMGT, an allele is a polymorphic variant of a gene, which is

characterized by the mutations of its sequence at the nucle-
otide level, identified in its core coding sequence and com-
pared to the gene reference sequence designated as allele *01.
Identical sequences at the amino acid level may therefore cor-
respond to different alleles in the IMGT domain reference
directory.
3. The Smith-Waterman algorithm is a well-known algorithm for
performing local sequence alignment (36). The algorithm
determines identical regions between two nucleotide or pro-
tein sequences. Instead of looking at the complete sequence,
the Smith-Waterman algorithm compares segments of all pos-
sible lengths and optimizes the similarity measure. The higher
the score, the better the alignment in a series of results.
Selecting a higher score in IMGT/DomainGapAlign corre-
sponds to a display of higher sequence similarities.
4. The Expect value (E-value) is a parameter that describes the
number of hits one can “expect” to see by chance, for a given
score of the match, when searching a database of a particular
size. Each hit is associated to a score and an E-value. For exam-
ple, an E-value of 1 assigned to a match “means” that in a
database of a particular size one can expect 1 hit with a similar
score simply by chance. The lower the E-value, or the closer it
is to zero, the more “significant” the match is. A way of mea-
suring the significance of a score of an alignment is to consider
its E-value. Decreasing the E-value parameter for the align-
ment corresponds to a higher selection of the results to
display.
5. The alignment is performed by a modified Smith-Waterman
algorithm that considers the IMGT gap as a full amino acid
and which discriminates the creation of gaps in the IMGT ref-
erence sequence. The asymmetry management of insertions
between the user query and the reference sequences allows the
user to modify one or the other gap penalty.
6. The amino acid (AA) changes are described according to the
IMGT AA hydropathy, volume, and physicochemical classes
(18). For example, Q1 > E (++−) means that in the AA change
(Q > E), the two amino acids belong to the same hydropathy
(+) and volume (+) classes but to different physicochemical
properties (−) classes. Four types of AA changes are identified
in IMGT: very similar (+++), similar (++−, +−+), dissimilar
(−−+, −+−, +−−) and very dissimilar (−−−).
7. Anchors are positions that belong to strands and represent
anchors for the loops of the V and C domains (and by exten-
sion to the transverse CD strand of the C domain that does not
have the C¢-C″ loop). Positions 26 and 39 are anchors of
the BC-LOOP (CDR1-IMGT in V-DOMAIN). Positions 55
632 F. Ehrenmann and M.-P. Lefranc

and 66 are anchors of the C¢-C″ loop (CDR2-IMGT in

V-DOMAIN), whereas positions 45 and 77 are anchors of the
CD-STRAND of the C domain. Positions 104 and 118 are
anchors of the FG-LOOP (CDR3-IMGT in V-DOMAIN).
Anchor positions are shown as squares in IMGT Colliers de
Perles.
8. MhSF proteins other than MH only include RPI-MH1Like
proteins (there is no “RPI-MH2Like” identified so far) (11,
22, 24). The RPI-MH1Like proteins in humans comprise:
AZGP1, CD1A to CD1E, FCGRT, HFE, MICA and MICB,
MR1, PROCR (previously EPCR), RAET1E, RAETG, and
RAET1L (22, 24).

References
1. Lefranc M-P, Giudicelli V, Ginestoux C, receptors and Ig-like domains. Immunologist
Jabado-Michaloud J, Folch G, Bellahcene F, 7:132–136
Wu Y, Gemrot E, Brochet X, Lane J, Regnier 9. Lefranc M-P, Pommié C, Ruiz M, Giudicelli V,
L, Ehrenmann F, Lefranc G, Duroux P (2009) Foulquier E, Truong L, Thouvenin-Contet V,
IMGT®, the international ImMunoGeneTics Lefranc G (2003) IMGT unique numbering
information system®. Nucleic Acids Res for immunoglobulin and T cell receptor vari-
37:D1006–D1012 able domains and Ig superfamily V-like
2. Ehrenmann F, Kaas Q, Lefranc M-P (2010) domains. Dev Comp Immunol 27:55–77
IMGT/3Dstructure-DB and IMGT/ 10. Lefranc M-P, Pommié C, Kaas Q, Duprat E,
DomainGapAlign: a database and a tool for Bosc N, Guiraudou D, Jean C, Ruiz M, Da
immunoglobulins or antibodies, T cell recep- Piedade I, Rouard M, Foulquier E, Thouvenin
tors, MHC, IgSF and MhcSF. Nucleic Acids V, Lefranc G (2005) IMGT unique numbering
Res 38:D301–D307 for immunoglobulin and T cell receptor con-
3. Giudicelli V, Lefranc M-P (1999) Ontology for stant domains and Ig superfamily C-like
immunogenetics: IMGT-ONTOLOGY. domains. Dev Comp Immunol 29:185–203
Bioinformatics 15:1047–1054 11. Lefranc M-P, Duprat E, Kaas Q, Tranne M,
4. Lefranc M-P, Giudicelli V, Ginestoux C, Bosc Thiriot A, Lefranc G (2005) IMGT unique
N, Folch G, Guiraudou D, Jabado-Michaloud numbering for MHC groove G-DOMAIN and
J, Magris S, Scaviner D, Thouvenin V, Combres MHC superfamily (MhcSF) G-LIKE-
K, Girod D, Jeanjean S, Protat C, Yousfi DOMAIN. Dev Comp Immunol 29:917–938
Monod M, Duprat E, Kaas Q, Pommié C, 12. Lefranc M-P, Lefranc G (2001) The immuno-
Chaume D, Lefranc G (2004) IMGT- globulin FactsBook. Academic Press, London,
ONTOLOGY for Immunogenetics and UK, pp 1–458
Immunoinformatics. In Silico Biol 4:17–29. 13. Lefranc M-P, Lefranc G (2001) The T cell
http://www.imgt.org receptor FactsBook. Academic Press, London,
5. Lefranc M-P (2004) IMGT-ONTOLOGY and UK, pp 1–398
IMGT databases, tools and web resources for 14. Giudicelli V, Chaume D, Lefranc M-P (2005)
immunogenetics and immunoinformatics. Mol IMGT/GENE-DB: a comprehensive database
Immunol 40:647–660 for human and mouse immunoglobulin and T
6. Duroux P, Kaas Q, Brochet X, Lane J, cell receptor genes. Nucleic Acids Res
Ginestoux C, Lefranc M-P, Giudicelli V (2008) 33:D256–D261
IMGT-Kaleidoscope, the formal IMGT- 15. Ruiz M, Lefranc M-P (2002) IMGT gene
ONTOLOGY paradigm. Biochimie 90: identification and Colliers de Perles of human
570–583 immunoglobulins with known 3D structures.
7. Lefranc M-P (1997) Unique database number- Immunogenetics 53:857–883
ing system for immunogenetic analysis. 16. Kaas Q, Lefranc M-P (2007) IMGT Colliers de
Immunol Today 18:509 Perles: standardized sequence-structure repre-
8. Lefranc M-P (1999) The IMGT unique sentations of the IgSF and MhcSF superfamily
numbering for immunoglobulins, T cell domains. Curr Bioinform 2:21–30
 33 IMGT/DomainGapAlign: The IMGT® Tool for the Analysis… 633

17. Kaas Q, Ehrenmann F, Lefranc M-P (2007) 27. Lefranc M-P (2009) Antibody databases and
IG, TR and IgSf, MHC and MhcSF: what do tools: The IMGT® experience. In: Zhiqiang An
we learn from the IMGT Colliers de Perles? (ed) Therapeutic monoclonal antibodies: from
Brief Funct Genomic Proteomic 6:253–264 bench to clinic (chapter 4). John Wiley & Sons,
18. Pommié C, Levadoux S, Sabatier R, Lefranc M-P Inc., Hoboken, NJ, USA, pp 91–114
(2004) IMGT standardized criteria for statistical 28. Ehrenmann F, Duroux P, Giudicelli V, Lefranc
analysis of immunoglobulin V-REGION amino M-P (2010) Standardized sequence and struc-
acid properties. J Mol Recognit 17:17–32 ture analysis of antibody using IMGT®. In:
19. Duprat E, Kaas Q, Garelle V, Lefranc G, Kontermann R, and Dübel S (eds) Antibody
Lefranc M-P (2004) IMGT standardization for engineering (chapter 2), vol 2, Springer-Verlab,
alleles and mutations of the V-LIKE-DOMAINs Berlin, Heidelberg, pp 11–31
and C-LIKE-DOMAINs of the immunoglob- 29. Kaas Q, Ruiz M, Lefranc M-P (2004)
ulin superfamily. In: Pandalai SG (ed) Recent IMGT/3Dstructure-DB and IMGT/Structural-
research developments in human genetics Query, a database and a tool for immunoglobu-
Research Signpost, Trivandram, Kerala, vol 2. lin, T cell receptor and MHC structural data.
pp 111–136 Nucleic Acids Res 32:D208–D210
20. Garapati VP, Lefranc M-P (2007) IMGT 30. Kaas Q, Lefranc M-P (2005) T cell receptor/
Colliers de Perles and IgSF domain standard- peptide/MHC molecular characterization and
ization for T cell costimulatory activatory standardized pMHC contact sites in
(CD28, ICOS) and inhibitory (CTLA4, IMGT/3Dstructure-DB. In Silico Biol 5:
PDCD1 and BTLA) receptors. Dev Comp 505–528
Immunol 31:1050–1072 31. Kaas Q, Duprat E, Tourneur G, Lefranc M-P
21. Bernard D, Hansen JD, du Pasquier L, Lefranc (2008) IMGT standardization for molecular
M-P, Benmansour A, Boudinot P (2005) characterization of the T cell receptor/
Costimulatory receptors in jawed vertebrates: peptide/MHC complexes. In: Schoenbach C,
conserved CD28, odd CTLA4 and multiple Ranganathan S, Brusic V (eds) Immunoin-
BTLAs. Dev Comp Immunol 31:255–271 formatics. Immunomics Reviews, Series of
22. Duprat E, Lefranc M-P, Gascuel O (2006) A Springer Science and Business Media LLC
simple method to predict protein binding from (chapter 2). Springer, New York, USA,
aligned sequences-application to MHC super- pp 19–49
family and beta2-microglobulin. Bioinformatics 32. Lefranc M-P (2009) Antibody databases:
22:453–459 IMGT®, a French platform of world-wide
23. Bertrand G, Duprat E, Lefranc M-P, Marti J, interest. Bases de données anticorps: IMGT®,
Coste J (2004) Characterization of human une plate-forme française d’intérêt mondial
FCGR3B*02 (HNA-1b, NA2) cDNAs and “Anticorps monoclonaux en thérapeutique”.
IMGT standardized description of FCGR3B Med Sci 25:1020–1023
alleles. Tissue Antigens Cover 64:119–131 33. Lefranc M-P (2011) Antibody nomenclature:
24. Frigoul A, Lefranc M-P (2005) MICA: standard- from IMGT-ONTOLOGY to INN definition.
ized IMGT allele nomenclature, polymorphisms MAbs 3:1–2
and diseases. In: Pandalai SG (ed) Recent research 34. Jefferis R, Lefranc M-P (2009) Human immu-
developments in human genetics, vol 3. Research noglobulin allotypes: possible implications for
Signpost, Trivandrum, Kerala, pp 95–145 immunogenicity. MAbs 1(4):332–338
25. Brochet X, Lefranc M-P, Giudicelli V (2008) 35. Lefranc M-P, Lefranc G (2012) Human Gm,
IMGT/V-QUEST: the highly customized and Km and Am allotypes and their molecular
integrated system for IG and TR standardized characterization: a remarkable demonstration
V-J and V-D-J sequence analysis. Nucleic Acids of polymorphism. In: Tait B, Christiansen F
Res 36:W503–W508 (eds) Methods in Molecular Biology –
26. Lefranc M-P, Giudicelli V, Regnier L, Duroux P Immunogenetics, Chapter 34
(2008) IMGT®, a system and an ontology that 36. Smith TF, Waterman MS (1981) Identification
bridge biological and computational spheres in of common molecular subsequences. J Mol
bioinformatics. Brief Bioinform 9:263–275 Biol 147(1):195–197
 Chapter 34

Human Gm, Km, and Am Allotypes and Their Molecular

Characterization: A Remarkable Demonstration
of Polymorphism
Marie-Paule Lefranc and Gérard Lefranc

Abstract
Human immunoglobulin allotypes are antigenic determinants (or “markers”) determined serologically,
classically by hemagglutination inhibition, on the human immunoglobulin (IG) heavy and light chains.
The allotypes have been identified on the gamma1, gamma2, gamma3, and alpha2 heavy chains (they are
designated as G1m, G2m, G3m, and A2m allotypes, respectively), and on the kappa light chain (Km allo-
types). Gm-Am allotypes are inherited in fixed combinations, or Gm-Am haplotypes, owing to the linkage
of the human IGHC genes (IGHG3, IGHG1, IGHA1, IGHG2, IGHG4, IGHE, and IGHA2 from 5¢ to
3¢ in the IGH locus on chromosome 14). Gm and Am allotypes have been one of the most powerful tools
in population genetics and very instrumental in molecular characterization of the human IGHC genes
(gene conversion, copy number variation, gene order). They represent a major system for understanding
immunogenicity of the polymorphic IG chains, in relation with amino acid and conformational changes.
The correlation between G3m allotypes and amino acid changes has been possible with the sequencing of
many alleles of the IGHG3 gene, from individuals from different populations and with known allotypes.
In this chapter, we integrate genetics and sequence data and provide an updated overview of the Gm-Am
haplotypes and Km allotypes. We propose, for the first time, a complete elucidation of the G3m allotypes,
illustrated by the “IMGT G3m allele butterfly” concept that allows a graphical representation of the
G3m alleles (variants of a gene expressing a given set of allotypes). Knowledge of allotypes is important in
antibody engineering and humanization of monoclonal antibodies to improve immunotherapy.

Key words: IMGT, Immunogenetics, Allotype, Haplotype, Gm, Km, Am, Gm-Am, Immuno-
informatics, Immunoglobulin, Antibody, Allelic polymorphism, IMGT-ONTOLOGY

1. Introduction

Allotypes are allelic antigenic determinants identified in humans on

the immunoglobulin (IG) gamma1, gamma2, gamma3, and alpha2
heavy chains (they are designated as G1m, G2m, G3m, and A2m allo-
types, respectively), and on the kappa light chain (Km allotypes) (1).

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9_34, © Springer Science+Business Media New York 2012

635
636 M.-P. Lefranc and G. Lefranc

Recently, allotypes regained a lot of attention, owing to the

development of therapeutical monoclonal antibodies (2) and their
potential immunogenicity (3, 4). IMGT®, the international
ImMunoGeneTics information system® (http://www.imgt.org)
(5, 6), was built on the concepts of IMGT-ONTOLOGY (7–9),
providing a standardized characterization of the allotypes. The
concepts of classification (nomenclature of genes and alleles
(10–13)) and the concepts of numerotation (IMGT unique num-
bering for constant (C) domains (14)) are used in the IMGT gene
database, IMGT/GENE-DB (15), the three-dimensional (3D)
database, IMGT/3Dstructure-DB (16, 17), the IMGT online tools,
IMGT/Collier-de-Perles (18, 19), IMGT/DomainGapAlign (17),
and the IMGT Repertoire Web resources (“Protein displays,”
“Alignments of alleles,” “Allotypes,” etc.) (5, 6).
The first allotype was identified by Grubb in 1956 (20, 21),
and for a period of 20 years, Gm, Am, and Km allotypes were dis-
covered and characterized (20–46). Currently, 26 allotypes are
known. Gm and Am allotypes are inherited in fixed combinations,
or Gm-Am haplotypes (47–53), owing to the linkage of the cor-
responding encoding human IG heavy constant (IGHC) genes
(IGHG3, IGHG1, IGHG2, IGHA2, respectively, from 5¢ to 3¢ in
the IGH locus on chromosome 14) (1, 11). This system is unique
in its ability to characterize human populations by specific sets of
haplotypes, and for years, the Gm haplotypes have been the most
powerful tools for the characterization of different populations
(47–87). The Gm system is very instrumental for population
genetics research, as it is highly polymorphic and unbalanced with
respect to linkage. Interestingly also, it is one of the cheapest and
one of the most informative systems. Vast screenings of human
populations worldwide have uncovered considerable variability
both in the contents of Gm haplotypes and in their frequencies,
which enabled the investigators to examine valuable data with
regard to gene admixture, genogeography, ethnoanthropology,
evolutionary, and population genetics. It is one of the most power-
ful tools to characterize the polymorphism between individuals.
Prior to the development of DNA fingerprinting techniques, Gm
haplotypes were used for follow-up of bone marrow transplants
(88), forensic medicine, and paternity testing.
Gm analysis has thrown light on the molecular characterization
and evolution of the human IGHC genes (1, 47–53, 89–113),
including evidence for “silent” IGHG3 on three exceptional haplo-
types, in seven homozygous individuals in two related Lebanese and
Syrian families (48), gene conversion (100, 102, 103, 106), gene
duplications and deletions, and gene copy number variations (CNV)
(97–99, 102, 105, 107, 108, 110, 111). It is worthwhile to note
that it is the absence of the G1m allotypes in a healthy consanguine-
ous Tunisian woman that led to the first description of a large dele-
tion of 150 kilobases encompassing several IGHC genes on both
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 637

chromosomes 14 and of the simultaneous and complete absence of

the IgG1, IgG2, IgG4 and IgA2 subclasses in a healthy individual
(97–99). This multigene deletion contributed to the identification
of the order of the human IGHC genes (114, 115) and was the first
demonstration of a CNV of the IGHC genes in humans (97–99).
Correlations between Gm and restriction fragment length poly-
morphism (RFLP) have been carried out for haplotypes from
diverse populations (1, 116–122). In this chapter, we first present
an overview on the basic information on allotypes and isoallotypes,
then we describe Gm, Am, and Km allotypes with their localization
in the C domains and correlation with amino acid changes, and
Gm, Am, and Km alleles with their correspondence to the IGHG,
IGHA, and IGKC gene alleles. The first complete molecular char-
acterization of the G3m alleles, illustrated by the “IMGT G3m
allele butterfly” representation, is provided. This information
confirms the importance of conformational configuration in the
expression of allotypes and will be useful for understanding immu-
nogenicity of therapeutical antibodies that may be used with various
isotypes and allotypes in different populations.

2. Basic
Information
on Allotypes
and Isoallotypes Allotypy within IgG was first described by Grubb who showed
that certain human sera would agglutinate erythrocytes sensitized
2.1. Human IG Allotype with human “incomplete” anti-Rh antibody (20, 21) (see Note 1).
Discovery Polymorphism of the C region of human IG heavy gamma and
alpha and light kappa chains was subsequently recognized by sero-
logical typing using the classical reaction of inhibition of hemag-
glutination (20–46). Thus, the discovery of this polymorphism
demonstrated that exposure of an individual to IgG or IgA of a
nonself allotype can induce an anti-allotype response.

2.2. Gm, Am, and Km Allotypes of IG are unique antigenic determinants recognized by
Allotype and Allele specific antibodies. They are IG markers that, in terms of immuno-
Definition genicity, represent B cell epitopes. Allotypes correspond to sero-
logically detected amino acid changes that characterize the
polymorphism of a chain within a given isotype. By definition, allo-
types are shared among individuals within populations. Allotypes
have been identified on the C region of the human IG heavy
gamma1, gamma2, gamma3, and alpha2 chains of the IgG1, IgG2,
IgG3, and IgA2 subclasses, respectively, and on the C region of the
human IG light kappa chains (see Note 2). They are designated as
Gm (Gamma marker), Am (Alpha marker), and Km (Kappa
marker), with a number for the subclass: G1m, G2m, and G3m for
allotypes of the gamma1, gamma2, and gamma3 chains, and A2m
for allotypes of the alpha2 chains (1).
638 M.-P. Lefranc and G. Lefranc

A combination of allotypes, encoded by one or several alleles

of a given gene (IGHG1, IGHG2, IGHG3, IGHA2, or IGKC)
(see Note 3), is defined as a G1m, G2m, G3m, A2m, or Km allele,
respectively.

2.3. Gm-Am Haplotype Gm allotypes are inherited in fixed combinations called “Gm haplo-
Definition types” (see Subheading 7) or “Gm-Am haplotypes” (if A2m allo-
types are tested), owing to the linkage of the IGHC genes within a
cluster in the IGH locus (47–53). Haplotypes have a low frequency
of crossovers; however, crossover events and gene conversions
(100, 102, 103, 106) have occurred during evolution resulting in
characteristic haplotypes present in diverse populations, hence the
usefulness of the allotype system in population studies. Equal or
unequal crossovers, the later generating gene duplication (or
expansion) or, in contrast, gene deletion (or contraction) in the
IGH locus has been demonstrated (97–99, 102, 105, 107, 108,
110, 111) and they are now commonly named CNV.

2.4. Allotype Allotypes are determined by serological typing using a classical reac-
Determination tion of inhibition of hemagglutination. The methodology uses
human O Rh+ erythrocytes (red blood cells) sensitized (coated)
2.4.1. Hemagglutination
with “incomplete” anti-Rh IgG antibodies of known allotypes
Inhibition Methodology
(e.g., G1m1), and human reagents (polyclonal IgG specific for a
given allotype, e.g., anti-G1m1). The polyclonal IgG reagents are
obtained from multiparous women, multiple transfused individuals,
and normal blood donors. If the tested serum is G1m1-negative, the
anti-G1m1 reagent binds the G1m1-positive antibodies coating
the erythrocytes and the hemagglutination occurs. In contrast, if
the tested serum is G1m1-positive, the anti-G1m1 reagent binds the
G1m1-positive antibodies contained in the serum, and the hemag-
glutination is inhibited. The reactions are performed with different
dilutions of the reagents and of the tested sera (47, 48, 52, 53). The
determination of some allotypes (G2m23, A2m1, and A2m2) is
more delicate owing to the rarity of the reagents and, in particular,
the absence of available anti-Rh antibodies for the coating. In those
cases, myeloma proteins are coupled using chromic chloride (123).
It is highly recommended to provide the list and source of the anti-
Rh, myeloma proteins, polyclonal IgG, and eventually their dilu-
tions, the list of tested allotypes, and the number of individuals tested
for each allotype, for allowing standardized result comparison.

