Cell-Free

Circulating DNA
Purification and Analysis Techniques
 Cell-Free
Circulating DNA
Purification and Analysis Techniques

Editors

Kristina Warton
University of New South Wales, Australia

Goli Samimi
National Cancer Institute, USA

World Scientific
NEW JERSEY • LONDON • SINGAPORE • BEIJING • SHANGHAI • HONG KONG • TAIPEI • CHENNAI • TOKYO
 Published by
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601
UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

CELL-FREE CIRCULATING DNA

Purification and Analysis Techniques
Copyright © 2022 by World Scientific Publishing Co. Pte. Ltd.
All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means,
electronic or mechanical, including photocopying, recording or any information storage and retrieval
system now known or to be invented, without written permission from the publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance
Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy
is not required from the publisher.

ISBN 978-981-124-467-4 (hardcover)

ISBN 978-981-124-468-1 (ebook for institutions)
ISBN 978-981-124-469-8 (ebook for individuals)

For any available supplementary material, please visit

https://www.worldscientific.com/worldscibooks/10.1142/12494#t=suppl

Typeset by Stallion Press

Email: [email protected]

Printed in Singapore

Devi - 12494 - Cell-Free Circulating DNA.indd 1 6/7/2022 9:59:25 am

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_fmatter

Preface

Over the past decade a paradigm shift has occurred in the field of
circulating diagnostics and personalized oncology that has been
enabled by a dramatic expansion in high sensitivity technologies
and instrumentation. This shift has its foundations in a series of
discoveries that span more than a century and a half of clinical
research. The initial observation of circulating tumour cells (CTCs)
in the peripheral blood by the Australian physician Thomas Ash-
worth was published in 1869. The second fundamental discovery
was the presence of small fragments of cell-free DNA (cfDNA)
in blood by Mandel and Metais in 1948. Following on from this
Leon (1977) and Stroun (1989) determined that a fraction of the
cfDNA present in the blood of cancer patients was linked to their
tumour. Together, these biological findings delivered the critical
understanding that important genomic information is present in
the peripheral blood and has potential diagnostic application.
Although the early implementation of both CTCs and cfDNA
was focused on cancer prognostication and the detection of fetal
DNA, it wasn’t until 2010 when Pantel and Alix-Panabieres pro-
posed the term liquid biopsy that its broad diagnostic utility as
a minimally-invasive decision making tool was fully appreciated.
The liquid biopsy concept quickly evolved to include any analyte

b4444_FM.indd 5 08-Jul-22 6:12:30 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

vi Cell-Free Circulating DNA: Purification and Analysis Techniques

present in different biological fluids that could be harnessed for

diagnostic purposes. At the same time, the rise in the availability of
highly sensitive detection instrumentation, including next genera-
tion sequencing and digital PCR (dPCR), has enhanced the abil-
ity to detect and interrogate these analytes. This has opened the
way for liquid biopsies to be increasingly integrated into clinical
diagnostic workflows alongside traditional invasive tissue biopsy.
From a cancer perspective, the liquid biopsy of choice is circu-
lating tumour DNA (ctDNA); it can be applied in different can-
cer scenarios including early detection, minimal residual disease,
recurrence and treatment response monitoring, and can provide a
representative cross-section of tumour heterogenicity.
The rapid advances in ctDNA technologies have resulted in
pronouncements that we are now in the golden age of liquid biopsy
research. However, there are still many biological, technical stan-
dardisation and other pre-analytical and clinical diagnostic issues
(false negative/positive rates) that need to be addressed before the
promise of liquid biopsy achieves its full potential. In particular,
the issues surrounding technical variation and standardisation of
workflows are important considerations for all in the field. The
analytical validity of ctDNA assays has been mostly achieved, but
there are still areas remaining that require attention such as clin-
ical diagnostic assay validation and accreditation, along with the
demonstration of assay clinical utility before implementation into
routine clinical practice is possible. This book edited by Warton
and Samimi is a timely and practical compilation of articles that
will provide an important aid for researchers and clinicians wish-
ing to utilize ctDNA for translational research, clinical trials and
diagnostic purposes.
The first chapter focusing on the biology of cfDNA is a fun-
damental area of research interest. The understanding of cfDNA

b4444_FM.indd 6 08-Jul-22 6:12:30 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preface vii

biology is important as new knowledge on cfDNA origin, shed-

ding, fragmentation and dynamics can advance the development
of new analytical approaches and potentially improve pre-ana-
lytical processing. The subsequent three chapters reinforce the
need to be aware of critical pre-analytical factors, including sam-
ple type, as well as sample collection, processing and storage that
may significantly affect the reproducibility and analytical accuracy
of downstream assays. Quality assurance aspects involving inter-
nal and external controls, and reference materials are covered,
as well as considerations for biobanking in the context of clinical
trials. Pre-planning and careful attention to numerous logistical
and sample variables, including extraction methods is required to
ensure sufficient sample is available for biomarker evaluation and
assay development and is clearly set out by the authors in these
chapters. A detailed examination of factors influencing extraction
performance and an extensive review of extraction methods and
kits is provided in Chapter Four and should serve as a useful guide
for choosing the most appropriate method or kit. Another pre-an-
alytical factor that has been under-considered is blood DNases
and their effects on cfDNA samples and is presented in Chapter
Seven.
Knowing the intended analytical technique is important as
it influences the extraction method chosen, for example targeting
methylated cfDNA, mitochondrial cfDNA or total cfDNA. How-
ever, the final assay to be performed is not always known, especially
for biobanked samples earmarked for future unspecified research,
or due to the development of new analytical techniques. Chapters
Five and Six consider the analytical phase of cfDNA liquid biop-
sies and focus on qPCR, dPCR and next generation sequencing.
In addition, potential for PCR and sequencing biases exists and
require careful consideration of sample preparation approaches.

b4444_FM.indd 7 08-Jul-22 6:12:30 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

viii Cell-Free Circulating DNA: Purification and Analysis Techniques

Tailoring these techniques for the analysis of cfDNA and the criti-
cal factors involved is discussed in detail by the authors.
The final chapters reflect on circulating extracellular vesicles
(EV) and exosomes in addition to the application of methylated
DNA analysis. The authors summarise the origin and potential
for these EV/exosome based biomarkers and describe some of
the issues and challenges that still exist. The epigenetic analysis
of circulating methylated DNA is highly topical and has generated
great interest in the liquid biopsy community, especially cell-of-or-
igin methylation pattern analysis. The technology is now available
that allows for the extraction and analysis of small quantities of
fragmented methylated DNA. However, its application requires a
good understanding of methylated DNA biology and these aspects
are clearly and concisely described by the authors.
The growth of liquid biopsy research involving both genetic
and non-genetic molecular targets has resulted in important new
insights and a critical understanding of circulating biomarkers and
their diagnostic utility. This book makes an important contribution
to the field and contains a wealth of up-to-date information on cir-
culating DNA techniques and summarises the major advances that
are essential for the implementation of liquid biopsies into routine
clinical practice.

Kevin J Spring

b4444_FM.indd 8 08-Jul-22 6:12:30 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Contents

Preface v
List of Contributorsxi

Chapter 1 Biology of Circulating DNA

in Health and Disease 1
Ahuva Odenheimer-Bergman, Havell Markus and
Muhammed Murtaza

Chapter 2 Preanalytics and Quality Control

of Different Substrates for Circulating
DNA Studies 21
Fay Betsou and Wim Ammerlaan

Chapter 3 Considerations in Biobanking Blood

Samples and Plasma Volume for
Circulating DNA Studies 47
Kate Mahon, Goli Samimi and Kristina Warton

ix

b4444_FM.indd 9 08-Jul-22 6:12:30 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

x Cell-Free Circulating DNA: Purification and Analysis Techniques

Chapter 4 Maximizing Yield from Plasma

Circulating DNA Extraction 63
Florence Mauger and Jean-François Deleuze

Chapter 5 Optimal Design of PCR Assays for

Circulating DNA 113
Rikke F. Andersen

Chapter 6 Preparation of Next-Generation

Sequencing Libraries for Sequencing
Circulating DNA 139
Katrin Heider, Florent Mouliere and
Christopher G. Smith

Chapter 7 Blood Nucleases Affecting Circulating

DNA in Serum and Plasma 175
Gustavo Barra

Chapter 8 Isolating Circulating Exosomes as

Biomarkers: Challenges and
Opportunities209
Alexander Semaan and Anirban Maitra

Chapter 9 DNA Methylation Analysis of

Circulating DNA Biomarkers 247
Kristina Warton, Clare Stirzaker,
Goli Samimi and Susan Clark

Index279

b4444_FM.indd 10 08-Jul-22 6:12:30 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

List of Contributors

Ahuva Odenheimer-Bergman
Translational Genomics Research Institute, Phoenix, AZ, USA

Dr. Havell Markus

Penn State College of Medicine, Hershey, PA, USA

Prof. Muhammed Murtaza

Translational Genomics Research Institute, Phoenix, AZ; Depart-
ment of Surgery and Center for Human Genomics and Precision
Medicine, University of Wisconsin-Madison, Madison, WI, USA

Prof. Fay Betsou

Laboratoire National de Santé, Luxemburg

Wim Ammerlaan
Integrated Biobank of Luxemburg, Luxemburg

Dr. Kate Mahon

Chris O’Brien Lifehouse, Sydney, Australia

Dr. Goli Samimi

National Cancer Institute, Bethesda, MD, USA

xi

b4444_FM.indd 11 08-Jul-22 6:12:30 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

xii Cell-Free Circulating DNA: Purification and Analysis Techniques

Dr. Kristina Warton

University of New South Wales, Sydney, Australia

Dr. Florence Mauger

Université Paris-Saclay, CEA, Centre National de Recherche en
Génomique Humaine, Evry, France

Dr. Jean-François Deleuze

Université Paris-Saclay, CEA, Centre National de Recherche en
Génomique Humaine, Evry, France

Dr. Rikke F. Andersen

Lillebaelt Hospital, Vejle, Denmark

Dr. Katrin Heider

University of Cambridge, Cambridge, UK

Dr. Florent Mouliere

University of Cambridge, Cambridge, UK

Dr. Christopher G. Smith

University of Cambridge, Cambridge, UK

Dr. Gustavo Barra

Sabin Medicina Diagnóstica, Brasilia, Brazil

Dr. Alexander Semaan

MD Anderson Cancer Center, Houston, USA

Prof. Anirban Maitra

MD Anderson Cancer Center, Houston, USA

b4444_FM.indd 12 08-Jul-22 6:12:30 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

List of Contributors xiii

A/Prof. Clare Stirzaker

Garvan Institute of Medical Research and St. Vincent’s Clinical
School, Faculty of Medicine, UNSW, Sydney, Australia

Prof. Susan Clark

Garvan Institute of Medical Research and St. Vincent’s Clinical
School, Faculty of Medicine, UNSW, Sydney, Australia

b4444_FM.indd 13 08-Jul-22 6:12:30 PM

 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0001

Chapter 1
Biology of Circulating DNA
in Health and Disease
Ahuva Odenheimer-Bergman,
Havell Markus and Muhammed Murtaza

Abstract

Circulating tumor DNA (ctDNA) analysis holds tremendous diagnostic

value throughout cancer management. Potential applications of ctDNA
analysis include minimally invasive tumor genotyping, monitoring treat-
ment response, tracking subclonal evolution, detecting recurrence or
minimal residual disease, and early detection. While some applications
are already clinically available, numerous clinical studies are currently
ongoing across cancer subtypes and disease stages and several groups
are developing and adapting molecular analysis methods ideally suited
for each application. While excitement about potential clinical value of
ctDNA analysis has mounted, our understanding of the biology of cir-
culating DNA has lagged. Mechanisms of circulating DNA release from
cells, degradation, and clearance remain poorly understood. Recent
studies have shown that leveraging insights into circulating DNA biol-
ogy when developing biomarkers greatly improves accuracy and clini-
cal relevance. In this chapter, we describe current understanding and

b4444_Ch-01.indd 1 08-Jul-22 5:56:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

2 Cell-Free Circulating DNA: Purification and Analysis Techniques

unanswered questions about the biology of circulating DNA, particularly

highlighting areas with relevance to preanalytical and analytical tech-
niques and methods for biomarker development.

Introduction
The presence of extracellular DNA fragments in serum was first
described by Mandel and in 1948.1 Hundreds of studies have
since evaluated circulating DNA as biomarkers using increasingly
sophisticated molecular analysis methods, across multiple diseases
and physiological conditions. Initial studies compared total circu-
lating cell-free DNA (cirDNA) levels in serum or plasma between
healthy individuals and patients with acute illnesses (such myo-
cardial infarction, stroke, infections, or trauma) or with chronic
diseases (such as cancer or autoimmune diseases). While cirDNA
levels are generally higher in patients, they can rarely differentiate
adequately between healthy and diseased individuals due to over-
lapping distributions. Enabled by advances in molecular analysis
and sequencing technologies, recent efforts have focused on mea-
suring disease-specific cirDNA fragments. Noninvasive prenatal
diagnostics (NIPD) is the first successful example of such efforts,
focusing on fetal DNA in plasma from pregnant mothers. Since
the first definitive report detecting the presence of chrY DNA
fragments (specific to male fetuses) in maternal plasma published
in 1997,2 advances in this field have already led to clinically vali-
dated and commercially available screening tests for fetal aneu-
ploidies and changed the diagnostic paradigm in prenatal testing.3
The field of circulating tumor DNA (ctDNA) analysis hopes
to emulate the success in NIPD. While several aspects of circulat-
ing DNA biology learnt from prenatal testing are relevant, there
are unique challenges and opportunities in patients with cancer.

b4444_Ch-01.indd 2 08-Jul-22 5:56:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 3

ctDNA analysis relies on measuring cell-free DNA fragments that

carry somatic genomic alterations specific to cancer cells.4 These
include single nucleotide variants, small insertions and deletions,
rearrangement breakpoints, as well as cancer-specific DNA meth-
ylation changes. In addition, ctDNA analysis can also measure
differences in cell-free DNA abundance across genomic loci,
reflecting somatic copy number aberrations.
The presence of cancer-specific mutations in cirDNA was first
reported in the 1970s5 and its clinical relevance was demonstrated
in 2004.6 Recent technological advances in molecular analysis and
DNA sequencing have enabled several potential diagnostic appli-
cations including noninvasive tumor genotyping,7 treatment mon-
itoring,8 tracking clonal evolution in advanced cancer patients,9,10
recurrence monitoring and minimal residual disease detection in
early stage cancer patients,11,12 as well as early detection of can-
cer in pre-symptomatic individuals.13,14 Among these applications,
noninvasive tumor genotyping for patients with metastatic cancers
is furthest in its development and available for clinical testing from
multiple commercial service providers today.
Cancer-specific mutations identified in ctDNA from plasma
samples are generally concordant with similar analysis performed
on tumor tissue samples from the same patients with metastatic
cancer, except when affected by tumor heterogeneity.15 However,
cancer mutations identified in tumor tissue can often be missed in
plasma samples, highlighting the challenge of detecting low levels
of ctDNA in plasma and the need for high-sensitivity analytical
methods. Ongoing efforts in this area are focused on establishing
consensus criteria for analytical and clinical validity and evaluating
the clinical utility of ctDNA-based genotyping. Several examples
in the literature demonstrate the relevance of ctDNA analysis for
other diagnostic scenarios, but these generally still require further

b4444_Ch-01.indd 3 08-Jul-22 5:56:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

4 Cell-Free Circulating DNA: Purification and Analysis Techniques

technical/analytical improvements and definitive evaluation of

clinical validity.16
While excitement regarding clinical applications of ctDNA
analysis across cancer types and diagnostic scenarios has grown,
new insights into cell-free DNA biology have been limited.
Recent studies have demonstrated how biological characteristics
of cirDNA can be leveraged to enable novel diagnostic applica-
tions for cancer patients. For example, genomic positioning of
cell-free DNA fragments in plasma was recently shown to reflect
dynamic nucleosome positioning in the cells of origin, with nucle-
osome binding postulated to protect the DNA from digestion
by plasma nucleases. This observation led to two novel poten-
tial applications in metastatic cancer patients, where ctDNA
levels in plasma are often very high. First, it enabled inference
of histological tumor cell type, with potential relevance for can-
cer patients with unknown primary tumor sites.17 Second, with
deeper sequencing and in a fraction of patients with the highest
ctDNA levels, differential plasma DNA abundance at and around
the transcription start sites enabled inference of gene expres-
sion in the tumor cells.18 Another example of a biological insight
enabling better diagnostics is observation of differences in ctDNA
fragment size. Several studies have shown that while most plasma
DNA fragments are mono-nucleosomal in size (~167 bp), fetal
and tumor-derived DNA fragments tend to be shorter, poten-
tially lacking nucleosome linker DNA regions due to differences
in mechanisms of DNA degradation and release. Recent efforts
are now relying on size selection of cell-free DNA fragments to
enrich for tumor signal, either during sequencing library prepa-
ration or using informatics tools.19
To enable further such advances, in this chapter, we summa-
rize current understanding and unanswered questions in circulating

b4444_Ch-01.indd 4 08-Jul-22 5:56:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 5

cell-free DNA biology, with a specific focus on areas relevant to

diagnostic development.

Total Circulating Cell-Free DNA Concentration

Several studies have evaluated the diagnostic value of total
cirDNA levels across disease types. Although significantly higher
levels are often observed when comparing cancer patients with
healthy volunteers, overlapping observations of total cirDNA
levels between the two groups limit their diagnostic accuracy,
at least in univariate analysis.20 In addition, total cirDNA levels
vary between individuals and are affected by physiological dif-
ferences such as body mass index as well as disease states such as
autoimmune and rheumatic disorders.21,22 They are also affected
by transient physiological conditions such as exercise or fasting.
The effects of diet, food and fluid intake or fasting, physical activ-
ity, and time of the day (circadian rhythm) on total cirDNA and
tumor-specific ctDNA remain poorly understood. For example,
several studies have found exercise increases total cirDNA con-
centrations in healthy volunteers.23–29 However, the source of the
additional cirDNA released during exercise is unclear, and it is
not known whether exercise or physical stress will transiently
dilute tumor-specific mutation levels in cancer patients. Total
cirDNA levels are also affected heavily by preanalytical factors
such as delays between venipuncture and processing of blood
samples, and differences between blood collection protocols and
DNA extraction protocols.30 Delays in blood processing or inap-
propriate blood collection protocols can lead to lysis of peripheral
blood cells, leading to artifactual increases in total cirDNA. These
unanswered questions further limit the utility of total cirDNA
levels for cancer diagnostics.

b4444_Ch-01.indd 5 08-Jul-22 5:56:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

6 Cell-Free Circulating DNA: Purification and Analysis Techniques

cirDNA levels do not appear to be influenced by circadian

rhythm; however, the data available is limited. In healthy vol-
unteers, a comparison of total cirDNA levels in plasma samples
obtained at different times of the day showed no significant differ-
ence.31,32 In cancer patients, one recent study evaluated multiple
plasma samples obtained from nine colorectal cancer patients at
four different times of the day and suggested total cirDNA lev-
els were lowest and tumor-specific ctDNA was most detectable at
mid-night.33 Based on these preliminary results, they inferred that
daytime activity could potentially dilute tumor-derived DNA lev-
els in circulation and an early morning sample may be most infor-
mative. Another study compared histone modifications in cirDNA
between different times of blood collection and found no signif-
icant differences in 5mC or H3K9me3 histone modification lev-
els.34 Despite limitations, it is conceivable that total cirDNA levels
may add value as part of multiparametric assessments, where total
cirDNA levels are interpreted together with other variables such
as tumor-specific ctDNA levels, fragmentation patterns, and size
distributions.35
Total cirDNA levels also determine the amount of DNA
available for analysis from a clinically obtainable volume of
blood, a key variable that affects clinical assay sensitivity, accu-
racy, and precision. Despite several recent advances in detection
of low abundance mutations using next-generation sequencing
or digital PCR, actual sensitivity is often limited by low total
cirDNA amounts available from patient samples and by loss of
analyte due to technical inefficiencies. In healthy volunteers
with no known cancers, total cirDNA concentration is typically
5–7 ng/mL of plasma (equivalent to 1500–2100 haploid genome
copies or hGCs).30 A 10 mL blood sample typically yields about
4 mL plasma, resulting in ~6000–8400 hGCs. Due to sampling

b4444_Ch-01.indd 6 08-Jul-22 5:56:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 7

variation, any molecular assay is expected to have limited sen-

sitivity for individual mutations at <0.1% AF in 10 mL blood
samples (<6–8 mutated DNA fragments available on average
for detection for each mutation). Total cirDNA concentrations
are expected to be similar in pre-symptomatic cancer patients (a
screening or early detection population) or in localized cancer
patients.11,12 In contrast, plasma samples from metastatic cancer
patients have higher total cirDNA concentrations and yield more
analyte, circumventing concerns for sampling errors.20

Clinical Tumor Characteristics Affecting Circulating Tumor

DNA Levels
ctDNA levels can be affected by biology of the tumor or clinical
variables such as ongoing treatments. Differences in tumor dou-
bling time, number of proliferating cells, and level of tumor loss
can lead to different ctDNA levels across cancer types. For exam-
ple, in one study, colorectal cancer was shown to have a doubling
time of 90 days, cell loss factor of 96%, and level of cell prolifera-
tion of 15%, and patients with colorectal cancer have some of the
highest levels of ctDNA.15,36 In contrast, a different study showed
thyroid cancer has a tumor doubling time of 212 days and ctDNA
levels observed for thyroid cancer are typically much lower.37,38
Within each tumor type, several studies have reported associa-
tion of ctDNA levels with mitotic rate or Ki67 levels39,40 as well as
tumor vascularity and hypoxia.41–43
In recent studies of lung cancer, ctDNA levels have been cor-
related with tumor volume although such data are limited across
most tumor types.44 In addition, whether a linear correlation holds
when tumor volumes are extremely low and undetectable on imag-
ing is unclear. For example, a linear model of ctDNA levels prior

b4444_Ch-01.indd 7 08-Jul-22 5:56:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

8 Cell-Free Circulating DNA: Purification and Analysis Techniques

to treatment with tumor volumes suggested ctDNA levels would

be below limit of detection post-operatively using the same assay
in one study.39 However, ctDNA was detected in several cases with
residual disease, suggesting that the model may not hold across
tumor volumes or throughout stages of treatment and response.
A potential explanation for a higher rate of ctDNA shedding from
residual disease may be a more aggressive, growth phenotype in
the systemic residual tumor upon resection of the primary site.
Recent studies identify minimum tumor volumes where ctDNA
is detectable, but these calculations are bound to be tumor-type
specific and affected by the limit of detection for assays used.45
Tumor-derived ctDNA levels can be affected acutely due to
treatment-related events. For example, ctDNA levels tend to spike
within hours of surgical manipulation of the tumor or infusion of
chemotherapy.46,47 While a relationship with intravenously infused
cytotoxic chemotherapy has been observed, how oral molecularly
targeted agents or immune checkpoint inhibitors affect ctDNA
levels acutely is not clear. In some studies, acute increases in
tumor-specific ctDNA have been useful as markers of treatment
response.48,49