2.4.2. Phenotype Positive results (corresponding to a hemagglutination inhibition)

and Genotype Deduction and negative results (corresponding to a hemagglutination)
obtained for the tested allotypes allow assignment of the “Gm
phenotype” (or “Gm-Am phenotype,” if the A2m allotypes are
tested) of an individual (see Subheading 7.3). In populations
where the main Gm (or Gm-Am) haplotypes are already known,
it is usually possible to deduce, from the phenotype the Gm
(or Gm-Am) genotype that is the two haplotypes that contribute
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 639

to it (47–53). However, it is not always possible to deduce the

genotype. In those cases, familial studies with allotype typing of
several members are needed.

2.4.3. Other Methodologies To compensate for the rarity of some reagents, attempts were made
to obtain monoclonal antibodies; however, the main difficulty
resides in the characterization of their specificity (124, 125).
Molecular biology was first used for the determination of A2m2 by
RFLP (101) (see Subheading 8.3). This method is particularly
interesting given the rarity of the reagents and was the first proto-
col to determine allotypes in the absence of serum (e.g., from cell
lines). Polymerase chain reaction (PCR) amplification methods,
using allele-specific oligonucleotides (ASO) or specific restriction
sites, were subsequently developed for the determination of the
Km allotypes (126) and of some Gm allotypes (4, 127–130), but
their application remains limited.

2.5. Allotype At present, 26 human allotypes are known, 20 Gm (18 “classical”

Characterization ones and two “surnumerary” ones), three A2m, and three Km
allotypes (Table 1). The 18 “classical” Gm allotypes comprise four
2.5.1. Allotype
G1m, G1m (1, 2, 3, 17), one G2m, G2m (23), and thirteen G3m,
Nomenclature
G3m (5, 6, 10, 11, 13, 14, 15, 16, 21, 24, 26, 27, 28). The two
“surnumerary” allotypes, G1m27 and G1m28, correspond to allo-
types demonstrated to be on gamma1 chains in Negroid popula-
tions, instead of being on gamma3, as expected (96). The A2m
and Km comprise A2m (1, 2, 3) and Km (1, 2, 3), respectively.

2.5.2. Allotype Localization The localization of the allotypes has been determined by inhibition
studies with Fab and Fc fragments (obtained by papain digestion),
and with pF’c fragments (obtained by pepsin digestion) of
Gm-positive or Gm-negative myeloma proteins (131–136). Amino
acid sequence analysis of peptides obtained from Fd (part of the heavy
chain from a Fab) and from pF’c have revealed amino acid changes,
allowing establishment of correlations between serologically defined
allotypes and amino acid sequences. Except for G1m3 and G1m17
located on the CH1 of gamma1, all other Gm allotypes are localized
on the Fc (on CH2 or on CH3) (Table 1). Thus, for example, the
G3m21 allotype was detectable on the Fc fragment of gamma3 chains
but not on isolated CH3 domains, and therefore was localized on the
CH2 (134). G3m10, G3m11, and G3m13 were localized on the
CH3 domain of the gamma3 chain (134). G3m28 was localized on
the CH3 domain of a G3m28 myeloma protein (38). However, the
first detailed correlation between G3m allotypes and amino acid
changes has only been possible following the complete nucleotide
sequencing of many IGHG3 alleles from individuals homozygous for
well-characterized G3m alleles (104, 106, 113). G3m allotypes and
their localization, and correspondence with G3m alleles, illustrated
by the “IMGT G3m allele butterfly” representation, are further
defined in this chapter (see Subheading 5.2).
640 M.-P. Lefranc and G. Lefranc

Table 1
Nomenclature of the Gm and Km allotypes

Localizationa Nomenclature

IG heavy Domain WHO Previous

chain nomenclatureb designation
H-GAMMA1 CH3 G1m1 G1m(a) Grubb (20), Grubb and Laurell (21)
CH3 G1m2 (x) Harboe and Lundevall (22)
CH1 G1m3 (f) Steinberg and Wilson (23), Gold et al.
(24, 25)
CH1 G1m17 (z) Litwin and Kunkel (26)
CH3 G1m27c van Loghem et al. (96)
c
CH3 G1m28 van Loghem et al. (96)
H-GAMMA2 CH2 G2m23 G2m(n) Kunkel et al. (27)
H-GAMMA3 CH3 G3m5 G3m(b1) Harboe (28)
CH3 G3m6 (c3) Steinberg et al. (29)
CH3 G3m10 (b5) Ropartz et al. (30)
CH3 G3m11 (b0) Ropartz et al. (30)
CH3 G3m13 (b3) Steinberg and Goldblum (31)
CH3 G3m14 (b4) Steinberg and Goldblum (31)
CH3 G3m15 (s) Martensson et al. (32)
CH2 G3m16 (t) Martensson et al. (32)
CH2 G3m21 (g1) Natvig (33)
CH3 G3m24 (c5) van Loghem and Martensson (34)
CH3 G3m26 (u) van Loghem and Grobbelaar (35)
CH3 G3m27d (v)d Schanfield and Fudenberg (36)
d d
CH3 G3m28 (g5) Blanc et al. (37)
H-ALPHA2 CH1 A2m1 A2m1 Vyas and Fudenberg (39), Kunkel et al. (40)
CH1 A2m2 A2m2 van Loghem et al. (41)
CH3 A2m3e
(continued)
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 641

Table 1
(continued)

Localizationa Nomenclature

IG light Domain WHO Previous

chain nomenclatureb designation
L-KAPPA C-KAPPA Km1 Km1 Ropartz et al. (43)
C-KAPPA Km2 Km2 Ropartz et al. (42)
C-KAPPA Km3 Km3 Steinberg et al. (44)
a
IMGT® labels for chains and domains are written in capital letters (concepts of description of IMGT-ONTOLOGY) (7, 8)
b
The World Health Organization (WHO) nomenclature (45, 46) has been adopted by the WHO/International Union
of Immunological Societies (IUIS)/IMGT nomenclature (12, 13) (see Note 4)
c
Observed in Negroid populations (96) and in rare unusual haplotypes with “surnumerary” Gm27 and Gm28 in other
populations (52, 53, 94, 95, 102). G1m27 and G1m28 are detected with the same reagents as G3m27 and G3m28 (see
Subheading 5.1.7)
d
In Caucasoid and Mongoloid populations, Gm27 and Gm28 are on gamma3 (G3m27 and G3m28); however, in rare
unusual haplotypes, “surnumerary” Gm27 and Gm28 may be explained by the presence of G1m27 and G1m28 (52,
53, 94, 95, 102). In Negroid populations, Gm27 can be on the gamma3 chains (G3m27), but also on the gamma1
chains (G1m27), Gm28 is always “surnumerary” and on the gamma1 chains (G1m28) (96)
e
Identified with TOU II-5 sera (IGHA2*02 allele) (98, 100, 101) (unpublished data, van Loghem, GL and MPL)
(see Subheading 8.5)

Table 2
Nomenclature and distribution of the isoallotypes

Nomenclature Distribution on IG heavy chains

WHO Previous
nomenclature designation gamma1 gamma2 gamma3 gamma4 alpha1 alpha2
nG1m1 nG1m(a) Allo Iso Iso Isoa – –
nG1m17 nG1m(z) Allo – Iso Iso – –
nG3m5 nG3m(b1) Isob Iso Allo – – –
nG3m11 nG3m(b0) Iso Iso Allo – – –
nG3m21 nG3m(g) – Iso Allo – – –
nG4m(a) nG4m(a) Iso – Iso Allo – –
nG4m(b) nG4m(b) – Iso – Allo – –
nA2m3c nA2m(2) – – – – Iso Allo
a
Isoallotype only detected by some antisera
b
In Negroid populations, the G1m28 allotype can be expressed instead of nG3m5 (see Subheading 5.1.7)
c
nA2m3 has been renamed as this isoallotype is located on the CH3 domain and is antithetical to the A2m3 allele
(it is not antithetical to A2m2 located on the CH1) (see Subheading 8.5)
642 M.-P. Lefranc and G. Lefranc

2.6. Isoallotypes By definition, allotypes are found on chains within one IG isotype
(encoded by one given IG gene). However, the same amino acids
may be found in chains of other isotypes (encoded by other IG
genes), but without being polymorphic in these isotypes. If these
amino acids are detected in vitro by antibody reagents, they are
referred as “isoallotypes” (designated with the letter “n” preced-
ing the allotype name, e.g., nG1m1). Seven isoallotypes have
been identified for the gamma chains (137–141) and one for the
alpha chains identified on the CH3 domain (142) (Table 2) (see
Note 5).

2.7. Protein Displays Protein displays are standardized IMGT representations of the amino
and IMGT Colliers de acid sequences of the coding regions of the genes (11). Protein dis-
Perles plays of C domains show the sequences per domain, using the IMGT
unique numbering for C domain (14). They allow a standardized
localization of the amino acids involved in the allotypes in relation to
the strands and loops of the domains. Figure 1 provides the Protein
displays of the C domains (CH1, CH2, and CH3) of the IGHG and
IGHA genes (5, 6, 17). Only the first allele is shown. Other alleles are
available in IMGT/DomainDisplay (5, 6) and in IMGT/GENE-DB
(15) (IMGT®, http://www.imgt.org). IMGT/DomainGapAlign
(17) allows gaps to be inserted in C domains of the IGHG, IGHA,
and IGKC genes according to the IMGT unique numbering (14).
Correspondences with other numberings are available in the IMGT
Scientific chart (http://www.imgt.org). Standardized representa-
tions or IMGT Colliers de Perles of the C domains can be obtained
using the IMGT/Collier-de-Perles tool (18, 19).

3. G1m

3.1. G1m Allotypes G1m1 (previously G1m(a)) was the first discovered allotype (20).
and nG1m Allotypes In 1956, Grubb noticed that 60% of normal human sera could
inhibit the agglutination of human O Rh+ erythrocytes sensitized
3.1.1. G1m1 and nG1m1
by means of certain “incomplete” anti-Rh sera, the factor respon-
sible for the inhibition was called Gm(a) (now G1m1 allotype).
Grubb and Laurel demonstrated that Gm(a) was transmitted as a
dominant autosomal Mendelian trait (21).
By inhibition studies with IgG1 fragments, it was established
that G1m1 is located on the CH3 domain and was associated with
aspartate 356 and leucine 358 (Eu numbering) (131). According
to the IMGT unique numbering for C domain (14), the G1m1
allotype corresponds to IGHG1 CH3 Asp D12 and Leu L14
(Table 3) (Figs. 2 and 3).
In the G1m1-negative gamma1 chains and in the gamma
chains of the other IgG subclasses, glutamate (E) and methionine
(M) are found, respectively, at positions 12 and 14 of the CH3
 34
Human Gm, Km, and Am Allotypes and Their Molecular Characterization…

Fig. 1. Protein display of the human IGHG and IGHA constant (C) domains. (a) Human IGHG CH domains. (b) Human IGHA CH domains. Only allele *01 is shown (see Note 3). Other
alleles are available in IMGT/DomainGapAlign (5, 6) and in IMGT/GENE-DB (15) (IMGT®, http://www.imgt.org). The alignments are based on the IMGT unique numbering for C domain
(14). Hinge and CHS regions are not shown. Domains are numbered with [D1] being the variable domain (not shown).
643
 Table 3
G1m allotypes and isoallotypes

Amino acid positionsa

G1m allotypes
and isoallotypes Domain CH1 CH3b
IMGT 120 12 14 110
Eu 214 356 358 431
G1m1 Asp D12 Leu L14
nG1m1 Glu E12 Met M14
G1m2 Gly G110
G1m3 Arg R120 + Ileu I103
G1m17 Lys K120
nG1m17 Arg R120
a
Positions in bold are according to the IMGT unique numbering for C domain (14), and in italics, Eu numbering
(IMGT Scientific chart, http://www.imgt.org) (5, 6)
b
G1m27 and G1m28 found in Negroid populations (96) are not shown. G1m27 most probably corresponds to
IGHG1 CH3 Ileu I101, and G1m28 to IGHG1 CH3 Arg R115, Tyr Y116

Fig. 2. IMGT Collier de Perles of the IGHG1 CH3 domain. The CH3 domain is from the b12 antibody (IMGT/3Dstructure-DB,
code PDB:1hzh, http://www.imgt.org) (16, 17). The IMGT Collier de Perles is shown on two layers, with hydrogen bonds
shown as green lines online. Hatched positions correspond to gaps according to the IMGT unique numbering (14). The
aspartate D12 and leucine L14 (strand A) correspond to G1m1, whereas E12 and methionine M14 (not shown) correspond
to nG1m1. In G1m2-negative chain, as that of b12, there is an alanine at position 110. A glycine at position 110 would
correspond to G1m2. The amino acids glycine (G)-lysine (K) at positions 129 and 130 represent the CHS in secreted IG.
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 645

Fig. 3. Three-dimensional structure of the IGHG1 CH3 domain. The CH3 domain is from the
b12 antibody (IMGT/3Dstructure-DB, code PDB:1hzh, http://www.imgt.org) (16, 17).
Positions 12 and 14 of the G1m1/nG1m1 allotype, and position 110 of the G1m2/- allotype
in the IGHG1 CH3 domain are shown. The aspartate D12 and leucine L14 correspond to
G1m1, whereas alanine A110 corresponds to nG1m2. Glutamate E12 and methionine M14
correspond to nG1m1, whereas a glycine G110 corresponds to G1m2.

domain (131). These amino acid changes can be determined, by

specific antisera, on the G1m1-negative gamma1, gamma2, and
gamma3 chains (Table 2). This epitope corresponds to the isoal-
lotype nG1m1. The gamma4 chains also express CH3 E12 and
M14, but the corresponding epitope is only detected by certain
antisera (143). This restricted accessibility of the nG1m1 epitope
of the gamma4 chains has been correlated with the presence, at
position 11 of CH3, of a glutamine (Q) instead of an arginine (R)
as found in the other subclasses (11, 15) (Fig. 1a). In Old World
Monkeys (OWM) (144), the IGHG1 CH3 sequence contains E12
and L14 and it has therefore be postulated that two independent
single amino acid changes may have led to the G1m1 allotype
(CH3 E12 > D) and to the nG1m1 isoallotype (CH3 L14 > M).

3.1.2. G1m2 The G1m2 allotype was discovered in 1959 by Harboe and Lundevall
(22). The G1m2 allotype was detected on the CH3 domain (134).
It corresponds to a glycine at position 110 (IGHG1 CH3 Gly
G110), whereas the absence of the allotype correlates to alanine at
that position (145) (see Note 6) (Table 3) (Figs. 2 and 3).
646 M.-P. Lefranc and G. Lefranc

Fig. 4. IMGT Collier de Perles of the IGHG1 CH1 domain. The CH1 domain is from the b12 antibody (IMGT/3Dstructure-DB,
code PDB:1hzh, http://www.imgt.org) (16, 17). The IMGT Collier de Perles is shown on two layers, with hydrogen bonds
shown as green lines online. Hatched positions correspond to gaps according to the IMGT unique numbering (14). The
lysine at position 120 (K120) corresponds to the G1m17 allotype. The isoleucine I103 is specific for the gamma1 chain
isotype. If an arginine is expressed at position 120 (R120), the simultaneous presence of R120 and I103 corresponds to the
expression of the G1m3 allotype.

3.1.3. G1m3, G1m17, G1m3 was first identified by Steinberg and Wilson in 1963 (23),
and nG1m17 then further characterized by Gold et al. (24, 25). G1m3, located
on the CH1 domain, corresponds to an arginine at position 120
(IGHG1 CH1 Arg R120) (132) (Figs. 4 and 5). The expression of
G1m3 requires the presence of an amino acid specific for gamma1,
because the CH1 Arg R120 is also present on the gamma3 and
gamma4 chains (Fig. 1a). It seems likely from the analysis of 3D
structures of Fab and of the b12 antibody, the only complete
human IG so far crystallized (IMGT/3Dstructure-DB, code
PDB:1hzh) (16, 17) (see Note 7) that the isoleucine at position
103 (IGHG1 CH1 Ile I103) is the amino acid involved in the
expression of the G1m3 allotype (Figs. 4 and 5).
G1m17 (26), located in the CH1 domain, corresponds to
lysine at position 120 (IGHG1 CH1 Lys K120) (132). G1m17 is
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 647

Fig. 5. Three-dimensional structure of the IGHG1 CH1 domain. The CH1 domain is from the
b12 antibody (IMGT/3Dstructure-DB, code PDB:1hzh, http://www.imgt.org) (16, 17).
The lysine K120 (strand G) and the isoleucine I103 (strand F) are shown. The K120
corresponds to the G1m17 allotype. The simultaneous presence of I103 (specific of
the gamma1 isotype) and of arginine R120 corresponds to the G1m3 allotype. The R120
corresponds to the nG1m17 isoallotype.

mutually exclusive with (“antithetical to”) G1m3 and is present on

gamma1 chains which are G1m3-negative.
The isoallotype nG1m17 corresponds to arginine at position
120 in the CH1 (IGHG1 CH1 Arg R120). It is detectable on
isolated gamma3 and gamma4 chains where R120 is present but
without the gamma1-specific determinant CH1 Ileu I103 (threo-
nine T103 is found instead in the other subclasses) (Fig. 1a).

3.1.4. G1m27 and G1m28 G1m27 and G1m28 have only been demonstrated to be present
on the gamma1 chains in Negroid populations (96); however, it is
not excluded that these “surnumerary” allotypes may explain some
uncommon haplotypes found in other populations (52, 53, 94, 95,
102). G1m27 most probably corresponds to IGHG1 CH3 Ileu
I101 (resulting from an amino acid change V101 > I), and G1m28
to IGHG1 CH3 Arg R115, Tyr Y116 (with an amino acid change
H115 > R) (see Subheading 5.1.7).
648 M.-P. Lefranc and G. Lefranc

Fig. 6. G1m allotypes localizations on gamma1 chains. The CH1, CH2, and CH3 domains of the b12 gamma1 chains are
shown (IMGT/3Dstructure-DB, code PDB:1hzh, http://www.imgt.org) (16, 17), with the positions involved in the G1m
allotypes. The CH2 position 45.1 is not related to the G1m allotypes, but indicates the amino acid position that should be
responsible for the G2m23 allotype, or of its absence (G2m..), on a gamma2 chain.

3.2. Correspondence The heavy gamma1 chains of IgG1 may express four typical G1m
Between G1m Alleles alleles (combinations of G1m allotypes): G1m3, G1m3,1, G1m17,1,
and IGHG1 Alleles and G1m17,1,2 (and three additional G1m alleles, Gm17,1,27,
Gm17,1,28, and Gm17,1,27,28, the last two identified in Negroid
populations (96)). The C region of the G1m3,1, G1m17,1, and
G1m17,1,2 chains differs from that of the G1m3 chains by two,
three, and four amino acids, respectively. The structural correlations
with amino acids are illustrated in Fig. 6. The correspondence between
the G1m alleles and IGHG1 alleles is shown in Table 4. Thus,
IGHG1*01 and IGHG1*02 (see Note 3) are G1m17,1, IGHG1*03
is G1m3, IGHG1*04, IGHG1*05, and IGHG1*06 are G1m17,1,27,
G1m17,1,28, and G1m17,1,27,28, respectively, IGHG1*07 is
G1m17,1,2, and IGHG1*08 is G1m3,1. In Table 4, amino acids
corresponding to G1m allotypes are shown in bold. The nG1m1
Table 4
Correspondence between the G1m alleles and IGHG1 alleles

Amino acid positionsb

34

G1m allelesa IGHG1 alleles Domain CH1 CH3

IMGT 103 120 12 14 110
Exon (82) (97) (16) (18) (91)
Eu 199 214 356 358 431
G1m17/nG1m17c G1m1/nG1m1 G1m2/-
G1m3c
G1m17,1 IGHG1*01, IGHG1*02 Ile I103 atc Lys K120 aaa Asp D12 gat Leu L14 ctg Ala A gct
G1m17,1,27 IGHG1*04
G1m17,1,28 IGHG1*05p
G1m17,1,27,28 IGHG1*06p
G1m3 IGHG1*03 Ile I103 atc Arg R120 aga Glu E gag Met M atg Ala A gct
nG1m1
nG1m17
G1m17,1,2 IGHG1*07pd Ile I103 atc Lys K120 (aaa) Asp D12 (gat) Leu L14 (ctg) Gly G110 (ggt)
d
G1m3,1 IGHG1*08p Ile I103 atc Arg R120 (aga) Asp D2 (gat) Leu L14 (ctg) Ala A (gct)
nG1m17
a
In Negroid populations, the G1m17,1 allele frequently includes G1m27 and G1m28, leading to new G1m alleles, G1m17,1,28 and G1m17,1,27,28, as demonstrated
serologically (96) (see Subheading 5.1.7). They were assigned to IGHG1*05p and IGHG1*06p, respectively, following the sequencing of IGHG1*04 (164) (IMGT/
GENE-DB, http://www.imgt.org). The letter “p” indicates that these alleles have not yet been sequenced at the nucleotide level, and therefore are not shown in IMGT
Repertoire, Alignments of alleles: Homo sapiens IGHG1 (http://www.imgt.org) (5, 6). Amino acid changes and codons for G1m27 (CH3 Ileu I101) and G1m28 (most
probably CH3 Arg R115, Tyr Y116) are not shown
b
Positions in bold are according to the IMGT unique numbering for C domain (14); between parentheses, exon numbering (11), and in italics, Eu numbering (correspondence
between C numberings, in IMGT Scientific chart, http://www.imgt.org) (5, 6)
c
The presence of R120 is detected by anti-nG1m17 antibodies, whereas the simultaneous presence of I103 and R120 in the gamma1 chains is detected by anti-Gm3 antibodies
Human Gm, Km, and Am Allotypes and Their Molecular Characterization…

d
IGHG1*07p and IGHG1*08p amino acids, and codons between parentheses, are expected (GL and MPL) (11). The letter “p” indicates that these alleles have not yet been
sequenced at the nucleotide level, and therefore are not shown in IMGT Repertoire, Alignments of alleles: Homo sapiens IGHG1 (http://www.imgt.org) (5, 6)
649
650 M.-P. Lefranc and G. Lefranc

and nG1m17 isoallotypes present on the Gm1-negative and Gm-17

negative gamma-1 chains (and on other gamma chains, Table 2) are
shown in italics.

4. G2m

4.1. G2m Allotype G2m23 (27) is the only allotype shown on the IgG2 heavy chains,
and the gamma2 chains are either G2m23 or G2m.. (two dots
indicate that a specimen was tested and found to be negative for
G2m23 (48, 52, 53)) (see Note 8). G2m23 is localized on the
CH2 domain (detectable on the Fc of G2m23 myeloma proteins
but not on isolated CH3 domains) (134). Amino acid sequence
and 3D structure comparisons show that the G2m23 allotype is
correlated with methionine 45.1 (‘.1’ for first position in the trans-
verse CD strand (14)) in the CH2 (IGHG2 CH2 Met M45.1),

Fig. 7. Three-dimensional structure of the IGHG2 CH2 domain. CH2 position 45.1 (first
position of the transversal CD strand) corresponds to the G2m23/G2m.. allotype. Valine
V45.1 corresponds to G2m.., whereas a methionine corresponds to G2m23.
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 651

Table 5
Correspondence between the G2m alleles and IGHG2 alleles

Amino acid positiona

G2m alleles IGHG2 alleles Domain CH2

IMGT 45.1
Exon (52)
Eu 282
G2m23/G2m..
G2m23 IGHG2*02 Met M45.1 atg
G2m.. IGHG2*01, IGHG2*03, IGHG2*04, IGHG2*05, IGHG2*06 Val V45.1 gtg
a
Positions in bold are according to the IMGT unique numbering for C domain (14); between parentheses, exon
numbering (11), and in italics, Eu numbering (correspondence between C numberings, in IMGT Scientific chart,
http://www.imgt.org) (5, 6)

whereas the absence of the allotype (G2m..) is correlated with

valine 45.1 (IGHG2 CH2 Val V45.1) (11, 15) (Fig. 7).
The G2m23-positive gamma2 chains are also characterized by
the presence of threonine 92 in the CH1 domain (IGHG2 CH1
Thr T92) (11, 15). In contrast, the G2m23-negative chains and
the gamma chains of other IgG subclasses have proline 92 in the
CH1 domain (IGHG2 CH1 Pro P92) (11, 15) (Fig. 1a). Being
located on the CH1 domain, this amino acid change is not involved
in the expression of the G2m23 allotype, but owing to the strong
linkage on the same chain, the CH1 Thr T92 codon has been used
for the molecular characterization of the G2m23 chains (130).

4.2. Correspondence The G2m alleles are characterized by the presence or absence of
Between the G2m the G2m23 allotype. Only the IGHG2*02 allele is G2m23. The
Alleles and IGHG2 other alleles IGHG1*01, IGHG2*03, IGHG2*04, IGHG2*05
Alleles and IGHG2*06 are G2m23-negative (or G2m..) (Table 5) (see
Note 9).

5. G3m

5.1. G3m Allotypes The G3m allotypes make the gamma3 chain the most polymorphic
IG chains in humans. Thirteen G3m allotypes are characterized:
5.1.1. G3m Allotypes
G3m5, G3m6, G3m10, G3m11, G3m13, G3m14, G3m15,
and IGHG3 Sequences
G3m16, G3m21, G3m24, G3m26, G3m27, G3m28 (Table 1).
Three isoallotypes (nG3m5, nG3m11, and nG3m21) have also
been characterized (Table 2).
652 M.-P. Lefranc and G. Lefranc

Amino acids which could give rise to subclass-specific epitopes

and to G3m allotypes were identified following the first complete
nucleotide sequence of the IGHG3 gene, by Huck et al. (104).
The IGHG3 gene was from a healthy Tunisian individual (EZZ,
TOU II-4) homozygous for a multigene IGHC deletion (encom-
passing IGHG1 to IGHG4) and homozygous for the G3m5* allele
(G3m5,10,11,13,14,26,27, or G3mb0,b1,b3,b4,b5,u,v) (104).
The IGHG3 EZZ translation was compared with amino acid
sequences of heavy chain disease (HCD) gamma3 proteins (ZUC
(146), Wis (147), OMM (148)), of myeloma gamma3 chains
(Goe ( 149 ) and JIR (150 )), and of gamma chains of other
subclasses (104). Although only G3m5*, G3m21*, and G3m16*
(see Subheading 5.2) could be compared, the analysis confirmed
that the number of amino acid changes was lower than the number
of allotypes and suggested that the conformational structure of
a combination of amino acids was required to explain several
allotypes.
In 1989, Huck et al. published the sequence of a new allele
from a healthy Tunisian individual (LAT) homozygous for the
G3m24* allele (G3m5,6,11,24,26 or G3mb0,b1,c3,c5,u) and
demonstrated that this allele results from a gene conversion event
(106). In 2001, Dard et al. sequenced 51 full-length genomic
IGHG3 alleles from healthy individuals from African, Siberian,
West Asian, and European population samples whose sera have
been typed for the Gm allotypes (113). Different levels of molecu-
lar diversity were observed for the G3m alleles. EZZ (104) and
LAT (106) were included in the analysis. Analysis of 19 DNA
sequences of the G3m5* allele yielded 11 different IGHG3 alleles;
similar analysis of 10 DNA sequences of the G3m21* allele yielded
4 distinct IGHG3 alleles; in contrast, the 9 DNA sequences of
the G3m24* allele were monomorphic (113). These data allowed
the first identification of the amino acids involved in the G3m
allotypes; however, detailed correlations remained to be defined.
In the following paragraphs, we report, for the first time, the full
elucidation of the G3m allotypes and amino acid correlations,
taking into account conformational structures that may involve
two or even three amino acids (Fig. 8).

5.1.2. G3m6, G3m10, A first mosaic of G3m allotypes on the CH3 can be defined around
G3m11, nG3m11, G3m13, G3m11. G3m11 is characterized by a serine at position 44 of the
G3m24, G3m27 CH3 (IGHG3 CH3 Ser S44), whereas nG3m11 depends on an
asparagine at the same position (IGHG3 CH3 Asn N44) (113).
Haplotype analysis shows that four allotypes, G3m6, G3m10,
G3m13, and G3m24, depends on the presence of G3m11 (49–53).
Extensive analysis of the sequence (104, 106, 113) and genetic
data demonstrate that:
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 653

Fig. 8. G3m allotypes localizations on gamma3 chains. G3m allotypes are described in Subheading 5.1. G3m16 (tryptophan
Trp W83) and G3m21 (leucine Leu L82), nG3m21 (proline Pro P82) are located on the CH2. The other G3m allotypes form
two mosaics on the CH3. G3m26 (R115), G3m5 (R115, F116), G3m28 (R115, Y116), nG3m5 (H115, Y116), G3m14
(M84, R115, F116) and G3m15 (M39, H115, Y116) form a first mosaic. G3m11 (S44), nG3m11 (N44), G3m10 (S44, I101),
G3m24 (S44, V101), G3m27 (I101), G3m6 (S44, E98), G3m13 (S44, Q98) form a second mosaic.

– G3m10 corresponds to the simultaneous expression of S44

(G3m11) with an isoleucine at position 101 (IGHG3 CH3 Ile
I101).
– G3m24 corresponds to the simultaneous expression of S44
(G3m11) with a valine at position 101 (IGHG3 CH3 Val
V101).
– G3m6 corresponds to the simultaneous expression of S44
(G3m11) with a glutamate at position 98 (IGHG3 CH3 Glu
E98).
654 M.-P. Lefranc and G. Lefranc

– G3m13 corresponds to the simultaneous expression of S44

(G3m11) with a glutamine at position 98 (IGHG3 CH3 Gln
Q98).
These data confirm that G3m27 corresponds to isoleucine at
position 101 in the CH3 domain (IGHG3 CH3 Ileu I101).
G3m27 is expressed with G3m10 but not G3m24, which is in
agreement with the observation that G3m24 and G3m27 are anti-
thetical in genetic analysis.
These data are also in agreement with the genetic data that
show that the allotypes G3m6 and G3m13 are antithetical (owing
to an amino acid change at the same position, E98 for G3m6, and
Q98 for G3m13).

5.1.3. G3m5, nG3m5, The second mosaic, also observed on the CH3 domain of IGHG3,
G3m28, G3m26 defines three G3m allotypes (G3m5, G3m28, G3m26) and one
isoallotype (nG3m5). Extensive analysis of sequences (104, 106,
113) and genetic data (49–53, 89–96, 102) demonstrate that:
– G3m5 corresponds to the simultaneous expression of arginine
at position 115 (IGHG3 CH3 Arg R115) and phenylalanine
at position 116 (IGHG3 CH3 Phe F116).
– The nG3m5 isoallotype corresponds to the simultaneous
expression of histidine at position 115 (IGHG3 CH3 His
H115) and tyrosine at position 116 (IGHG3 CH3 Tyr Y116).
– G3m28 corresponds to the simultaneous expression of argin-
ine at position 115 (IGHG3 CH3 Arg R115) and tyrosine at
position 116 (IGHG3 CH3 Tyr Y116). This observation is
in agreement with the genetic data that show that G3m5 and
G3m28 are antithetical and mutually exclusive on the gamma3
chains.
These data confirm that G3m26 corresponds to arginine at
position 115 (IGHG3 CH3 Arg R115). This explains the high
frequency of G3m26, present on all Gm5-positive (R115, F116)
and G3m28-positive (R115, Y116) gamma3 chains, and its
absence, the arginine being replaced by a histidine, on G3m15-
positive gamma3 chains which are nG3m5 (H115,Y116).

5.1.4. G3m14 and G3m15 G3m14 has been the subject of discussion concerning its localiza-
tion on either the CH2 or CH3 domain with contradictory sero-
logical data (discussed in (50)). Extensive analysis of previously
published data of usual and uncommon haplotypes, supported by
familial studies (47–53, 89–95, 102), led us to postulate that
G3m14 corresponds to the simultaneous presence on CH3 of a
methionine at position 84 (IGHG3 CH3 Met M84) with the
G3m5 mosaic (CH3 Arg R115, Phe F116).
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 655

Gm15 was reported to be located first on the CH2 domain.

In 1983, Matsumoto et al. postulated that G3m15 correlates with
histidine at position Eu 435 (IGHG3 CH3 H115) and G3m16
with methione at position Eu 379 (IGHG3 CH3 M39), on JIR,
a G3m15,16-positive myeloma gamma3 protein (150). In 1986,
Matsumoto et al. reported that the G3m15 is located on the CH3
domain of Kam, another G3m15,16-positive myeloma gamma3
protein (151). Dard et al. (113) showed that His H115 alone can-
not be responsible for G3m15, as it is found in all the nG3m5
chains, and suggested a role for Met M39. Based on these obser-
vations, and in a similar way as for Gm14, we postulate that
G3m15 corresponds to the simultaneous presence on CH3 of
Met M39 (IGHG3 Met M39) with the nG3m5 mosaic (CH3 His
H115, Tyr Y116).
Interestingly, they are the two allotypes, G3m14 and G3m15,
localization of which have been controversial for a long time, that
require an amino acid at a position that interferes with the binding
site but is not directly located at the epitope site. This emphasizes
the importance of the conformational configuration, but at a
degree that was not suspected.

5.1.5. G3m16 and G3m21 Two allotypes, G3m16 and G3m21, are localized on the CH2
domain. G3m16 correlates with a tryptophane at position 83
(IGHG3 CH2 Trp W83), whereas G3m16-negative chains have
an arginine at that position (113). G3m21 correlates with a leucine
at position 82 (IGHG3 CH2 Leu L82), whereas nG3m21 corre-
lates with Pro P82 (113) (IMGT Repertoire, Alignments of alleles,
Homo sapiens IGHG3, http://www.imgt.org (5, 6)).

5.1.6. “Silent” G3m Among data obtained by Dard et al. (113), one sequence was
Allotypes noted as unusual as it presents an asparagine (N44) (IGHG3*08 in
IMGT Repertoire, Alignments of alleles, Homo sapiens IGHG3,
http://www.imgt.org (5, 6)), whereas a serine (S44) was expected,
given the G3m5* phenotype. The individual, Mand114, is
heterozygous for a normal G3m5* haplotype associated to the
unusual haplotype (the one that was sequenced) (113). We postu-
late that the unusual haplotype corresponds to G3m5,14,26, with
an absence of G3m (10,11,13), as previously demonstrated in the
Tunisian family 275 (95, 102). Two nucleotide substitutions in
codon 44 (as a result of mutations or of a genic conversion) are
the most probable explanation for these silent G3m (10, 11,13)
allotypes.

5.1.7. “Surnumerary” In Caucasoid and Mongoloid populations, G3m28 is frequently

Gm27 and Gm28 Allotypes associated with G3m21, although exceptions have been shown
(38, 52, 53, 94, 95, 102). In contrast, Negroid populations are
G3m21-negative and, interestingly, for individuals who are
Gm28-positive, this allotype appears as “surnumerary.” It has
 Table 6
656

Correlation between G3m allotypes and isoallotypes and amino acids

Amino acid positiona

G3m allotypes
and isoallotypes Domain CH2 CH3
IMGT 82 83 39 44 84 98 101 115 116
Exon (61) (62) (39) (44) (57) (79) (82) (95) (96)
Eu 291 292 379 384 397 419 422 435 436
G3m5 Arg R cgc Phe F ttc
nG3m5 His H cac Tyr Y tac
M.-P. Lefranc and G. Lefranc

G3m6 Ser S agc Glu E gag

G3m10 Ser S agc Ileu I atc
G3m11 Ser S agc
nG3m11 Asn N aat
G3m13 Ser S agc Gln Q cag
G3m14 Met M atg Arg R cgc Phe F ttc
G3m15 Met M atg His H cac Tyr Y tac
G3m16 Trp W tgg
G3m21 Leu L ctg
nG3m21 Pro P ccg
G3m24 Ser S agc Val V gtc
G3m26 Arg R cgc
G3m27 Ileu I atc
G3m28 Arg R cgc Tyr Y tac
a
Positions in bold are according to the IMGT unique numbering for C domain (14); between parentheses, exon numbering (11), and in italics, Eu numbering (correspondence
between C numberings, in IMGT Scientific chart, http://www.imgt.org (5, 6))
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 657

Table 7
The six most prevalent G3m alleles

WHO/IUIS/IMGT nomenclature Simplified form

G3m5,10,11,13,14,26,27 G3m5*
G3m5,6,10,11,14,26,27 G3m6*
G3m5,6,11,24,26 G3m24*
G3m10,11,13,15,27 G3m15*
G3m10,11,13,15,16,27 G3m16*
G3m21,26,27,28 G3m21*

been demonstrated in Lat-IV-5 (homozygous for G3m24*) and

Sno (homozygous for G3m6*) that the surnumerary allotype
Gm28 is expressed on the gamma1 chains (96). Thus, in Negroid
populations, Gm28 represents a G1m28 allotype. The amino acid
change on the gamma1 chain, nG3m5 (H115, Y116) to G1m28
(R115, Y116), most probably corresponds to a single nucleotide
mutation (a344 > g, H115 > R) (Table 6).
The presence of surnumerary Gm27 has also been demon-
strated on the gamma1 chain of Lat IV-5 and Sno representing a
G1m27 allotype (96). The amino acid change on the gamma1
chain most probably corresponds to a single nucleotide mutation
(g301 > a, V101 > I).
Both Gm27 and Gm28 are qualified as “alloallotype,” being
an allotype for two different gamma chains, gamma1 (G1m27,
G1m28) and gamma3 (G3m27, G3m28), depending on the
populations.

5.1.8. IGHG3 Hinge CNV For the gamma3 chains, an additional polymorphism results from
Exon Polymorphism differing numbers (CNV) of hinge exons. The hinge is encoded by
2–5 exons, depending on the alleles (104, 106, 112, 113). Thus,
the hinge region can vary from 27 to 83 amino acids and can
influence structural conformations.

5.2. G3m Alleles The thirteen G3m allotypes are inherited in different combinations
and the “IMGT G3m or G3m alleles. The six most prevalent G3m alleles are shown in
Allele Butterfly” Table 7 and illustrated in Fig. 9, as “IMGT G3m alleles butterfly”
Representation representation. For convenience, these most common G3m alleles
can be written in a simplified form indicated with an asterisk (see
Note 10 for correspondence between the WHO/IUIS/IMGT
nomenclature and previous designation).

5.3. Correspondence The correspondence between the G3m alleles and IGHG3 alleles
Between the G3m is shown in Table 8.
Alleles and IGHG3
Alleles
658 M.-P. Lefranc and G. Lefranc

Fig. 9. “IMGT G3m allele butterfly” representation. The six most prevalent G3m alleles (Table 7) are shown. The first G3m
mosaic on the CH3 domain of IGHG3 (detailed in Subheading 5.1.2) is illustrated by the top part of each butterfly represen-
tation and corresponds to G3m11/nG3m11, G3m27, and the antithetical G3m10/G3m24, and G3m13/G3m6 allotypes. The
second G3m mosaic on the CH3 domain of iGHG3 (detailed in Subheading 5.1.3) is illustrated by the bottom part of each
butterfly representation and corresponds to G3m26, and the antithetical G3m5/nG3m5/G3m28 allotypes. Amino acids
involved in the allotype expression (Table 6) and their position according to the IMGT unique numbering for C domain (14)
are indicated. Two allotypes are on the CH2 domain, G3m16 (Trp W83) and G3m21 (Leu L82). For each G3m allele butterfly,
G3m alleles are indicated with the simplified form (e.g., G3m5*) and the full nomenclature (Table 7). Haplotypes to which
the G3m alleles belong are indicated between square brackets (Table 10) (see Note 11).
 Table 8
Correspondence between the G3m alleles and IGHG3 alleles

Amino acid positiona, b

G3m alleles
and IGHG3 alleles Domain CH2 CH3

IMGT 82 83 39 44 84 98 101 115 116

Exon (61) (62) (39) (44) (57) (79) (82) (95) (96)
Eu 291 292 379 384 397 419 422 435 436
G3m5* (G3m5,10,11,13,14,26,27) Pro P ccg Arg R cgg Val V gtg Ser S agc Met M atg Gln Q cag Ileu I atc Arg R cgc Phe F ttc
nG3m21
IGHG3*01, IGHG3*05, IGHG3*06,
IGHG3*07, IGHG3*09, IGHG3*10,
IGHG3*11, IGHG3*12
G3m6* (G3m5,6,10,11,14,26,27) Pro P ccg Arg R cgg Val V gtg Ser S agc Met M atg Glu E gag Ileu I atc Arg R cgc Phe F ttc
nG3m21
IGHG3*13
G3m24* (G3m5,6,11,24,26) Pro P ccg Arg R cgg Val V gtg Ser S agc Val V gtg Glu E gag Val V gtc Arg R cgc Phe F ttc
nG3m21
IGHG3*03
G3m15* (G3m10,11,13,15,27) Pro P ccg Arg R cgg Met M atg Ser S agc Val V gtg Gln Q cag Ileu I atc His H cac Tyr Y tac
nG3m5, nG3m21
IGHG3*17
G3m16* (G3m10,11,13,15,16,27) Pro P ccg Trp W tgg Met M atg Ser S agc Val V gtg Gln Q cag Ileu I atc His H cac Tyr Y tac
nG3m5, nG3m21
IGHG3*18, IGHG3*19
G3m21* (G3m21,26,27,28) Leu L ctg Arg R cgg Val V gtg Asn N aat Met M atg Gln Q cag Ileu I atc Arg R cgc Tyr Y tac
nG3m11
IGHG3*14, IGHG3*15, IGHG3*16
a
Positions in bold are according to the IMGT unique numbering for C domain (14); between parentheses, exon numbering (11), and in italics, Eu numbering (correspondence between
C numberings, in IMGT Scientific chart, http://www.imgt.org (5, 6))
b
The amino acid change asparagine/lysine at position 79 in the CH3 domain (IGHG3 CH3 Asn/Lys N79/K) (exon numbering (52), Eu numbering 392) is not reported in the table
as no specific antibody (and therefore no allotype) has been characterized. It should be noted that the asparagine Asn N79, if present, belongs to a N-glycosylation site in alleles
IGHG3*01 to *05, *08 to *12, *14, and *16)
660 M.-P. Lefranc and G. Lefranc

6. nG4m
Isoallotypes
6.1. nG4m(a) No allotype has been defined for the gamma4 chains of the IgG4
and nG4m(b) subclass. The only serologically defined polymorphism corresponds
to the isoallotypes nG4m(a) and nG4m(b), described on the CH2
domain (138). These antithetical determinants of the gamma4
chains behave as allotypes in the IgG4 subclass, but they are pres-
ent on the other subclasses and therefore must be considered as
isoallotypes.
It has been postulated that nG4m(a) was correlated to leucine
309 (IGHG4 CH2 Leu L92) and isoallotype nG4m(b) to a dele-
tion at that position (152). However, comparison with the transla-
tion of IGHG4 sequences did not confirm that deletion and instead
showed that it was an amino acid change of leucine into valine
(IGHG4 CH2 Val V92) which explained the “disappearance” of
the leucine and was responsible for the expression of nG4m(b)
chains (1, 104). The nG4m(a) epitope (IGHG4 CH2 Leu L92) is
expressed on the gamma1 and gamma3 chains, whereas the
nG4m(b) epitope (IGHG4 CH2 Val V92) is expressed on the
gamma2 chains (Table 2).