Half-Life and ctDNA Clearance

Fetal DNA is cleared from maternal plasma using a two-phased
clearance mechanism, which begins with a rapid clearance in the
first 2-hr postpartum, and then a second slower phase, which takes
up to 2 days postpartum to clear the remaining fetal DNA from
maternal plasma.50 Tumor-derived DNA has been observed to
have a similar half-life of 114 min, observed in colorectal cancer
patients after surgical resection.6,47 Experimental studies evaluat-
ing the half-life of cirDNA have made similar observations. One
group studied clearance of mono-nucleosomes with radiolabeled

b4444_Ch-01.indd 8 08-Jul-22 5:56:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 9

histones injected intravenously into mice and found that within

10 min, 70%–85% of the signal was detected in the liver and <1%
found in the kidneys.29,51
To be finally cleared from the body, cirDNA degradation
products are likely excreted in urine. However, mononucleosomes
are unlikely to filter into urine through the glomeruli (90-kDa-
sized mononucleosomes and a maximum size cutoff of 70 kDa).
The negative charge of DNA likely lowers the permeability fur-
ther. Although several groups have suggested observing cell-free
DNA fragments of ~150 bp size in urine, whether such DNA is
trans-renal or locally shed by the uro-epithelium is unclear and
a mechanism that could facilitate filtration of mono-nucleosomes
into urine is unknown. Other groups have observed much shorter
DNA fragments in urine, suggesting further degradation of
cirDNA prior to filtration and within the genitourinary tract.50,52,53
The ends of cirDNA fragments in plasma and urine reveal
distinct but conserved sequence motifs across multiple studies,
suggesting the role of distinct nucleases in DNA digestion in
plasma and in urine.53–55 Recent studies have shown that DNAse1-
like3 is likely responsible for degrading chromatin in plasma while
DNAse1 is more active in urine and acts on DNA not bound to
nucleosomes. The sequence motifs on plasma DNA fragments
show a preference for cytosine at the fragment ends, while those
on urine DNA fragments show a preference for thymine.53 These
observations are consistent with the sequence motifs related to
corresponding nucleases in plasma and urine.56,57

Circulating DNA Fragment Size and Positioning

Circulating DNA in plasma is predominantly fragmented to ~167
bp, corresponding to DNA length wrapped in mono-nucleosomes.
Additional modal sizes are observed consistent with di-nucleosomes

b4444_Ch-01.indd 9 08-Jul-22 5:56:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

10 Cell-Free Circulating DNA: Purification and Analysis Techniques

and tri-nucleosomes. In addition, the size distribution shows a

10 bp step pattern, suggesting enzymatic degradation.58 These
observations suggest cirDNA is shed during cell turnover by
apoptosis, although alternative mechanisms for enzymatic degra-
dation cannot be ruled out, such as secondary digestion of longer
DNA fragments shed by necrosis by enzymes in the bloodstream,
or release of fragmented DNA by phagocytes that engulf and
digest dying cells.6,59–61 Longer DNA fragments are observed in
some blood samples, but it is generally assumed that these are
contributed by lysis of cells in the bloodstream, either in vivo
or during venipuncture and sample processing. Delays in and
differences in preanalytical sample processing have been shown
to increase fraction of longer DNA fragments in plasma sam-
ples and to lower measures of tumor-derived ctDNA.30 In some
patients with metastatic cancers with high tumor burden and
necrotic tumors, it is likely that longer fragments are contributed
by necrotic tumor cells.62
Assaying short fragments is a key consideration during circulat-
ing DNA analysis. Conventional DNA extraction approaches often
prioritize intact over “degraded” DNA and this results in lost yield.
Newer cirDNA-focused extraction methods can overcome this lim-
itation. Similarly, any negative selection of shorter fragments down-
stream (e.g., during sequencing library preparation and cleanup)
can lower effective yield. Due to ongoing degradation, sequenc-
ing library preparation that can capture single-stranded DNA has
been shown to capture a larger fraction of cell-free DNA fragments
shorter than 100 bp.63 Short fragment size also limits the size of
amplicons designed for target-specific PCR analysis (digital PCR or
amplicon sequencing). Assuming cirDNA is fragmented randomly,
the maximum fraction of DNA fragments captured by both PCR
primers decreases with increasing amplicon size, predictable using

b4444_Ch-01.indd 10 08-Jul-22 5:56:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 11

the exponential distribution. For example, a 120-bp amplicon will

capture ~45% of 150-bp DNA fragments, while a 71-bp amplicon
will capture ~62% of ~150-bp DNA fragments.30
Recent studies have shown cirDNA fragmentation is not com-
pletely random and is consistent with protection from complete
degradation by nucleosomes.53,64 Using whole genome sequencing,
studies found relative levels and periodicity in cirDNA abundance
across genomic loci were consistent with micrococcal nuclease
digestion, which cleaves DNA at inter-nucleosomal sites and has
traditionally been used to map nucleosome positioning.65 Actively
transcribed genes have a nucleosome-depleted region at tran-
scription start sites and in relative sequencing coverage in cirDNA
at these genomic loci is correlated with gene expression in tissues
that contribute cirDNA. In patients with advanced metastatic can-
cers where tumor-derived ctDNA fractions are high, sequencing
coverage and fragment size at transcription start sites and known
transcription factor-binding loci can be leveraged to identify pri-
mary tumor cell types and infer gene expression.17,18,66 To maxi-
mize analytical yield and potentially enrich tumor-derived signal,
design of PCR assays can be further informed by the nonrandom
fragmentation of cirDNA. However, due to differences expected
between contributing cell types, dynamic changes in nucleosome
positioning, and inability to predict nucleosome positioning pre-
cisely for most genomic loci, no published ctDNA studies have
leveraged nonrandom positioning to aid PCR assay design so far.
Several studies have observed that circulating DNA frag-
ments shed by fetal cells or tumor cells tend to be shorter than
maternal cells or healthy wild-type cells. In prenatal diagnostics
studies, fetal-specific DNA has been found to have modal size of
~147 bp, about 20 bp shorter than maternal-specific DNA ~166 bp,
and the relative fraction of these sizes could be used to infer fetal

b4444_Ch-01.indd 11 08-Jul-22 5:56:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

12 Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA fraction in circulation.58,67 A follow-up study demonstrated

that cirDNA has preferential end sites (PES), genomic positions
where fragments are more likely to be cleaved, associated with
long and short fragments.68 Using this set of long and short PES,
it was shown that maternal cirDNA is more likely to be cleaved at
long PES, and fetal cirDNA is more likely to be cleaved at short
PES. Long PES associated with maternal DNA tend to fall in the
linker region, while short PES associated with fetal DNA tend to
fall closer to or within the nucleosome core. This observation sug-
gests fetal DNA may be more accessible to nuclease cleavage than
maternal DNA, perhaps due to more active transcription across
the genome in rapidly proliferating cells. Selecting fragments to
analyze based on the PES enabled a 90-fold enhancement in the
resolution of the fetal genome.69
Similarly, relative shortening of circulating DNA has also
been observed in tumor-derived DNA from plasma70 and cerebro-
spinal fluid,71 and selection of shorter fragments before or after
sequencing library preparation has been leveraged to enrich tumor
signal.19,72 Using in silico size selection, Jiang et al. showed a signifi-
cant positive correlation between the proportion of fragments less
than 150 bp and tumor DNA concentration in plasma of patients
with hepatocellular carcinoma, while fragments between 150 and
180 bp showed no correlation.70 Mouliere et al. selected fragments
between 90 and 150 bp in plasma samples of patients with various
cancer types and achieved a median 2-fold enrichment of ctDNA.
Through size selection, they were also able to identify clinically
relevant/actionable somatic mutations and copy number alteration
events that were otherwise not observed.19
Cell-type-specific DNA organization and differences in processes
underlying cell death can result in cirDNA with cell-type-specific

b4444_Ch-01.indd 12 08-Jul-22 5:56:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 13

nonrandom fragmentation signatures.53 Chandrananda et al.

described that cirDNA fragmentation in healthy plasma samples is
not random as the sequences of first and last 10 bp of cirDNA frag-
ments show a conserved pattern for nucleotide preference.54 Chen
et al. followed this observation and used the nucleotide motif pres-
ent in the first and last 10 bp of fragments to build a machine learn-
ing classifier to distinguish cirDNA from plasma and urine, and
samples from formalin-fixed paraffin embedded (FFPE) tumor
DNA samples.55 Although they did not find a conserved fragmen-
tation pattern in cirDNA from cerebrospinal fluid, this could be
due to a limited sample set or different processing methods. The
study further showed that these fragmentation patterns are sim-
ilar in every fragment size (long or short), thus suggesting short
fragments present in circulation are not due to degradation of
long fragments. Findings of these studies suggest cirDNA carries
important structural and functional information that could poten-
tially be used to improve bioinformatics algorithms to distinguish
ctDNA among wild-type cirDNA based on molecular features in
addition to fragment sizes.
It is unclear how much of the cirDNA in circulation is
enclosed within membrane-bound vesicles or bound to cell
surface.73 Extracellular vesicles have been shown to contain mul-
tiple cellular components including proteins, microRNAs and
messenger RNAs, as well as single- and double-stranded DNA.74
In cancer patients, exosomes derived from tumors may carry dou-
ble-stranded tumor-derived and mutated DNA.75,76 This suggests
selection of vesicles using membrane and cell-surface markers
may enrich tumor-derived ctDNA signal, above background noise
in molecular assays. However, any such gains in tumor fraction
will have to be balanced against loss of material associated with
enrichment strategies.74,76,77

b4444_Ch-01.indd 13 08-Jul-22 5:56:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

14 Cell-Free Circulating DNA: Purification and Analysis Techniques

Summary
Analysis of cell-free DNA has already transformed diagnostics in
multiple areas of medicine, without much initial regard to bio-
logical determinants of cirDNA in health and disease. However,
diagnostic development has arrived at stiff limits of detection for
methods based on single nucleotide variants and polymorphisms
that can no longer be overcome with greater depth of sequenc-
ing while analyzing feasible volumes of blood. This has spurned
rapid progress into analysis of biological features of cirDNA that
could be leveraged to improve diagnostic performance, making
it the current frontier in this field. There are several unanswered
questions that remain in cell-free DNA biology, which could have
profound impacts on the choices and accuracy of preanalytical and
analytical techniques used for cell-free DNA analysis.

References
1. Mandel P, Metais P. (1948) Les acides nucléiques du plasma sanguin chez
l’homme. C R Seances Soc Biol Fil 142: 241–243.
2. Lo YM, Corbetta N, Chamberlain PF, et al. (1997) Presence of fetal DNA in
maternal plasma and serum. Lancet 350: 485–487.
3. Wong FC, Lo YM. (2016) Prenatal diagnosis innovation: Genome sequenc-
ing of maternal plasma. Annu Rev Med 67: 419–432.
4. Wan JC, Massie C, Garcia-Corbacho J, et al. (2017). Liquid biopsies come of
age: Towards implementation of circulating tumour DNA. Nat Rev Cancer
17: 223–238.
5. Leon S, Shapiro B, Sklaroff D, Yaros M. (1977) Free DNA in the serum of
cancer patients and the effect of therapy. Cancer Res 37: 646–650.
6. Diehl F, Li M, Dressman D, et al. (2005) Detection and quantification of
mutations in the plasma of patients with colorectal tumors. Proc Natl Acad
Sci USA 102: 16368–16373.
7. Forshew T, Murtaza M, Parkinson C, et al. (2012) Noninvasive identifica-
tion and monitoring of cancer mutations by targeted deep sequencing of
plasma DNA. Sci Transl Med 4: 136ra68–ra68.

b4444_Ch-01.indd 14 08-Jul-22 5:56:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 15

8. Dawson SJ, Tsui DW, Murtaza M, et al. (2013) Analysis of circulating tumor
DNA to monitor metastatic breast cancer. N Engl J Med 368: 1199–1209.
9. Murtaza M, Dawson SJ, Pogrebniak K, et al. (2015) Multifocal clonal evo-
lution characterized using circulating tumour DNA in a case of metastatic
breast cancer. Nat Commun 6: 8760.
10. Murtaza M, Dawson SJ, Tsui DW, et al. (2013) Non-invasive analysis
of acquired resistance to cancer therapy by sequencing of plasma DNA.
Nature 497.
11. McDonald BR, Contente-Cuomo T, Sammut SJ, et al. (2019) Personalized
circulating tumor DNA analysis to detect residual disease after neoadjuvant
therapy in breast cancer. Sci Transl Med 11: eaax7392.
12. McDonald BR, Contente-Cuomo T, Sammut S-J, et al. (2018) Detection of
residual disease after neoadjuvant therapy in breast cancer using personal-
ized circulating tumor DNA analysis. bioRxiv.
13. Phallen J, Sausen M, Adleff V, et al. (2017) Direct detection of early-stage
cancers using circulating tumor DNA. Sci Transl Med 9: eaan2415.
14. Cohen JD, Li L, Wang Y, et al. (2018) Detection and localization of surgi-
cally resectable cancers with a multi-analyte blood test. Science eaar3247.
15. Bettegowda C, Sausen M, Leary RJ, et al. (2014) Detection of circulating
tumor DNA in early- and late-stage human malignancies. Sci Transl Med 6:
224ra24.
16. Merker JD, Oxnard GR, Compton C, et al. (2018) Circulating tumor DNA
analysis in patients with cancer: American Society of Clinical Oncology and
College of American Pathologists joint review. Arch Pathol Lab Med 142:
1242–1253.
17. Snyder MW, Kircher M, Hill AJ, et al. (2016) Cell-free DNA comprises
an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164:
57–68.
18. Ulz P, Thallinger GG, Auer M, et al. (2016) Inferring expressed genes by
whole-genome sequencing of plasma DNA. Nat Genet 48: 1273–1278.
19. Mouliere F, Chandrananda D, Piskorz AM, et al. (2018) Enhanced
detection of circulating tumor DNA by fragment size analysis. Sci Transl
Med 10.
20. Farooq M, Egan JB, McDonald B, et al. (2018) Detection of copy number
aberrations in cholangiocarcinoma using shallow whole genome sequencing
of plasma DNA. Am Soc Clin Oncol (36)4, Abstract 293.

b4444_Ch-01.indd 15 08-Jul-22 5:56:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

16 Cell-Free Circulating DNA: Purification and Analysis Techniques

21. Duvvuri B, Lood C. (2019) Cell-free DNA as a biomarker in autoimmune

rheumatic diseases. Front Immunol 10: 502.
22. Vora NL, Johnson KL, Basu S, et al. (2012) A multifactorial relationship
exists between total circulating cell-free DNA levels and maternal BMI.
Prenat Diagn 32: 912–914.
23. Hummel EM, Hessas E, Muller S, et al. (2018) Cell-free DNA release
under psychosocial and physical stress conditions. Transl Psychiatry 8:
236.
24. Tug S, Helmig S, Deichmann ER, et al. (2015) Exercise-induced increases
in cell free DNA in human plasma originate predominantly from cells of the
haematopoietic lineage. Exerc Immunol Rev 21: 164–173.
25. Fruhbeis C, Helmig S, Tug S, et al. (2015) Physical exercise induces rapid
release of small extracellular vesicles into the circulation. J Extracell Vesicles
4: 28239.
26. Breitbach S, Tug S, Simon P. (2012) Circulating cell-free DNA: An up-coming
molecular marker in exercise physiology. Sports Med 42: 565–586.
27. Velders M, Treff G, Machus K, et al. (2014) Exercise is a potent stimulus for
enhancing circulating DNase activity. Clin Biochem 47: 471–474.
28. Chevion S, Moran DS, Heled Y, et al. (2003) Plasma antioxidant status
and cell injury after severe physical exercise. Proc Natl Acad Sci USA 100:
5119–5123.
29. Pokrywka A, Zembron-Lacny A, Baldy-Chudzik K, et al. (2015) The influ-
ence of hypoxic physical activity on cfDNA as a new marker of vascular
inflammation. Arch Med Sci 11: 1156–1163.
30. Markus H, Contente-Cuomo T, Farooq M, et al. (2018) Evaluation of
pre-analytical factors affecting plasma DNA analysis. Sci Rep 8: 7375.
31. Korabecna M, Horinek A, Bila N, Opatrna S. (2011) Circadian rhythmicity
and clearance of cell-free DNA in human plasma. In Circulating Nucleic
Acids in Plasma and Serum.Springer Netherlands, Dordrecht, pp. 195–198.
32. Marie Korabecna AH, Nikola B, Sylvie O. (2010) Circadian rhythmicity and
clearance of cell-free DNA in human plasma. Circulating Nucleic Acids in
Plasma and Serum.
33. Toth K, Patai AV, Kalmar A, et al. (2017) Circadian rhythm of methylated
septin 9, cell-free DNA amount and tumor markers in colorectal cancer
patients. Pathol Oncol Res 23: 699–706.

b4444_Ch-01.indd 16 08-Jul-22 5:56:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 17

34. Rasmussen L, Herzog M, Romer E, et al. (2016) Pre-analytical variables

of circulating cell-free nucleosomes containing 5-methylcytosine DNA or
histone modification H3K9Me3. Scand J Clin Lab Invest 76: 448–453.
35. Mouliere F, El Messaoudi S, Pang D, et al. (2014) Multi-marker analysis
of circulating cell-free DNA toward personalized medicine for colorectal
cancer. Mol Oncol 8: 927–941.
36. Ota DM, Drewinko B. (1985) Growth kinetics of human colorectal carci-
noma. Cancer Res 45: 2128–2131.
37. Thierry AR, El Messaoudi S, Gahan PB, et al. (2016) Origins, structures,
and functions of circulating DNA in oncology. Cancer Metastasis Rev 35:
347–376.
38. Rossing RM, Jentzen W, Nagarajah J, et al. (2016) Serum thyroglobulin dou-
bling time in progressive thyroid cancer. Thyroid 26: 1712–1718.
39. Abbosh C, Birkbak NJ, Wilson GA, et al. (2017) Phylogenetic ctDNA anal-
ysis depicts early-stage lung cancer evolution. Nature 545: 446–451.
40. Riva F, Bidard FC, Houy A, et al. (2017) Patient-specific circulating tumor
DNA detection during neoadjuvant chemotherapy in triple-negative breast
cancer. Clin Chem 63: 691–699.
41. Weidner N. (2002) New paradigm for vessel intravasation by tumor cells.
Am J Pathol 160: 1937–1939.
42. Lee SY, Chao-Nan Q, Seng OA, et al. (2012) Changes in specialized blood
vessels in lymph nodes and their role in cancer metastasis. J Transl Med 10:
206.
43. Qian CN, Tan MH, Yang JP, Cao Y. (2016) Revisiting tumor angiogenesis:
Vessel co-option, vessel remodeling, and cancer cell-derived vasculature
formation. Chin J Cancer 35: 10.
44. Avanzini S, Kurtz DM, Chabon JJ, et al. (2020) A mathematical model of
ctDNA shedding predicts tumor detection size. Sci Adv. 6(50):eabc4308.
45. Chabon JJ, Hamilton EG, Kurtz DM, et al. (2020) Integrating genomic fea-
tures for non-invasive early lung cancer detection. Nature 580: 245–251.
46. Siravegna G, Bardelli A. (2014) Genotyping cell-free tumor DNA in the
blood to detect residual disease and drug resistance. Genome Biol 15:
449.
47. Diehl F, Schmidt K, Choti MA, et al. (2008) Circulating mutant DNA to
assess tumor dynamics. Nat Med 14: 985–990.

b4444_Ch-01.indd 17 08-Jul-22 5:56:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

18 Cell-Free Circulating DNA: Purification and Analysis Techniques

48. Husain H, Melnikova VO, Kosco K, et al. (2017) Monitoring daily dynamics
of early tumor response to targeted therapy by detecting circulating tumor
DNA in urine. Clin Cancer Res 23: 4716–4723.
49. Xi L, Pham TH, Payabyab EC, et al. (2016) Circulating tumor DNA as an
early indicator of response to T-cell transfer immunotherapy in metastatic
melanoma. Clin Cancer Res 22: 5480–5486.
50. Lo YM, Zhang J, Leung TN, et al. (1999) Rapid clearance of fetal DNA from
maternal plasma. Am J Hum Genet 64: 218–224.
51. Gauthier VJ, Tyler LN, Mannik M. (1996) Blood clearance kinetics and liver
uptake of mononucleosomes in mice. J Immunol 156: 1151–1156.
52. Cheng THT, Jiang P, Tam JCW, et al. (2017) Genomewide bisulfite sequenc-
ing reveals the origin and time-dependent fragmentation of urinary cfDNA.
Clin Biochem 50: 496–501.
53. Markus H, Zhao J, Contente-Cuomo T, et al. (2021) Analysis of recurrently
protected genomic regions in cell-free DNA found in urine. Sci Transl Med
13.
54. Chandrananda D, Thorne NP, Bahlo M. (2015) High-resolution charac-
terization of sequence signatures due to non-random cleavage of cell-free
DNA. BMC Med Genomics 8: 29.
55. Chen S, Liu M, Zhang X, et al. (2017) A study of cell-free DNA fragmen-
tation pattern and its application in DNA sample type classification. IEEE/
ACM Trans Comput Biol Bioinform.
56. Han DSC, Ni M, Chan RWY, et al. (2020) The biology of cell-free DNA
fragmentation and the roles of DNASE1, DNASE1L3, and DFFB. Am J
Hum Genet 106: 202–214.
57. Serpas L, Chan RWY, Jiang P, et al. (2019) Dnase1l3 deletion causes aber-
rations in length and end-motif frequencies in plasma DNA. Proc Natl Acad
Sci USA 116: 641–649.
58. Lo YM, Chan KC, Sun H, et al. (2010) Maternal plasma DNA sequencing
reveals the genome-wide genetic and mutational profile of the fetus. Sci
Transl Med 2: 61ra91.
59. Mouliere F, Robert B, Arnau Peyrotte E, et al. (2011) High fragmentation
characterizes tumour-derived circulating DNA. PLoS One 6: e23418.
60. Tamkovich SN, Cherepanova AV, Kolesnikova EV, et al. (2006) Circulating
DNA and DNase activity in human blood. Ann N Y Acad Sci 1075: 191–
196.

b4444_Ch-01.indd 18 08-Jul-22 5:56:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biology of cirDNA in Health and Disease 19