6.2. Correspondence The nG4m(a) allele corresponds to IGHG4*01, IGHG4*03, and

Between nG4m Alleles IGHG4*04, whereas the nG4m(b) allele corresponds to IGHG4*02
and IGHG4 Alleles (Table 9).

Table 9
Correspondence between the nG4m alleles and IGHG4 alleles

Amino acid positionsa

G4m alleles IGHG4 alleles Domain CH2

IMGT 92
Exon (79)
Eu 309
nG4m(a)/nG4m(b)
nG4m(a) IGHG4*01, IGHG4*03, IGHG4*04 Leu L92 ctg
nG4m(b) IGHG4*02 Val V92 gtg
a
Positions in bold are according to the IMGT unique numbering for C domain (14); between paren-
theses, exon numbering (11), and in italics, Eu numbering (correspondence between C numberings,
in IMGT Scientific chart, http://www.imgt.org (5, 6))
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 661

7. Gm Haplotypes

7.1. Description The G1m, G2m, and G3m alleles are inherited in fixed combina-
of the Main Gm tions or Gm haplotypes. Table 10 shows the eleven most prevalent
Haplotypes Gm haplotypes. The nomenclature of the Gm haplotypes takes
into account the IGHG gene order in the locus (97, 114, 115).
The Gm allotypes are written in the linkage order of the IGHG
subclass genes, i.e., IGHG3, IGHG1, and IGHG2, with semico-
lons separating the subclasses and comas separating the allotypes;
G2m23 is the only allotype defined on gamma 2, two dots being
used to indicate that a specimen was tested and found to be nega-
tive for G2m23 (47, 48, 52, 53).

Table 10
Prevalent Gm haplotypes

Prevalent Gm haplotypesa

Populations Gm haplotypes Simplified formb Complete description

Caucasoid A Gm5*;3;23 Gm5,10,11,13,14,26,27;3;23

B Gm5*;3;.. Gm5,10,11,13,14,26,27;3;..
Caucasoid and C Gm21*;17,1;.. Gm21,26,27,28;17,1;..
Mongoloid
D Gm21*;17,1,2;.. Gm21,26,27,28;17,1,2;..
Negroid E Gm5*;17,1;.. Gm5,10,11,13,14,26,27;17,1;.. (+G1m28)c
F Gm6*;17,1;.. Gm5,6,10,11,14,26,27;17,1;.. (+G1m28)c
Gc Gm24*;17,1;.. Gm5,6,11,24,26;17,1;.. (+G1m28)c
Khoisand H Gm15*;17,1;.. Gm10,11,13,15,27;17,1;..
Mongoloid I Gm16*;17,1;.. Gm10,11,13,15,16,27;17,1;..
J Gm5*;3,1;23 Gm5,10,11,13,14,26,27;3,1;23
K Gm5*;3,1;.. Gm5,10,11,13,14,26,27;3,1;..
a
Prevalent haplotypes are characteristic of given populations. This does not mean that they are totally absent from other
populations. They can be found to frequency usually <0.01 and represent rare alleles without metissage
b
The current simplified form of the Gm haplotypes only contains one number for the G3m allele. 24* was previously
designated as 6,24*, and 16* as 15,16*
c
The presence of the G1m28 allotype (96) (see Subheading 5.1.7) is frequently observed in association with haplotypes
E, F, and G. In contrast, the presence of G1m27 can only be confirmed in homozygotes for haplotype G (haplotype G
is negative for G3m27, whereas haplotypes E and F are positive for G3m27) (96, 106). As the assignment to one or the
other haplotype in heterozygous individuals is not possible in usual typing, G1m28 allotype is usually indicated as
“+G1m28,” in genotype description. In homozygous individuals, as demonstrated in (96), the haplotype is
Gm24*;17,1,27,28;.. in Lat IV-5, and Gm6*;17,1,27,28;.. in Sno. The G1m28 also explains the uncommon haplo-
types found in Tunisian families (94, 95, 102)
d
The Khoisan population regroups the San (or Bushmen), native people, hunters/gatherers of southwestern Africa, and
the Khoikhoi (or Hottentots), pasturalists who had lived in southern Africa since the fifth century AD
662 M.-P. Lefranc and G. Lefranc

In haplotypes A and B, the G3m5* allele is inherited with the

G1m3 allele, with or without G2m23, respectively. In haplotypes
C and D, the G3m21* allele is inherited with G1m17,1 or
G1m17,1,2, respectively (without G2m23). In haplotypes E, F, G,
and H, the G3m alleles (G3m5*, G3m6*, G3m24*, and G3m15*)
differentiate the four haplotypes since they are all associated with
the G1m17,1 allele (without G2m23). The haplotype I corre-
sponds to Gm16*;17,1;.. whereas the haplotypes J and K corre-
spond to the G3m5* allele inherited with the G1m3,1 allele, with
or without G2m23, respectively (Table 10). Less frequent haplo-
types include, for example, haplotypes L (Gm5*;17,1,2;..) and M
(Gm21*;17,1;23), identified in Lebanese and Tunisian popula-
tions (47, 52, 53) (see Note 11 for correspondence between the
WHO/IUIS/IMGT nomenclature and previous designation for
the Gm haplotypes).

7.2. Prevalent Gm The eleven most prevalent Gm haplotypes are differently repre-
Haplotypes in Different sented in the Negroid, Khoisan, Caucasoid, and Mongoloid popu-
Populations lations (Table 10). Thus, Gm5*;3;23 [A] and Gm5*;3;.. [B] are
only typical of the Caucasoid populations. Two haplotypes
Gm21*;17,1;.. [C] and Gm 21*;17,1,2;.. [D] are shared by the
Caucasoid and Mongoloid populations. A unique set of haplotypes
characterizes the Negroid populations, Gm5*;17,1,.. [E],
Gm6*;17,1;.. [F], and Gm24*;17,1,.. [G]. The haplotype
Gm15*;17,1;.. [H] (with or without G3m16) is characteristic of
the Khoisan population of southern Africa. Three Gm haplotypes
only occur in the Mongoloid populations, Gm16*;17,1;.. [I]
largely in the northern hemisphere, Gm5*;3,1;23 [J] and
Gm5*;3,1;.. [K] in the southern hemisphere.
Haplotypes E, F, and G in the Negroid populations have been
found with G1m27 and G1m28 on gamma1 (demonstrated on
isolated chains of Lat-IV-5 (haplotype G) and Sno (haplotype F))
(96) (see Subheading 5.1.7).

7.3. From Gm Gm phenotypes and deduced Gm genotypes and haplotypes from

Phenotypes to Gm the Tunisian population (53) are shown as examples in Table 11.
Genotypes and The observed diversity (26 phenotypes, and 28 deduced geno-
Haplotypes types, contributed by 8 haplotypes) is explained by the geograph-
ical localization of Tunisia at the carrefour of many population
migrations and civilizations (Table 11).

7.4. Gm Haplotype Gm haplotype frequencies from different populations are shown,

Frequency in Different as examples, in Table 12.
Populations
34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 663

Table 11
Gm phenotypes observed in the Tunisian population
and deduced Gm haplotypes

Gm phenotypesa Gm genotypesb Observedc

5,10,11,13,14,26,27;3;23 A/A ou A/B 117

5,10,11,13,14,26,27;3;.. B/B 7
5,10,11,13,14,21,26,27,28;1,3,17;23 A/C 93
5,10,11,13,14,21,26,27,28;1,3,17;.. B/C 17
5,10,11,13,14,26,27;1,3,17;23 A/E 45
5,10,11,13,14,26,27;1,3,17;.. B/E 8
5,6,10,11,13,14,24,26,27;1,3,17;23 A/G 3
5,6,10,11,13,14,24,26,27;1,3,17;.. B/G 1
5,6,10,11,13,14,26,27;1,3,17;23 A/F 4
21,26,27,28;1,17;23 C/M 2
21,26,27,28;1,17;.. C/C 18
5,10,11,13,14,21,26,27,28;1,17;23 E/M 2
5,10,11,13,14,21,26,27,28;1,17;.. C/E 13
5,10,11,13,14,21,26,27,28;1,2,3,17;23 A/D 17
5,10,11,13,14,21,26,27,28;1,2,3,17;.. B/D 2
21,26,27,28;1,2,17;23 D/M 1
21,26,27,28,1,2,17;.. D/D ou C/D 6
5,10,11,13,14,21,26,27,28;1,2,17;.. D/E 3
5,10,11,13,14,26,27;1,17;.. E/E 1
5,6,10,11,13,14,24,26,27;1,17;.. E/G 1
5,6,10,11,13,14,26,27;1,17;.. E/F 1
5,6,11,21,24,26,27,28;1,17;.. C/G 1
5,6,10,11,14,21,26,27,28;1,17;..;23 C/F 1
10,11,13,15,16,21,26,27,28;1,17,.. C/I 1
5,10,11,13,14,15,16,26,27;1,3,17 A/I 1
5,10,11,13,14,15,16,26,27;1,3,17;.. B/I 1
Total 367
a
In Gm phenotypes, allotypes are given in increasing numbers, in the order G3m, G1m,
G2m, and separated by semicolons between chains of different subclasses
b
Gm haplotypes (A to M) are described in Subheading 7 (Table 10) (see Note 11)
c
Five unusual phenotypes, each one found in one individual, were the starting point of
familial studies and are not included in this table (53)
 664

Table 12
Gm haplotype frequencies among Middle Eastern, European, African, and Asian populations

Gm haplotypes

Caucasoid
Caucasoid Mongoloid Negroid Khoisan Mongoloid
M.-P. Lefranc and G. Lefranc

Populationsa A,B C D E F,G H I J,K References

Tunisian 0.602–0.618 0.226–0.243 0.025–0.051 0.083–0.101 0.018–0.022 – 0.004–0.007 – (53)
Lebanese 0.692–0.748 0.137–0.197 0.007–0.033 0.007–0.053 0.000–0.037 – 0.013–0.042 – (47, 49, 52)
Hungary, 0.776–0.777 0.138–0.147 0.049–0.076 0.005–0.013 0.005–0.013 – 0.005–0.014 – (69, 70)
Czechoslovakia
Austria, Germany, 0.725–0.831 0.120–0.190 0.049–0.085 – – – – – (60, 66, 73)
Italy Sardinia
Ethopia Sidamos 0.157 0.236 0.018 0.400 0.135 0.027 0.027 – (65)
Angola – – – 0.793 0.207 – – – (59)
Mozambique – – – 0.716 0.260 0.024 – – (59)
San – 0.120 – 0.295 0.030 0.555 – – (71)
Khoikhoi – – 0.048 0.429 0.190 0.333 – – (71)
Japan – 0.437 0.151 – – – 0.261 0.151 (90)
a
More information on these populations and other populations are available in (1) and in the original references. See also (47–87)
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 665

8. A2m Allotypes

8.1. A2m1 and A2m2 Two allotypes of the IGHA2 gene have been described, A2m1,
identified by two groups, independently (39, 40), and A2m2 (41).
A2m1 and A2m2 are antithetical and located on the CH1 domain.
A2m1 corresponds to a proline at position 115 (IGHA2 CH1 Pro
P115) and is also correlated to a proline at position 124 (IGHA2
CH1 Pro P124) (153) (Fig. 1b). A2m2 corresponds to a serine
at position 115 (IGHA2 CH1 Ser S115) and is also correlated to
an arginine at position 124 (IGHA2 CH1 Arg R124) (154, 155)
(Fig. 10a).

8.2. A2m1 and A2m2 The frequency of the A2m1 and A2m2 allotypes varies a lot
Allotype Frequency between and within most populations (53, 77, 82, 90, 94). In
Caucasoid populations, almost all individuals are homozygous for
the A2m1 allotype. The A2m2 allotype, rare in Caucasoid popula-
tions, is present with a frequency of 0.40–0.75 in Mongoloid
populations (78) and 0.60–0.85 in Negroid populations (79).

8.3. A2m1 and A2m2 A2m1 and A2m2 allotypes can be determined at the DNA level by
Allotype Determination RFLP using appropriate restriction enzymes (101). The amino
by RFLP acid responsible for the A2m2 allotype, serine 115, is encoded by
nucleotides which are part of an EcoRI restriction site (101)
(Table 13). Due to a nucleotide substitution, this site is absent in
the A2m1 allele. Thus, whereas the restriction enzyme PstI yields
two fragments containing the IGHA1 (1.2 kb) and IGHA2 (2 kb)
genes (97, 114) when DNA samples are probed with a Ca probe,
double digests EcoRI–PstI show two different patterns: one simi-
lar to the PstI one for the A2m1 allele and the other with a new
0.9 kb band for the A2m2 allele due to the existence of the EcoRI
site (101). The determination of A2m alleles by RFLP is particu-
larly useful when reagents are not readily available for the sero-
logical determination of the allotypes. Moreover, it is the only way
to identify A2m alleles when no serum is available (for instance
cell lines).

8.4. Sequence Identity The sequence of the A2m1 chain is identical to the sequence of
and Conformational the alpha 1 chain at positions 115 (CH1 Pro P115) and 124 (CH1
Difference for A2m1 Pro P124) (109) (Fig. 1b). However, the IGHA2 A2m1 allotype
and the Alpha1 and the IGHA1 isotypic epitope are recognized by different anti-
Isotypic Epitope bodies (Erna van Loghem, GL, MPL), due to a difference in the
disulfide bridge. Indeed, in IgA1 (as in IgA2 A2m2), the cystein
(CH1 Cys C123) is normally involved in a heavy-light (H-L)
disulfide bridge. In contrast, in IgA2 A2m1, there is no H-L
disulfide bridge to the cysteine (CH1 Cys C123) due to a confor-
mational hindrance of the Pro P124, and the two light chains are
directly linked to each other (102).
666 M.-P. Lefranc and G. Lefranc
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 667

8.5. A2m3 Allotype The C-terminal region (CH3 domain and/or CHS region) of the
Antithetical to nA2m3 IGHA2*02 allele (TOU II-5) (11) (Fig. 10b) was shown to be
immunogenic and to correspond to a new allotype-designated
A2m3, as it was antithetical to the nA2m3 isoallotype (previously
called nA2m(2)) (unpublished data, Erna van Loghem, GL and
MPL). One or several of the following IGHA2 amino acid changes
may be involved in the A2m3 allotype/nA2m3 isoallotype expres-
sion, Tyr Y85.2/Phe F85.2, Glu E100/Asp D100, Met M124/Leu
L124 (in the CH3), Ileu I134/Val V134, or Ala A143/Val V143
(in the CHS region) (IMGT Repertoire, Alignments of alleles, Homo
sapiens IGHA2, http://www.imgt.org (5, 6)) (Fig. 1b).

8.6. Correspondence The correspondence between A2m alleles and IGHA2 alleles is
Between A2m Alleles shown in Table 13. Three A2m alleles, A2m1, A2m2,3, and A2m2,
and IGHA2 Alleles are defined based on their allotypes. A2m1 corresponds to the
IGHA2*01 allele, A2m2,3 to the IGHA2*02 (TOU II-5) allele,
and A2m2 to the IGHA2*03 allele.

9. Km Allotypes

9.1. Definition Allotypes have been identified for the human IGKC gene and are
designated as Km (for “kappa marker”) (previously Inv) (42, 43).
There are three kappa chain allotypes designated Km1, Km2,
and Km3 that define three alleles. In 1961, Ropartz described the
first IG light chain allotype Inv(a) (42), now called Km2. One year
later, the Inv(l) allotype, now called Km1, was described by the
same group (43). In 1962, Steinberg et al. described the third Km
allotype, Inv(b) (44), now called Km3. They found that sera nega-
tive for Km1 and Km2 were always positive for Km3, and that sera
negative for Km3 were always positive for Km1 and mostly also for
Km2. Thus, the three Km allotypes define three alleles, Km3,
Km1,2, and Km1 (detailed in Subheading 9.3)

Fig. 10. IMGT Colliers de Perles of the IGHA2*02 CH1 and CH3 domains. The CH1 and CH3 domains are from IGHA2*02
(IMGT/DomainDisplay, IMGT/GENE-DB (11, 15)) (IMGT®, http://www.imgt.org). The IMGT Colliers de Perles are shown on
two layers. Hatched positions correspond to gaps according to the IMGT unique numbering (14). (a) CH1 domain. The
serine at position 115 (S115) corresponds to A2m2, with the arginine at position 124 (R124) also correlating to that allotype
(11, 101, 155). A proline at positions 115 (P115) and 124 (P124) would correspond to the A2m1 allotype (11, 101, 155).
(b) CH3 domain. One or several of the following amino acids in the CH3 domain, tyrosine Tyr at position 85.2 (Y85.2),
glutamate Glu at position 100 (E100), methionine Met at position 124 (M124), and/or in the CHS (not shown), isoleucine
Ileu I134, alanine Ala A143 may be involved in the A2m3 allotype (11, 15, 155). One or several of the following amino acids,
in the CH3 domain, phenylalanine Phe F85.2, aspartate Asp D100, leucine Leu L124, and/or in the CHS (not shown), valine
Val V134 and valine Val V143 are involved in the antithetical nA2m3 allotype (11, 15, 155).
 668

Table 13
Correspondence between A2m alleles and IGHA2 alleles

Amino acid positionsa

M.-P. Lefranc and G. Lefranc

A2m alleles IGHA2 alleles Domain CH1 CH3 CHS

IMGT 115 123 124 85.2 100 124 134 143

Exon (93) (101) (102) (70) (87) (110) (117) (126)
Bur 212 220 221
A2m1 IGHA2*01 Pro P ccc Cys C tgc Pro P cca Phe F ttc Asp D gac Leu L ttg Val V gtc Val V gtg
nA2m3
A2m2,3b IGHA2*02 Ser S tcc Cys C tgc Arg R cga Tyr Y tac Glu E gag Met M atg Ile I atc Ala A gcg
A2m2 IGHA2*03 Ser S tcc Cys C tgc Arg R cga Phe F ttc Asp D gac Leu L ttg Val V gtc Val V gtg
nA2m3
a
Positions in bold are according to the IMGT unique numbering for C domain (14), between parentheses: exon numbering (11) and in italics: Bur numbering (164)
(correspondence between C numberings, in IMGT scientific chart, Homo sapiens IGHA2, http://www.imgt.org (5, 6))
b
The allotype A2m3 was identified serologically on the C-terminal region (CH3 and/or CHS) of TOU II-5 (Erna van Loghem, GL and MPL)
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 669

9.2. Km Allotype The three Km allotypes, Km1, Km2, and Km3, are determined by
Determination, Km the hemagglutination inhibition technique (47, 50, 51, 53). Given
Phenotypes, and Km the rarity of anti-Km2 and anti-Km3 reagent antibodies, only the
Genotypes Km1 allotype is tested in most of the population studies. Km geno-
types and observed phenotypes depend on the number of tested
allotypes. In Table 14, are shown:
– The six Km genotypes defined by their Km alleles.
– The five Km phenotypes observed when sera are tested for the
three allotypes Km1, Km2, and Km3.
– The three Km phenotypes observed when sera are only tested
for allotypes Km1 and Km2.
– The two Km phenotypes observed when sera are only tested
for allotype Km1.

9.3. Km Phenotypes The two allotypes Km1 and Km2 mostly cooccur and are present
and Km Allele in 10–20% of Caucasoid (47, 53, 54, 67), 40–60% of Negroid
Frequencies in (79), and 30–60% of Mongoloid (78) populations (Table 15) (1).
Different Populations Km allele frequency in the different populations is given in Table 16.
Details on extensive Km analysis in populations from Lebanon

Table 14
Km genotypes and observed Km phenotypes depending
on the tested Km allotypes

Km phenotypes

Sera tested Sera only tested Sera only

for allotypes Km1, for allotypes tested for
Km genotypes Km2, and Km3 Km1 and Km2 allotype Km1

Km1/Km1 Km1 Km1,-2 Km1

Km1/Km3 Km1,3
Km1/Km1,2 or Km1,2 Km1,2
Km1,2/Km1,2
Km1,2/Km3 Km1,2,3
Km3/Km3 Km3 Km-1,-2 Km-1

Table 15
Km phenotype

Populations

Km phenotype frequency Caucasoid Negroid Mongoloid

Km3 80–90 40–60 40–70
Km1,2 10–20 40–60 30–60
670 M.-P. Lefranc and G. Lefranc

Table 16
Km allele frequency

Populations

Km allele frequency Caucasoid Negroid Mongoloid

Km3 0.90–0.94 0.80–0.50 0.80–0.50
Km1,2 0.06–0.08 0.20–0.50 0.20–0.50
Km1 0.01 0.01 0.01

(6,076 alleles studied in 8 different communities) and Tunisia (756

alleles) (47, 52, 53) are available in IMGT Repertoire, Allotypes,
Homo sapiens IGKC, http://www.imgt.org (5, 6)).

9.4. Correspondence Km allotypes correspond to amino acid changes in the C region of

Between Km Alleles the kappa chain (C-KAPPA) at position IMGT 45.1 and 101
and IGKC Allele Names (156–158) (Fig. 11) (IMGT Repertoire, Alignments of alleles,
Homo sapiens IGKC, http://www.imgt.org) (5, 6)).
Km1 correlates with valine at position 45.1 (‘.1’ for first posi-
tion in the transverse CD strand (14)) (IGKC Val V45.1) and leu-
cine at position 101 (IGKC Leu L101), Km1,2 correlates with
alanine at position 45.1 (IGKC Ala A45.1) and leucine at position
101 (IGKC Leu L101), and Km3, the most frequent allotype, cor-
relates with alanine at position 45.1 (IGKC Ala A45.1) and valine
at position 101 (IGKC Val V101). Correspondence between Km
alleles and IGKC allele names is shown in Table 17.
Km3 corresponds to four IGKC alleles, IGKC*01, *02, *03,
and *05, Km1,2 to the IGKC*04 allele, and Km1 to the IGKC*06
allele.

10. Notes

1. An “incomplete” antibody binds to erythrocytes or bacteria

but does not produce agglutination.
2. No allotypes have been found on the gamma4 and alpha1
chains of the IgG4 and IgA1 subclasses, respectively. Only one
allotype, Em1, has been discovered on the epsilon chain of the
IgE (159). As the concentration of IgE is very low, Em1 can-
not be typed using the hemagglutination inhibition method, as
the other allotypes, but instead using a radioimmunoassay with
a monoclonal antibody (124). Em1 is very common in all pop-
ulations and therefore not very informative. It has not been
characterized at the amino acid level and currently is not deter-
mined. Allotypic markers have not been reported for IG lambda
 Fig. 11. Three-dimensional structure of IGKC*01. The C-KAPPA domain is from the b12
antibody (IMGT/3Dstructure-DB, code PDB:1hzh, http://www.imgt.org) (16, 17). The ala-
nine Ala A45.1 (strand CD) and the valine Val V101 (strand F) are shown. The A45.1 and
V101 correspond to the Km3 allotype. The A45.1 and L101 correspond to Km1,2, and the
V45.1 and L101 to Km1.

Table 17
Correspondence between Km alleles and IGKC alleles

Km alleles IGKC alleles Amino acid positionsa

IMGT 45.1 101
Exon (46) (84)
Eu/Kabat 153 191
Km3 IGKC*01, IGKC*02, IGKC*03, IGKC*05 Ala gcc Val gtc
Km1,2 IGKC*04 Ala gcc Leu ctc
b
Km1 IGKC*06 Val (gtc) Leu (ctc)
a
Positions in bold are according to the IMGT unique numbering for C domain (14); between parentheses, IMGT exon
numbering (11), and in italics, Eu/Kabat numbering (correspondence between C numberings, in IMGT Scientific
chart, http://www.imgt.org) (5, 6))
b
IGKC*06 (not sequenced at the nucleotide level) corresponds to Km1 allele and has amino changes A45.1 > V and
V101 > L (compared to IGKC*01). The expected codons (shown between parentheses) are gtc (V45.1) and ctc (L101),
respectively (IMGT Repertoire, http://www.imgt.org, GL and MPL, 13/05/2003)
672 M.-P. Lefranc and G. Lefranc

(IGL) chains; however, there are multiple lambda chain isotypes

and the number of IGLC genes vary between individuals (11,
160–163).
3. Alleles of a gene, or “gene alleles,” correspond to polymorphic
variants of a gene that differ by at least one nucleotide in their
coding region (10, 11). Gene alleles are part of the concepts of
classification of IMGT-ONTOLOGY (7, 8). They include the
IMGT gene symbol with an asterisk, followed by a number
starting from *01, e.g., IGHG1*01, IGHG1*02, etc. Gene
alleles are available in IMGT/GENE-DB (15) and in IMGT
Repertoire, Alignments of alleles, http://www.imgt.org (5, 6).
4. The WHO numeric nomenclature (45, 46) was established in
1976. For the G3m, the number is based on the chronological
order of their discovery. The correspondence with the previous
alphanumeric designation is the following: G3m5 or G3m(b1),
G3m6 or G3m(c3), G3m10 or G3m (b5), G3m11 or G3m(b0),
G3m13 or G3m(b3), G3m14 or G3m(b4), G3m15 or G3m(s),
G3m16 or G3m(t), G3m21 or G3m(g1), G3m24 or G3m(c5),
G3m26 or G3m(u), G3m27 or G3m(v), G3m28 or G3m(g5).
Specificities for which there is no more antisera include Gm (7,
9, 18, 19, 20, 22), whereas Gm (8) is a marker of uncertain
status. Correspondence with older designations is available in
reference (50) and in IMGT Repertoire, Allotypes, http://
www.imgt.org (5, 6).
5. As isoallotypes are present in several subclasses, typing of isoal-
lotypes can only be done on isolated proteins. This is the case of
the nG4m(a) and nG4m(b) isoallotypes (see Subheading 6).
6. Alanine is present at position 110 of CH3 in G1m2-negative
gamma1 chains and in the gamma chains of the other IgG
subclasses (Fig. 1a); however, CH3 Ala A110 is not an isoal-
lotype as no antibody reagent has been characterized.
7. The presence of CH1 Ala A121 (IMGT numbering) in the 3D
structure is a file error in PDB. It should be a valine (V) as in
the b12 Fab (IMGT/3Dstructure-DB code PDB:1n0x_H).
The sequence of the C region of the b12 heavy gammal chain
(1hzh_H) should be IGHG1*01 100% in its entirety
(IMGT/3Dstructure-DB entry card for 1hzh, http://www.
imgt.org) (16, 17).
8. G2m23 (27) was detected by using an antiserum produced in
a nonhuman primate. Since then no polyclonal human anti-
G2m23 has been found (124).
9. A G2m23-positive serum can be from a homozygous individual
G2m23/G2m23 or from a heterozygous individual G2m23/
G2m… Because there is no antiserum to detect G2m.., the
hemagglutination inhibition method cannot distinguish between
sera from homozygotes or from heterozygotes.
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 673

10. Correspondence between the WHO/IUIS/IMGT nomenclature

and the previous designation for the G3m alleles.
WHO/IUIS/IMGT nomenclature Previous designation

Simplified forma Complete description Simplified formb Complete description

G3m5* G3m5,10,11,13,14,26,27 G3m(b)* G3m b0,b1,b3,b4,b5,u,v
G3m6* G3m5,6,10,11,14,26,27 G3m(c3)* G3m b0,b1,b4,b5,c3,u,v
G3m24* G3m5,6,11,24,26 G3m(c5)* G3m b0,b1,c3,c5,u
G3m15* G3m10,11,13,15,27 G3m(s)* G3m b0,b3,b5,s,v
G3m16* G3m10,11,13,15,16,27 G3m(t)* G3m b0,b3,b5,s,t,v
G3m21* G3m21,26,27,28 G3m(g)* G3m g1,g5,u,v
a
G3m24* was also designated as G3m6,24*, and G3m16* as G3m15,16*
b
G3m(c5)* was also designated as G3m(c3,c5)*, and G3m(t)* as G3m(s,t)*

11. Correspondence between the WHO/IUIS/IMGT nomencla-

ture and the previous designation for the Gm haplotypes. Only
the simplified form is indicated. The correspondence for the
detailed description of the G3m alleles is given in Note 10.
Gm haplotypes WHO/IUIS/IMGT nomenclaturea Previous designationb

A Gm5*;3;23 Gmb*;f;n
B Gm5*;3;.. Gmb*;f;..
C Gm21*;17,1;.. Gmg*;z,a;..
D Gm21*;17,1,2;.. Gmg*;z,a,x;..
E Gm5*;17,1;.. Gmb*;z,a;..
F Gm6*;17,1;.. Gmc3*;z,a;..
G Gm24*;17,1;.. Gmc5*;z,a;..
H Gm15*;17,1;.. Gms*;z,a;..
I Gm16*;17,1;.. Gmt*;z,a;..
J Gm5*;3,1;23 Gmb*;f,a;n
K Gm5*;3,1;.. Gmb*;f,a;..
L Gm5*;17,1,2;.. Gmb*;z,a,x;..
M Gm21*;17,1;23 Gmg*;z,a;n
a
24* was previously designated as 6,24*, and 16* as 15,16*
b
c5* was previously designated as c3,c5*, and t* as s,t*
674 M.-P. Lefranc and G. Lefranc

Acknowledgments

We thank Géraldine Folch for the “IMGT G3m allele butterfly”

representation, Chantal Ginestoux and Saida Hadi-Saljoqi for their
help, and François Ehrenmann for the 3D structure figures. We are
very grateful to the IMGT® team for its expertise and constant
motivation. This chapter is a tribute to Erna van Loghem, Liliane
Rivat and Claude Ropartz.

References

1. Lefranc M-P, Lefranc G (1990) Molecular that bridge biological and computational
genetics of immunoglobulin allotype expres- spheres in bioinformatics. Brief Bioinform
sion. In: Shakib F (ed) The human IgG sub- 9:263–275
classes: molecular analysis of structure, 10. Lefranc M-P (2000) Nomenclature of the
function and regulation (chapter 4). Pergamon, human immunoglobulin genes. In: Coligan JE,
Oxford, UK, pp 43–78 Bierer BE, Margulies DE, Shevach EM, Strober
2. Lefranc M-P (2011) Antibody nomenclature: W (eds) Current protocols in immunology.
from IMGT-ONTOLOGY to INN definition. Wiley, Hoboken, NJ, pp A.1P.1–A.1P.37
MAbs 3(1):1–2 11. Lefranc G, Lefranc M-P (2001) The immuno-
3. Jefferis R, Lefranc M-P (2009) Human immu- globulin FactsBook. Academic Press, London,
noglobulin allotypes: possible implications for UK, pp 1–458
immunogenicity. MAbs 1(4):332–338 12. Lefranc M-P (2007) WHO-IUIS nomencla-
4. Magdelaine-Beuzelin C, Vermeire S, Goodall ture subcommittee for immunoglobulins and
M, Baert F, Noman M, Van Assche G, T cell receptors report. Immunogenetics
Ohresser M, Degenne D, Dugoujon JM, 59:899–902
Jefferis R, Rutgeerts P, Lefranc M-P, Watier H 13. Lefranc M-P (2008) WHO-IUIS nomenclature
(2009) IgG1 heavy chain-coding gene poly- subcommittee for immunoglobulins and T cell
morphism (G1m allotypes) and development receptors report August 2007, 13th Inter-
of antibodies-to-infliximab. Pharmacogenet national Congress of Immunology, Rio de Janeiro,
Genomics 19:383–387 Brazil. Dev Comp Immunol 32:461–463
5. Lefranc M-P (2008) IMGT®, the international 14. Lefranc M-P, Pommié C, Kaas Q, Duprat E,
ImMunoGeneTics information system® for Bosc N, Guiraudou D, Jean C, Ruiz M, Da
immunoinformatics. Methods for querying Piédade I, Rouard M, Foulquier E, Thouvenin
IMGT® databases, tools and Web resources in V, Lefranc G (2005) IMGT unique number-
the context of immunoinformatics. Mol ing for immunoglobulin and T cell receptor
Biotechnol 40:101–111 constant domains and Ig superfamily C-like
6. Lefranc M-P, Giudicelli V, Ginestoux C, domains. Dev Comp Immunol 29:185–203
Jabado-Michaloud J, Folch G, Bellahcene F, 15. Giudicelli V, Chaume D, Lefranc M-P (2005)
Wu Y, Gemrot E, Brochet X, Lane J, Regnier IMGT/GENE-DB: a comprehensive database
L, Ehrenmann F, Lefranc G, Duroux P (2009) for human and mouse immunoglobulin and T
IMGT®, the international ImMunoGeneTics cell receptor genes. Nucleic Acids Res
information system®. Nucleic Acids Res 33:D256–D261
37:D1006–D1012 16. Kaas Q, Ruiz M, Lefranc M-P (2004)
7. Giudicelli V, Lefranc M-P (1999) Ontology IMGT/3Dstructure-DB and IMGT/
for immunogenetics: IMGT-ONTOLOGY. StructuralQuery, a database and a tool for
Bioinformatics 15:1047–1054 immunoglobulin, T cell receptor and MHC struc-
8. Duroux P, Kaas Q, Brochet X, Lane J, tural data. Nucleic Acids Res 32:D208–D210
Ginestoux C, Lefranc M-P, Giudicelli V 17. Ehrenmann F, Kaas Q, Lefranc M-P (2010)
(2008) IMGT-Kaleidoscope, the formal IMGT/3Dstructure-DB and IMGT/
IMGT-ONTOLOGY paradigm. Biochimie DomainGapAlign: a database and a tool for
90:570–583 immunoglobulins or antibodies, T cell recep-
9. Lefranc M-P, Giudicelli V, Regnier L, Duroux tors, MHC, IgSF and MhcSF. Nucleic Acids
P (2008) IMGT®, a system and an ontology Res 38:D301–D307
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 675

18. Kaas Q, Lefranc M-P (2007) IMGT Colliers 33. Natvig JB (1966) Gm(g)–a “new” gamma-
de Perles: standardized sequence-structure globulin factor. Nature 211:318–319
representations of the IgSF and MhcSF super- 34. van Loghem E, Martensson L (1967) Genetic
family domains. Curr Bioinform 2:21–30 (Gm) determinants of the g2c (Vi) subclass of
19. Kaas Q, Ehrenmann F, Lefranc M-P (2007) human IgG immunoglobulins. A study with
IG, TR and IgSf, MHC and MhcSF: what do special reference to Gm(c3) and Gm(c5), and
we learn from the IMGT Colliers de Perles? their relationship with the Gm(b) determinants.
Brief Funct Genomic Proteomic 6:253–264 Vox Sang 13:369–392
20. Grubb R (1956) Agglutination of erythrocytes 35. van Loghem E, Grobbelaar BG (1971) A new
coated with “incomplete” anti-Rh by certain genetic marker of human IgG3 immunoglob-
rheumatoid arthritic sera and some other sera. ulin. Evolutionary dissociation of Gm allo-
The existence of human serum groups. Acta types. Vox Sang 21:405
Pathol Microbiol Scand 39:195–197 36. Schanfield MS, Fudenberg HH (1974) Gm
21. Grubb R, Laurell AB (1956) Hereditary sero- (Ray), a new allotypic marker on human IgG3.
logical human serum groups. Acta Pathol Vox Sang 26:133–140
Microbiol Scand 39:390–398 37. Blanc M, Dugoujon JM, Ruffié J (1976) [A
22. Harboe M, Lundevall J (1959) A new type in new genetic marker for human IgG3 usually
the Gm system. Acta Pathol Microbiol Scand associated with Gm (21) factor found in the
45:357–370 caucasian and mongolian populations]. In
23. Steinberg AG, Wilson JA (1963) Studies on French. Un nouveau marqueur génétique des
hereditary gamma globulin factors: evidence IgG3 humaines habituellement associé au fac-
that Gm (b) in Whites and Negroes is not the teur Gm(21) dans les populations caucasoides
same and that Gm-like is determined by an et mongoloides. C R Acad Sci Paris
allele at the Gm locus. Am J Hum Genet 283:1461–1463
15:96–105 38. van Loghem E, Blanc M, de Lange G (1977)
24. Gold ER, Ropartz C, Rivat L, Rousseau PY Human IgG3 allotypes, with special reference
(1965) Gm(f)–a determinant of human to a new allotype related to G3m(g)
gamma-globulin. Vox Sang 10:299–302 (G3m(21)). J Immunogenet 4:371–383
25. Gold ER, Mandy WJ, Fudenberg HH (1965) 39. Vyas GN, Fudenberg HH (1969) Am(1), the
Relation between Gm(f) and the structure of first genetic marker of human immunoglobu-
the gamma-globulin molecule. Nature lin A. Proc Natl Acad Sci USA
207(5001):1099–1101 64:1211–1216
26. Litwin SD, Kunkel HG (1966) A gamma- 40. Kunkel HG, Smith WK, Joslin FG, Natvig JB,
globulin genetic factor related to Gm(a) but Litwin SD (1969) Genetic marker of the
localized to a different portion of the same gamma-A2 subgroup of gA immunoglobulins.
heavy chains. Nature 210:866 Nature 223:1247–1248
27. Kunkel HG, Yount WJ, Litwin SD (1966) 41. van Loghem E, Wang AC, Shuster J (1973) A
Genetically determined antigen of the Ne sub- new genetic marker of human immunoglobu-
group of gamma-globulin: detection by pre- lins determined by an allele at the a2 locus.
cipitin analysis. Science 154(752):1041–1043 Vox Sang 24:481–488
28. Harboe M (1959) A new hemagglutinating 42. Ropartz C, Lenoir J, Rivat L (1961) A new
substance in the Gm system, anti-Gmb. Acta inheritable property of human sera: the Inv
Pathol Microbiol Scand 47:191–198 factor. Nature 189:586–587
29. Steinberg AG, Giles BD, Stauffer R (1960) A 43. Ropartz C, Rivat L, Rousseau PY (1962) Deux
Gm-like factor present in Negroes and rare or nouveaux facteurs dans les systèmes héréditaires
absent in whites: its relation to Gma and Gmx. de gammaglobuline: le Gm(e) et l’Inv(l).
Am J Hum Genet 12:44–51 Proceedings of the 9th Society of blood
30. Ropartz C, Rivat L, Rousseau PY (1963) Le Transfusion. Karger Baset, Mexico, p 455
Gm(b) et ses problèmes. Vox Sang 8:717–723 44. Steinberg AG, Wilson JA, Lanset S (1962) A
31. Steinberg AG, Goldblum R (1965) A genetic new human gamma globulin factor deter-
study of the antigens associated with the mined by an allele at the Inv locus. Vox Sang
Gm(b) factor of human g-globulin. Am J Hum 7:151–156
Genet 17:133–147 45. WHO (1976) Review of the notation for the
32. Martensson L, van Loghem E, Matsumoto H, allotypic and related markers of human immu-
Nielsen J (1966) Gm(s) and Gm(t): genetic noglobulins. WHO meeting on human immu-
determinants of human gamma-globulin. Vox noglobulin allotypic markers. J Immunogenet
Sang 11:393–418 3:357–362
676 M.-P. Lefranc and G. Lefranc

46. WHO. Review of the notation for the allo- the Inv and the ISf systems in Danes. Hum
typic and related markers of human immuno- Hered 20:275–280
globulins (1976) WHO meeting on human 59. Jenkins T, Zoutendyk A, Steinberg AG (1970)
immunoglobulin allotypic markers, held July Gammaglobulin groups (Gm and Inv) of vari-
16–19, 1974, Rouen, France. Report amended ous Southern African populations. Am J Phys
June 1976. Eur J Immunol 6:599 Anthropol 32:197–218
47. Lefranc G, Loiselet J, Rivat L, Ropartz C 60. Mayr WR, Mickerts D (1970) Der menschliche
(1976) Gm, Km and ISf allotypes in the Gammaglobulin-Polymorphismus. Acta Biol
Lebanese population. Acta Anthropogenet Med Germ 25:475–482
1:34–45 61. Bonné B, Godber M, Ashbel S, Mourant AE,
48. Lefranc G, Rivat L, Rivat C, Loiselet J, Tills D (1971) South-Sinai Beduin. A prelimi-
Ropartz C (1976) Evidence for “deleted” or nary report on their inherited blood factors.
“silent” genes homozygous at the locus cod- Am J Phys Anthropol 34:397–408
ing for the constant region of the g3 chain. 62. Ropartz C, Rivat L, Rousseau PY, Legueult
Am J Hum Genet 28:51–61 LC (1972) Frequency of Gm, Inv and ISf
49. Lefranc G, Rivat L, Salier JP, van Loghem E, phenotypes in the population of 4 Yugoslavian
Aydenian H, Zalzal P, Chakhachiro L, Loiselet villages. Hum Hered 22:508–518
J, Ropartz C (1977) Recombination, muta- 63. Salzano FM, Steinberg AG, Tepfenhart MA
tion, or constitutive expression at a Gm locus (1973) Gm and Inv allotypes of Brazilian Cayapo
and familial hypergammaglobulinemia. Am J Indians. Am J Hum Genet 25:167–177
Hum Genet 29:523–536
64. Steinberg AG (1973) Contribution of the Gm
50. Lefranc G (1978) Allotypes et haplotypes des and Inv allotypes to the characterization of
immunoglobulines dans les communautés human populations. Israel J Med Sci
libanaises: intérêt exceptionnel en immunogé- 9:1249–1256
nétique et séro-anthropologie [in French].
65. Steinberg AG (1973) Gm and Inv allotypes of
PhD Thesis. Université de Rouen, France
some Sidamo Ethiopians. Am J Phys Anthropol
51. Lefranc G (1978) Immunoglobulin allotypes 39:403–408
and haplotypes in the Lebanese communities:
66. Wiebecke D (1973) Die frequenczen von Gm
discussion of unusual genetic events and pop-
(1,2,3,5) und Inv (1) in der Bevölkerung der
ulation structure. PhD Thesis. Université de
Bundesrepublik Deustchland. Humangenetik
Rouen, France, pp 1–137
18:175–180
52. Lefranc G, Rivat L, Serre JL, Lalouel JM, Pison 67. Fraser GR, Volkers WS, Bernini LF, van
G, Loiselet J, Ropartz C, de Lange G, van Loghem E, Meera Khan P, Nijenhuis LE
Loghem E (1978) Common and uncommon (1974) Gene frequencies in a Dutch popula-
immunoglobulin haplotypes among Lebanese tion. Hum Hered 24:435–448
communities. Hum Genet 41:197–209
68. Daveau M, Rivat L, Langaney A, Afifi N, Bois E,
53. Lefranc G, de Lange G, Rivat L, Langaney A, Ropartz C (1975) Gm and Inv allotypes in
Lefranc M-P, Ellouze F, Sfar G, Sfar M, van French Guiana Indians. Hum Hered 25:88–92
Loghem E (1979) Gm, Am and Km immuno-
globulin allotypes of two populations in 69. Schanfield MS, Gergely J, Fudenberg HH
Tunisia. Hum Genet 50:199–211 (1975) Immunoglobulin allotypes of European
populations. I. Gm and Km(Inv) allotypic mark-
54. Ropartz C, Rivat L, Rousseau PY, Baitsch H, ers in Hungarians. Hum Hered 25:370–377
van Loghem J (1963) The Gm and Inv systems
in Europe. Acta Genet Stat Med 13:109–123 70. Schanfield MS, Herzog P, Fudenberg HH
(1975) Immunoglobulin allotypes of
55. Ropartz C, Rousseau PY, Rivat L (1963) Les European populations. II.Gm, Am and
groupes de gamma-globulines Gm et Inv : Km(Inv) allotypic markers in Czechoslovakians.
Etude génétique et anthropologique. Bull Soc Hum Hered 25:382–392
Anthropol Paris 4(Xie ser.):458–469
71. Steinberg AG, Jenkins T, Nurse GT,
56. Steinberg AG, Matsumoto H (1964) Studies Harpending HC (1975) Gammaglobulin
on the Gm, Inv, Hp and Tf serum factors of groups of the Khoisan peoples of Southern
Japanese populations and families. Hum Biol Africa: evidence for polymorphism for a
36:77–85 Gm1,5,13,14,21 haplotype among the San. Am J
57. Taleb N, Ruffie J (1968) Hémotypologie des Hum Genet 27:528–542
populations jordaniennes. Bull Mém Soc 72. Lalouel JM, Loiselet J, Lefranc G, Chaiban D,
Anthrop Paris 3(XIIe sér.):269–282 Chakhachiro L, Rivat L, Ropartz C (1976)
58. Ropartz C, Rivat L, Rousseau PY, Lauridsen Genetic differentiation among Lebanese com-
UB, Persson I (1970) A survey of 9 Gm-factors, munities. Acta Anthropogenet 1:15–33
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 677