61. Rostami A, Lambie M, Yu CW, et al. (2020) Senescence, necrosis, and apop-
tosis govern circulating cell-free DNA release kinetics. Cell Rep 31: 107830.
62. Ungerer V, Bronkhorst AJ, Van den Ackerveken P, et al. (2021) Serial profil-
ing of cell-free DNA and nucleosome histone modifications in cell cultures.
Sci Rep 11: 9460.
63. Burnham P, Kim MS, Agbor-Enoh S, et al. (2016) Single-stranded DNA
library preparation uncovers the origin and diversity of ultrashort cell-free
DNA in plasma. Sci Rep 6: 27859.
64. Murtaza M, Caldas C. (2016) Nucleosome mapping in plasma DNA pre-
dicts cancer gene expression. Nat Genet 48: 1105–1106.
65. Schones DE, Cui K, Cuddapah S, et al. (2008) Dynamic regulation of nucle-
osome positioning in the human genome. Cell 132: 887–898.
66. Ulz P, Perakis S, Zhou Q, et al. (2019) Inference of transcription factor bind-
ing from cell-free DNA enables tumor subtype prediction and early detec-
tion. Nat Commun 10: 4666.
67. Yu SC, Chan KC, Zheng YW, et al. (2014) Size-based molecular diagnostics
using plasma DNA for noninvasive prenatal testing. Proc Natl Acad Sci USA
111: 8583–8588.
68. Chan KC, Jiang P, Sun K, et al. (2016) Second generation noninvasive fetal
genome analysis reveals de novo mutations, single-base parental inheritance,
and preferred DNA ends. Proc Natl Acad Sci USA 113: E8159–E8168.
69. Sun K, Jiang P, Wong AIC, et al. (2018) Size-tagged preferred ends in mater-
nal plasma DNA shed light on the production mechanism and show utility in
noninvasive prenatal testing. Proc Natl Acad Sci USA 115: E5106–E5114.
70. Jiang P, Chan CW, Chan KC, et al. (2015) Lengthening and shortening of
plasma DNA in hepatocellular carcinoma patients. Proc Natl Acad Sci USA
112: E1317–E1325.
71. Mouliere F, Mair R, Chandrananda D, et al. (2018) Detection of cell-free
DNA fragmentation and copy number alterations in cerebrospinal fluid
from glioma patients. EMBO Mol Med 10: e9323.
72. Underhill HR, Kitzman JO, Hellwig S, et al. (2016) Fragment length of
circulating tumor DNA. PLoS Genet 12: e1006162.
73. Bryzgunova OE, Tamkovich SN, Cherepanova AV, et al. (2015) Redistribu-
tion of free- and cell-surface-bound DNA in blood of benign and malignant
prostate tumor patients. Acta Naturae 7: 115–118.

b4444_Ch-01.indd 19 08-Jul-22 5:56:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

20 Cell-Free Circulating DNA: Purification and Analysis Techniques

74. Bang C, Thum T. (2012) Exosomes: New players in cell-cell communication.

Int J Biochem Cell Biol 44: 2060–2064.
75. Melo SA, Luecke LB, Kahlert C, et al. Glypican-1 identifies cancer exo-
somes and detects early pancreatic cancer. Nature 523: 177–182.
76. Thakur BK, Zhang H, Becker A, et al. (2014) Double-stranded DNA in
exosomes: A novel biomarker in cancer detection. Cell Res 24: 766–769.
77. van der Pol E, Boing AN, Harrison P, et al. (2012) Classification, functions,
and clinical relevance of extracellular vesicles. Pharmacol Rev 64: 676–705.

b4444_Ch-01.indd 20 08-Jul-22 5:56:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0002

Chapter 2
Preanalytics and Quality
Control of Different Substrates
for Circulating DNA Studies
Fay Betsou and Wim Ammerlaan

Abstract

In this chapter, we review the most critical preanalytical factors in

cirDNA analysis by sequencing, dPCR or BEAMing digital PCR, and
methylation-specific PCR, with a focus on applications other than pre-
natal testing. We also review mitigation of preanalytical risks by the use
of quality control procedures.

Introduction
cirDNA is used as a diagnostic, prognostic, or predictive biomarker.
Diagnostic noninvasive prenatal testing (NIPT) routinely uses
maternal and fetal cirDNA to test fetal aneuploidies via several
available FDA-approved and CE-marked kits.1,2 Analysis of cirDNA
for clinical applications such as prognosis of cancer recurrence or
metastasis, prediction of cancer therapy success, and adaptation of
targeted therapies holds great promise.3 cirDNA is being evaluated

21

b4444_Ch-02.indd 21 08-Jul-22 5:56:52 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

22 Cell-Free Circulating DNA: Purification and Analysis Techniques

in cancer clinical trials as a biomarker and has already reached the

stage of being used in clinical practice (e.g., by Guardant Health
in the U.S. or Biocartis assays; https://www.genomeweb.com/pcr/
biocartis-gets-ce-marking-two-colorectal-cancer-liquid-
biopsy-tests#.WvgVaGcR1vA). In disease areas other than cancer,
such as nonalcoholic fatty liver disease (NAFL) and nonalcoholic
steatohepatitis (NASH), the use of cirDNA as a diagnostic or prog-
nostic biomarker is at the exploratory research stage, for example,
in the clinical evolution of NAFL disease.4 cirDNA methylation is
also being explored as a biomarker in colorectal cancer5 and NIPT,6
and a diagnostic test for colorectal cancer based on the methylation
of SEPTIN97,8 is already approved by FDA.
As for any biomarker, fitness for purpose of the specimens
being used for cirDNA analysis is necessary for the accuracy of the
analytical results.9,10 Preanalytical handling and storage of blood,
plasma, and purified cirDNA samples affect parameters such as
DNA quantity, size distribution, and proportion of total DNA
sample that is derived from the tumor. Uncontrolled and undoc-
umented preanalytics may introduce catastrophic bias, invalidate
clinical results, or lead to irreproducible research publications.
Where samples collected from patients with a disease condition
are compared to healthy control samples, it is particularly import-
ant that preanalytical variables such as blood storage and handling
and also biobanking of plasma or cirDNA samples are standard-
ized. This will often require advanced planning at the sample col-
lection stage of a study.

Critical Preanalytical Factors

There are four major preanalytical factors that are critical for
cirDNA extraction and preparation. The first critical factor is the
processing of blood specimens, which carries two main challenges

b4444_Ch-02.indd 22 08-Jul-22 5:56:52 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 23

with regard to cirDNA extraction. The first challenge is the low

concentration of around 1000 genome equivalents (GEs) per mL
blood11 cirDNA (a few ng per mL of plasma in healthy individu-
als), introducing the need for highest yield to ensure best sensi-
tivity of downstream analysis. The second challenge with regard
to cirDNA extraction is the possible “contamination” of cirDNA
by bigger size cellular genomic DNA fragments, introducing the
need for either prompt processing of blood samples, which is not
always feasible in clinical settings, or for white blood cell (WBC)
stabilization to ensure best specificity of downstream analysis.
With regard to blood collection, the question of serum ver-
sus plasma translates into the question of quantity or quality?
Extraction of cirDNA consistently gives 5–8-fold higher yields
in serum than in plasma.12,13 Moreover, the team of Warton et al.
have shown13 that when blood collection tubes (BCT) are spiked
with high-molecular-weight (MW) DNA, this spiked DNA is not
recovered or detected electrophoretically in the serum cirDNA
extracted from the blood collected in the tube. On the contrary,
it is detected in its high-MW form in the plasma cirDNA extract,
indicating that the spiked DNA may have been degraded during
the serum processing. Therefore, the absence of high-MW DNA
in cirDNA serum extracts may not guarantee the absence of con-
tamination by cellular DNA fragments that may not be visible
electrophoretically.
In another study evaluating cirDNA outputs in serum ver-
sus plasma, Parpart-Li et al. found significantly more total GEs,
but significantly lower mutant allele fractions, in serum than in
EDTA plasma from cancer patients. The fragment sizes in the
cirDNA fraction extracted from serum ranged from 150 to 2000
bp, while the samples from plasma showed the single typical peak
at around 150 bp.14 Release of cellular DNA, therefore, seems to
be taking place during the blood clotting process from lysing cells,

b4444_Ch-02.indd 23 08-Jul-22 5:56:52 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

24 Cell-Free Circulating DNA: Purification and Analysis Techniques

thus explaining the higher yield of cirDNA obtained from serum.

As a result, anticoagulated blood is the preferred option for any
cirDNA-based assay. A summary of the studies having compared
serum and plasma in the scope of cirDNA is shown in Table 1.
A number of anticoagulant options for the collection of plasma
exist, namely EDTA, citrate, and heparin. A 2014 survey of Euro-
pean laboratories showed that the overwhelming majority of stud-
ies that collect blood for cirDNA research use EDTA tubes, with
only 2 out of 70 groups using sodium citrate tubes.15 No group
reported using heparin tubes, which may be due to concerns
with residual heparin inhibiting downstream PCR applications.16
EDTA has been shown to produce slower cell lysis than citrate
or heparin,17 and commercialized diagnostic kits require EDTA
plasma as primary material.
Another important preanalytical factor for cirDNA collection
is the time and temperature between blood collection and plasma
isolation. Concerning the maximum allowable time between blood
collection and processing by centrifugation for plasma separation,
many biospecimen research studies indicate that EDTA blood
should be processed in the 3–6 hr following collection, if it is
stored at room temperature (RT). If EDTA blood is stored at 4°C,
the delay can safely be extended to 8 hr13 or to 24 hr.14,18–23 Table 2
summarizes cirDNA concentration kinetics in EDTA BCT.
In the event that prompt blood processing is not possible, the
next question is whether the anticoagulated blood should be stabi-
lized and with what type of stabilizer. The standard Paxgene tubes
that have been used for years for blood cell nucleic-acid-based
analyses are inadequate for cirDNA isolation because of extensive
cell lysis.24 Cell lysis also occurs in EDTA tubes, upon prolonged
incubation of the blood, especially at RT. Following such WBC
lysis, cellular genomic DNA fragments as well as DNases are

b4444_Ch-02.indd 24 08-Jul-22 5:56:52 PM

b4444_Ch-02.indd 25

9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Table 1. Preanalytical Factors in Circulating DNA Extraction in Plasma and Serum
Plasma Serum
Fold storage clotting DNA extraction DNA quantification
Ref Plasma yield Serum yield change time time method method
Lee et al. 50 1 × 103 copies/mL 6 × 103 copies/mL 6x 2 hr 2 hr HIV Monitor assay Semi-qPCR
(HLA-DQ)
Lui et al. 51 1195.1 genome 16344.8 GE/mL 14x 6-hr RT 6-hr RT QIAamp Blood kit qPCR
equivalent (SKY, β-globin)
(GE)/mL

Preanalytics and Quality Control for cirDNA Studies 25

Thijssen Healthy, 4.8 ± Healthy, 12.9 ± 3x <1 hr <1 hr Puregene (Gentra) qPCR (albumin) +
et al. 52 3.6 ng/mL 10.7 ng/mL +QIAquick PCR Picogreen
Cancer, 10.6 ± Cancer, 47.4 ± purification kit
14.2 ng/mL 46.1 ng/mL
Holdenrieder ~110 μg/L ~200 μg/L 2x NA NA No extraction CellDeath Detection
et al. 53 ELISA
Warton 3.6 ± 0.5 ng/mL 32.7 ± 19.9 ng/mL 10x 30 min RT 30 min RT QIAamp Circulating qPCR
et al. 13 Nucleic Acids kit (SFN1, SFTA3)
Ammerlaan 5 ± 0.5 ng/mL 23.6 ± 2.9 ng/mL 5x 5 min RT 45 min RT QIAamp Circulating Quant-IT Picogreen
et al.12 Nucleic Acids kit
Parpart-Li ~5000 GE/2 mL ~100,000 GE/2 mL 20x 4-hr RT 30 min RT QIAamp Circulating Qubit HS assay +
et al.,14 (size 155 bp) (size 150–2000 bp) Nucleic Acids kit ddPCR (KRAS)
08-Jul-22 5:56:52 PM
b4444_Ch-02.indd 26

26 Cell-Free Circulating DNA: Purification and Analysis Techniques


Table 2. Circulating DNA Concentration Kinetics (EDTA Blood Collection Tubes)

b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Ref Baseline cirDNA 4 hr 8 hr 24 hr 48 hr 72 hr 120 hr 7 days 14 days
50 3
Lee et al. 1 × 10 copies/mL NA NA NA NA NA NA 3 × 104 copies/mL
Norton ~1000 copies/mL NA NA stable 2x ↑ 12x ↑ NA 49x ↑ 204x ↑
et al. 22
Wong et al.54 Normalized NA NA 1454 2631 NA NA 7590 7995
medians
(total genomic
copies/mL)
Warton 4.4–4.9 ng/mL stable stable 6.5 ±2.2 ng/mL 10.8 ± NA NA NA NA
et al.13 4.5 ng/mL
Parpart-Li NA NA NA ~25x ↑ genome NA NA ~500x ~1300x NA
et al.14 equivalent ↑ GE/mL ↑ GE/mL
(GE)/mL
Warton 1.82–4.8 ng/mL NA NA NA NA NA NA NA 57–98 ng/mL
et al.55

9”x6”
08-Jul-22 5:56:52 PM
 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 27

released. While it has been shown that EDTA can inhibit endog-
enous DNase activity to some degree, DNA degradation may still
occur.25 Blood cell lysis is avoided by stabilization solutions, used
in Streck, CellSave, Roche, Norgen Biotek, or Paxgene cirDNA
tubes.26–28 Stability claims of these products have generally been
confirmed: 7 days at RT for Streck BCT and Paxgene cirDNA
tubes. The stability of the cirDNA collection tube from Roche
Diagnostics GmBH seems to be lower.29 Yields and performance
of cirDNA are comparable from Streck BCT and Paxgene cirDNA
tubes (Ammerlaan, unpublished). Recently, an abstract from Qia-
gen showed that Streck tubes contain formaldehyde, which can
induce DNA deaminations and introduce bias in cirDNA methyl-
ation analyses.30 Furthermore, Paxgene cirDNA tubes allow unbi-
ased quantification of methylated sequences to be performed and
are therefore fit-for-purpose for downstream cirDNA methylation
analyses.31 Table 3 summarizes comparison data between EDTA
and the most commonly used BCT with stabilizers.
The consensus on plasma separation is for double spun
plasma, with the second high-speed centrifugation at 16,000g
before cirDNA extraction. However, while it is intuitive that a
plasma centrifugation step will remove any WBC that were inad-
vertently collected and this can be visually corroborated with the
observation of a pellet following the second spin, there is no pub-
lished research that systematically examines the effects of plasma
centrifugation to determine the optimum time and speed. A study
examining the distribution of extracellular DNA in plasma vesi-
cles reported an apoptotic body-associated DNA fraction that can
be collected by a 1200g spin for 30 min.32 It is of concern that
this, and other microvesicle-associated fractions, may be removed
during plasma centrifugation steps. A further consideration in
plasma centrifugation is that non-DNA plasma components such

b4444_Ch-02.indd 27 08-Jul-22 5:56:52 PM

b4444_Ch-02.indd 28

28 Cell-Free Circulating DNA: Purification and Analysis Techniques


Table 3. Circulating DNA Preanalytical Factors in Most Commonly Used Blood Collection Tubes with Stabilizers
Ref Endpoints (method) Conditions EDTA Streck BCT PAXgene cirDNA
Norton Yield/size (ddPCR) Up to 14 days pDNA, 49x ↑ Day 7 pDNA, stable Day 7 and 4x NA

b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

et al.22 RT and 204x ↑ Day 14 ↑ Day 14
gDNA, 80x ↑ Day 7 gDNA, stable Day 7 and 2x
and 456x ↑ Day 14 ↑ Day 14
Wong Total and fetal copies/mL Up to 7 days at Fetal fraction ↓ Day Fetal fraction stable Day 7, NA
et al.54 (FQA assay) different 7, from 0.16 to 0.03 but ↓ at T°>32°C
temperatures
Parpart-Li Genome equivalent (GE)/ Up to 7 days RT 1300% ↑ GE/mL 6% ↑ GE/mL NA
et al.14 mL (ddPCR) 50%–100% ↓ GE/mL Stable
MAF (ddPCR)
Parpart-Li GE/mL (ddPCR) Up to 3 days 4°C Stable Stable NA
et al.14 MAF (ddPCR) 21% ↓ GE/mL 4% ↓ GE/mL
Medina Yield of short/long DNA Up to 5 days RT, NA All endpoints stable, but NA
Diaz et fragments (qPCR) also 4°C and ↑ long/short fragments
al.56 Mutational changes 40°C from 0.2–0.3 to 0.5–0.6 at
(BEAMing, SafeSeqS) the extreme temperatures
Warton Yield (qPCR) 4 days RT 9–22x > yield Stable yield Stable yield
et al.55 size (qPCR+gel) 247 bp/115 bp ↑ 247 bp/115 bp ↓ 247 bp/115 bp ↑
High MW, intensively High MW, just visible on No high MW on gel
visible on gel gel

9”x6”
08-Jul-22 5:56:52 PM
 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 29

Fig. 1: Blood processing to create plasma aliquots for biobanking: double

centrifugation removes contaminating cell debris; mixing the double-spun
plasma prior to aliquoting creates uniform samples for reproducible results.

as lipids will form layers during the additional high-speed spin,

and these layers will lead to nonuniform samples when the plasma
is aliquoted directly from the centrifuged tube. The effect of this
variation on cirDNA stability and extractability is not known. For
this reason, some laboratories transfer the plasma to a fresh tube
after the second spin, and mix it again, in order to obtain an even
distribution of plasma components and uniform aliquots for long-
term storage (Fig. 1) (Warton, pers. comm.).
The cirDNA extraction kit used is another critical preanalyt-
ical factor. Specific extraction methods are considered in detail in
Chapter 4. In broad terms, standard genomic DNA extraction kits
are unsuitable for cirDNA purification, as they are configured to
bind high-MW DNA and exclude small-size fragments. This results
in a purified sample that underrepresents the low-MW DNA that
forms the bulk of cirDNA, resulting in an underestimate of the
total cirDNA concentration, and an incorrect size distribution.

b4444_Ch-02.indd 29 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

30 Cell-Free Circulating DNA: Purification and Analysis Techniques

For downstream cirDNA methylation analyses, fitness-for-

purpose of the cirDNA has been shown, although downstream
analysis protocols may require modification.13 If bisulfite conversion
is used, the bisulfite conversion kit is another critical preanalytical
factor. This methodology is covered in detail in Chapter 9.
The final critical preanalytical factor is cirDNA storage after
extraction. Limited data have been published regarding the sta-
bility of cirDNA in storage. Sozzi and colleagues evaluated the
stability of cirDNA in plasma stored at −80°C and of purified
cirDNA stored at −20°C.33 They found that storage of either sam-
ple type leads to a loss of cirDNA at a rate of about 30% per
year. However, these data need to be considered in the light of
subsequent advances in the understanding of cirDNA proper-
ties. Firstly, as no specialized cirDNA purification kit was com-
mercially available at the time of that study, the authors used the
QIAamp DNA Mini kit, which biases the purified sample against
the low-MW fragments that form the bulk of the cirDNA pool.
Hence, the 30% annual loss of cirDNA from frozen plasma may
reflect the degradation of long DNA fragments more accurately
than the degradation of short DNA fragments. Secondly, the puri-
fied DNA was eluted in water rather than buffer, and it is possible
that acid hydrolysis of the unbuffered sample contributed to the
30% annual loss of DNA stored after extraction. Finally, stabil-
ity data may not be transferrable between samples purified using
different kits, as the final DNA sample likely contains different
components (e.g., different elution buffer, carrier RNA vs. no
carrier RNA, variation in contaminant carryover). No information
regarding the influence of freeze-thaw cycles on sample stability
is currently published.
Commercial protocols for whole genome amplification
(WGA) from plasma are available (Sigma modified WGA2 or

b4444_Ch-02.indd 30 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 31

modified WGA4 protocols, https://www.sigmaaldrich.com/

technical-documents/protocols/biology/whole-genome-amplifica-
tion-serum-or-plasma.html); however, no independent validation
has been published. cirDNA fragments of around 150 bp are too
short for WGA by currently available methods (Johanna Trouet,
unpublished).
A first set of evidence-based guidelines for processing spec-
imens for cirDNA analyses has been published by El Messaoudi
et al.34 More evidence-based procedures and recommendations
have been prepared by the European consortium CANCER-ID
(https://www.cancer-id.eu/), in the context of the Innovative Med-
icines Initiative (IMI).35 The European program SPIDIA (http://
www.spidia.eu/) has developed a CEN Technical Specification
standard for the preanalytical phase of cirDNA.36 This standard
includes requirements on the traceability of the most critical pre-
analytical factors.

Quality Control
Quality assurance of cirDNA processing methods includes method
validation, documentation, internal quality control, and external
quality control. A new ISO technical standard has been prepared
by the ISO Technical Committee on Biotechnology (TC276),
which includes requirements for validation of biospecimen pro-
cessing methods.37 This validation plan includes experimental
plans for optimization of protocols relative to yield and purity of
the produced cirDNA. It verifies fitness for purpose, feasibility,
and the accuracy of specific downstream analyses, such as cirDNA
sequencing. It sets the targets to be met for reproducibility and
robustness to specific preanalytical variables, such as freeze-thaw
cycles. Reproducibility and robustness are evaluated, based on

b4444_Ch-02.indd 31 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

32 Cell-Free Circulating DNA: Purification and Analysis Techniques

selected analytical endpoints of scientific interest. These are usu-

ally quantitative, such as yield of cirDNA, size of cirDNA, or vari-
ant allele frequencies.
Recording and checking the preanalytical data associated
with each sample via the Standard PREanalytical Code (SPREC)
supports quality assurance and control. The latest version of the
SPREC has been updated to include cirDNA-relevant options and
codes.38 More specifically, Table 4 includes some of the options
and corresponding codes that are relevant in cirDNA-related pro-
cessing of samples.
Internal quality control of a processing method, such as
cirDNA extraction, is performed through assays for biomarkers of

Table 4. Circulating DNA Processing and Corresponding Codes

Type of sample
Blood (whole) BLD
Plasma, single spin PL1
Plasma, double spin PL2
Serum SER
Type of primary container
Acid citrate dextrose ACD
Chemical additives/stabilizers ADD
Serum tube without clot activator CAT
Citrate phosphate dextrose CPD
EDTA and gel EDG
Lithium heparin HEP
Lithium heparin and gel LHG
PAXgene® blood RNA+ PAX
Potassium EDTA PED
PAXgene® blood DNA PXD

b4444_Ch-02.indd 32 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 33

Table 4.   (Continued )

Type of primary container
RNA Later® RNL
Sodium citrate SCI
Non-aldehyde-based stabilizer for cell-free nucleic acids SCK
Sodium EDTA SED
Sodium heparin SHP
Serum separator tube with clot activator SST
Pre-centrifugation (delay between collection and processing)
RT <30 min A1
2°C–10°C <30 min B1
RT <2 hr A
2°C–10°C <2 hr B
RT 2–4 hr C
2°C–10°C 2–4 hr D
RT 4–8 hr E
2°C–10°C 4–8 hr F
RT 8–12 hr G
2°C–10°C 8–12 hr H
RT 12–24 hr I
2°C 10°C 12–24 hr J
RT 24–48 hr K
2°C–10°C 24–48 hr L
RT >48 hr M
2°C–10°C >48 hr N
Centrifugation
RT 10–15 min <3000g no braking A
RT 10– 15 min <3000g with braking B
2°C–10°C 10–15 min <3000g no braking C
2%–10°C 10–15 min <3000g with braking D
(Continued )

b4444_Ch-02.indd 33 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

34 Cell-Free Circulating DNA: Purification and Analysis Techniques

Table 4. (Continued )
Second centrifugation
RT 10–15 min >10,000g with braking I
2°C–10°C 10– 15 min >10,000g with braking J
Post-centrifugation delay
<1 hr 2°C– 10°C A
<1 hr RT B
1–2 hr 2°C–10°C C
1–2 hr RT D
Long-term storage
PP tube 0.5–2 mL −85°C to −60°C A
PP tube 0.5–2 mL −35°C to (−18)°C B
PP tube 0.5–2 mL <−135°C V
Cryotube 1–2 mL LN C
Cryotube 1–2 mL −85°C to −60°C D
Cryotube 1–2 mL Programmable freezing to <−135°C E
Plastic cryo straw LN F
Straw −85°C to −60°C G
Straw −35°C to −18°C H
Straw Programmable freezing to <−135°C I
PP tube ≥ 5 mL −85°C to −60°C J
PP tube ≥ 5 mL −35°C to −18°C K
Cryotube 1–2 mL LN after temporary −85°C to −60°C N

sample quality, such as electrophoretic assessment of cirDNA con-

tamination by WBC DNA. Internal quality control is also achieved
through usage of in-process quality control materials, such as
spiked samples, which are processed in every extraction run.
External quality control of a processing method is done through
participation in corresponding external quality assurance process-
ing schemes.

b4444_Ch-02.indd 34 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 35

Analytical methods have their own set of internal quality control

samples, which are tested in every run, such as a well-characterized
cirDNA sample, and their own relevant external quality assurance
testing schemes, such as quantification of cirDNA, or cirDNA
sequencing and genotyping.

Internal Quality Controls

In-process QC Materials
Several spike-in sources can be applied. Synthetic ultramers,
DNA fragments up 200 bases in length, represent an economical
approach to the in-house production of in-process QC materials.
Ultramers can be synthesized and contain any sequence/variant
of interest and can be spiked in a blood tube that will undergo
extraction in the same run with the plasma samples. However,
synthetic ultramers are single-stranded naked DNA, and they do
not anneal 100%; therefore, they do not wholly represent cirDNA.
Alternatively, synthetic “gBlock Gene Fragments” can be designed.
The “gBlock Gene Fragments” are synthetic dsDNA fragments
ranging from 125 to 3000 bp. Synthesis allows for custom design
sequences, for example, to create EGFR T790M and wt gBlocks
in the cirDNA fragment size of approximately 150 bp, to be used
as In-Process Controls (IPC). They have the advantage over the
ultramers of being dsDNA, therefore not requiring annealing.
More physiological and thus reflective of cirDNA is the ultrasonic
shearing (e.g., Covaris) of purified DNA from appropriate cancer
cell lines. This creates fragments of appropriate length (approxi-
mately 170 bp) and more natural sequence distribution by using
the complete genome.
A disadvantage of these methods for in-process QC is that
they produce naked DNA strands, which do not mimic the

b4444_Ch-02.indd 35 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

36 Cell-Free Circulating DNA: Purification and Analysis Techniques

secondary structure of native cirDNA molecules, which are

mononucleosomal DNA, in which the DNA is associated with
histones. Mononucleosomal DNA preparations are, therefore,
preferred as in-process QC materials. In the Integrated Biobank
of Luxembourg (IBBL), such mononucleosomal DNA for spiking
purposes is prepared from cell lines, with the following materials
and method:

• A specific cancer cell line, as appropriate

• RPMI-1640 supplemented with 10% Fetal Bovine Serum
(Gibco)
• PBS (Gibco)
• Trypsin (Lonza)
• Glycerol (Sigma)
• Nucleosome Preparation kit (Active Motif; cat# 53504)
• QIAquick PCR Purification Kits (QIAgen, cat# 28104)

Cells are collected with Trypsin and counted using CASY

counter. Nucleosomes are prepared following the instruction
manual from Active Motif (Nucleosome Preparation Kit; Version
1a, Active Motif). DNA is isolated from a part of the prepared
nucleosomes, following the manufacturer’s instruction (Nucleo-
some Preparation Kit; Version 1a, Active Motif). The Active Motif
method for Mononucleosomes (MNs) preparations first isolates
nuclei of a selected cell line or tissue, in the presence of prote-
ase inhibitors. The purified nuclei are digested by an Active Motif
undisclosed enzymatic mixture, again in the presence of protease
inhibitors. Digestion is stopped by adding EDTA. Prior to DNA
quantification, the Mononucleosomal preparation is digested by
Proteinase K, followed by a DNA cleanup with the QIAquick
PCR purification kit. MN quantity equivalent to µg of DNA can

b4444_Ch-02.indd 36 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 37

be prepared from a single run starting with 10 million cells. The

DNA is analyzed on a DNA high sensitivity HSD 5000 chip (4200
Tapestation) to verify the size is around 200 nucleotides and finally
analyzed by dPCR for the mutations of interest.

Quality Control Assays

A review of relevant quality control assays for different types of
biospecimens has been published by the ISBER Biospecimen Sci-
ence Working group39 and an online tool has been made available
(www.findmyassay.com). Assays relevant to the context of cirDNA
research are described as follows.

Quality of EDTA Plasma

The hemolytic index quantifies the level of hemoglobin in plasma.
High levels of hemoglobin indicate hemolysis, that is, the lysis of
red blood cells (RBCs). Although RBCs do not contain DNA and
therefore there is no risk of contamination of the cirDNA fraction
by RBC cellular DNA, hemolysis may indicate concomitant WBC
lysis. The degree of hemolysis is measured either by hemoglobin
ELISA, with a cutoff for significant hemolysis in plasma being 20
mg/L, or by simple spectrophotometric measurement at 414 nm
with a cutoff for significant hemolysis being OD414>0.2.40
When preanalytical data have not been documented, for
example in the case of plasma legacy collections, the following
quality control assays can be applied:

(i) Measurement of the Lacascore (ratio of ascorbic to lactic acid)

in EDTA plasma can give an indication of the preanalytical
quality of the plasma and more specifically indicates if the pre-
centrifugation delay had been longer than 3 hr at RT.41

b4444_Ch-02.indd 37 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

38 Cell-Free Circulating DNA: Purification and Analysis Techniques

(ii) Measurement of IL-16 indicates if the precentrifugation delay

of EDTA had been longer than 24 hr at RT (cutoff 313 pg/mL)
or longer than 48 hr at RT (cutoff 897 pg/mL).42

Quality of Extracted cirDNA

Measurement of the concentration of extracted cirDNA can be
done by high-sensitivity spectrofluorimetry or by qPCR or dPCR.
A qPCR method targeting 90 bp and 222 bp fragments of LIPA2
has successfully been used for THE quantification of cirDNA
in 1:40 diluted EDTA plasma, without previous extraction.43
High-sensitivity microfluidic electrophoresis chips are used to
assess contamination by high-MW cellular DNA.
A differential amplicon length PCR, such as the Kapa hgDNA
quantification and QC kit, has been suggested as a relevant QC
assay to assess the degree of contamination by high-MW cellular
DNA. This kit amplifies the fragments of 41 bp, 129 bp, and 305
bp length, and a ratio of 305 bp/41 bp > 0.3 would be indicative
of cell lysis.29 Another differential amplicon PCR that can be used
is an Alu qPCR with primer sets targeting fragments of 115 bp
(Alu115) and 247 bp (Alu247). The cirDNA integrity can then be
calculated as the ratio of Alu257/Alu115, which should be lower
than 0.5.44 A third differential amplicon dPCR assay that has been
described is based on multiplex amplification of five single copy
genomic loci of average length 71 bp and of four single copy loci of
average length 471 bp. The cirDNA concentration was estimated
as the difference between short amplicon-based GEs and long
amplicon-based GEs.45
A qPCR targeting a genomic locus, present at one copy per
haploid genome, such as RPPH1, GAPDH, NAGK or ERV3, can
be used to measure extracted cirDNA copies/mL. However, more
than one reference target should be used because the proportion

b4444_Ch-02.indd 38 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 39

of each locus in the cirDNA fraction is not necessarily the same

as the corresponding proportion in cellular genomic DNA, with
telomeric loci being more abundant in cirDNA.46

External Quality Controls

Proficiency Testing
IBBL has developed a proficiency testing program, which allows
participating laboratories to assess the efficiency of their cirDNA
extraction methods. The processing items that are distributed to
participants are mononucleosomal DNA-spiked blood samples in
Streck BCT or Paxgene cirDNA tubes.47 Evaluation is based on
cirDNA recovery and the presence or absence of contamination
by high-MW genomic DNA.
The clinical laboratory EQA program providers AIOM,
EMQN, ESP EQA, GenQA (part of UK NEQAS), and Gen&-
Tiss provide a pilot cirDNA EQA scheme to assess the standard of
testing EGFR gene mutations in circulating free DNA in plasma
(https://www.emqn.org/exciting-news-new-ctdna-scheme-plasma-
testing-lung-cancer/).

Reference Materials
Reference materials are needed for formal analytical method val-
idation according to ISO17025 accreditation requirements. There
are only a few reference materials commercially available for these
purposes.
Seracare commercialize a SeraseqTM circulating tumor DNA
Reference Material. This is a mixture of human genomic DNA,
fragmented into 170-bp fragments, in a commutable matrix (sim-
ulated plasma whose use is compatible with many different plat-
forms and, therefore, allows comparison of different analytical

b4444_Ch-02.indd 39 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

40 Cell-Free Circulating DNA: Purification and Analysis Techniques

platforms). The material contains several SNV and INDEL vari-

ants at allele frequencies of 5%–0.125%. The intended uses are
targeted NGS, qPCR, and dPCR assays, and the Seraseq mate-
rial can be used either as an internal QC material in the earlier
mentioned assays or as an in-process QC material to monitor the
performance of the extraction method as well as the performance
of the analysis.
Horizon commercialize cirDNA HDx Reference Standards
for method validation purposes (research use only). These mate-
rials are provided as mechanically sheared, fragmented DNA
(average size 160 bp) from engineered human cell lines and con-
tain variants at allelic frequencies down to 0.1%,48 (https://www.
horizondiscovery.com/media/resources/Application%20Notes/
reference-standards/independent-dPCR-study-of-horizon-
cirDNA-reference-standards.pdf ).
Devonshire et al. have developed and validated spike mate-
rials that allow users to measure cirDNA extraction efficiency,
fragment size bias, and yield.46 These materials are based on a lin-
earized and digested plasmid, containing the Arabidopsis thaliana
alcohol dehydrogenase gene (ADH) and fragments have sizes of
189 bp and roughly 3kb.
A homemade reference material that has been used for the
analytical validation of a cirDNA NGS assay has been described
and is based on overlapping extension PCR for site-directed muta-
genesis to obtain fragments of 537–2030 bp, containing specific
mutation sequences.49

Conclusion
The use of cirDNA as a biomarker has extended from NIPT to
oncologic indications, and growing research and development
efforts on cirDNA are targeting other disease areas. Standardization

b4444_Ch-02.indd 40 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 41

of the cirDNA preanalytical phase is key for the accuracy of

analytical results and the reproducibility of research conclusions.
The commercialization of BCT with stabilizers that avoid blood
cell lysis has increased the robustness of cirDNA testing. Exter-
nal quality assurance programs for both cirDNA processing and
analytical methods provide a means for laboratory benchmarking.
Reference materials in combination with dPCR methods provide
metrological traceability and the possibility of accreditation. With
the mentioned developments, we are at a stage where we know
which substrates to use, how to process them for each specific
cirDNA-based application, and how to control for the quality of
the substrates, of the cirDNA and of the methods applied.

References
1. Liu L, Li K, Fu X, et al. (2016) A forward look at noninvasive prenatal test-
ing. Trends Mol Med 22: 958–968.
2. Skrzypek H, Hui L. (2017) Noninvasive prenatal testing for fetal aneuploidy
and single gene disorders. Best Pract Res Clin Obstet Gynaecol 42: 26–38.
3. Cree IA, Uttley L, Buckley Woods H, et al. (2017) The evidence base for
circulating tumour DNA blood-based biomarkers for the early detection of
cancer: A systematic mapping review. BMC Cancer 17: 697.
4. Karlas T, Weise L, Kuhn S, et al. (2017) Correlation of cell-free DNA plasma
concentration with severity of non-alcoholic fatty liver disease. J Transl Med
15: 106.
5. Worm Orntoft MB. (2018) Review of blood-based colorectal cancer screen-
ing: How far are circulating cell-free DNA methylation markers from clini-
cal implementation? Clin Colorectal Cancer 17: e415–e433.
6. Lim JH, Lee DE, Kim KS, et al. (2014) Non-invasive detection of fetal tri-
somy 21 using fetal epigenetic biomarkers with a high CpG density. Clin
Chem Lab Med 52: 641–647.
7. Potter NT, Hurban P, White MN, et al. (2014) Validation of a real-time
PCR-based qualitative assay for the detection of methylated SEPT9 DNA
in human plasma. Clin Chem 60: 1183–1191.

b4444_Ch-02.indd 41 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

42 Cell-Free Circulating DNA: Purification and Analysis Techniques

8. Church TR, Wandell M, Lofton-Day C, et al. (2014) Prospective evaluation

of methylated SEPT9 in plasma for detection of asymptomatic colorectal
cancer. Gut 63: 317–325.
9. Bartak BK, Kalmar A, Galamb O, et al. (2019) Blood collection and cell-free
DNA isolation methods influence the sensitivity of liquid biopsy analysis for
colorectal cancer detection. Pathol Oncol Res 25: 915–923.
10. Bronkhorst AJ, Ungerer V, Diehl F, et al. (2020) Towards systematic nomen-
clature for cell-free DNA. Hum Genet 140(4): 565–578.
11. Chiu RW, Poon LL, Lau TK, et al. (2001) Effects of blood-processing pro-
tocols on fetal and total DNA quantification in maternal plasma. Clin Chem
47: 1607–1613.
12. Ammerlaan W, Trezzi JP, Lescuyer P, et al. (2014) Method validation for
preparing serum and plasma samples from human blood for downstream
proteomic, metabolomic, and circulating nucleic acid-based applications.
Biopreserv Biobank 12: 269–280.
13. Warton K, Lin V, Navin T, et al. (2014) Methylation-capture and next-gen-
eration sequencing of free circulating DNA from human plasma. BMC
Genomics 15: 476.
14. Parpart-Li S, Bartlett B, Popoli M, et al. (2017) The effect of preservative
and temperature on the analysis of circulating tumor DNA. Clin Cancer Res
23: 2471–2477.
15. Malentacchi F, Pizzamiglio S, Verderio P, et al. (2015) Influence of storage
conditions and extraction methods on the quantity and quality of circulating
cell-free DNA (ccfDNA): The SPIDIA-DNAplas external quality assess-
ment experience. Clin Chem Lab Med 53: 1935–1942.
16. Farnert A, Arez AP, Correia AT, et al. (1999) Sampling and storage of blood
and the detection of malaria parasites by polymerase chain reaction. Trans
R Soc Trop Med Hyg 93: 50–53.
17. Lam NY, Rainer TH, Chiu RW, Lo YM. (2004) EDTA is a better anticoag-
ulant than heparin or citrate for delayed blood processing for plasma DNA
analysis. Clin Chem 50: 256–257.
18. Xue X, Teare MD, Holen I, et al. (2009) Optimizing the yield and utility
of circulating cell-free DNA from plasma and serum. Clin Chim Acta 404:
100–104.

b4444_Ch-02.indd 42 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 43

19. Page K, Guttery DS, Zahra N, et al. (2013) Influence of plasma processing
on recovery and analysis of circulating nucleic acids. PLoS One 8: e77963.
20. Jung M, Klotzek S, Lewandowski M, et al. (2003) Changes in concentration
of DNA in serum and plasma during storage of blood samples. Clin Chem
49: 1028–1029.
21. Board RE, Williams VS, Knight L, et al. (2008) Isolation and extraction of
circulating tumor DNA from patients with small cell lung cancer. Ann N Y
Acad Sci 1137: 98–107.
22. Norton SE, Lechner JM, Williams T, Fernando MR. (2013) A stabilizing
reagent prevents cell-free DNA contamination by cellular DNA in plasma
during blood sample storage and shipping as determined by digital PCR.
Clin Biochem 46: 1561–1565.
23. Barrett AN, Zimmermann BG, Wang D, et al. (2011) Implementing prena-
tal diagnosis based on cell-free fetal DNA: Accurate identification of factors
affecting fetal DNA yield. PLoS One 6: e25202.
24. Toro PV, Erlanger B, Beaver JA, et al. (2015) Comparison of cell stabilizing
blood collection tubes for circulating plasma tumor DNA. Clin Biochem 48:
993–998.
25. Barra GB, Santa Rita TH, de Almeida Vasques J, et al. (2015) EDTA-medi-
ated inhibition of DNases protects circulating cell-free DNA from ex vivo
degradation in blood samples. Clin Biochem 48: 976–981.
26. van Dessel LF, Beije N, Helmijr JC, et al. (2017) Application of circulating
tumor DNA in prospective clinical oncology trials - standardization of pre-
analytical conditions. Mol Oncol 11: 295–304.
27. Sherwood JL, Corcoran C, Brown H, et al. (2016) Optimised pre-analyti-
cal methods improve KRAS mutation detection in circulating tumour DNA
(ctDNA) from patients with non-small cell lung cancer (NSCLC). PLoS
One 11: e0150197.
28. Kang Q, Henry NL, Paoletti C, et al. (2016) Comparative analysis of circu-
lating tumor DNA stability In K3EDTA, Streck, and CellSave blood collec-
tion tubes. Clin Biochem 49: 1354–1360.
29. Nikolaev S, Lemmens L, Koessler T, et al. (2018) Circulating tumoral DNA:
Preanalytical validation and quality control in a diagnostic laboratory. Anal
Biochem 542: 34–39.

b4444_Ch-02.indd 43 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

44 Cell-Free Circulating DNA: Purification and Analysis Techniques

30. Groelz D, Viertler C, Pabst D, et al. (2018) Impact of storage conditions

on the quality of nucleic acids in paraffin embedded tissues. PLoS One 13:
e0203608.
31. Schmidt B, Reinicke D, Reindl I, et al. (2017) Liquid biopsy—Performance
of the PAXgene(R) blood ccfDNA tubes for the isolation and characteri-
zation of cell-free plasma DNA from tumor patients. Clin Chim Acta 469:
94–98.
32. Lazaro-Ibanez E, Sanz-Garcia A, Visakorpi T, et al. (2014) Different gDNA
content in the subpopulations of prostate cancer extracellular vesicles:
Apoptotic bodies, microvesicles, and exosomes. Prostate 74: 1379–1390.
33. Sozzi G, Roz L, Conte D, et al. (2005) Effects of prolonged storage of whole
plasma or isolated plasma DNA on the results of circulating DNA quantifi-
cation assays. J Natl Cancer Inst 97: 1848–1850.
34. El Messaoudi S, Rolet F, Mouliere F, Thierry AR. (2013) Circulating cell
free DNA: Preanalytical considerations. Clin Chim Acta 424: 222–230.
35. Lampignano R, Neumann MHD, Weber S, et al. (2020) Multicenter eval-
uation of circulating cell-free DNA extraction and downstream analyses for
the development of standardized (Pre)analytical work flows. Clin Chem 66:
149–160.
36. CEN/TS 16835-3 Venous Whole Blood - Part 3: Isolated circulating cell free
DNA from plasma 2015.
37. ISO 21899:2020 Biotechnology – Biobanking – General requirements for
the validation and verification of processing methods for biological material
in biobanks.
38. Betsou F, Bilbao R, Case J, et al. (2018) Standard PREanalytical code ver-
sion 3.0. Biopreserv Biobank 16: 9–12.
39. Betsou F, Bulla A, Cho SY, et al. (2016) Assays for qualification and quality
stratification of clinical Biospecimens used in research: A technical report
from the ISBER Biospecimen science working group. Biopreserv Biobank
14: 398–409.
40. Appierto V, Callari M, Cavadini E, et al. (2014) A lipemia-independent
Nano­Drop((R))-based score to identify hemolysis in plasma and serum sam-
ples. Bioanalysis 6: 1215–1226.
41. Trezzi JP, Bulla A, Bellora C, et al. (2016) LacaScore: A novel plasma sample
quality control tool based on ascorbic acid and lactic acid levels. Metabolom-
ics 12: 96.

b4444_Ch-02.indd 44 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Preanalytics and Quality Control for cirDNA Studies 45

42. Kofanova O, Henry E, Aguilar Quesada R, et al. (2018) IL8 and IL16 levels
indicate serum and plasma quality. Clin Chem Lab Med 56: 1054–1062.
43. Breitbach S, Tug S, Helmig S, et al. (2014) Direct quantification of cell-free,
circulating DNA from unpurified plasma. PLoS One 9: e87838.
44. Fawzy A, Sweify KM, El-Fayoumy HM, Nofal N. (2016) Quantitative anal-
ysis of plasma cell-free DNA and its DNA integrity in patients with met-
astatic prostate cancer using ALU sequence. J Egypt Natl Canc Inst 28:
235–242.
45. Markus H, Contente-Cuomo T, Farooq M, et al. (2018) Evaluation of
pre-analytical factors affecting plasma DNA analysis. Sci Rep 8: 7375.
46. Devonshire AS, Whale AS, Gutteridge A, et al. (2014) Towards standard-
isation of cell-free DNA measurement in plasma: Controls for extraction
efficiency, fragment size bias and quantification. Anal Bioanal Chem 406:
6499–6512.
47. Gaignaux A, Ashton G, Coppola D, et al. (2016) A biospecimen proficiency
testing program for biobank accreditation: Four years of experience. Bio-
preserv Biobank 14: 429–439.
48. Amit H, Wei SH, Armisen-Garrido J, et al. (2015) Using cell free DNA
reference standards to evaluate the analytical performance of circulating
tumor DNA testing and solid organ transplant health surveillance. BioTech-
niques 59: 248–250.
49. Yang X, Chu Y, Zhang R, et al. (2017) Technical validation of a next-gen-
eration sequencing assay for detecting clinically relevant levels of breast
cancer-related single-nucleotide variants and copy number variants using
simulated cell-free DNA. J Mol Diagn 19: 525–536.
50. Lee TH, Montalvo L, Chrebtow V, Busch MP. (2001) Quantitation of
genomic DNA in plasma and serum samples: Higher concentrations of
genomic DNA found in serum than in plasma. Transfusion 41: 276–282.
51. Lui YY, Chik KW, Chiu RW, et al. (2002) Predominant hematopoietic origin
of cell-free DNA in plasma and serum after sex-mismatched bone marrow
transplantation. Clin Chem 48: 421–427.
52. Thijssen MA, Swinkels DW, Ruers TJ, de Kok JB. (2002) Difference
between free circulating plasma and serum DNA in patients with colorectal
liver metastases. Anticancer Res 22: 421–425.
53. Holdenrieder S, Stieber P, Chan LY, et al. (2005) Cell-free DNA in serum and
plasma: comparison of ELISA and quantitative PCR. Clin Chem 51: 1544–1546.

b4444_Ch-02.indd 45 08-Jul-22 5:56:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

46 Cell-Free Circulating DNA: Purification and Analysis Techniques

54. Wong D, Moturi S, Angkachatchai V, et al. (2013) Optimizing blood collec-

tion, transport and storage conditions for cell free DNA increases access to
prenatal testing. Clin Biochem 46: 1099–1104.
55. Warton K, Yuwono NL, Cowley MJ, et al. (2017) Evaluation of streck BCT
and PAXgene stabilised blood collection tubes for cell-free circulating DNA
studies in plasma. Mol Diagn Ther 21: 563–570.
56. Medina Diaz I, Nocon A, Mehnert DH, et al. Performance of streck cfDNA
blood collection tubes for liquid biopsy testing. PLoS One 11: e0166354.

b4444_Ch-02.indd 46 08-Jul-22 5:56:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0003

Chapter 3
Considerations in Biobanking
Blood Samples and Plasma
Volume for Circulating
DNA Studies
Kate Mahon, Goli Samimi and Kristina Warton

Abstract

Clinical circulating DNA studies are contingent on the availability of

appropriate biospecimens and linked patient data. Incorporating cir-
culating DNA analysis in clinical trial design adds value to the trial, as
well as allowing future retrospective analysis of markers that may be
unknown at the time of the trial. However, this requires forward plan-
ning and commitment to sample curation. This chapter considers circu-
lating DNA in the broader context of biobanking.
How much plasma does this assay need? If not the first question in a
circulating DNA study, it is perhaps the one asked with the most conster-
nation. Unless project-specific, prospective biobanking is undertaken,
the volumes of clinical plasma samples available for circulating DNA
studies are often small. This chapter also examines the implications of
the typically low biological circulating DNA concentrations and provides

47

b4444_Ch-03.indd 47 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

48 Cell-Free Circulating DNA: Purification and Analysis Techniques

an overview of the plasma volumes used in circulating DNA research in

cancer.