73. Piazza A, van Loghem E, de Lange G, Curtoni 85. Sukernik RI, Osipova LP, Schanfield MS
ES, Ulizzi L, Terrenato L (1976) (1992) Distribution of GM allotypes and
Immunoglobulin allotypes in Sardinia. Am J deleted IGHG1 haplotypes in the Sel’kups in
Hum Genet 28:77–86 comparison to the Forest Nentsi of north-
74. Gershowitz H, Neel JV (1978) The immuno- western Siberia. Exp Clin Immunogenet
globulin allotypes (Gm and Km) of twelve 9:15–23
Indian tribes of Central and South America. 86. Schanfield MS (1992) Immunoglobulin
Am J Phys Anthropol 49:289–301 allotypes (GM and KM) indicate multiple
75. Rivat L, Rivat C, Cook CE, Steinberg AG founding populations of Native Americans:
(1978) Gm(28). Un nouvel allotype du sys- evidence of at least four migrations to the New
tème Gm présent sur les IgG3: Intérêt particu- World. Hum Biol 64:381–402
lier de son étude dans les populations négroides. 87. Yacoubi Loveslati B, Sanchez-Mazas A,
Ann Immunol Inst Pasteur 129:33–45 Ennafaa H, Marrakchi R, Dugoujon J-M,
76. Steinberg AG, Cook CE (1981) The distribu- Lefranc M-P, Ben Ammar Elgaaied A (2001)
tion of the human immunoglobulin allotypes. A study of Gm allotypes and immunoglobulin
Oxford University Press, Oxford heavy gamma chain IGHG genes in Berbers,
77. Chaabani H, Helal AN, van Loghem E, Arabs and sub-Saharan Africans from Jerba
Langaney A, Benammar Elgaaied A, Rivat Island, Tunisia. Eur J Immunogenet
Peran L, Lefranc G (1984) Genetic study of 28:531–538
Tunisian Berbers. I. Gm, Am and Km immu- 88. Geha RS, Malakian A, Lefranc G, Chaiban D,
noglobulin allotypes and ABO blood groups. Serre JL (1976) Immunologic reconstitution
J Immunogenet 11:107–113 in severe combined immunodeficiency follow-
78. de Lange G, Zhong FM, Henke J, Feng ZC, ing transplantation with parental bone mar-
Bernhardt R, van Leeuwen F, van Eede P row. Pediatrics 58:451–455
(1985) Immunoglobulin allotypes in a Chinese 89. van Loghem E, Natvig JB (1970) Uncommon
population: comparison of haplotype frequen- Gm gene complexes. Vox Sang 18:421–434
cies with other Asian groups. J Immunogenet 90. van Loghem E, Natvig JB, Matsumoto H
12:191–195 (1970) Genetic markers of immunoglobulins
79. van Loghem E (1986) Immunoglobulin mark- in Japanese families. Inheritance of associated
ers. In: Cavalli-Sforza LL (ed) African pygmies markers belonging to one IgA and three IgG
(chapter 14). Academic Press, Orlando, p 201 subclasses. Ann Hum Genet 33:351–359
80. Helal AN, Rivat-Peran L, van Loghem E, de 91. Lefranc G, Dumitresco SM, Salier JP, Rivat L,
Lange G, Langaney A, Lefranc G (1988) de Lange G, van Loghem E, Loiselet J (1979)
Further contribution of common Gm-Am Familial lack of the IgG3 subclass: gene
haplotypes and Km alleles in the characteriza- elimination or turning off expression and neu-
tion of the Tunisian population. Exp Clin tral evolution in the immune system. J
Immunogenet 5:1–14 Immunogenet 6:215–221
81. Helal AN, de Lange G, van Loghem E, 92. Salier JP, Rivat L, Daveau M, Lefranc G,
Lefranc G (1988) Uncommon Gm haplotypes Breton P, De Menibus CH, Henocq A,
in the Tunisian population: further contribu- Fudenberg HH (1980) Quantitative studies
tion to the genetics of the IgG immunoglobu- of Gm allotypes. V. Simultaneous presence of
lins. Exp Clin Immunogenet 5:24–38 latent Gm allotypes and deficient Gm genes in
82. Soua Z, Ghanem N, Salem B, Lefranc G, a family with hypogammaglobulinaemic
Lefranc M-P (1989) Frequencies of the human probands. J Immunogenet 7:123–135
immunoglobulin IGHA2*M1 and 93. van Loghem E, Sukernik RI, Osipova LP,
IGHA2*M2 alleles corresponding to the Zegers BJM, Matsumoto H, De Lange G,
A2m(1) and A2m(2) allotypes in the French, Lefranc G (1980) Gene deletion and gene
Lebanese, Tunisian and Black African popula- duplication within the cluster of human heavy
tions. Nucleic Acids Res 17:3625 chain genes. Selective absence of IgG sub-
83. Zhao TM, Lee TD (1989) Gm and Km allo- classes. J Immunogenet 7:285–299
types in 74 Chinese populations: a hypothesis 94. Lefranc M-P (1980) Les allotypes Gm, Am et
of the origin of the Chinese nation. Hum Km des immunoglobulines humaines dans
Genet 83:101–110 deux échantillons de la population Tunisienne
84. Blanc M, Sanchez-Mazas A, Van Blyenburgh (Mahdia et Sfax): contribution à la connais-
NH, Sevin A, Pison G, Langaney A (1990) sance séro-anthropologique de cette popula-
Interethnic genetic differentiation: Gm poly- tion et à la génétique des immunoglobulines
morphism in eastern Senegal. Am J Hum [in French] Thesis, Série 3° cycle n°16,
Genet 46:383–392 Université Paris Sud, France, pp 1–108
678 M.-P. Lefranc and G. Lefranc

95. Lefranc G, Lefranc M-P, Helal AN, Boukef K, (Gmb0, b1, c3, c5, u) with an IGHG4 con-
Chaabani H, Gandoura MS, van Loghem E verted region and three hinge exons.
(1982) Unusual heavy chains of human IgG Immunogenetics 30:250–257
immunoglobulins: rearrangements of the CH 107. Keyeux G, Lefranc G, Lefranc M-P (1989) A
domain exons. J Immunogenet 9:1–9 multigene deletion in the human IGH constant
96. van Loghem E, de Lange G, van Leeuwen region locus involves highly homologous hot
AM, van Eede PH, Nijenhuis LE, Lefranc spots of recombination. Genomics 5:431–441
M-P, Lefranc G (1982) Human IgG allotypes 108. Lefranc M-P, Hammarström L, Smith CI,
co-occurring in more than one IgG subclass. Lefranc G (1991) Gene deletions in the
Vox Sang 43:301–309 human immunoglobulin heavy chain constant
97. Lefranc M-P, Lefranc G, Rabbitts TH (1982) region locus: molecular and immunological
Inherited deletion of immunoglobulin heavy analysis. Immunodefic Rev 2:265–281
chain constant region genes in normal human 109. Bensmana M, Chuchana P, Lefranc G, Lefranc
individuals. Nature 300:760–762 M-P (1991) Sequence of the CH1 and hinge-
98. Lefranc M-P, Lefranc G, de Lange G, Out TA, CH2 exons of the human immunoglobulin
van den Broek PJ, van Nieuwkoop J, Radl J, IGHA2 A2m(2) allele: comparison with the
Helal AN, Chaabani H, van Loghem E, nonallelic and allelic IGHA genes. Cytogenet
Rabbitts TH (1983) Instability of the human Cell Genet 56:128
immunoglobulin heavy chain constant region 110. Lefranc M-P, Hammarström L, Smith CIE,
locus indicated by different inherited chromo- Lefranc G (1993) Gene deletions in the
somal deletions. Mol Biol Med 1:207–217 human immunoglobulin heavy chain constant
99. Lefranc G, Chaabani H, van Loghem E, region locus: molecular and immunological
Lefranc M-P, de Lange G, Helal AN (1983) analysis. In: Rosen FS, Seligmann M (eds)
Simultaneous absence of the human IgG1, Immunodeficiencies (chapter 5). Harwood
IgG2, IgG4 and IgA1 subclasses: immuno- Academic Publishers, Newark, New Jersey,
logical and immunogenetical considerations. USA, pp 99–109
Eur J Immunol 13:240–244 111. Wiebe V, Helal A, Lefranc M-P, Lefranc G
100. Flanagan JG, Lefranc M-P, Rabbitts TH (1994) Molecular analysis of the T17 immu-
(1984) Mechanisms of divergence and con- noglobulin CH multigene deletion (del
vergence of the human immunoglobulin A1-GP-G2-G4-E). Hum Genet 93:520–528
alpha1 and alpha2 constant region gene 112. Dard P, Huck S, Frippiat JP, Lefranc G,
sequences. Cell 36:681–688 Langaney A, Lefranc M-P, Sanchez-Mazas A
101. Lefranc M-P, Rabbitts TH (1984) Human (1997) The IGHG3 gene shows a structural
immunoglobulin heavy chain A2 gene allo- polymorphism characterized by different
type determination by restriction fragment hinge lengths: sequence of a new 2-exon hinge
length polymorphism. Nucleic Acids Res gene. Hum Genet 99:138–141
12:1303–1311 113. Dard P, Lefranc M-P, Osipova L, Sanchez-
102. Lefranc M-P (1984) Recombinaisons, délé- Mazas A (2001) DNA sequence variability of
tions et conversions de gènes de la région con- IGHG3 alleles associated to the main G3m
stante des chaînes lourdes des immunoglobulines haplotypes in human populations. Eur J Hum
humaines [in French]. PhD Thesis. Université Genet 9:765–772
Montpellier 2, France 114. Flanagan JG, Rabbitts TH (1982) Arrangement
103. Lefranc M-P, Helal AN, de Lange G, Chaabani of human immunoglobulin heavy chain con-
H, van Loghem E, Lefranc G (1986) Gene stant region genes implies evolutionary dupli-
conversion in human immunoglobulin g locus cation of a segment containing g, e and a
shown by unusual location of IgG allotypes. genes. Nature 300:709–713
FEBS Lett 196:96–102 115. Rabbitts TH, Flanagan JG, Lefranc M-P
104. Huck S, Fort P, Crawford DH, Lefranc M-P, (1983) Flexibility and change within the
Lefranc G (1986) Sequence of a human immu- human immunoglobulin gene locus. In:
noglobulin g3 heavy chain constant region Chater KF et al (eds) Genetic rearrangement.
gene: comparison with the other human Cg Croom Helm, London, UK, pp 143–154
genes. Nucleic Acids Res 14:1779–1789 116. Ghanem N, Lefranc M-P, Lefranc G (1988)
105. Lefranc M-P, Lefranc G (1987) Human immu- Definition of the RFLP alleles in the human
noglobulin heavy-chain multigene deletions in immunoglobulin IGHG gene locus. Eur
healthy individuals. FEBS Lett 213:231–237 J Immunol 18:1059–1065
106. Huck S, Lefranc G, Lefranc M-P (1989) A 117. Ghanem N, Dugoujon JM, Bensmana M,
human immunoglobulin IGHG3 allele Huck S, Lefranc M-P, Lefranc G (1988)
 34 Human Gm, Km, and Am Allotypes and Their Molecular Characterization… 679

Restriction fragment haplotypes in the human 128. Balbin M, Grubb A, de Lange GG, Grubb R
immunoglobulin IGHG locus and their cor- (1994) DNA sequences specific for Caucasian
relation with the Gm polymorphism. Eur J G3m (b) and (g) allotypes: allotyping at the
Immunol 18:1067–1072 genomic level. Immunogenetics 39:187–193
118. Ghanem N, Dugoujon JM, Lefranc M-P, 129. Brusco A, de Lange GG, Boccazzi C, Carbonara
Lefranc G (1989) BstEII restriction fragment AO (1995) Molecular characterization of G2m
alleles and haplotypes of the human IGHG (n+) and G2m (n-)allotypes. Immunogenetics
genes with reference to the BamHI/SacI 42:414–417
RFLPs and to the Gm polymorphism. Exp 130. Hougs L, Svejgaard A, Barington T (2001)
Clin Immunogenet 6:39–54 The first constant-domain (CH1) exon of
119. Ghanem N, Bensmana M, Dugoujon JM, human IGHG2 is polymorphic and in strong
Constans J, Lefranc M-P, Lefranc G (1989) linkage disequilibrium with the (CH2) exon
BamHI and SacI RFLPs of the human immuno- polymorphism encoding the G2m(n+) allotype
globulin IGHG genes with reference to the Gm in Caucasians. Immunogenetics 52:242–248
polymorphism in African people. Evidence for a 131. Frangione B, Franklin EC, Fudenberg HH,
major polymorphism. Hum Genet 83:37–4418 Koshland ME (1966) Structural studies of
120. Hammarström L, Ghanem N, Smith CIE, human gG myeloma proteins of different anti-
Lefranc G, Lefranc M-P (1990) RFLP of genic subgroups and genetic specificities.
human immunoglobulin genes. Exp Clin J Exp Med 124:715–732
Immunogenet 7:7–19 132. Press EM, Hogg NM (1970) The amino-acid
121. Dard P, Sanchez-Mazas A, Dugoujon JM, de sequences of the Fd fragments of two human
Lange G, Langaney A, Lefranc M-P, Lefranc g1 heavy chains. Biochem J 117:641–660
G (1996) DNA analysis of the immunoglobu- 133. Burton RM, Deutsch HF (1970) The struc-
lin IGHG loci in a Mandenka population from ture of peptides of Fc-fragments from gG-
eastern Senegal: correlation with Gm haplo- globulins of known allotypic specificity.
types and hypotheses for the evolution of the Immunochemistry 7:145–156
Ig CH region. Hum Genet 98:36–47 134. Natvig JB, Turner MW (1971) Localization of
122. Osipova LP, Posukh OL, Wiebe VP, Miyazaki Gm markers to different molecular regions of the
T, Matsumoto H, Lefranc G, Lefranc M-P Fc fragment. Clin Exp Immunol 8:685–700
(1999) BamHI-SacI RFLP and Gm analysis of 135. Michaelsen TE, Frangione B, Franklin EC
the immunoglobulin IGHG genes in the (1977) The amino acid sequence of a human
Northern Selkups (west Siberia): new haplo- immunoglobulin G3m(g) pFc’ fragment.
types with deletion, duplication and triplica- J Immunol 119:558–563
tion. Hum Genet 105:530–541
136. Pardo AG, Rosenwasser ES, Frangione B
123. Gold ER, Fudenberg HH (1967) Chromic (1978) The primary structure of a human
chloride: a coupling reagent for passive hemag- immunoglobulin G2 (IgG2) pFc’fragment.
glutination reactions. J Immunol 99(5): J Immunol 121:1040–1044
859–866
137. Natvig JB, Kunkel HG, Joslin FG (1969)
124. de Lange G (1988) Monoclonal antibodies Delineation of two antigenic markers, “non a”
against human immunoglobulin allotypes. and “non g” related to the genetic antigens of
PhD Thesis. London, Amsterdam, pp 1–175 human gamma globulins. J Immunol
125. Zelaschi D, Newby C, Parsons M, van West B, 102:611–617
Cavalli-Sforza LL, Herzenberg LA, 138. Kunkel HG, Joslin FG, Penn GM, Natvig JB
Herzenberg LA (1983) Human immuno- (1970) Genetic variants of gG4 globulin. A
globulin allotypes: previously unrecognized unique relationship to other classes of gG
determinants and alleles defined with mono- globulin. J Exp Med 13:508–520
clonal antibodies. Proc Natl Acad Sci USA
80:3762–3766 139. van Loghem E, de Lange G (1972) ‘non-b’,
antigenic determinant of IgG. Vox Sang
126. Kurth JH, Bowcock AM, Erlich HA, Nevo S, 22:193–199
Cavalli-Sforza LL (1991) Km typing with
140. Gaarder PI, Natvig JB (1972) Two new anti-
PCR: application to population screening. Am
gens of human IgG, “non b0” and “non b1”,
J Hum Genet 48(3):613–620
related to the Gm system. J Immunol
127. Balbin M, Grubb A, Abrahamson M, Grubb R 108:617–621
(1991) Determination of allotypes G1m (f) and
141. van Loghem E, de Lange G (1976) The first
G1m (z) at the genomic level by subclass-specific
example of an isoallotype of human IgG
amplification of DNA and use of allele-specific
located on the Fd fragment. Scand J Immunol
probes. Exp Clin Immunogenet 8:88–95
5:1015–1019
680 M.-P. Lefranc and G. Lefranc

142. van Loghem E, de Lange G, Koistinen J 154. Toraño A, Putnam FW (1978) Complete
(1976) The first isoallotype of human IgA amino acid sequence of the alpha 2 heavy
proteins. An antigenic determinant occurring chain of a human IgA2 immunoglobulin of
as allotype in the IgA2 subclass and as isotype the A2m (2) allotype. Proc Natl Acad Sci USA
in the IgA1 subclass. Scand J Immunol 75(2):966–969
5:161–164 155. Flanagan JG, Lefranc M-P, Rabbitts TH
143. van Loghem E (1971) Formal genetics of the (1984) Mechanisms of divergence and con-
immunoglobulin systems. Ann N Y Acad Sci vergence of the human immunoglobulin alpha
190:136–149 1 and alpha 2 constant region gene sequences.
144. Wang AC, Shuster J, Fudenberg HH (1969) Cell 36(3):681–688
Evolutionary origin of the Gm“a” peptide of 156. Baglioni C, Alescio Zonta L, Cioli D, Carbonara
immunoglobulins. J Mol Biol 41:83–86 A (1966) Allelic antigenic factor Inv(a) of the
145. Cook CE, Steinberg AG (1979) An amino light chains of human immunoglobulins:
acid substitution in the g1 chain of human chemical basis. Science 152:1517–1519
immunoglobulin G associated with the Gm 157. Steinberg AG, Milstein CP, McLaughlin CL,
(2) allotype. Mol Immunol 16:555–558 Solomon A (1974) Structural studies of an
146. Wolfenstein-Todel C, Frangione B, Prelli F, Inv (1, -2) kappa light chain. Immunogenetics
Franklin EC (1976) The amino acid sequence 2:1–10
of “heavy chain disease” protein ZUC. Structure 158. Milstein CP, Steinberg AG, McLaughlin CL,
of the Fc fragment of immunoglobulin G3. Solomon A (1974) Amino acid sequence
Biochem Biophys Res Commun 71:907–914 change associated with genetic marker Inv(2)
147. Frangione B, Rosenwasser E, Prelli F, Franklin of human immunoglobulin. Nature
EC (1980) Primary structure of human g3 248:160–161
immunoglobulin deletion mutant: g3 heavy- 159. van Loghem E, Aalberse RC, Matsumoto H
chain disease protein Wis. Biochemistry (1984) A genetic marker of human IgE heavy
19:4304–4308 chains, Em(1). Vox Sang 46:195–206
148. Alexander A, Steinmetz M, Barritault D, 160. Ghanem N, Dariavach P, Bensmana M,
Frangione B, Franklin EC, Hood L, Buxbaum Chibani J, Lefranc G, Lefranc M-P (1988)
JN (1982) Gamma heavy chain disease in Polymorphism of immunoglobulin lambda
man: cDNA sequence supports partial gene constant region genes in populations from
deletion model. Proc Natl Acad Sci USA France, Lebanon and Tunisia. Exp Clin
79(10):3260–3264 Immunogenet 5:186–195
149. Recht B, Frangione B, Franklin E, van Loghem 161. Lefranc M-P, Lefranc G (1992) Les
E (1981) Structural studies of a human g3 Immunoglobulines humaines. In: Breton-
myeloma protein (GOE) that binds Staph Gorius J, Reyes F, Rochant H, Rosa J, Vernant
protein A. J Immunol 127:917–923 JP (eds) L’hématologie de Bernard Dreyfus
150. Matsumoto H, Ito S, Miyazaki T, Ohta T (chapter 20). Médecine-Sciences Flammarion,
(1983) Structural studies of a human g3 pp 197–254
myeloma protein (JIR) bearing the allotypic 162. Lefranc M-P, Pallarès N, Frippiat J-P (1999)
marker Gm(st). J Immunol 131: 1865–1870 Allelic polymorphisms and RFLP in the human
151. Matsumoto H, Ito S, Miyazaki T, Kawai N immunoglobulin lambda light chain locus.
(1986) Structural studies of another human Hum Genet 104:361–369
IgG3 myeloma protein (Kam) carrying the 163. Lefranc M-P, Lefranc G (2004) Immunoglo-
allotypic markers Gm(s,t) and its alteration bulin lambda (IGL) genes of human and
induced by chemical modification. In: mouse. In: Honjo T, Alt FW, Neuberger MS
Brinkmann B, Henningsen K (eds) Advances (eds) Molecular biology of B cells. Academic
in forensic haemogenetics, Springer Verlag, Press, Elsevier Science Amsterdam, The
Berlin-Heidelberg, Germany vol I. p 153 Netherlands, pp 37–59
152. Abel CA, Despont JP (1974) An amino acid 164. Goetze AM, Zhang Z, Liu L, Jacobsen FW,
deletion associated with the IgG4b allotype of Flynn GC (2011) Rapid LC-MS screening for
human IgG4 myeloma proteins. J IgG Fc modifications and allelic variants in
Immunogenet 1:79–82 blood. Mol Immunol 49:338–352
153. Tsuzukida Y, Wang CC, Putnam FW (1979) 165. Liu YS, Low TL, Infante A, Putnam FW
Structure of the A2m(1) allotype of human (1976) Complete covalent structure of a
IgA. A recombinant molecule. Proc Natl Acad human IgA1 immunoglobulin. Science
Sci USA 76:1104–1108 193(4257):1017–1020
 INDEX