Considerations in Biobanking Blood Samples for Studies

of Circulating DNA in Plasma
Biobanking involves collection, storage, maintenance, and analy-
sis of biological samples and clinical data for future clinical and
research purposes.1 In recent years, biobanking has become a
standard requirement for clinical trials and is increasingly identi-
fied as an ideal standard for routine clinical care. While this is an
admirable goal, multiple issues arise in such large-scale collections
and biobank procedures must be well researched and planned
from the outset.
As the number and scale of biobanks has expanded over
recent years, it has become imperative to define and classify bio-
banks. Many classification systems have been developed, and with
the complexity and evolution of biobanking, these classification
systems are likely to evolve with time. A detailed classification sys-
tem including multiple elements proposed by Watson and Barnes2
has been adopted internationally. This schema utilizes four func-
tional elements to define individual biobanks. These are:

1. Donor/participant — describes the defining features of the test

subjects such as disease state (healthy/at risk/specific disease),
alive or deceased, age categorization.
2. Design — including accrual plan (prospective/retrospective),
collection timepoints (single/multiple), scale (single or multiple
research studies), extent of linked data collection.
3. Biospecimens — including type of biospecimen (e.g., blood,
tissue, fluid, cells), method of preservation.

b4444_Ch-03.indd 48 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 49

4. Brand — including leadership (researchers, groups, institutions),

sponsors and users (single, several, or many).

The elements broadly detailed in this classification schema

highlight the need to consider and define these aspects in the opti-
mal design and setup of a biobank to ensure that it is optimally fit
for purpose from the outset.

Purpose of Biobank
The specific purpose of a biobank should be the initial consider-
ation. Biobanks for circulating tumor DNA (ctDNA) may focus on
early detection and diagnosis of cancer, detecting disease recurrence
in patients following curative intent treatment, or predicting and/or
monitoring response to therapy. This purpose will define the study
population, location, and method of accrual. The timing of sample
collection is also an important consideration. Clearly, this must fit the
proposed purpose of the test but must also be practical in the clinical
setting. Inconvenient sampling such as requiring patients to attend for
tests when they would not normally do so or at multiple time points
will hamper the ability to recruit and retain study subjects. Further-
more, if a test proves to be clinically useful, translation into clinical
practice will be challenging if the testing procedure is complicated
and impractical.
Ideally, the scale of a biobank and the proposed number of
users should also be defined from the outset. The number of users
can range from a single research group or institution to multiple
groups across international sites. With greater numbers of users
and sites, the need for clearly detailed standardized operating pro-
cedures (SOPs) for specimen collection and processing and con-
sidered data management becomes particularly important. Large

b4444_Ch-03.indd 49 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

50 Cell-Free Circulating DNA: Purification and Analysis Techniques

biobanks may require samples to be moved long distances for

storage and these procedures also need to be thoroughly detailed
to avoid degradation of samples. Recognition that preanalytical
inconsistencies may cause greater variability in results than the true
biological differences3 has led to standardized frameworks (includ-
ing Standard PREanalytical Code4 and Biospecimen Reporting
for Improved Study Quality5) to ensure thorough documentation
and reporting of essential elements of preanalytical processing. A
sample processing trail (e.g., dates, times, temperature, location,
operator) should be detailed to flag any potential issues in sample
quality.6 Reliable labeling systems to track sample locations are
imperative to avoid lost or misidentified samples. Ongoing investi-
gator training and audit procedures are also required to maintain
consistency. International organizations such as the National Can-
cer Institute (NCI) provide best practice documents encompass-
ing small to very large scale biobanks to guide essential aspects of
setup and management.7 While ensuring best practice, complying
with internationally recognized protocols may also facilitate col-
laboration and integration with other biobanks in the future.

Ethical Considerations
Ethical considerations and informed consent are paramount in
biobank management, particularly in the proposed genetic analy-
sis of stored samples. International biobanks include an extra layer
of complexity with a wide range of legal requirements in differ-
ent jurisdictions. Difficulties arise when the nature of a proposed
test cannot clearly be defined at the time of consent. Obtaining
consent for each new research project arising from a biobank is
impractical and often impossible so a “broad consent” encom-
passing as yet unspecified future uses is required. The legal and

b4444_Ch-03.indd 50 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 51

ethical implications of such a consent have long been debated.8

The range of consent procedures in biobanks internationally is
wide.9 Many biobank consent procedures allow subjects to specify
exactly how their specimens may be used including genetic testing
and future unspecified research. Consent processes should inform
subjects about long-term storage of samples and potential shar-
ing of data and specimens with other researchers. Confidentiality
procedures must also be clearly detailed such as de-identification
of samples and clinical data including how data may be shared
with other research groups (e.g., aggregated data versus individ-
ual patient data). While most patient information includes details
around the physical risks of sample collection (e.g., venipuncture
or biopsy), potential legal and psychosocial risks associated with
possible results should also be communicated. Patient access to
genetic information arising from study results must be consid-
ered. Genomic findings may not only impact the subject but may
also have implications for family members. Plans to communicate
positive results back to participants or family members must be
determined and should be clearly explained at the time of con-
sent. Obligations around reporting results to external organiza-
tions (e.g., government agencies, insurance companies) must also
be detailed.10

Clinical Data
Accurate, thorough, and accessible clinical data management is
crucial for the analysis of accompanying biospecimens. A mini-
mum clinical data set should be ascertained at the outset and data
collection time points should be established. In some instances,
clinical data will only need to be documented at the time of spec-
imen collection; however, most cancer research will benefit from

b4444_Ch-03.indd 51 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

52 Cell-Free Circulating DNA: Purification and Analysis Techniques

ongoing followup. Clear plans must be made around when, how,

and by whom ongoing clinical details will be obtained (e.g., at spe-
cific clinic visits, by phone call or mail outs). Dedicated staff are
usually required to obtain long-term follow-up data and maintain
database integrity.
Reliable, flexible, and secure information technology systems
are imperative. Clinical data should be linked with biospecimen
collection information and results of analyses. Ideally, databases
will be easily accessible by all relevant members of the research
team and will allow extraction of data, which may be usefully shared
between research groups for future collaboration. Maintaining
data accessibility, such as in web-based systems, while ensuring
data security to maintain patient confidentiality, is an additional
challenge. Defined security levels are required so that users can
only access the information required for their specific tasks and
research sites. While multiple systems are available, both com-
mercially and through academic institutions, no system provides
a perfect comprehensive solution. Future integration of biobank
databases may be better facilitated by standardization of minimum
data sets, ontologies, and data formats.10

Considerations in Selecting Assay Plasma Volume

When considering biobanking of plasma for circulating DNA
(cirDNA) studies, collecting an adequate volume can maximize
the usefulness of the samples for downstream analysis. The vol-
ume of plasma needed for a biomarker assay is a key question
when designing experiments or planning clinical studies. Too little,
and there won’t be enough DNA target for reliable detection; too
much, and the reaction may be crippled by blood-derived enzyme
inhibitors. This issue is particularly pertinent when designing
assays intended for the early diagnosis of cancer, where detection

b4444_Ch-03.indd 52 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 53

of very low concentrations of target, corresponding to very low

tumor volume, is required for the assay to be clinically useful.
The optimal volume of plasma for an assay or a protocol is
dependent on a number of factors; these being the total volume of
plasma available, the concentration of tumor-derived DNA in the
plasma, the sensitivity of the detection assay, and the presence of
blood-derived inhibitors. We will consider each of these in turn,
with a focus on PCR, since it is the most commonly used detection
method, and underpins many other types of molecular analyses
(e.g., sequencing).

Available Plasma Volume

Limitations stemming from the volume of available plasma are
the easiest to consider. Blood samples collected from patients in a
clinical context are generally quite low volume, with a single large-
draw blood tube yielding at best around 10 mL of blood and 5 mL
of plasma, while standard blood tubes yield around 4 mL of blood
and 2 mL of plasma. Unless the blood collection is specifically car-
ried out for a particular project, the sample may be further split
between different studies. The challenge is more acute for rare
cancer types, since a small number of biospecimens means that
less sample is spread between projects. Another issue arises when
blood from patients is being compared against healthy controls for
the purpose of biomarker discovery or validation, as an equivalent
volume of matched control sample needs to be available. In addi-
tion to the work of collecting, processing, and storing the plasma,
patient samples also require the collection and curation of sam-
ple-linked clinical information. All in all, this makes patient plasma
samples a much valued and limited resource.
The low volume of plasma available can be a hurdle to bio-
marker development. Volumes of 500 µL or less are reported in

b4444_Ch-03.indd 53 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

54 Cell-Free Circulating DNA: Purification and Analysis Techniques

studies, presumably reflecting the amount of biobanked sample

that could be obtained.11–13 Early stages of biomarker development
are most likely to be carried out on retrospective biobanked sam-
ples,14 since early stage biomarker discovery is considered risky
and generally not seen to warrant a prospective plasma collec-
tion effort. Ironically, it is the early stages that often require the
most plasma, in particular if they involve repeated measurements
or evaluation of multiple biomarker candidates to identify ones
that have adequate sensitivity and specificity.15 Publicly funded
biobanks may be unwilling to release large volumes of plasma for
early stage testing of unvalidated biomarkers. Thus, a situation
arises where the largest number of biomarker candidates needs to
be tested at the project stage where the lowest volume of plasma
is readily available.
The highest volume of plasma will be available from patients
once the test is utilized in a clinical setting, and a similarly large
volume will be available from samples collected in any prospective
trials planned as part of biomarker validation, in which the entire
blood sample can be allocated for the study. The question then is
whether to design assays based on the volume of patient plasma
that will be available from individual patients in a clinical setting
or based on the likely more limited volume of plasma that will be
available to validate the biomarker in retrospective cohorts. This
question can only be answered within the practical and budget
constraints of each individual project and depends largely on the
funds allocated to sourcing or collecting biospecimens.

Concentration of cirDNA in Plasma

The concentration of cirDNA in plasma has been noted to vary
widely, even in healthy individuals,16 and determination of the

b4444_Ch-03.indd 54 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 55

true concentration is hampered by samples being particularly sen-

sitive to handling and measurement artefacts. For example, stor-
age of collected blood prior to plasma separation leads to lysis of
leukocyte and contamination of the cirDNA pool with leukocyte
genomic DNA, thus increasing the apparent concentration.17,18 A
large, multi-center study prospectively examining the relation-
ship between cirDNA concentration and cancer risk identified
the research center carrying out the blood collection and process-
ing as by far the single most influential variable in determining
cirDNA levels, demonstrating the impact that handling protocols
have on derivation of cirDNA.19 In addition to the effects of sam-
ple handling, different extraction methods have large differences
in yield,20–22 inevitably leading to differences in apparent cirDNA
concentration reported by studies applying the different method-
ologies. There is also some indication that cirDNA in plasma and
purified DNA samples degrade during storage, further decreas-
ing the substrate available for analysis.23,24 From the perspective of
healthy variation, physical activity has profound effects on cfDNA
levels, with transient increases of up to ninefold observed imme-
diately after exercise.25

Quantitation of cirDNA from Plasma

There is a range of techniques available to quantitate cirDNA
purified from plasma. These include spectrophotometry-based
techniques such as Nanodrop (ThermoFisher), techniques that
rely on fluorescent dye binding such as Picogreen Kit and the
Picogreen-based QuBit fluorometer (both from ThermoFisher),
and various types of qPCR. The different quantitation methods
produce different concentration results. For example, QuBit has
been shown to measure around several-fold the apparent yield of

b4444_Ch-03.indd 55 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

56 Cell-Free Circulating DNA: Purification and Analysis Techniques

qPCR,26 most likely due to PCR not detecting DNA fragments

below the size of the target amplicon. For the purposes of this
chapter, we will consider the quantity of cirDNA in plasma to refer
to the quantity of DNA that is detectable by PCR, since this is the
amount accessible to analysis by PCR-based methods. While this
introduces some consistency, there is still a level of artefactual vari-
ation, since the fragmented nature of cirDNA means that using a
shorter PCR assay results in a higher apparent cirDNA concentra-
tion27 (see Chapter 5), and the positioning of nucleosomes relative
to the PCR amplicon is also important.28,29
With the above considerations in mind, most current esti-
mates place the cirDNA concentration in healthy individuals
between 1 and 10 ng per milliliter of plasma.30 The level in cancer
patients is higher, with the increase due to tumor DNA as well as
an increased contribution from nontumor cells.31 The proportion
of total cirDNA derived from the tumor depends on the tumor
volume and stage,31,32 and possibly on the type of cancer.33 The
upper end of the range, measured in advanced tumors, can be
quite high, with the most tumor DNA-rich samples containing
~65% ctDNA in cirDNA from advanced ovarian cancer patients,33
27% in patients with metastatic colorectal cancer,31 and ~90% in
breast cancer.34 Different subclones of tumor can contribute dif-
ferent amounts of ctDNA. For example, Murtaza and colleagues
found that mutations already present in the primary tumor of a
breast cancer patient ranged from 3.8% to 34.9% in plasma sam-
pled at multiple time points over ~16 months, while mutations
limited to metastatic tumor deposits ranged from 2.5% to 19%
over the same time period.35 Evidently, the lower range of ctDNA
concentration in cancer patients cannot be determined below the
sensitivity of the analysis techniques applied, but mutant allele
fractions of 0.01% have been measured in early stage colorectal

b4444_Ch-03.indd 56 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 57

cancer patients by a variation on digital PCR technique (BEAM-

ing — Beads Emulsion Amplification and Magnetics)31 and by
massively parallel sequencing.36
The cirDNA concentration, together with the ctDNA frac-
tion, allows some estimate of the ctDNA amount available for an
assay and thus of the minimum plasma input requirements for an
assay of a given sensitivity. For example, with 10 ng of cirDNA
per milliliter, and 5% tumor DNA, an assay needs to detect 500
pg of target when 1 mL of plasma is used, and with 0.5% tumor
DNA it needs to detect 50 pg of target. Numbers of ctDNA mol-
ecules in plasma samples from cancer patients have been directly
quantitated by sequencing, with techniques able to detect less
than 1 mutant molecule per milliliter of plasma.32,36 For some
applications, detecting the smallest possible ctDNA amount is
less important than for others. For example, ctDNA assays can
be applied to patient monitoring during drug treatment, without
stringent sensitivity requirements. This variable application is
analogous to protein biomarkers, where sensitivity and specificity
requirements for cancer monitoring are much less stringent than
for cancer diagnosis. For example, CA125 and CEA are routinely
used for ovarian and colorectal cancer patient monitoring respec-
tively but are unsuitable for use in diagnostic cancer screening.37–39

Reporting Standards
Consistency in reporting cfDNA studies has repeatedly been
called for in the field to allow interpretation and replication.40
Where authors do not explicitly state the plasma volume corre-
sponding to a particular amount of cirDNA used in a PCR or other
downstream assay, and it is not always possible to calculate it from
the methods information provided. In order to allow readers to

b4444_Ch-03.indd 57 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

58 Cell-Free Circulating DNA: Purification and Analysis Techniques

calculate the volume of plasma effectively analyzed in each assay,

the authors need to provide the following information: volume of
plasma used in the cirDNA extraction; DNA elution/resuspension
volume; and the seeding volume of the PCR. Where the DNA is
bisulfite converted and used for methylation analysis, authors also
need to provide the volume of purified DNA used in the bisulfite
conversion reaction, and the elution/resuspension volume of the
bisulfite-converted DNA, followed by the PCR seeding volume.
Where the methods of published studies describe only the total
amount (i.e., nano- or picograms) of cirDNA that was used in a
PCR reaction, it may not be possible to calculate back to the cor-
responding volume of plasma without the plasma concentration of
cirDNA also being provided. Unfortunately, in published research
papers, the necessary information about plasma volumes used and
recovered at each step during cirDNA extraction/conversion is the
exception rather than the rule.

Blood-Derived PCR Inhibitors

If the volume of plasma used as starting material in an assay is not
limited by the total amount of plasma available, then it may be
limited by the carryover of blood-derived PCR inhibitors. PCR
inhibitors are the compounds present in a sample that have either
carried over from the biospecimen extracted or been introduced
during DNA purification or bisulfite treatment. Blood-derived
PCR inhibitors include hemoglobin and immunoglobulin G,41,42
while introduced inhibitors include heparin, ethanol, and deter-
gent.43,44
For this reason, provided there is a sufficient volume of
plasma available, choosing a purification method that halves the
carryover of PCR inhibitors is as good as doubling the yield from

b4444_Ch-03.indd 58 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 59

the purification protocol, as it would allow the PCR reaction to be

seeded with double the amount of eluted DNA without encoun-
tering the problem of inhibitors.

Summary
Biobanks provide opportunities for identifying circulating DNA
mutations for diagnosis, prognosis, and therapeutic decision-
making. Circulating DNA mutations can be difficult to identify,
often occurring in small numbers of individuals and across cancer
types. Integration of biobanks can overcome this issue to increase
patient numbers and allow analysis of data across tumor-specific
collections. Biobanking adequate plasma volumes will help to
maximize the usefulness of the samples. To avoid the multitude
of issues that may arise, thorough planning and consideration of
biobank procedures at the outset are essential.

References
1. Smith ME, Aufox S. (2013) Biobanking: The melding of research with clin-
ical care. Curr Genet Med Rep 1: 122–128.
2. Watson PH, Barnes RO. (2011) A proposed schema for classifying human
research biobanks. Biopreserv Biobank 9: 327–333.
3. Ellervik C, Vaught J. (2015) Preanalytical variables affecting the integrity of
human biospecimens in biobanking. Clin Chem 61: 914–934.
4. Betsou F, Lehmann S, Ashton G, et al. (2010) Standard preanalytical coding
for biospecimens: Defining the sample PREanalytical code. Cancer Epide-
miol Biomarkers Prev 19: 1004–1011.
5. Moore HM, Kelly AB, Jewell SD, et al. (2011) Biospecimen reporting for
improved study quality (BRISQ). Cancer Cytopathol 119: 92–101.
6. Vaught J, Abayomi A, Peakman T, et al. (2014) Critical issues in interna-
tional biobanking. Clin Chem 60: 1368–1374.
7. Insititute NC. (2016) NCI Best Practices for Biospecimen Resources.

b4444_Ch-03.indd 59 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

60 Cell-Free Circulating DNA: Purification and Analysis Techniques

8. Hansson MG, Dillner J, Bartram CR, et al. (2006) Should donors be allowed
to give broad consent to future biobank research? Lancet Oncol 7: 266–269.
9. Wolf LE, Bouley TA, McCulloch CE. (2010) Genetic research with stored
biological materials: Ethics and practice. IRB 32: 7–18.
10. Watson RW, Kay EW, Smith D. (2010) Integrating biobanks: Addressing the
practical and ethical issues to deliver a valuable tool for cancer research. Nat
Rev Cancer 10: 646–651.
11. Olsson E, Winter C, George A, et al. (2015) Serial monitoring of circulating
tumor DNA in patients with primary breast cancer for detection of occult
metastatic disease. EMBO Mol Med 7: 1034–1047.
12. Alborelli I, Generali D, Jermann P, et al. (2019) Cell-free DNA analysis in
healthy individuals by next-generation sequencing: A proof of concept and
technical validation study. Cell Death Dis10: 534.
13. Zemmour H, Planer D, Magenheim J, et al. (2018) Non-invasive detection
of human cardiomyocyte death using methylation patterns of circulating
DNA. Nat Commun 9: 1443.
14. Mahon KL, Qu W, Devaney J, et al. (2014) Methylated Glutathione S-transferase
1 (mGSTP1) is a potential plasma free DNA epigenetic marker of prognosis
and response to chemotherapy in castrate-resistant prostate cancer. Br J
Cancer 111: 1802–1809.
15. Lofton-Day C, Model F, Devos T, et al. (2008) DNA methylation biomark-
ers for blood-based colorectal cancer screening. Clin Chem 54: 414–423.
16. Thierry AR, El Messaoudi S, Gahan PB, et al. (2016) Origins, structures,
and functions of circulating DNA in oncology. Cancer Metastasis Rev 35:
347–376.
17. El Messaoudi S, Rolet F, Mouliere F, Thierry AR. (2013) Circulating cell
free DNA: Preanalytical considerations. Clin Chim Acta 424: 222–230.
doi:10.1016/j.cca.2013.05.022. Epub May 30.
18. Warton K, Lin V, Navin T, et al. (2014) Methylation-capture and next-gen-
eration sequencing of free circulating DNA from human plasma. BMC
Genomics 15: 476. doi:10.1186/471-2164-15-476.
19. Gormally E, Hainaut P, Caboux E, et al. (2004) Amount of DNA in plasma
and cancer risk: A prospective study. Int J Cancer 111: 746–749.
20. Diefenbach RJ, Lee JH, Kefford RF, Rizos H. (2018) Evaluation of
commercial kits for purification of circulating free DNA. Cancer Genet
228–229: 21–7.