A AmpliTaq™ Gold DNA polymerase...............32, 33, 38, 41,

51, 145, 146, 162, 514
Abacavir..... ........................................................................ 29 Ancestral haplotypes (AHs) ............................................ 160
Acute lymphocytic leukaemia (ALL) .............. 107, 201, 408 Angiotensin type 1 receptor ............................................ 272
Acute myeloid leukaemia (AML)........................... 201, 408, Ankylosing spondylitis ......................................................29
469–470, 475, 493–495, 531 Antigen presenting cell (APC) ................................ 512, 513
AFBACs. See Affected family based controls (AFBACs) Anti-human globulin (AHG) .........................291, 359, 361,
Affected family based controls (AFBACs) .............. 259–260 368–369, 376
Affected sib pair method ......................................... 258, 259 Anti-thymocyte globulin (ATG) ............................ 124–127,
Agarose gel .................................... 10, 17, 18, 52, 57, 71–74, 274, 361, 363, 365, 369, 374, 441, 471
76–77, 131, 134, 135, 138, 139, 145–146, 148, ANTT. See Allele Name Transition Tool (ANTT)
154, 157, 185, 422, 428, 454, 458–460, 521, 553, APC. See Antigen presenting cell (APC)
555, 557, 558, 563, 564 Apheresis.......................................... 272, 495–497, 503, 504
Agencourt® AMPure kit (Beckman Coulter, Beverly, Aplastic anaemia.............................................................. 531
MA, USA) ..................................... 436, 463, 465 Applied Biosystems™
Agencourt® Clean SEQ kit (Beckman CopyCaller™ software ...................................... 162, 168
Coulter) ................................................. 443, 461 MicroAmp® optical adhesive film ............................. 163
Agencourt® SPRIPlate magnetic plate MicroAmp® reaction plates ....................................... 163
(Beckman Coulter) ................................ 436, 452 Aqua Dest™ ..................................... 145, 148, 149, 154, 155
AHG. See Anti-human globulin (AHG) Arlequin (Integrated software for population genetics
AHs. See Ancestral haplotypes (AHs) analysis) .......................... 218, 232, 234, 237, 238
Alemtuzumab .......................................................... 613, 624 Assign ATF™ ..................................................... 75, 81, 121
ALL. See Acute lymphocytic leukaemia (ALL) Assign SBT™ (Conexio Genomics, Applecross,
Allele WA, Aust.) ................... 75, 81, 87–121, 452, 462
frequency by direct counting.............................. 217, 219 ATG. See Anti-thymocyte globulin (ATG)
frequency estimates from carrier Autoantibody ................................................... 360, 370, 372
frequencies ............................................. 219, 220
Allele Name Transition Tool (ANTT) .................... 204, 206 B
Allophycocyanin (APC) .......................................... 312, 472
BD FACSCanto™ flow cytometer
Alloreactivity ................................... 271, 309, 310, 339–340,
(BD Biosciences) ........................... 331, 481, 484
408–409, 416, 469–475, 477–488, 492–494, 513
BD FACSDiva™ software
Allorecognition ........................................................ 474, 544
(BD Biosciences) ........................... 331, 481, 484
Allotypes............................208, 319, 395, 426, 435, 470, 635
BD GolgiStop™ (BD Biosciences) ......................... 480, 487
Am allotypes............................................................ 635–673
Behcet’s disease................................................................ 144
Amicon Ultra™ centrifugal filters (Millipore, Billerica,
Betaine.............................. 436–438, 440, 442, 453–456, 464
MA, USA) ..................................................... 463
Big Dye® Terminator (BDT ) Applied
AML. See Acute myeloid leukaemia (AML)
Biosystems .............................38, 74, 79–80, 147,
AmpErase® uracil-N-glycosylase (UNG)................ 176–178
184, 443, 514
Amplicon... .................................... 10–13, 18, 19, 22, 23, 76,
Biorad DYAD™ ................................................................ 71
77, 128, 133, 150, 424–426, 428, 432–434, 437,
Biorad Gel Doc 2000 Transilluminator™ ................... 74, 76
438, 440, 442–444, 446, 448, 450, 454–456,
BioRobot EZ1 workstation (Qiagen, Valencia,
458–460, 463–466, 515–519, 553, 559, 560, 562,
CA, USA) .............................................. 442, 457
564, 565

Frank T. Christiansen and Brian D. Tait (eds.), Immunogenetics: Methods and Applications in Clinical Practice,
Methods in Molecular Biology, vol. 882, DOI 10.1007/978-1-61779-842-9, © Springer Science+Business Media New York 2012

681
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
682 Index

Biotage®...........................................................................130 Complementarity determining regions

BioTaq™ (Bioline UK) ................................................... 427 (CDR) ...................................577, 579, 581, 583,
BMDW. See Bone Marrow Donors Worldwide (BMDW) 588–593, 596, 598, 599, 601, 602, 606, 607, 610,
Bone Marrow Donors Worldwide 613, 619, 621, 624, 630–632
(BMDW) ...................................... 532, 535, 542 Complement genes
Bonferroni correction ................................................ 28, 248 C4A..... .......................................160–162, 164, 168–170
Bortezomib ...................................................................... 297 C4B..................................... 160–162, 164, 166, 168–170
Brefeldin A (Sigma) ........................................ 317, 327, 334 Conditional haplotype method (CHM) .................. 253–254
Conexio Genomics Pty Ltd.............................70–72, 74–75,
C 81, 83, 88, 90, 452
Cancer Confidence intervals (CI) ................................219–220, 252,
breast......................................................................... 493 255, 268–270
colon.......................................................................... 493 Contingency tables ............... 27–28, 247–254, 256, 259–260
hepatocellular............................................................. 493 Copy number variants (CNVs)............................5, 162, 202,
melanoma .................................................................. 493 257, 262, 415, 636–638, 657
neuroblastoma ................................................... 493, 494 Cord blood ...................................... 510, 532, 533, 535, 538,
ovarian....................................................................... 493 541–543, 545, 546
Carbamazepine ............................................................ 28, 29 C-region........................................... 637, 648, 667, 670, 672
Carboxyfluorescein diacetate (CFDA) ............ 306, 362, 376 Crossmatch
Case-control studies .........................247–257, 259, 550–551 allogeneic .................... 306, 365–368, 370, 372, 494–495
CD107 autologous................................... 306, 365, 366, 371, 372
cytotoxicity assay ................................479, 483–484, 486 flow cytometric .................................. 272–274, 379–389
flow cytometric assay ......................................... 477–488 interpretation .............. 306, 364–367, 370, 380, 386, 389
C-domain.........................................607–608, 610, 612, 615, lymphocytotoxic
619, 622, 624–626, 630–632, 636, 637, 642–644, B cell .................................................... 271, 273, 275
649, 656, 658–660, 668, 671 T cell.................................................... 273–275, 359
Cell processing ........................................................ 497, 498 Cryopreservation .............. 320, 324, 333, 351–357, 470, 498
CellQuest™ analysis software CTS. See Collaborative Transplant Study (CTS)
(Becton Dickinson)......... 317, 318, 328, 331, 383 Cytokines
CFDA. See Carboxyfluorescein diacetate (CFDA) IFN γ... ........................ 328, 329, 391, 491–492, 552, 554
Chaga’s disease ................................................................ 503 IL6............................................................. 552, 554, 561
Chido (Ch) determinanats .............................................. 161 IL10............................................ 331, 334, 552, 554, 562
Chi-square test ............................ 27–28, 221–223, 227, 228, IL13........................................................................... 552
247–248, 298 IL1 β........................................................... 552, 554, 559
Chloroform.............................................................. 556–557 Interleukin-2 (IL2) ....................................331, 332, 334,
CHM. See Conditional haplotype method (CHM) 392, 492, 550, 552, 554, 560
51
Chromium (Cr) release assay ........................ 310, 347, 479 TGF β 1 ............................. 491–492, 550, 552, 554, 560
CI. See Confidence intervals (CI) TNF α.................................................331, 552, 554, 560
CliniMACS® (Miltenyi Biotech, Bergisch, Gladbach,
D
Germany)................................471, 496–500, 503
Cluster of differentiation (CD) molecules ....................... 392 Data
CNVs. See Copy number variants (CNVs) management ...................................................... 197–212
COBE 2991 Cell Processor (CaridianBCT, Lakewood, master data file................................................... 198–202
Colorado, USA) ............................. 496, 498–502 standards .................................................... 197, 199, 201
Coding region.......................... 2, 5, 124–126, 128, 129, 134, storage................................................................ 198, 202
140, 432, 452, 461, 463, 642, 672 ddNTPs. See Dideoxynucleotides (ddNTPs)
Codon.......................................... 78, 95, 107, 109, 124–129, Denaturation ....................................... 16, 19, 41, 49, 52–53,
133–135, 403, 441, 523, 560, 574, 577, 581, 583, 74–75, 80, 133, 138, 146, 151, 372, 422, 455,
586–587, 593, 600, 607–610, 649, 651, 655, 671 467, 523, 563
Cold ischemia time.......................................... 267–269, 303 Dendritic cell ............................................393, 409, 513, 514
Collaborative Transplant Study (CTS) ................... 267–276, Deoxyribonucleic acid (DNA).......................5, 9, 47, 67, 87,
362, 550, 553–555, 557–563 128, 144, 161, 173, 184, 198, 224, 248, 332, 341,
Complement 354, 417, 432, 492, 513, 532, 570, 636
rabbit...................................................361, 364, 371–374 Dideoxynucleotides (ddNTPs) .................................... 29, 70
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
Index
683
Dimethyl sulphoxide (DMSO) .................31, 341, 352–354, Ewens–Watterson homozygosity statistic........ 220, 228–229
356, 436–438, 440, 443, 453, 454, 460, 464, 467, Exonuclease ..................................... 131, 188, 417, 514, 521,
480, 481 570, 576, 578
Diplotypes............................................................... 209, 261 Exo SAP-IT® USB Corporation ....................38, 74, 76, 77,
Dithiothreitol (DTT) ..............................296, 360, 366–368, 147, 154, 188, 193
375, 376 Expectation-maximization (EM)
DMSO. See Dimethyl sulphoxide (DMSO) algorithm ....................................... 139, 224–226
DNA. See Deoxyribonucleic acid (DNA)
DNA polymerase............................. 9, 10, 15–18, 20, 31–33, F
38, 39, 70, 73, 75, 128, 131, 137, 145, 146, 162, FACSCalibur flow cytometer
435, 442, 454–456, 514, 523, 555 (BD Biosciences) ........................... 331, 381, 385
DNAzol® (Molecular Research Center) ......................... 428 FACSDiva™ software (Becton
Donor lymphocyte infusion (DLI) .................. 494, 512–514 Dickinson) ..................................... 331, 481, 484
Donor search Family based studies ................................................ 257–260
related....................................................... . 494, 497, 503 6FAM™ dye .................................................... 162, 174, 175
unrelated .....................................531–533, 535, 540, 541 FASTA format ........................ 118, 461, 462, 571, 574, 581,
D-region gene ..........................................575, 577, 578, 581 585, 610, 619
Driftcon Temperature Verification System (CYCLERtest FCS. See Foetal calf serum (FCS)
Landgraaf, Netherlands) ................................ 465 Fermentas™ ............................... 49, 130, 131, 137, 139, 555
Dynabeads® (Invitrogen)................................. 334, 341, 344 Fibroblasts. .............................................................. 183, 511
Ficoll-Paque Plus® (GE Healthcare).......................314, 319,
E 320, 332, 352, 353, 380, 381, 480–482
EasySep™ beads (Stem Cell Technologies) ............ 360, 362 FITC. See Fluorescein isothiocyanate (FITC)
Eculizumab...................................................................... 297 Flow cytometry............................ 13, 30, 272–275, 279–287,
EDTA. See Ethylenediaminetetracetic acid (EDTA) 289, 291, 292, 294, 303, 306, 312, 313, 317, 318,
Electrophoresis .............................. 15, 16, 21, 33, 38–41, 48, 328, 329, 331, 341, 344, 345, 348, 379–389, 416,
49, 52, 57, 58, 70–74, 76–77, 81, 82, 130, 131, 471, 477–488, 498, 501–503, 553
134, 138, 140, 145–146, 148–151, 154, 157, 161, FlowJo™ software (Treestar Inc, Ashland,
186, 187, 417, 422, 426, 428, 436, 454, 465, 519, OR, USA)...............................317, 318, 328, 331
522, 555–556, 563–565 Fluorescein isothiocyanate (FITC)..................312, 326, 327,
ELISA. See Enzyme linked immunosorbent assay (ELISA) 341, 342, 347, 348, 381–389, 473, 481, 483, 485
EM algorithm. See Expectation-Maximization (EM) Fluorinated ethylene propylene (FEP) bags (AFC, Gaithersburg,
algorithm Maryland, USA) ............................ 496, 497, 501
EMDIS. See European Marrow Donor Information System FluoroSpot™ technique .......................................... 311, 312
(EMDIS) Foetal calf serum (FCS)...................................314, 316, 324,
Entrez Gene (National Center for Biotechnology 340–346, 348, 352–354, 356, 360, 363, 480, 486
Information, NCBI) ...................................... 600 Forward scatter (FSC) .............. 329, 383, 386, 387, 484, 485
Enzyme linked immunosorbent assay Frameshifts ...................................................... 403, 577, 581
(ELISA)......................... 161, 271, 272, 291–295, Framework regions .......................... 577, 579, 581, 583, 588,
304, 341–343, 345–346, 348, 349, 553 591, 593, 596, 599, 606, 607, 613, 619, 621, 630
Enzyme linked immunospot (ELISPOT) FSC. See Forward scatter (FSC)
ELISPOT reader (Autoimmun Diagnostika GmbH,
G
Germany)....................................... 311, 316, 326
Eppendorf Mastercycler Pro™ .......................................... 71 Gamma Cell 3000 ELAN gamma irradiator
Epstein–Barr virus (EBV).. .............................310, 313, 318, (MDS, Nordion, Canada) .............................. 480
321, 322, 324, 328, 330, 339, 340, 343, 348, 479, GCN. See Gene copy number (GCN)
482–483, 486, 495 G-CSF. See Granulocyte colony stimulating factor (G-CSF)
Ethidium bromide ........................ 16, 19, 145, 436, 454, 556 GelDoc 2000 DNA visualization system
Ethylenediaminetetracetic acid (EDTA) .................... 19, 20, (Biorad).......................................................... 519
23, 61, 71, 74, 79, 81, 130, 140, 145, 147, 163, GelRed™ (Biotium) ........................................................ 423
175, 176, 184, 189–190, 373, 384, 442, 452, 464, GenAlEx. See Genetic analysis in Excel (GenAlEx)
496, 514, 519, 555, 556, 565 Gene Amp®....................................................... 71, 514, 520
European Marrow Donor Information System Gene copy number (GCN) ..................................... 159–170
(EMDIS) ....................................... 535, 536, 542 Gene pulser™ (Biorad).................................................... 341
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
684 Index