b4444_Ch-03.indd 60 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Biobanking Blood Samples and Plasma Volume 61

21. Devonshire AS, Whale AS, Gutteridge A, et al. (2014) Towards standard-
isation of cell-free DNA measurement in plasma: Controls for extraction
efficiency, fragment size bias and quantification. Anal Bioanal Chem 406:
6499–6512.
22. Warton K, Graham LJ, Yuwono N, Samimi G. (2018) Comparison of 4
commercial kits for the extraction of circulating DNA from plasma. Cancer
Genet 228–229: 143–150.
23. Sozzi G, Roz L, Conte D, et al. (2005) Effects of prolonged storage of whole
plasma or isolated plasma DNA on the results of circulating DNA quantifi-
cation assays. J Natl Cancer Inst 97: 1848–1850.
24. Sato A, Nakashima C, Abe T, et al. Investigation of appropriate pre-analyt-
ical procedure for circulating free DNA from liquid biopsy. Oncotarget 9:
31904–31914.
25. de Sousa MV, Madsen K, Fukui R, et al. Carbohydrate supplementation
delays DNA damage in elite runners during intensive microcycle training.
Eur J Appl Physiol 112: 493–500.
26. Ponti G, Maccaferri M, Manfredini M, et al. (2018) The value of fluorimetry
(Qubit) and spectrophotometry (NanoDrop) in the quantification of cell-
free DNA (cfDNA) in malignant melanoma and prostate cancer patients.
Clin Chim Acta 479: 14–19.
27. Andersen RF, Spindler KL, Brandslund I, et al. (2015) Improved sensitivity
of circulating tumor DNA measurement using short PCR amplicons. Clin
Chim Acta 439: 97–101.
28. Snyder MW, Kircher M, Hill AJ, et al. (2016) Cell-free DNA comprises
an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164:
57–68.
29. Johansson G, Andersson D, Filges S, et al. (2019) Considerations and qual-
ity controls when analyzing cell-free tumor DNA. Biomol Detect Quantif
17: 100078.
30. Wan JCM, Massie C, Garcia-Corbacho J, et al. (2017) Liquid biopsies come
of age: Towards implementation of circulating tumour DNA. Nat Rev Can-
cer 17: 223–238.
31. Diehl F, Li M, Dressman D, et al. (2005) Detection and quantification of
mutations in the plasma of patients with colorectal tumors. Proc Natl Acad
Sci USA 102: 16368–16373.

b4444_Ch-03.indd 61 08-Jul-22 5:57:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

62 Cell-Free Circulating DNA: Purification and Analysis Techniques

32. Bettegowda C, Sausen M, Leary RJ, et al. (2014) Detection of circulating

tumor DNA in early- and late-stage human malignancies. Sci Transl Med 6:
224ra224.
33. Forshew T, Murtaza M, Parkinson C, et al. (2012) Noninvasive identifica-
tion and monitoring of cancer mutations by targeted deep sequencing of
plasma DNA. Sci Transl Med 4: 136ra168.
34. Jahr S, Hentze H, Englisch S, et al. (2001) DNA fragments in the blood
plasma of cancer patients: Quantitations and evidence for their origin from
apoptotic and necrotic cells. Cancer Res 61: 1659–1665.
35. Murtaza M, Dawson SJ, Pogrebniak K, et al. (2015) Multifocal clonal evo-
lution characterized using circulating tumour DNA in a case of metastatic
breast cancer. Nat Commun 6: 8760.
36. Cohen JD, Li L, Wang Y, et al. (2018) Detection and localization of sur-
gically resectable cancers with a multi-analyte blood test. Science 359:
926–930.
37. Hall C, Clarke L, Pal A, et al. (2019) A review of the role of carcinoembry-
onic antigen in clinical practice. Ann Coloproctol 35: 294–305.
38. Blyuss O, Burnell M, Ryan A, et al. (2018) Comparison of longitudinal
CA125 algorithms as a first-line screen for ovarian cancer in the general
population. Clin Cancer Res 24: 4726–4733.
39. Nash Z, Menon U. (2020) Ovarian cancer screening: Current status and
future directions. Best Pract Res Clin Obstet Gynaecol 65: 32–45.
40. Bronkhorst AJ, Ungerer V, Diehl F, et al. (2020) Towards systematic nomen-
clature for cell-free DNA. Hum Genet 40(4): 565–578.
41. Al-Soud WA, Radstrom P. (2001) Purification and characterization of
PCR-inhibitory components in blood cells. J Clin Microbiol 39: 485–493.
42. Sidstedt M, Hedman J, Romsos EL, et al. (2018) Inhibition mechanisms of
hemoglobin, immunoglobulin G, and whole blood in digital and real-time
PCR. Anal Bioanal Chem 410: 2569–2583.
43. Schrader C, Schielke A, Ellerbroek L, Johne R. (2012) PCR inhibitors –
Occurrence, properties and removal. J Appl Microbiol 113: 1014–1026.
44. Yokota M, Tatsumi N, Nathalang O, et al. Effects of heparin on polymerase
chain reaction for blood white cells. J Clin Lab Anal 13: 133–140.

b4444_Ch-03.indd 62 08-Jul-22 5:57:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0004

Chapter 4
Maximizing Yield from Plasma
Circulating DNA Extraction
Florence Mauger and Jean-François Deleuze

Abstract

Circulating cell-free DNA (cirDNA) has showed great promise as a

sensitive and relevant noninvasive biomarker for liquid biopsies, but
it is highly fragmented and often present in low-abundance in plasma/
serum. Consequently, the extraction method needs to be very efficient to
recover all fragment sizes and in particular, the short-length fragments.
It is important to choose the most appropriate extraction method for
each application and to optimize extraction steps to maximize the yield
and the recovery of all fragment sizes of isolated cirDNA from plasma.
A number of modified phenol–chloroform protocols, column-based,
and magnetic bead commercial kits have been developed and com-
pared for isolating cirDNA from different volumes of healthy individual,
cancer patient, or maternal plasma. Although, there is not a universal
method for all applications, the QIAamp Circulating Nucleic Acid kit,
which is the most commonly used kit, appears to be the most effec-
tive and versatile kit to maximize the yield of total and tumor-derived as
well as fetal cirDNA allowing genetic analysis from a few milliliters of

63

b4444_Ch-04.indd 63 08-Jul-22 5:58:07 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

64 Cell-Free Circulating DNA: Purification and Analysis Techniques

plasma. Epigenetic marks of cirDNA can also be analyzed and a meth-

ylation-on-beads protocol has been established to extract and bisulfite
convert cirDNA; however, cirDNA for methylation analysis is typically
isolated using conventional extraction methods.
Furthermore, an enrichment step may improve the analysis of fetal
or methylated cirDNA. Extraction methods could be replaced by direct
analysis of unpurified plasma by digital PCR or cirDNA enrichment step
from a limited amount of plasma.
The cirDNA extraction method has to be very efficient, especially
in the case of a limited amount of input material, and an automated
extraction system is useful for clinical studies to ensure optimal repro-
ducibility of samples.
In addition, standard operating procedures, Laboratory Informa-
tion Management System, and ISO International Standards should be
applied to ensure traceability, standardization of processing, and analyt-
ical reproducibility of the extraction of cirDNA.
Furthermore, the yield of extracted cirDNA can be quantified by
fluorescent assay or qPCR and digital PCR using specific short-size
fragments, and the presence of low- and/or high-molecular weight in
extracted cirDNA can be identified by microfluidic electrophoresis
or qPCR and digital PCR. A complete specific liquid biopsy workflow
including collection, transport, biobanking, pre-analytical process, sam-
ple volume, extraction method, quantification method, and fragment-size
analysis for each kind of plasma should be established to maximize the
extraction yield and improve the analysis of low-abundance cirDNA.

Introduction
In the context of precision medicine, circulating cell-free DNA
(cirDNA) shows great promise as a sensitive biomarker for dis-
ease prediction, diagnosis and prognosis, and also for monitoring
the response to therapeutic treatment of cancer patients and oth-
ers with complex diseases (Chapter 1). As the powerful clinical

b4444_Ch-04.indd 64 08-Jul-22 5:58:07 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 65

applications of cirDNA have evolved and improved, the number

of protocols and commercial kits available for cirDNA isolation has
increased. cirDNA in plasma/serum is present at low-abundance,
and the tumor-derived cirDNA or fetal cirDNA that is most often
of interest in diagnostic tests is only a small fraction of the total.
The amount of cirDNA in plasma depends on a number of differ-
ent biological parameters such as the age and sex of the patient,
the stage and the type of cancer (for tumor-derived cirDNA), and
the week of pregnancy (for fetal cirDNA). In addition, the yield of
isolated cirDNA from plasma is dependent on experimental vari-
ables such as the type of liquid biopsy (Chapter 2), conditions of
biobanking, pre-analytical processes (Chapters 2 and 3), and the
extraction method. During the extraction process, cirDNA can be
lost during binding, washing, and elution steps and, consequently,
lower yields of extracted cirDNA are isolated. Various extraction
methods from published protocols or commercial kits have been
established to isolate cirDNA in the forms of total cirDNA, mito-
chondrial cirDNA, methylated cirDNA, tumor-derived cirDNA,
donor-derived cirDNA, and fetal cirDNA from plasma. The
extraction methods from plasma/serum should have enough
efficiency and specificity to isolate all kinds of fragment sizes to
improve the recovery of isolated cirDNA. The low quantity of
cirDNA obtainable from clinical samples can impact downstream
analytical techniques that require a minimum amount of sample
to yield reliable data, and for this reason, the choice of extraction
method to maximize yield can be critical to the success of a study.
In this chapter, we outline the principles behind cirDNA isolation
techniques, summarize the published literature comparing differ-
ent commercial kits and in-house protocols, and review selected
clinical studies with regard to the effect that the choice of cirDNA
purification protocol has on experiment outcome.

b4444_Ch-04.indd 65 08-Jul-22 5:58:07 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

66 Cell-Free Circulating DNA: Purification and Analysis Techniques

Principles of cirDNA Extraction from Biofluids

DNA in plasma and serum is associated with protein complexes
such as nucleosomes, and a proportion of this DNA is enclosed in
extracellular vesicles.1 To obtain pure DNA, the proteins must be
removed by denaturation and digestion, and the DNA is extracted
either by differential partitioning between aqueous and organic
solvents or by selective binding to a silica-based matrix. These
principles also form the basis of DNA extraction from cells and
tissues; however, cirDNA presents a number of additional chal-
lenges. First, the cirDNA is heavily fragmented, occurring pre-
dominantly as ~170 bp fragments, with smaller populations with
sizes in multiples of 170 bases. The size of the pieces corresponds
to the length of DNA wrapped around a single nucleosome or mul-
tiple nucleosomes. The small size can have an impact on recovery
and leads to low yields unless protocols are specifically configured
for low-molecular-weight DNA. Second, cirDNA is very dilute,
with concentrations as low as just a few ng per mL. With so little
starting material, nonspecific binding to plastic ware during han-
dling can contribute significantly to sample loss, and in matrix-
based protocols, the elution volume must be kept low in order to
maintain a reasonable concentration for downstream applications.
There are two main approaches to the extraction of cirDNA: liq-
uid–liquid extraction and matrix binding. We will consider each of
these in turn.

Matrix Binding
Many commercially available cirDNA purification kits are based
on column extraction using silica-based resin (Fig. 1). Binding
to silica as a method of DNA purification has been applied for
several decades2,3; however, the nature of the interaction is still

b4444_Ch-04.indd 66 08-Jul-22 5:58:07 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 67

Fig. 1: Purification of cfDNA by binding to a silica matrix column. DNA is

separated from nucleosomes by protease digestion at 60°C–65°C, helped by
protein denaturation with detergent and chaotropic salt. Following the diges-
tion, the DNA is captured on a silica matrix, washed, and eluted.

being elucidated.4 Silica, with its low isoelectric point of ~pH 3, is

negatively charged at neutral pH, as is DNA, creating an electro-
static repulsion. Overcoming the repulsion and binding is depen-
dent on chaotropes that disrupt the hydrogen-bonded structure
of water and appears to involve hydrophobic interactions as well
as attraction between the phosphate groups in the DNA back-
bone and silanol.4 The strength of the binding is influenced by
multiple factors such as pH, chaotrope concentration5,6 the size
and shape of the silica particles, and the type of silanol groups on
the surface.7 Furthermore, the silica can be modified with surface
molecules such as amines, metal ions, or lysine to modulate the
binding affinity of different sized DNA.8
The advantages of matrix binding methods are that
the extraction takes only a few hours, and it can be run in a
high-throughput mode. Various kits have been optimized to

b4444_Ch-04.indd 67 08-Jul-22 5:58:09 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

68 Cell-Free Circulating DNA: Purification and Analysis Techniques

isolate viral or genomic cirDNA from blood that is medium or high

molecular weight, respectively. There are also specific extraction
kits based on column or magnetic beads to isolate highly frag-
mented (<300 bp) cirDNA from plasma/serum.

Evaluation of Matrix Binding Protocols

As discussed earlier, many variations of matrix binding chemistry
are possible, and at least 26 different methods based on matrix
binding have been evaluated for cirDNA extraction efficiency
(Tables 1 and 2). Comparison of extraction methods was per-
formed using different extraction protocols, different samples,
and volume of input, quantification, and fragment analysis meth-
ods (Table 1).
The Qiagen QIAamp DNA Blood Mini (QIAamp DBM) kit,
originally developed for the purification of genomic DNA from
leukocytes in whole blood, was one of the earliest kits used for
cirDNA extraction. As a popular and robust method of DNA
extraction from whole blood, it was an obvious choice to extend its
use to cirDNA extraction from plasma, and it has been used in a
number of studies with clinical samples.9,10 However, the chemis-
try used in the kit targets the extraction of a pure sample of high-
molecular-weight genomic DNA, and it does not perform well in
purifying cirDNA, which is heavily fragmented. Every study that
included the QIAamp DBM kit was able to identify an alternative
method that gave better yields (Tables 1 and 2).
The QIAamp Circulating Nucleic Acids (QIAamp CNA)
kit from Qiagen was released in the market in 2009 and was first
evaluated in comparison with other cirDNA extraction methods
by Page and colleagues in 2013.11 This study and all but one sub-
sequent study (n = 8) identified the QIAamp CNA kit as the best

b4444_Ch-04.indd 68 08-Jul-22 5:58:09 PM

b4444_Ch-04.indd 69

Table 1. Workflow of cirDNA Extraction Method Comparison from Plasma/Serum of Healthy Individual and/or Cancer

9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Patients
Fragment analysis
Extraction method Plasma or serum Quantification method method Reference
Phenol–chloroform method 1 mL plasma qPCR assay: — (14)
and QIAamp DBM kit Lung cancer and benign 135 bp of ERV-3 gene
lung disease
NucleoSpin and QIAamp 200 and 240 µL plasma Fluorescent assay Microfluidic (13)
DBM kits qPCR assays: electrophoresis
135 bp of EVR-3 gene Spike-in 50,10,150,
81 bp of b-globin gene 250, and 100 bp

Maximizing Yield from Plasma cirDNA Extraction 69

QIAamp DBM, QIAamp Virus, 600 µL plasma and qPCR assay: PCR assays: (32)
Invitrogen ChargeSwitch, Serum 77 bp of APP gene 272 bp of p53 gene
and Agencourt Genfind kits Healthy individual and 512 bp of BCl-2 gene
Small cell lung cancer
NaI, PCI–glycogen, and 2 mL pooled sera Fluorescent assay Electrophoresis: (16)
guanidine–resin methods Colorectal cancer qPCR assays: 200, 400, and 500 bp
QIAamp DBM, Invitrogen 115 bp of ACTB gene bands
ChargeSwitch, ZR, and 68 bp of CDH1 gene
Puregene kits
(Continued )
08-Jul-22 5:58:09 PM
b4444_Ch-04.indd 70

Table 1. (Continued )

70 Cell-Free Circulating DNA: Purification and Analysis Techniques


Fragment analysis
Extraction method Plasma or serum Quantification method method Reference
THP method and QIAamp 500 µL pooled plasma qPCR assay: Spike-in: (25)

b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DBM kit and serum hGAPDH gene 50 or 200 ng/mL of
Healthy individual bcl-2 plasmid DNA,
genomic DNA and
100 bp ladder
QIAamp DBM, NucleoSpin 1 mL plasma qPCR assays: — (17)
kits, and MagnaPure Lung cancer, benign 92 bp of GAPDH gene
automated system lung disease, 101 bp of b-globin gene
esophageal cancer 135 bp of ERV-3 gene
and non-Hodgkin
lymphoma cancer
PC method, PC phase lock 1 mL plasma Fluorescent assay — (29)
tubes method, and QIAamp Nonsmall cell lung qPCR assays:
MinElute cancer 77, 123 and 202 bp of
Virus Spin Kit SERPINA1 gene
EGFR gene
Schmidt and Hufnagl‘s PC 1 mL plasma qPCR assay: PCR assays: (27)
protocols, NucleoSpin, Colon cancer, breast 98 bp of hTERT gene 100, 200, 300, 400,
QIAamp DBM, and Maxwell cancer, and healthy and 1000 bp
16LEV kits individual

9”x6”
08-Jul-22 5:58:09 PM
b4444_Ch-04.indd 71

9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

FitAmp, NucleoSpin, QIAamp 1 mL plasma Fluorescent assay NGS sequencing of (11)
DBM, and QIAamp CNA Pooled healthy female qPCR assays: 207 amplicons:
kits and metastatic breast 291 bp of GAPDH gene between 111 and 187
cancer 67 bp of ACTBL2 gene bp
64 bp of HPRT1 gene
Spike-in: l/hindIII: 50, 5,
0.5, and 0.05 ng
Unpurified plasma, PCI 100 µL plasma qPCR assays: Microfluidic (26)
method, and QIAamp DBM Unpurified plasma 90 and 222 bp of L1PA2 electrophoresis
kit dilution 1/40 gene
Healthy individual 88 bp of MSTN gene

Maximizing Yield from Plasma cirDNA Extraction 71

and coronary heart 1400 to 75 copies/ 2 µL of
disease murine plasma
FitAmp, NucleoSpin, 0.5 and 1 mL plasma qPCR assays: qPCR assays: (12)
QIAamp DBM, and Pooled healthy female 79 bp of TERT gene 115 bp of ADH gene
QIAamp CNA kits 64 bp of RPPH1 gene 461 bp of Adhb gene
ALUJ sequence 1448 bp of Adhδ gene
135 bp of ERV3 gene Spike-in: Fragmented
94 bp of GAPDH gene ADH plasmid 106
66 bp of NAGK gene copies/mL plasma
(Continued )
08-Jul-22 5:58:09 PM
b4444_Ch-04.indd 72

Table 1. (Continued )

72 Cell-Free Circulating DNA: Purification and Analysis Techniques


Fragment analysis
Extraction method Plasma or serum Quantification method method Reference
CGD method 20–200 µL plasma Fluorescent assay — (36)

b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

and QIAamp CNA kits cancer qPCR assays:
KRAS, BRAF
NRAS and PIK3CA genes
Spike-in: 5 and 20 ng/mL
standard reference DNA
THP, Hufnagl, Yuan and 200 to 1000 µL plasma Fluorescence assay qPCR assays: (28)
Schmidt’s PC methods Healthy individual, qPCR assays: 67 and 180 bp of APP
QIAamp CNA, NucleoSpin, metastatic cancer 55 bp of Kpn sequence gene
Chemagic NA, (colon, pancreatic, 165 bp of DHFRP2 PCR assays:
Norgen cDNA and cfDNA breast and lung sequence 400 and 800 bp of
kits cancer) P53 gene
PME-free, QIAamp CNA 2 mL plasma qPCR assays: — (22)
and QIAamp DSP Virus/ Pooled individual 87 bp of Rnase P gene
Pathogen kits Nonsmall cell lung KRAS gene
cancer

QIAamp CNA kit, Maxwell 1 mL plasma Fluorescent assay Microfluidic (40)

RSC and MagaPure Lung and colon cancer Digital PCR assays: electrophoresis
automated systems EGFR and KRAS genes

9”x6”
08-Jul-22 5:58:09 PM
b4444_Ch-04.indd 73

9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

QIAamp CNA, PME, Maxwell 0.5 mL plasma Fluorescent assay qPCR assays: (37)
RSC, EpiQuick, NEXTprep- Pancreatic pathology Digital droplet PCR of 105 and 236 bp of
Mag cfDNA v1 and 2 kits KRAS gene GAPDH gene

Unpurified plasma 1-2 mL plasma Fluorometric method — (41)

and QIAamp CNA kit Metastatic colorectal Digital PCR assays:
cancer KRAS, NRAS, and BRAF
genes
QIAamp CNA, QIAamp DBM, 1 mL plasma qPCR assays: Zebrafish DNA: (21)
QIAamp Ultrasens Virus and Pooled healthy 144 bp of cxcr4b gene fragmented and intact
QIASymphony DSP Virus individual 115 and 247 bp of Alu 20 ng/mL of plasma

Maximizing Yield from Plasma cirDNA Extraction 73

kits sequences
105 bp of TP53 gene

PIBEX method 5 mL Plasma qPCR assays: Microfluidic (39)

and QIAamp CNA kit Healthy individual TERT, RPPH1, electrophoresis
GAPDH, and NAGK genes

PME-free, QIAamp CNA, 250 µL and 1 mL Multiplex Digital PCR — (23)

Norgen cfDNA plasma assays:
Mag-Bind, MagMax, Maxwell Pooled 5 short PCRs and 4 long
RSC, ccfDNA NeoGenestar PCRs
kits
08-Jul-22 5:58:09 PM
  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”
74 Cell-Free Circulating DNA: Purification and Analysis Techniques
Direct analysis of unpurified plasma
optimized phenol-chloroform method
Liquid–liquid

Guanidine resin method

extraction

NaI method
PCI-glycogen method
phenol-chloroform phase lock tubes method
triton/heat/Phenol-Chloroform method
phenol chloroform method
PIBEX method
ccfDNA NeoGeneStar kit
Summary of cirDNA Extraction Method Comparison Studies

Mag-Bind cDNA kit

MagMAX cfDNA kit
NEXTprep-Mag cfDNA v2 kit
NEXTprep-Mag cfDNA v1 kit
Epiquick ccfDNA kit
Maxwell RSC automated system
BSCI SNAP method
QIAamp Ultrasens Virus kit
Akonni TruTip kit
Matrix-based extraction

NucliSens system
PME-free kit
Norgen cfDNA kit
Norgen CDNA kit
Chemagic NA kit
Promega Maxwell 16LEV kit
ZR serum DNA kit
Puregene kit
CGD method
Agencourt Genfind kit
QIAamp MinElute Virus Spin kit
MagnaPure automated system
FitAmp kit
Table 2:

Invitrogen ChargeSwitch kit

QIAamp DSP virus kit
Nucleospin (RP and HS) kit
QIAamp CNA kit
QIAamp DBM kit

(14)
(16)
(25)
(29)
(27)
(26)
(28)
Reference

circulating
Healthy

patient
cancer
and/or

DNA
b4444_Ch-04.indd 74 08-Jul-22 5:58:10 PM
b4444_Ch-04.indd 75

9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

(41)
(32)
(13)
(17)
(11)
(12)
(36)
(22)
(40)
(37)
(21)

Maximizing Yield from Plasma cirDNA Extraction 75

(23)
(39)
Fetal (66)
circulating (20)
DNA (70)*
(72)
(74)
method was included in study
method gave the highest DNA yield in study
criteria other than total DNA yield were used to evaluate method
*Akonni TruTip method had a lower total yield of DNA but a higher percentage fraction of fetal DNA.
08-Jul-22 5:58:10 PM
  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

76 Cell-Free Circulating DNA: Purification and Analysis Techniques

or joint-best-performing DNA extraction kit (Tables 1 and 2),

largely due to improved efficiency in extracting small DNA frag-
ments.12 While most published studies indicate that the QIAamp
CNA kit is most effective for extraction of cirDNA, there remain
many gaps in the literature (Tables 1 and 2). Of the 26 kit-based
methods evaluated, 15 have been tested in only a single study,
while 6 have been tested in only 2 studies. The remaining five eval-
uated kits (QIAamp DBM kit, QIAamp CNA kit, Macherey-Na-
gel Nucleospin kit, Maxwell RSC ccfDNA Plasma (Promega),
and PME-free-circulating DNA extraction (Analytik Jena)), have
been tested by >2 individual studies. Individual studies comparing
cirDNA extraction methods are described as follows.
Two commercial kits, the NucleoSpin plasma XS (Mache-
rey-Nagel) kit and the QIAamp DBM kit were compared using 200
µL (Rapid protocol: RP) or 240 µL (High-sensitivity protocol: HS)
of colorectal cancer plasma.13 The recovery of cirDNA was mea-
sured using a PicoGreen assay (Invitrogen) and two qPCR assays,
135 bp of ERV-3 gene14 and 81 bp of b-globin gene, and fragment
sizes were measured using the Agilent 2000 Bioanalyzer.15 The
NucleoSpin kit demonstrated a good recovery of spike-in of 50,
100, 150, 250, and 1000 bp of fragmented DNA compared to the
QIAamp DBM kit, with best recovery of the 50 bp fragment. In
addition, the b-globin assay showed PCR inhibition with the eluate
input of the QIAamp DBM kit.
In another study, seven methods were compared, including
a phenol–chloroform isoamyl with glycogen (PCI–glycogen), a
sodium iodide method (NaI), a guanidine–resin protocol, and four
commercial kits: QIAamp DBM, Invitrogen ChargeSwitch, ZR
Serum DNA and Puregene DNA Purification System Cell, and
Tissue kits.16 The PCI–glycogen, NaI method, and the QIAamp
DBM kit extracted higher yield of cirDNA from 2 mL of 12 pooled

b4444_Ch-04.indd 76 08-Jul-22 5:58:10 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 77

colorectal cancer sera quantified by a Quant-itTM dsDNA HS assay

(Invitrogen). The NaI method had the highest yield measured by
qPCR assays of a 68 bp CDH1 gene and of a 115 bp ACTB gene.
Moreover, for the NaI and PCI–glycogen methods, the size of the
smallest fragments (200 bp) was also detected.
Two column-based kits, the QIAamp DBM kit, The Nucle-
oSpin kit, and the automated MagNAPure isolation system
(Roche Diagnostics) were compared using 1 mL of plasma from
lung cancer, benign lung disease, esophageal cancer, and non-
Hodgkin lymphoma cancer patients.17 Extracted cirDNA was
quantified using three qPCR assays of 92, 101, and 135 bp tar-
geting the GAPDH, b-globin, and ERV-3 genes, respectively.14,18,19
The quantity of extracted cirDNA was between 1.6 and 2.9 ng/mL
with the QIAamp DBM kit, 4.9 and 8 ng/mL for the NucleoSpin
kit, and the highest yield, 12.4 and 28.1 ng/mL, was obtained for
the MagNA Pure system.
Four commercial kits, the QIAamp DBM kit, the QIAamp
CNA kit, the NucleoSpin kit, and the FitAmp Plasma/Serum DNA
isolation (Epigentek) kit, were compared using 1 mL of pooled
healthy female plasma.11 Three qPCR assays of the GAPDH (291
bp), ACTBL2 (67 bp), and HPRT1 (64 bp) genes, and a Qubit HS
dsDNA Assay (Life Technologies) were used to quantify cfDNA
and showed that the two QIAamp kits (mean 264.8 ng/mL for
QIAamp DBM and 239.85 ng/mL for QIAamp CNA) had a 10-fold
higher yield than the two other kits. Given the results of other
studies,12,20,21 it is somewhat unexpected that the QIAamp DBM
and QIAamp CNA kits recovered similar yields of cirDNA from
plasma. However, the level of cirDNA in the plasma samples used
was anomalously high for control plasma (~250 ng/mL), raising the
possibility that in addition to low-molecular-weight cirDNA, the
plasma also contained a large amount of high-molecular-weight

b4444_Ch-04.indd 77 08-Jul-22 5:58:10 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

78 Cell-Free Circulating DNA: Purification and Analysis Techniques

genomic DNA, and the QIAamp DBM and CNA kits recover
high-molecular-weight DNA with equal efficiency.12,21 A spike-in
of λ/HindIII DNA (Ambion) of 50–0.05 ng/plasma demonstrated
that the QIAamp CNA kit gave the highest yield of the smallest
564-bp fragment and gave similar recovery of the 23-bp fragment
compared to the QIAamp DBM kit.11 cirDNA was then isolated
from 1 mL of plasma from four metastatic breast cancer patients
using the two QIAamp kits and sequenced using the Ion AmpliS-
eqTM Cancer Hotspot Panel v2. The recovery fragment size and
sequencing coverage were similar for both kits.
Another study compared the same four kits using pooled and
individual healthy female plasma.12 cirDNA was quantified using
seven qPCR assays of 79 bp of the TERT, 64 bp of the RPPH1,
135 bp of the ERV3, 94 bp of the GAPDH, 66 bp of the NAGK
genes and ALUJ sequences, a commercial ValidPrimeTM assay, and
a digital droplet PCR. A fragmented ADH plasmid was spiked to
samples to estimate the recovery. qPCR assays of 115 bp of ADH,
461 bp of Adhb, and 1448 bp of Adhδ were used to estimate the
extraction efficiency. The QIAamp CNA kit had the better effi-
ciency (80%) for all fragment sizes and gave the best represen-
tation of smaller-size fragments (83%). The recovery efficiency,
compared to the QIAamp CNA kit, was 2-fold and 4.8-fold lower
for both the QIAamp DBM and the NucleoSpin kits, respectively.
The QIAamp DBM kit had a better efficiency than the Nucle-
oSpin kit, but the proportion of shorter fragments was lower
than for the QIAamp CNA and the NucleoSpin kits. QIAamp
kits had a good reproducibility in which the correlation variance
was approximately 10% for all assays. The largest difference mea-
sured was more than 2-fold higher between the lowest ERV3 and
highest TERT measurements. The mean quantity of total cirDNA
measured using all seven PCR assays was about 2500 copies/mL

b4444_Ch-04.indd 78 08-Jul-22 5:58:10 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 79

(8 ng/mL) for individual plasma. A confidence interval of mean

copy number based only on three reference genes — TERT,
RPPH1, and ERV3 could give sufficient estimation of the total
cirDNA. Furthermore, these two comparison studies, which used
the same four kits, showed that the QIAamp CNA kit gave a higher
yield of cirDNA and it was the best method to recover short frag-
ment lengths from 1 mL of plasma samples of healthy individuals
and cancer patients.11,12
Three commercial kits, the PME-free-circulating DNA
extraction (Analytik Jena), the QIAamp DSP Virus/Pathogen (DSP
virus/Pathogen) Midi (Qiagen), and the QIAamp CNA were also
compared using 2 mL plasma from pooled individual and NSCLC
patients.22 The cirDNA was quantified using the ABI TaqMan
Rnase P Detection Reagent kit (87 bp), and KRAS mutations were
analyzed using the rascreen KRAS RGQ kit (Qiagen). Extraction
was achieved using all three kits but the QIAamp CNA gave a
higher yield of cirDNA (3.03 ± 2.19 ng/µL): 3.6-fold and 5.7-fold
higher than Analytic Jena’s and DSP Virus/Pathogen, respectively.
This study also demonstrated the necessity of standardization of
pre-analytical steps such as sample collection tubes, incubation
time, centrifugation steps, input of plasma, and extraction meth-
ods to improve the detection of mutations and the yield of cirDNA.
Three column-based kits, the QIAamp CNA, the Plasma/
Serum Cell-free Circulating DNA Purification Midi (Norgen),
and the PME-free-circulating DNA extraction, and four mag-
netic-bead-based kits, the Mag-Bind circulating DNA (Omega
Bio-Tek), MagMAX Cell-free DNA isolation (Life Technolo-
gies), Maxwell RSC, and Circulating Cell Free DNA (NeoGene-
Star), were also compared starting with 1 mL of control pooled
plasma except for the Norgen and NeoGenestar kits (250 µL).23
The yield and fragment sizes of extracted cirDNA were then

b4444_Ch-04.indd 79 08-Jul-22 5:58:10 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

80 Cell-Free Circulating DNA: Purification and Analysis Techniques

analyzed by digital droplet PCR using a multiplexed assay targeting

nine single-copy genomic loci, which contained five PCRs (67–75
bp) labeled with FAM and four PCRs labeled with TET ranging
between 439 and 522 bp. The highest median yield (1.936 haploid
genome equivalent/mL) was obtained with QIAamp CAN kit with
89% of low-molecular-weight fraction. Although, the median yield
extracted from Norgen was not significantly different (1.760 cop-
ies/mL plasma, t-test p = 0.427), it was more variable; however, it
was adapted to extract low input samples (250 µL). MagMax was
the magnetic beads method that allowed the highest mean yield
even though it was significantly lower than the QIAamp CNA kit
(1.515 copies/mL) but the low-molecular-weight fraction is com-
parable (90%).
Finally, four QIAamp kits including the CNA, the DBM, the
Ultrasens Virus, and the QIASymphony DSP Virus (QS-DSP) were
compared using 1 mL of plasma pooled from healthy donors.21
High and low molecular weights of zebrafish DNA were spiked at
20 ng/mL of plasma and quantified using qPCR assays of 144 bp
targeting the cxcr4b gene. Three qPCR assays of 115 bp and 247
bp of Alu sequences and 105 bp of the TP53 gene were also used.
The QIAamp CNA and QS-DSP kits obtained 40% of recovery
of the spike-in and have a Alu247/Alu105 ratio of about 0.1–0.15.
Furthermore, the QIAamp DBM kit was more efficient in extract-
ing high-molecular-weight DNA, and the Ultrasens Virus KIT had
the lowest efficiency for both high and low molecular weights. The
authors also demonstrated that the carrier RNA improved yields
of cirDNA isolated from both QIAamp CNA and QS-DSP kits.

Evaluation of Phase Partitioning Protocols

Liquid–liquid extraction (also referred to as two-phase extraction
or phase partitioning) is based on the partitioning of molecules

b4444_Ch-04.indd 80 08-Jul-22 5:58:10 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 81

between a polar aqueous phase and a nonpolar organic phase,

typically phenol or a mixture of phenol and chloroform.24 DNA
is charged and will partition into the aqueous phase, while the
solubility of proteins in water is dependent on polar amino acid
residues facing to the outside of the folded protein. Unfolding
in the presence of phenol exposes the hydrophobic amino acid
side chains and the protein partitions irreversibly into the organic
phase. Phenol is denser than water, so the two phases can be sepa-
rated by centrifugation, and the DNA is finally collected from the
aqueous phase by precipitation (Fig. 2). Various modifications can
be introduced to improve the efficiency of the extraction process
including the addition of isoamyl alcohol to the phase-partitioning
mix,16 a heat denaturation step,25,26 and removal of bulk proteins by
a salt precipitation step prior to the two-phase extraction.14,27
A number of studies evaluated liquid–liquid extraction proto-
cols and compared the yield with matrix-based methods (Tables

Fig. 2: Two-phase extraction to purify cfDNA from plasma. The plasma

sample is agitated with a phenol–chloroform mixture to unfold proteins and
dissociate the cfDNA from nucleosomes. The cfDNA partitions into the
aqueous phase, which is separated from the organic phase by centrifugation.
The cfDNA is collected from the aqueous phase by precipitation.

b4444_Ch-04.indd 81 08-Jul-22 5:58:11 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

82 Cell-Free Circulating DNA: Purification and Analysis Techniques

1 and 2). All but one found that the liquid–liquid extraction
method gave a greater DNA yield than any commercial kit; how-
ever, there are two reasons why this does not necessarily mean that
phase partitioning is generally more efficient than matrix binding
for the purification of cirDNA. First, there may be an element
of publication bias in the tendency to publish in-house phenol–
chloroform extraction methods that perform better than estab-
lished commercial protocols but not vice-versa (i.e., the obser-
vation that an in-house phenol–chloroform extraction performs
worse than a commercial kit is not likely to be written up for pub-
lication). Second, and more importantly, only a single study com-
pared phase-partitioning methods against the best performing of
the matrix binding kits (QIAamp CNA kit), and this study found
that better yield was obtained with the QIAamp CNA kit, as well
as the Norgen Plasma/serum Circulating DNA Purification Mini
kit28. Hence, while it would be fair to say that two-phase extraction
is a better choice than the extensively evaluated QIAamp DBM
kit, two-phase extraction has not been extensively tested against
the best-performing matrix binding methods (Tables 1 and 2).
Individual studies of phase-partitioning methods are summarized
in Tables 1 and 2 and are discussed as follows.
A modified salting-out method was established by Schmidt
et al., using 1 mL of plasma and compared to the QIAamp DBM
Kit.14 The cirDNA was quantified using qPCR assay of 135 bp of
the ERV-3 gene. The mean yield of cirDNA was 7.8 and 4 ng/mL
from lung cancer patients and 15.8 and 3 ng/mL for benign lung
disease patients extracted by Schmidt’s protocol and the QIAamp
DBM kit, respectively. Thus, this protocol allowed up to a 5-fold
higher yield of cirDNA compared to the QIAamp DBM kit.
The mentioned protocol14 was further refined by Hufnagl
et al.27 by eliminating one of the DNA precipitation steps and

b4444_Ch-04.indd 82 08-Jul-22 5:58:11 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 83

using a slightly different salt concentration. It was compared to the

Maxwell 16 LEV DNA purification kit (Promega), the QIAamp
DBM kit, the NucleoSpin kit, and Schmidt’s protocol14 using 1 mL
of plasma from healthy individuals and colon and breast cancer
patients. The cirDNA was quantified using a qPCR assay of 98 bp
of the hTERT gene. Both published phenol–chloroform protocols
gave a higher yield of cirDNA than the commercial kits, but Hufna-
gl’s protocol27 gave a 4-fold higher yield (86.91 and 755.4 pg/µL for
1 mL of healthy and colon cancer patients, respectively) compared
to Schmidt’s protocol. The authors speculated that reducing the
DNA precipitation steps avoided some loss of DNA. Integrity of
cirDNA was evaluated using PCR products of 100, 200, 300, 400,
and 600 bp.30 These fragments were detected using all extraction
methods and the cirDNA from cancer patients was found to be
more fragmented compared to healthy patients. Mutation detec-
tions of KRAS codon 12–13 or BRAF V600E showed an increase
of the amount of cirDNA in cancer patients compared to healthy
plasma.
A Triton/Heat/Phenol–chloroform protocol (THP) was devel-
oped by Xue et al.25 It was compared to the QIAamp DBM kit
using 500 µL of pooled healthy plasma or serum spiked with 50
or 200 ng/ml of bcl-2 plasmid DNA or genomic DNA and Quick-
Load (New England Biolabs) 100 bp DNA ladder. The mean yields
of cirDNA, quantified using the hGAPDH qPCR assay,31 were
4.73 ng/ml and 1.67 ng/mL for the THP protocol and the QIAamp
DBM kit, respectively. The efficiency of the THP method was
higher (38.7%) compared to the QIAamp DBM kit (18.6%).
Yuan et al. evaluated the improvement in yield gained by
using phase-lock tubes to isolate cirDNA from 1 mL of nons-
mall-cell-lung cancer (NSCLC) patient plasma.29 The tubes con-
tain a gel that remains between the aqueous and the organic phase

b4444_Ch-04.indd 83 08-Jul-22 5:58:11 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

84 Cell-Free Circulating DNA: Purification and Analysis Techniques

following centrifugation and enables easy collection of the aqueous

phase. The tubes were compared to both the conventional phe-
nol–chloroform protocol and the QIAamp MinElute Virus Spin
Kit (Qiagen). The quantity of cirDNA was then measured using
three qPCR assays of 77, 123, and 202 bp of the SERPINA1 gene
and a NanoDrop Spectrometer. The efficiency of the extraction
method was also evaluated by mutation detection of exon 9 of
the EGFR gene using both mutant-enriched PCR coupled with
Sanger sequencing and the DxS EGFR mutation test kit (DxS
LtD). The phase-lock tubes allowed extraction of a higher yield
of cirDNA compared to the standard PCR protocol. The amount
of extracted cirDNA using this modified protocol was about 20%
higher, and the proportion of long fragments (>202 bp) was signifi-
cantly higher (54.2%) compared to the QIAamp Virus kit (32.2%).
In addition, both the detection of mutant EGFR and the quantity
of tumor-derived cirDNA were higher with this protocol.
Finally, Mauger et al. compared 11 different methods includ-
ing 5 liquid–liquid extractions and 6 matrix-based methods, quan-
titating the yield by qPCR.28 The protocol described by Hufnagl
et al.27 was identified as the most efficient of the phase-partitioning
extractions but overall less effective than the QIAamp CNA kit
or the cfDNA kit from Norgen.28 There was large variation in the
performance of the different methods on different sample types,
for example, cancer versus control, possibly reflecting differences
in the size distribution of the DNA in the samples. This was the
only study to include Qiagen’s CNA kit in comparisons with two-
phase extractions.
Liquid–liquid extraction protocols have several advantages:
they are relatively inexpensive compared with column-binding
kits, and they give more control over plasma input volume and
final sample volume allowing greater flexibility in concentrating

b4444_Ch-04.indd 84 08-Jul-22 5:58:11 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 85

the DNA sample. The disadvantages are that they are time con-
suming, taking up to 2 days, and involve considerable handling of
noxious reagents such as phenol.

Samples with Special Volume Requirements

Some protocols lend themselves better to small or large volumes
of plasma. Where the plasma volume is very small, disproportion-
ate DNA losses can occur through nonspecific binding to plastic
during handling, and the protocol must be compatible with a very
low elution or resuspension volume in order to minimize sample
dilution. Conversely, large volumes of plasma require specialized
plastic ware in order to accommodate sufficient sample volume to
avoid an unmanageable number of pipetting steps.

CirDNA Extraction from Low Plasma Volume Input

In the case of low input of plasma/serum, it is crucial to use the
most efficient extraction method to maximize the yield of isolated
cirDNA from a limited amount of sample. Several studies com-
pared extraction methods starting from small volumes of plasma
(Tables 1 and 2).
Four commercially available kits (QIAamp DBM, QIAamp
MiniElute Virus Spin, Agencourt Genfind Blood and Serum
Genomic DNA Isolation (Agencourt), and Invitrogen Charge-
Switch), were compared using 600 µL of plasma and serum from
healthy and small cell lung cancer (SCLC) patients.32 The quantity
of extracted cirDNA was measured using a 77-bp qPCR assay of
the AAT gene. The QIAamp Virus kit gave the highest yield of
cirDNA, and the cirDNA in SCLC patients was higher compared
to the healthy individuals in plasma (24.5 and 5.1 ng/mL, respec-
tively); however, there was no significant difference between

b4444_Ch-04.indd 85 08-Jul-22 5:58:11 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

86 Cell-Free Circulating DNA: Purification and Analysis Techniques

patients and controls in serum. The fragment size of the extracted

DNA was estimated using PCR of the p53 gene (272 bp) and the
BCl-2 gene (512 bp). The ratio of 272/77 bp fragments was higher
in SCLC patients compared to healthy controls (13% and 8%,
respectively), and the 512 bp fragment was not detected in the
healthy plasma samples. In addition, this study showed that the
plasma was more reliable than serum analysis of tumor-derived
cirDNA.
Another study used plasma to compare cirDNA extraction
from 11 methods and establish a complete workflow to isolate,
quantify, and characterize cirDNA isolated from small sample
volumes (as little as 200 µL).28 Five published protocols14,25,27,29 —
THP, Hufnagl, Yuan and Schmidt’s PC, and five kits — QIAamp
CNA, NucleoSpin (HS and R protocols), the Chemagic NA
extraction (Perkin-Elmer), and both Norgen Plasma/Serum Cir-
culating DNA Purification Mini (Norgen cDNA) and cell-free Cir-
culating DNA Purification Mini (Norgen cfDNA), were compared
using 10 plasma samples from metastatic cancer patients (colon,
pancreatic, lung, and breast) or healthy individuals. The cirDNA
was quantified by Qubit HS dsDNA Assay (Life Technologies) and
two qPCR assays using a standard curve of fragmented genomic
DNA. The qPCR assays targeted a repetitive sequence throughout
the genome, Kpn (55 bp), and a microsatellite sequence, DHFRP2
(165 bp). NucleoSpin, QIAamp CNA, both of the Norgen kits and
the two modified PC protocols27,29 were found to be suitable for
small volumes of plasma. Moreover, for the protocols that gave
higher amounts of cirDNA, all quantification methods gave sim-
ilar results. The Norgen cDNA and QIAamp CNA kits showed
the best accuracy and reproducibility. Previously, qPCR assays of
the APP gene (67 and 180 bp)33,34 and qPCRs (400 and 800 bp)
of the p53 gene35 showed that there was no bias of small-size

b4444_Ch-04.indd 86 08-Jul-22 5:58:11 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 87

fragments for the Norgen cDNA and QIAamp CNA kits. The Nor-
gen cDNA kit allowed extraction of the highest amount of cirDNA
from a small amount of plasma.
A new technology based on a proprietary, undisclosed enrich-
ment and recovery of cirDNA from droplet volumes of plasma
(CGD method) was reported by Spurgin and colleagues.36 CGD
extraction from 20 to 200 µL of cancer plasma was compared to the
QIAamp CNA kit. The cirDNA obtained was quantified by Qubit
HS and BR assays and by qPCR assays of KRAS, BRAF, PIK3CA,
and NRAS genes. The average yield of cirDNA was an astonishing
92.5 ng per µL of plasma for the CGD enrichment method com-
pared to 0.42 ng per µL of plasma for the QIAamp CNA kit, even
though the plasma input was 10-fold lower compared to QIAamp
CNA kit. The CGD method yielded size distributions from 100 to
500 bp. Furthermore, the comparison of Ct values of qPCR assays
showed that this protocol recovered at least 100-fold more ampli-
fiable DNA than the QIAamp CNA kit. Spike-in of 5 ng/mL of
standard 4 Reference DNA and 20 ng/mL of standard 6 Reference
DNA (Horizon Dx) showed 6/10 and 12/20 mutations detected
in cirDNA extracted from CGD method and no mutations were
detected in cirDNA isolated from the QIAamp CNA kit. Finally,
the sequencing analysis showed that the quality, the quantity of the
reads, and the detection of mutations were higher for this enrich-
ment technology compared to the QIAamp CNA kit.
Finally, QIAamp CNA, PME-free-circulating DNA
extraction, and magnetic beads methods (including the Maxwell
RSC ccfDNA (Promega), the EpiQuick Circulating Cell-Free
DNA Isolation (Epigentek) and two consecutive versions of the
NEXTprep-Mag cfDNA Isolation kit (Bioscientific)) were com-
pared using 0.5 mL of plasma. Ten plasma samples were ana-
lyzed, including five from patients with KRAS mutations related to

b4444_Ch-04.indd 87 08-Jul-22 5:58:11 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