GeneRuler Express (Fermentas™) .......................... 131, 139 Heterozygous ambiguity resolving primers
Genetic analysis in Excel (GenAlEx) ..................... 218, 219, (HARPs)..............................78, 94–95, 105, 107,
233, 234, 240 109, 112, 113, 115–118
Genetic polymorphism .................................... 224, 550–552 HLA. See Human leukocyte antigen (HLA)
Genome Diagnostics® (BV, Utrecht, HLA dictionary ........................................532, 538, 544, 545
Netherlands) .................................. 156, 520, 523 HLA disease associations
Genotyping.............................2, 3, 10, 11, 13, 14, 28, 30, 32, ankylosing spondylitis .................................................. 29
42, 49–53, 56, 58, 75, 84, 88, 113–115, 121, 133, coeliac disease .............................................................. 29
137, 144, 162, 164, 170, 173–181, 183–195, 199, drug hypersensitivity-Abacavir, Carbamazepine.......... 29
212, 217, 220, 221, 248, 261, 399–401, 415–418, narcolepsy .................................................................... 29
472, 478, 493, 497, 519, 532, 550, 554, 557, 558, rheumatoid arthritis ..................................................... 29
563, 564, 566, 638 type 1 diabetes ............................................................. 29
Gen-Probe® LIFECODES HLA-SSO™ ............ 50, 56–60 HLA Explorer software................................................... 543
Gen-Probe® LIFEMATCH Quick-Type™ ............... 59–60 HLA Fusion™ (One Lambda).................................... 55–56
Germline................................. 392, 581, 583–585, 591, 593, HLA Librarian™ software (Conexio Genomics,
599–601, 606, 609, 613, 630 Applecross, WA, Aust) .................................. 452
Gm allotypes ........................................................... 636–661 HLA microspheres ............................................................ 48
Gm-Am haplotypes................................................. 636, 638 Homozygosity
GMP. See Good manufacturing practice (GMP) normalized deviate of homozygosity ................. 229–230
Good manufacturing practice (GMP) ..................... 495–497 HPC. See Hematopoietic progenitor cell (HPC)
GPromo.R® ............................................................. 132, 137 HUGO. See Human Genome Organization (HUGO)
GPromo.S® ............................................................. 132, 137 Human Genome Organization (HUGO) ....................... 600
Graft versus host disease (GvHD) ...................183, 309, 310, Human genomes project.................................................. 178
407–409, 469–471, 478, 493, 495, 504, 511, 512, 550 Human leukocyte antigen (HLA)
Graft versus leukemia (GvL) ........................... 478, 511, 512 compatibility ..............................................267–268, 270,
Granulocyte colony stimulating factor 271, 472
(G-CSF) ................................................ 471, 491 HLA-A ..............................2, 3, 9, 11, 61, 72, 75, 76, 78,
Granzymes ...................................................... 331, 392, 479 143, 204, 268–271, 290, 303, 326, 393–394, 397,
GraphPad Prism® software ............................................. 248 416, 472, 478, 480, 511, 512, 524, 525, 531, 533,
GvHD. See Graft versus host disease (GVHD) 535, 536, 543, 545, 617
GvL. See Graft versus leukemia (GvL) HLA antibodies................. 267–276, 289–306, 359–360,
365–368, 371–372, 374, 379–381, 385, 386, 389
H clinical importance................................... 14–15, 302
HaploPrep™ kit (Qiagen, Valencia, CA, USA)............... 442 detection and characterization ..................... 289–306
Haplo.stats software ........................................................ 224 donor specific (DSA) .................................. 272, 359,
Haplotypes 366–368, 379, 380, 386, 389
haploidentical ............................................................ 472 HLA-B.................................... 11, 29, 31–33, 35, 38, 44,
haplo.stats package .................................................... 224 72, 75, 76, 78, 119, 183, 238, 269, 326, 340, 397,
rare estimated haplotypes .......................................... 225 405, 416, 470, 472, 477–478, 480, 511, 524, 525,
Hardy Weinberg equilibrium (HWE) 531, 539, 541, 543–545
exact test for HWEP ......................................... 222, 223 HLA-C ........................70, 72, 75, 76, 78, 143, 204, 261,
HWE proportions (HWEP) ............................. 220, 261 262, 396, 397, 406, 416, 470, 472, 474, 475, 477,
HARPs. See Heterozygous ambiguity resolving primers 478, 480, 531, 534, 537, 539, 541, 543, 544
(HARPs) HLA-DP................................................... 531, 543–544
Helmberg sequence compilation and rearrangement HLA-DQ ...............................................43–44, 299, 531
evaluation (SCORE) ..................................... 206 HLA-DR ................................... 9, 10, 28, 269–270, 531
Hemagglutination inhibition HLA-E ......................................................... 3, 143–157
methodology .................................. 638, 670, 672 HLA-G ..................................................... 123–140, 397
Hematopoietic progenitor cell isotypes
(HPC)..................... 432, 471, 531–532, 535, 539 IgA .............................................................. 271–272
Heterologous immunity................................................... 339 IgG ..............................................271–272, 326, 375,
Heterozygosity 380, 385, 481
HARPs ............................ 78, 94–95, 105, 107, 109, 112, IgM.............................................................. 271–272
113, 115–118 Human recombinant Interleukin-2 (IL-2) ...................... 314
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
Index
685
HWE. See Hardy Weinberg equilibrium (HWE) Internal positive control (IPC)........................ 296, 300, 399,
Hybridization .................................. 32, 41, 45, 58, 146, 151, 418, 420, 423, 424, 426, 427, 564
152, 426, 456, 457 International HapMap Project ........................................ 178
Hydropathy index .................................................... 588, 625 Intracellular cytokine staining (ICS) ...................... 312, 313,
Hyperacute rejection........................................ 272–273, 379 316–319, 326–328
Invitrogen™ .............................. 74, 130, 131, 137, 145, 314,
I 318, 334, 341, 352, 380, 384, 423, 435, 436, 442,
443, 463, 472, 480, 514, 519, 521
ICS. See Intracellular cytokine staining (ICS)
IPC. See Internal positive control (IPC)
IgSF. See Immunoglobulin superfamily (IgSF)
IPD-KIR database........................... 402, 404, 441, 452, 461,
IMGT/HLA database ............................14, 88, 95, 143, 157
462, 467
Immunogenes ..................................124, 197–212, 215–241,
Isoallotypes. ............................. 637–642, 644, 645, 647, 651,
245–263, 510–511
654, 656, 660, 667, 672
Immunogenetics information system®
Isoamyl alcohol ................................................ 442, 458, 463
IMGT/Automat ........................................ 574, 587, 596
IMGT/Colliers de Perles...........................574, 586, 602, J
606, 610, 612, 624–630, 632, 642, 667
J-region...... .............................. 574–576, 578, 581, 585, 586,
IMGT/DomainDisplay ..................................... 642, 667
596, 598, 600, 607, 609, 612, 613, 619, 624, 630
IMGT/DomainGapAlign .........................605–632, 636,
Junction analysis ......................................570, 572, 574–581,
642, 643
587–588, 590, 598, 601
IMGT/3D structure-DB...................613, 615, 629, 636,
644–648, 671, 672 K
IMGT/GENE-DB ...........................571, 600, 606, 636,
642, 643, 667, 672 Keratinocytes ................................................................... 511
IMGT/HighV-QUEST .................................... 569–602 Killer cell immunoglobulin like receptors (KIR)
IMGT/JunctionAnalysis ...........................570, 572, 574, genes.... .......................202, 225, 226, 246, 261, 394–396,
575, 577, 578, 580, 581, 587–588, 590, 598, 601 398–404, 407–408, 415–417, 425, 426, 431, 432,
IMGT/Kaleidoscope ......................................... 570, 606 434, 435, 443, 453, 462, 464, 466, 467, 470, 472,
IMGT/LIGM-DB database .....................576, 579, 580, 474, 492
582, 584, 586–589, 593 ligands........................ 246, 261, 262, 394–397, 405–408,
IMGT/PhyloGene .................................... 585, 594, 595 415, 416, 470, 474, 492
IMGT Repertoire Web resources .............................. 636 polymorphism............................ 202, 216, 224, 397, 402,
IMGT/Scientific chart ......................630, 642, 644, 651, 403, 416, 426, 435
656, 659, 668, 671 typing..........................207, 208, 212, 225, 261, 401, 416,
IMGT/V-QUEST .................................... 569–602, 610 419, 422, 423, 425–429, 431–467, 472, 474, 497
Immunogenomics data analysis working group KIR. See Killer cell immunoglobulin like receptors (KIR)
(IDAWG) .............................................. 204, 206 KIR haplotypes
Immunoglobulin centromeric ......................... 261, 399–401, 406, 425, 492
heavy chain ........................................................ 635, 641 gene content ...................................................... 398–401
IgG........................................... . 271–272, 279, 280, 284, telomeric .................................................... 401, 408, 492
287, 292–294, 296, 315, 324–326, 341, 361, 369, Km allotypes ............................. 635, 637, 640, 667, 669, 670
370, 372, 374–377, 379–382, 385, 386, 388, 472,
L
473, 481, 536, 636–638, 642, 648, 650, 651, 660,
670, 672 LabScan™100 Flow Analyser ..................................... 54, 59
IgM.................................... 271–272, 296, 341, 359–361, LABType® SSO .................................................... 48–56, 61
366, 369–372, 375–377, 379 Lasergene™ software programme (DNASTAR, Madison
light chain .................................................. 635, 665, 667 USA).............................................. 153, 155, 157
Immunoglobulin superfamily (IgSF) ....................... 605–632 LCL. See Lymphoblastic cell lines (LCL)
Immunoinformatics ......................................................... 570 LD. See Linkage disequilibrium (LD)
Immunotherapy ........................................416, 510, 512, 513 LDA. See Limiting dilution assay (LDA)
Indels (insertions/deletions) ............................2, 4, 103–107, Leukapheresis ...........................................471, 495, 497, 498
111, 128, 129, 202, 206 Lifecodes™............................................ ................30, 48, 50,
Influenza virus ......................................................... 310, 318 56–61, 64
Innate immunity .............................................................. 392 Life Technologies™ .............................................71, 74, 314,
Intergenic regions ................................................................5 316–318, 331
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
686 Index

Limiting dilution assay (LDA) ................................ 310, 313 MICA

Linkage disequilibrium (LD) antibodies .......................................................... 279–285
global LD statistics ............................................ 227–228 antigen beads ......................................284, 287, 368, 376
haplotype level LD statisitics ..................................... 227 epitopes.............................................................. 285, 312
measurement.............................................................. 227 short tandem repeats.................................................. 184
Locus........................................... 2, 11, 51, 67, 88, 204, 216, typing................................................................. 183–195
247, 301, 396, 425, 432, 536, 557, 587, 636 MicroAmp™ Optical Plate (Applied
Logistic regression ....................................254, 256–257, 263 Biosystems) .................................... 163–165, 185
LumaPlate-96 (Packard Biosciences-Perkin Microsatellites ...................... 4, 202, 206, 210, 225, 231, 234
Elmer USA)........................................... 315, 323 Minor histocompatibility antigens
Luminex® definition and classification ............................... 509–526
flow analyser .............................................. 48, 54–55, 59 typing................................................................. 509–526
fluoranalyser .............................................................. 146 Mixed lymphocyte reaction (MLR) ....................... 318–320,
fluorochromes ............................................................ 292 326, 332
multianalyte bead assay .............................................. 291 MLPA. See Multiple ligation dependent probe
pooled antigen panel beads ................................ 295, 300 amplification (MLPA)
single antigen beads ...................................279, 283, 284, MLR. See Mixed lymphocyte reaction (MLR)
287, 295 Molecules of equivalent soluble fluorochromes
Xmap............................................................. 13, 48, 146 (MESF) ......................................... 306, 388, 389
XY platform................................................................. 48 Monensin................................. 317, 327, 334, 480, 483, 487
Lymph node ....................................................351–357, 392, Monocytes................................................ 319, 320, 494, 495
428, 491 Monte Carlo (MC) simulation method .......... 222–223, 241
Lymphoblastic cell lines (B-LCL).......................... 319, 323, MUD. See Matched unrelated donor (MUD)
324, 326, 327, 331, 340, 479, 480, 482–484, 487 Multiple ligation dependent probe amplification
Lymphoblasts .................................................................. 473 (MLPA) ......................................................... 161
Lymphocytes MultiScreen®-HV (Millipore, Billerica, MA,
B cells................................................. 357, 360–363, 374 USA).............................................................. 185
freezing and storage ................................... 353, 356, 375
memory T cells .................................................. 312, 392 N
T cells........................................ 309, 310, 318, 329, 332, NanoDrop spectrophotometer (NanoDrop technologies inc.
357, 360–363, 374, 405, 570 Wilmington DE, USA) ................. 465, 554, 557
Lymphocytic choriomeningitis ........................................ 310 National marrow donor programme (NMDP)
Lymphocytotoxic assay .................................................... 273 cell repository ............................................................ 435
codes............................................................. . 90, 91, 115
M
Natural killer (NK) cells ...................183, 391–397, 491–504
Major histocompatibility complex (MHC) Nei genetic distance......................................... 237, 238, 240
gene content .............................................................. 3–4 Neubauer counting chamber............. 352, 353, 355, 356, 375
MHC tetramers NKG2D receptor............................................................. 183
tetramer staining ...................312, 317–319, 328–330 NMDP. See National marrow donor programme (NMDP)
organization ....................................................... 600, 641 Nomenclature
MALDI-TOF. See Matrix assisted laser desorption antigen recognition sequence (ARS) ......................... 204
ionization time-of-flight (MALDI-TOF) HLA............................................ .94, 191, 202–206, 532
MamuA................................................................... 617, 629 indels (insertions/deletions) ....................................... 2, 4
Markov chain........................................................... 223, 241 KIR.................................................................... 207–209
Matched unrelated donor (MUD)................... 510, 511, 531 microsatellites ............................................................ 206
Matrix assisted laser desorption ionization time-of-flight nomenclature management................................ 202–209
(MALDI-TOF)..................................... 212, 426 SNP.........................................................2, 202, 206, 207
McCoy’s tissue culture medium ....................... 360, 363, 371 Update NomenCLature tool (UNCL) .............. 204–205
Mean fluorescence intensity (MFI) ........................ 157, 280, Non-Hodgkin lymphoma........................................ 268, 269
283–287, 296–299, 301, 305, 366, 367, 376 Nucleotides (dNTPs) ........................................................ 70
Melanocytes .................................................................... 511
MESF. See Molecules of equivalent soluble fluorochromes O
(MESF) Odds ratio (OR) .................................28, 161, 252, 254–257
MFI. See Mean fluorescence intensity (MFI) OR. See Odds ratio (OR)
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
Index
687

P sequence specific primers (PCR-SSP) ........................ 87,

144–146, 148–150, 156, 416–419, 425–426, 429,
Panel reactive antibody (PRA) ........................270–272, 285, 472, 514, 515, 522, 523, 526, 553, 554, 557–563
295, 303, 374 Taq DNA polymerase............................9, 10, 15–18, 20,
Papilloma virus ................................................................ 407 31–33, 38, 39, 128, 131, 435, 442, 454–456
Paraformaldehyde ....................................313, 317, 328, 334, thin walled 96 well PCR plates (ISC BioExpress,
341, 344, 480, 483 Kaysville UT, USA) ....................................... 146
PBL. See Peripheral blood lymphocytes (PBL) Population statistics .................................227–228, 245, 246,
PBMC. See Peripheral blood mononuclear cells (PBMC) 254, 258, 583–585, 599
PCA. See Principal component analysis (PCA) Population stratification........................................... 247, 259
PCR. See Polymerase chain reaction (PCR) Post transplant lymphoproliferative disease
PE. See Phycoerythrin (PE) (PTLD) ......................................................... 495
Perforin............................................................ 331, 391, 479 PRA. See Panel reactive antibody (PRA)
Performance optimized polymer (POP-6)-Applied Pre-sensitization ...................................................... 270–273
Biosystems ..................................................... 184 Principal component analysis (PCA).............................. 218,
Peripheral blood lymphocytes (PBL) ..................... 381, 382, 232–235, 240
392, 479, 481–482, 495 Promoter region....................... 128, 131–132, 137–139, 333,
Peripheral blood mononuclear cells 392, 397, 402–404, 550, 559–562
(PBMC) ................................310, 311, 313–314, Pronase......................................................380, 382, 385–386
318–321, 324, 326, 329, 330, 333, 334, 340, 342, PSQ. See Pyrosequencing (PSQ)
343, 347, 348, 381, 473, 479, 481–483, 486, 487 PTLD. See Post transplant lymphoproliferative disease (PTLD)
PFGE. See Pulsed field gel electrophoresis (PFGE) Pulsed field gel electrophoresis (PFGE) .......................... 161
PHA. See Phytohemagglutinin (PHA) PyroMark™ ....................................................................... 13
Phenol.............................................. 343, 442, 458, 463, 497 Pyrosequencing (PSQ) .....................128–131, 133–137, 140
Phenotype................................ 173, 199, 210–212, 217, 219, Python for Population Genomics
247, 257, 261, 290–291, 295–296, 299–301, 305, (PyPOp) ..........218, 219, 223, 224, 228, 230, 247
312, 313, 326, 327, 391, 416, 472–473, 484, 485,
510, 538, 544, 551, 638–639, 655, 662, 663, 669 Q
Phycoerythrin (PE) ............................... 30, 48, 53, 279, 280,
QA. See Quality assurance (QA)
287, 294, 312, 326, 327, 341, 342, 344, 347, 348,
QC. See Quality control (QC)
381, 383, 385, 387, 472–473, 481, 484, 488
QIAamp® DNA Mini Kit (Qiagen, Valencia,CA,
Phylogenetic analysis ....................................... 218, 234–239
USA).............................................................. 464
PHYLogeny inference package (PHYLIP) ............. 236–239
QIAamp® spin column (Qiagen) ............................. 77, 435,
Phytohemagglutinin (PHA) ............................316, 324, 340,
452, 453
342, 343, 473, 475
Quality assurance (QA) ........................................... 497, 503
Platinum Taq™ Polymerase (Invitrogen, Carlsbad,
Quality control (QC)........................... 55, 56, 61, 63, 88, 89,
CA, USA) ......................................131, 137, 138,
118, 119, 121, 220, 283, 290, 300, 365, 374, 423,
435, 454
498, 501–503
PLINK software (whole genome association analysis
Quanti-Marker (GeneMate Inc Kaysville UT,
toolset).................................................. . 218, 257
USA)...................................................... 185, 187
Polymerase chain reaction (PCR)
dNTPs. ...............................10, 15, 16, 18, 29, 31, 32, 38, R
41, 69, 70, 73, 77, 128, 131, 137, 145, 188, 521
long template PCR ....................................434, 443, 455, Real time PCR ...................... 31, 32, 43, 161–163, 165–167,
459–460 173–174, 177, 179, 180, 401, 426, 514, 553
nested PCR .........433, 434, 442, 443, 455–456, 465, 466 Reference strand conformational analysis
PCR buffer ................................................15, 17, 31, 32, (RSCA).................................................. 212, 514
38, 41, 73, 137, 145, 184, 435, 437, 438, 440, 442, Registries... .................................................87, 223, 531–546
453–456, 514, 520 Relative predispositional effects (RPE) ................... 251–254
primers............................4, 28, 29, 36, 44, 68, 69, 73, 87, Relative risk (RR) ...............................28, 147, 253–256, 268
144, 184, 186, 188, 420–422, 436, 442, 443, 446, Reporter dye .................................................................... 326
514, 520, 521, 526, 553, 554 Resampling approximations .....................221–223, 239, 255
sequence specific oligonucleotides Residue@ position (IMGT)...................................... 608, 609
(PCR-SSO) ............. 11–14, 29, 48, 87, 128, 146, Resolver™ SBT assay ........................................................ 88
150–153, 156, 419 Restriction endonuclease ......................................... 138, 442
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
688 Index

Restriction enzyme ................... 433, 442–443, 457–458, 665 Side scatter (SSC) .................... 329, 383, 386, 426, 484, 485
Restriction fragment length polymorphism Single nucleotide polymorphisms (SNPs) .................. 2, 4, 5,
(RFLP) ............... 46–47, 144, 161, 637, 639, 665 30–32, 42–45, 124–129, 134–136, 173–181, 202,
RFLP. See Restriction fragment length polymorphism 206, 207, 210, 224, 225, 253, 256, 257, 523,
(RFLP) 550–552, 554, 559, 560, 562
RoboSep® automated magnetic cell separation Single strand conformational polymorphism
(stem cell technologies).................................. 361 (SSCP)................................................... 184, 426
Rodgers (Rg) determinant ............................................... 161 Single transfected K562 cells (SALs) ...................... 339–349
Rosette sep® (stem cell technologies) ..................... 333, 480, SLE. See Systemic lupus erythematosus (SLE)
482, 487 Smith-Waterman score............................................ 612, 631
ROX™ dye. ..................................................... 176, 178, 179 SNPs. See Single nucleotide polymorphisms (SNPs)
RPE. See Relative predispositional effects (RPE) Sodium azide ............................................294, 381, 386, 481
R-phycoerythrin-conjugated streptavidin Solid phase immunoassay .......................30, 41–42, 289–306
(SAPE) ....................... .32, 40, 48–50, 53–55, 58, Somatic hypermutations ...................570, 575, 581, 584–585
59, 62, 64, 146, 152, 156, 157 Southern blotting ............................................................ 161
RPMI cell culture medium ......................314–317, 319, 320, Spleen........ ...............................................351–357, 428, 491
322, 333, 348, 352–354, 356, 361, 368, 380–382, SSP. See Sequence specific priming (SSP)
480–482, 487 STATA software package ................................................ 257
RR. See Relative risk (RR) Statistical analysis system (SAS®) ............197, 218, 248, 257
RSCA. See Reference strand conformational analysis (RSCA) Statistical package for social sciences (SPSS®) ........ 248, 257
RunModule Rapidseq™ (Applied Biosystems) ............... 461 Streptavidin
RunOne™ electrophoresis unit (Embi Tee, San Diego, Streptavidin-horseradish peroxidase (HRP) ................ 19
CA,USA) ....................................................... 436 SYBR Safe™ (Invitrogen) ............................... 130, 134, 423
Systemic lupus erythematosus (SLE) ...................... 160, 161
S
T
Saponin (Sigma) .............................................. 317, 328, 334
SBT. See Sequence based typing (SBT) TAMRA (6-carboxy-tetramethyl-rhodamine).. ............. 162,
SBTEngine® software for allele assignment (genome 174, 176, 177
diagnostics) .................................... 520, 522, 523 Taqman® chemistry
Sephadex G-50 (GE healthcare) ..................... 514, 521, 522 copy number assay ...................... 162, 164, 167, 168, 170
Sequence based typing (SBT)..........................14, 23, 28, 29, copy number reference RNase P ................................ 167
31–39, 45, 67, 70, 71, 74, 75, 81, 84, 87–121, 129, genotyping master mix ...................32, 42, 162, 164, 170
144, 146–148, 153, 157, 183–195, 212, 452, 456, T cell receptor (TCR)...................... 309, 339, 340, 343, 348,
462, 514, 526 396, 513, 540, 569–602, 605
Sequence electropherograms ............................... 84, 89–106 TCR. See T cell receptor (TCR)
Sequence specific oligonucleotide (SSO) assay TDT. See Transmission disequilibrium test (TDT)
biotinylated oligonucleotide probes. ................13, 16, 29, TdT. See Terminal deoxynucleotidyl transferase (TdT)
30, 44, 48, 130, 150, 325 Tepnel®
DELFIA hybridization ............................. 30, 32, 41–42 beads.... ...................................................................... 287
hybridization buffer ..............................41, 146, 456, 457 Terasaki plates ................................................. 361, 363, 371
microbead hybridization typing .................281–283, 388, Terminal deoxynucleotidyl transferase (TdT) ........ 259–260,
496, 497, 499 570, 578
reverse SSO ..................................................... 13–19, 48 Thermal sealing film (ISC BioExpress)........................... 185
wash buffer .........................24, 32, 40, 41, 49, 52–54, 62, Thermocycler............................... 9, 15–18, 21, 40, 144, 146,
130, 146, 152, 156, 280, 292, 293, 386 148–152, 154, 155, 185, 187–190, 422
Sequence specific priming (SSP) .......................9–24, 28, 29, Thermowell® sealing tape ....................................... 281–282
31–34, 44, 45, 47–48, 84, 87, 144, 161, 184, 212, Thymoglobulin ................................................................ 297
415–429, 432, 464, 514, 553 TOPO TA Cloning® kit (Invitrogen) ..................... 443, 459
Sequencher™ sequence software (Ann Arbor, TrackItTM DNA ladder (Invitrogen) ................................... 74
MI, USA) .............................................. 452, 463 Transfection .......290, 319, 323, 326, 327, 331, 332, 339–349
Sequencing primer. ........................ 36–38, 69, 70, 73, 78, 83, Transmission disequilibrium test (TDT)......... 259–260, 551
84, 89, 97, 105, 116, 118–119, 129, 131, 139, 147, Transplantation
155, 157, 184, 186, 188, 189, 193, 195, 432, 443, haematopoietic stem cell (HSCT) ...............14, 407, 408,
444, 460, 463, 464, 466, 521 469, 471, 478, 483, 486
 IMMUNOGENETICS: METHODS AND APPLICATIONS IN CLINICAL PRACTICE
Index
689
heart................................................................... 273–276 Vic® dye....................................................162, 165, 174, 175
kidney.......................... 184, 267–271, 273–276, 379–389 Viral immunity ................................................................ 312
liver........................................................... . 273–276, 407 Vogt-Koyanagi-Harda (VKH) disease .................... 406–407
pretransplantation ...............................268, 272, 274, 275 V-region...................................................574–578, 580–588,
Transversions ............................ 583, 599, 607, 631, 650, 670 591–593, 596, 598, 599, 601, 602, 609, 612, 613,
Treponema pallidum .......................................................... 503 619, 621, 630
Trypan blue ..............................................314, 320, 481, 487
W
U
Whatman® 96 well plate ................................................. 281
Ultraviolet (UV) transilluminator..........................16, 22, 52, World Health Organization International Non proprietary
58, 130, 131, 134, 135, 138, 146, 427, 436, 555 name (INN) programme................................ 601
Untranslated region ......................................... 124, 441, 554 World Health Organization-International Union of
Immunological Societies
V (WHO-IUS) ......................................... 600, 641
Variable domain....................................................... 570, 643
Y
V-domain................................. 570, 572, 577, 601, 606–607,
609–610, 612, 613, 619, 621, 624, 625, 630–632 Yates’ correction ........................................... 27–28, 248–249