88 Cell-Free Circulating DNA: Purification and Analysis Techniques

pancreatic pathology.37 Digital droplet PCR was used to quantify

total cirDNA and KRAS mutations. cirDNA integrity measure-
ment was performed by qPCR using two amplicons of 105 and 236
bp targeting the GAPDH gene and the Qubit assay was used to
quantify cirDNA. QIAamp CNA and Maxwell RSC allowed simi-
lar average yields of 56.49 and 62.57 copies/µL, respectively, and
detection of KRAS mutations. Whereas the PME and the NEXT-
prep-Mag (version 2) kits enabled less yields of 4.06 and 6.08
copies/µL, respectively. CirDNA extracted from EpiQuick and
NEXTprep version 1 cannot be amplified by digital droplet PCR
due to the presence of inhibitors. qPCR analysis of both ampli-
cons showed that QIAamp CNA and Maxwell RSC were similar,
whereas efficiency of the other kits for isolation of short fragments
were lower.

cirDNA Extraction from High Plasma Volume Input

Given the low concentration of cirDNA in plasma, one approach
to increasing yield where sufficient sample volume is available is to
simply start with a higher input of plasma into the extraction pro-
tocol. However, the plastics supplied with commercial kits may not
easily accommodate sample input volumes above 1 or 2 mL. The
QIAamp CNA kit is designed to process up to 5 mL of sample, and
a single study has shown linear recovery of endogenous cirDNA
up to 17.5 mL of plasma input, with no column blockage.38 In this
context, one advantage of liquid–liquid extraction protocols is that
they scale up easily, and the purified cirDNA is recovered at the
final step by resuspension of a precipitated pellet, allowing many-
fold concentration of the sample. In contrast, column-based meth-
ods may require several consecutive elutions from the membrane
to maximally recover the cirDNA, leading to some sample dilution.

b4444_Ch-04.indd 88 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 89

An alternative extraction method called Pressure and Immis-

cibility-Based Extraction (PIBEX) that is centrifugation-free has
been developed and compared with the most commonly used
QIAamp CNA kit using 5 mL of plasma samples of seven healthy
individuals.39 The PIBEX method takes advantage of a polarity dif-
ference between liquids for a vacuum-driven flow system to oper-
ate under a low vacuum pressure throughout the entire process.
The total cirDNA concentration was quantified by qPCR using
TERT, RPPH1, GAPDH, and NAGK genes. The concentration of
extracted cirDNA from both methods was between 1.8 and 44 ng/
mL and the efficiency of the PIBEX method was higher than the
QIAMP CNA kit for three genes except for RPPH1 qPCR. The
size of cirDNA was also compared using microfluidic electropho-
resis (Bioanalyzer 2100, Agilent) and showed peaks at 169 and 172
bp for PIBEX method and QIAMP CNA kit, respectively.

cirDNA Extraction by Automated Systems

cirDNA extraction from plasma/serum can be performed using
commercially available automated systems that are based on mag-
netic bead isolation. Maxwell RSC automated systems showed sim-
ilar isolation efficiency from 0.5 mL of plasma but less efficiency
from 1 mL of plasma compared to the QIAamp CNA kit.23,37 The
MagNAPure automated system had 2–3-fold higher efficiency
than the NucleoSpin kit and 5–10-fold higher efficiency than the
QIAamp DBM kit.17
Another study compared these automated systems with the
most effective QIAamp CNA kit.40 cirDNA was extracted from
1 mL of plasma from lung and colon cancer patients followed
by quantification using Qubit fluorometer and fragment length
analysis by the Agilent 2100 Bioanalyzer prior to digital PCR

b4444_Ch-04.indd 89 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

90 Cell-Free Circulating DNA: Purification and Analysis Techniques

quantification of mutations. The median concentration of cirDNA

of 33 plasma samples was 1.25 ng/µL and 1.08 ng/µL per plasma
from Maxwell RSC system and QIAamp CNA kit, respectively.
cirDNA extraction of 26 plasma samples using both automated
systems showed that MagNAPure system extracted significantly
less cirDNA yield than Maxwell RSC. 88 % of cirDNA extracted
from automated systems showed nucleosome-bound DNA such
as mono-, di-, tri- nucleosomes and long-fragment cirDNA. The
recovery of fragments, between 150 and 200 bp, was higher for
MagNAPure compared to Maxwell RSC systems. In addition,
there was not a significant difference in the quantification of EGFR
and KRAS mutations according to extraction methods.

Direct Analysis of cirDNA from Unpurified Plasma

Direct analysis of plasma samples without any prior extraction
method or enrichment method may be used to improve the analy-
sis of cirDNA from a limited amount of plasma (Tables 1 and 2);
however, this method does not lend itself to downstream applica-
tions in which purified cirDNA is required.
A strategy using direct analysis of cirDNA without prior
extraction was compared to conventional cirDNA extraction meth-
ods including the QIAamp DBM kit and the Triton/Heat PCI-
based method using plasma from patients with coronary heart
disease and healthy plasma.26 The unpurified plasma was diluted
1:40 with water and 6.4 µL was divided between three PCR reac-
tions with a total volume of 48 µL, resulting in 0.05 µL final plasma
volume per PCR reaction. The quantity of cirDNA was measured
using 90 and 222 bp qPCR assays of the L1PA2 gene and an 88 bp
PCR assay of the MSTN gene. The quantity of cirDNA of untreated
plasma was 2.79-fold higher than the quantity of cirDNA extracted

b4444_Ch-04.indd 90 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 91

using the QIAamp DBM kit. The flow-through of QIAamp DBM

kit contained 36.7% of the total cirDNA from plasma. The liq-
uid–liquid extraction-based method performed well only with
high concentrations of cirDNA in which 87.4% of the amount of
cirDNA was extracted. The analysis of distribution size from 35 to
1500 bp using a Fragment Analyzer (Advanced Analytical) showed
different fragment lengths for both extraction methods depending
on the sample, but analysis of cirDNA extracted from the QIAamp
DBM kit showed an abundant peak around 170 bp. The direct
quantification of cirDNA from unpurified plasma using qPCR
assay of the L1PA2 gene showed that cirDNA from coronary heart
disease patients (20.1 ± 23.8 ng/mL) was 2-fold higher compared
to healthy individuals (9.7 ± 4.2 ng/mL).
Another study compared the mutation detection of KRAS,
NRAS, and BRAF using digital PCR in cirDNA from unpurified
plasma and extracted from the QIAamp CNA kit of 17 metastatic
colorectal cancer patients corresponding to 43 blood samples.41
cirDNA was extracted from 1 to 2 mL of plasma and quantified
using a fluorimetric method. A pre-amplification step is added
prior to digital PCR to obtain 200–2000 copies/µL of cirDNA
to perform digital PCR analysis. The detection rate was 93% in
extracted cirDNA and 88% in unpurified cirDNA and the con-
cordance rate between both groups was 91%. The mean value of
mutant allelic frequency was 16.9 ±18.9 for extracted cirDNA and
18.5 ± 18.9 for unpurified cirDNA and discordant cases have under
0.5% of mutant allelic frequency. The correlation coefficient r2 of
mutant allelic frequency between both groups was 0.82.
Together, the various comparisons of extraction methods,
which evaluated different input volumes and various sources of
plasma/serum, using different methods of quantification or fragment

b4444_Ch-04.indd 91 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

92 Cell-Free Circulating DNA: Purification and Analysis Techniques

analysis (Table 1), showed that the extraction method with the
highest yield had a better efficiency in extracting shorter fragment
lengths (Table 2). In addition, in four comparison studies, the
QIAamp CNA kit was the most efficient extraction method both in
yield and fragment size starting with mL volumes of plasma from
healthy individuals or patients.11,12,22,28 For limited input of plasma,
extraction methods could be also replaced by enrichment methods
or direct analysis of unpurified plasma to improve the analysis of
cirDNA.

Recovery of Shorter Fragment Lengths from Plasma

Next-Generation Sequencing (NGS) has provided an opportu-
nity for the investigation of various fragment lengths of cirDNA
through direct sequencing. For example, mitochondrial, micro-
bial, and tumor-derived cirDNA are differently fragmented and
shorter in length when compared to nuclear cirDNA.42
Jiang et al. studied the size distribution of cirDNA fragments
using the QIAamp DSP Blood (QIAamp DSP Blood) Mini Kit
(Qiagen) to extract cirDNA from 3 to 4.8 mL of plasma from hep-
tacellular carcinoma and chronic hepatitis B patients (with or with-
out cirrhosis) and healthy individuals prior to NGS sequencing.43
They showed that the nuclear cirDNA had a strong peak at 166
bp, mitochondrial cirDNA was shorter than the nuclear cirDNA,
and tumor-derived cirDNA was shorter than nontumor-derived
cirDNA.
Burnham et al., used the QIAamp CNA kit to extract cirDNA
from plasma of lung transplant recipients followed by a sin-
gle-stranded DNA library preparation to analyze different cirDNA
fragment sizes.44 There was higher yield of fragments <100 bp for
microbial cirDNA, mitochondrial cirDNA, and fragments between

b4444_Ch-04.indd 92 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 93

160 and 166 bp for nuclear cirDNA. The size of donor-specific

mitochondrial cirDNA was shorter than recipient mitochondrial
donors.
Efficient extraction methods of short fragment lengths cou-
pled with the improvement of analytical methods for short frag-
ment length analysis or NGS sequencing allowed cirDNA detec-
tion including tumor-derived cirDNA, mitochondrial cirDNA,
microbial cirDNA, and nuclear cirDNA.

Tumor-Derived cirDNA from Total cirDNA

The tumor-derived cirDNA extracted from cancer plasma rep-
resents only a small fraction of the total cirDNA (e.g. 0.01%–1.7%
for colorectal cancer).45 As such, both the extraction method and
the quantification method should focus on fragment length to
specifically detect tumor-derived cirDNA from total cirDNA for
optimal detection of low mutations in tumor-derived ccfDNA. In
describing the following studies, we note that where a purification
method that doesn’t efficiently capture, short cirDNA fragments
was used; the results achieved likely underestimate the sensitivity
of the target assay.
Diehl et al. have established an approach to detect
and quantify the tumor-derived cirDNA from total cirDNA
extracted from plasma.46 QIAamp MinElute Virus vacuum kit
(Qiagen) was used to extract cirDNA from 2 mL of plasma from
patients with colorectal cancer during therapy (before surgery).
Although the performance of this kit has never been compared
against the QIAamp CNA kit (Tables 1 and 2), it has been
compared against the QIAamp DNA Blood Mini Kit,32 and the
fold improvement in yield suggests that it does purify the frag-
mented cirDNA fraction with good efficiency. The total cirDNA

b4444_Ch-04.indd 93 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

94 Cell-Free Circulating DNA: Purification and Analysis Techniques

was quantified using qPCR assay of the LINE-1 gene and the
ratio of mutant to wild type was measured using the BEAMing
(Beads Emulsion Amplification Magnetics) assay to determine
the number of tumor-derived molecules per mL of plasma. The
median number of total cirDNA fragments was 4000 per mL of
plasma, the median percentage of mutant cirDNA was 0.18%,
and the median number of tumor-derived cirDNA fragments
was 39 per mL of plasma.
In a different study, total and tumor-derived cirDNA was
extracted from 500 µL of plasma from melanoma cancer patients
and healthy individuals using the QIAamp DSP Virus Kit and
quantified using qPCR assays of 67, 180, 306, and 476 bp.33 Three
integrity indexes (180/67, 306/67, and 476/67) were calculated,
with the 67-bp fragment assumed to represent the total cirDNA.
Reflecting a different cirDNA size distribution between cancer
patients and healthy controls, the integrity indexes were also dif-
ferent, with the 180/67 ratio best at distinguishing the two popu-
lations. The proportion of 180–306-bp fragments was higher for
melanoma cancer patients and the fraction of 67–180-bp frag-
ments was higher for healthy individuals.
The QIAamp DBM kit was used to extract total and
tumor-derived cirDNA from 200 µL of plasma from metastatic
colorectal cancer patients and healthy individuals.47 Fragmenta-
tion of cirDNA was estimated using qPCR assays of 60, 73, 101,
145, 185, 249, 300, 357, and 409 bp of intron 2 of the KRAS gene
when using the same reverse primer. The quantification was opti-
mal for the smaller 60–100-bp fragments. The concentration and
fragmentation of tumor-derived cirDNA were correlated with
tumor weight, and tumor-derived cirDNA had a higher fragmen-
tation pattern. The metastatic colorectal patients had a 5-fold
higher mean of cirDNA fragmentation compared to the healthy

b4444_Ch-04.indd 94 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 95

individuals. The investigators then established an allele-specific

blocker qPCR assay of KRAS and BRAF genes, based on small
fragments, for the simultaneous quantification of total cirDNA,
the analysis of the mutant allele and the determination of the
integrity index.48 The results confirmed that cirDNA is highly
fragmented and they estimated that 80% of cirDNA fragments
are below 145 bp in colorectal cancer patients.
Andersen et al. established a KRAS allele refractory muta-
tion system qPCR of short amplicon assay to analyze tumor-
derived cirDNA extracted using the QIAamp Virus/Pathogen midi
kit (Qiagen) from 2 mL of metastatic colorectal adenocarcinoma
plasma.49 The detection of KRAS mutation was compared using
short amplicon (85 bp) and long amplicon (120 bp) assays. KRAS
mutation is detected in 74% of samples using the short ampli-
con assay and in 61% of samples using long the amplicon assay.
The quantity of tumor-derived cirDNA was 3-fold higher in short
compared to long amplicons. Consequently, the analysis of tumor-
derived cirDNA was improved using short-fragment qPCR assays
(Chapter 5).
Finally, analysis of mutant allele frequency in tumor-derived
cirDNA using digital PCR assays showed no significant difference
between the QIAamp CNA kit, the MagNAPure, and Maxwell RSC
automated systems, and also between the QIAamp CNA kit and
direct analysis of unpurified plasma obtained a correlation coeffi-
cient of 0.82 for mutant allele frequencies above 0.5%.40,41 In addi-
tion, the quantification comparison of KRAS mutations in cirDNA,
isolated using the QIAamp CNA kit, from metastatic colorectal
cancer plasma samples by two digital quantification approaches
and an E-ice-COLD-PCR enrichment method, showed the same
range and concordance of mutation levels below the clinically rel-
evant threshold.50

b4444_Ch-04.indd 95 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

96 Cell-Free Circulating DNA: Purification and Analysis Techniques

Epigenetic Analysis of cirDNA from Plasma

Beyond classic genetic analysis, the epigenetic profile of cirDNA,
including nucleosome signature and methylation analysis, can also
be analyzed. Indeed, cirDNA maintains epigenetic signatures
whose characterization allows for the identification of the cellular
origin of the extracted fragments, thus further expanding the value
of cirDNA as a relevant biomarker beyond genetic analysis. For
example, the QIAamp DBM kit was used to extract cirDNA from
plasma of donors and breast cancer patients followed by NGS
sequencing to analyze the nucleosome signature.51

Methylated cirDNA from Plasma

Conventional extraction kits are used to extract total cirDNA prior
to cirDNA methylation analysis using NGS bisulfite sequencing.
For example, the QIAamp DNA Micro kit (Qiagen) was used to
extract 1 mL of pooled metastatic breast cancer plasma samples
prior to whole-genome sequencing or targeted bisulfite ampli-
con sequencing.52 The QIAamp DBM kit extracted cirDNA from
hepatocellular carcinoma plasma samples before bisulfite conver-
sion and a methylated CpG tandem amplification and sequenc-
ing method.53 The QIAamp DSP Blood kit also isolated cirDNA
from maternal plasma54,55 and from systemic lupus erythematosus
patients in fractions of IgG-bound or unbound plasma samples
before bisulfite sequencing.56 The QIAamp CNA kit has been used
to extract cirDNA from 2 mL of plasma or 1 mL of serum prior
to whole-genome bisulfite sequencing.57 The QIAamp CNA pro-
tocol was optimized using the following modifications: additional
wash, the elution buffer was heated at 40°C, and the elution was
performed in two steps. The average concentration of cirDNA was
17.7 ± 10.9 ng/ml of plasma from healthy individuals, 49.5 ± 55.2

b4444_Ch-04.indd 96 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 97

of plasma from lung cancer patients and 565.3 ± 1173.6 ng/ml of

serum from pancreatic neuroendocrine tumor patients. Fragment
sizes of cirDNA were analyzed using the Agilent 2100 Bioana-
lyzer, which showed, for some samples, contamination with high-
molecular-weight DNA of about 10,380 bp in length. An additional
step was performed to purify cirDNA using two size selections of
bead purifications: 0.5 X followed by 1.6 X.
While standard DNA extraction kits have been used to extract
methylated cirDNA as part of the total cirDNA pool, a specific
extraction method has also established for cirDNA methylation
analysis. Keeley et al. developed a methylation-on-beads (MOB)
protocol that used silica superparamagnetic beads to extract and
bisulfite convert cirDNA from large volumes (up to 2 mL) of
plasma or serum.58 It was compared to the conventional phenol–
chloroform method and the QIAamp CNA kit using a qPCR assay
of the b-actin gene to quantify methylated cirDNA. The extraction
efficiency of the MOB protocol was 1.5–5-fold higher compared to
the PC protocol and the QIAamp CNA kit and the sensitivity was
improved by 25-fold.
Another strategy for methylated cirDNA analysis is the
extraction of total cirDNA from plasma followed by an enrichment
step for methylated fragments prior to NGS sequencing. A meth-
yl-binding protein capture protocol (MBD-cap) used the QIAamp
CNA kit to extract cirDNA from 35 mL of plasma followed by a
MethylMiner kit (Invitrogen) to isolate methylated cirDNA prior
to a modified version of ChiP-Seq Illumina NGS sequencing.38
The total cirDNA was 6.9–10.7 ng/mL of plasma. The enrichment
using the methyl binding MBD2 of methylated cirDNA allowed
10.2%–14.9% of recovery and a fragment size around 180 bp.
Finally, a whole-genome sequencing method used the QIAamp
CNA kit followed by methyl-CpG immunoprecipitation EpiMark

b4444_Ch-04.indd 97 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

98 Cell-Free Circulating DNA: Purification and Analysis Techniques

(New England Biolabs) for the isolation of methylated cirDNA

from pregnant and nonpregnant patient plasma samples.59

Fetal cirDNA from Maternal Plasma

DNA purification methods are also used to extract fetal cirDNA
from maternal plasma for noninvasive pre-natal diagnosis test-
ing. The fetal cirDNA is present in low abundance in maternal
plasma, especially in the beginning of the pregnancy, and it must
represent a sufficient fraction of the total cirDNA for analysis to
be successful.

Fragment Lengths from Maternal Plasma

The main challenge of extraction methods is the efficiency in
extracting different sizes of fragments, especially in the case of
maternal plasma. As with the tumor DNA studies described ear-
lier, we note that extraction kits not configured for cirDNA purifi-
cation would have underestimated the fragmentation and the total
concentration of the DNA.
A size distribution study by Chan et al., of cirDNA extracted
using QIAamp DBM kit from 2 mL of plasma from pregnant and
nonpregnant individuals showed that cirDNA in pregnant samples
was longer than nonpregnant samples and that the fetal cirDNA
was more highly fragmented: 20% of fetal cirDNA was >193 bp
and 0% >313 bp as determined using targets within the SRY gene.60
The same group published another study using the QIAamp DSP
Blood kit to extract 4 mL of maternal plasma at 12 weeks of gesta-
tion.61 The NGS sequencing results showed high-abundance frag-
ments at 166 bp for maternal cirDNA, which was reduced to 143
bp for fetal cirDNA.

b4444_Ch-04.indd 98 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 99

The size distribution of fetal cirDNA was also compared to

total cirDNA by Fan et al., using the NucleoSpin Plasma F kit
(Macherey-Nagel) to extract total cirDNA from 1.6–2.4 mL of
maternal plasma (12th to 23rd weeks of gestation with a male foe-
tus) followed by NGS sequencing.62 The quantity of fetal cirDNA
was 0.7–5.6 ng/mL of plasma, measured using qPCR of DYS4 on
chromosome Y. The size distribution of cirDNA showed an abun-
dant peak at 162 bp and a low-abundance peak at 340 bp. The frac-
tion of fetal cirDNA was higher for shorter (< 150 bp) sequences
for chromosome Y.
Another study by Kimura et al. used QIAamp DBM kit to
extract total cirDNA from 1 mL of maternal plasma samples (17th
to 39th weeks of gestation).63 Fragment sizes were evaluated using
Y-STR and SRY genes (100–524 bp). The study showed that the
fetal cirDNA was more highly fragmented and the mean fragment
size detected were approximately 286 bp ± 28 bp.
In addition, the size distribution of cirDNA has been used to
separate fetal cirDNA from isolated total cirDNA.64 A protocol for
the enrichment of fetal cirDNA was established by Ramezanzadeh
et al., using size separation. The GenetBio genomic DNA isola-
tion kit was used to extract total cirDNA from 200 µL of maternal
plasma (12 weeks of gestation) that was separated on an electro-
phoresis gel and fragments below 300 bp were extracted and puri-
fied to be used for paternal mutation detection by allele-specific
qPCR assay.
Finally, size distributions in plasma and exosomes from non-
pregnant and pregnant donors were compared from 0.5 mL of
plasma extracted using the QIAamp CNA kit. The yield of differ-
ent fragment sizes was analyzed by four digital PCR b-actin gene
assays of 76, 135, 490, and 905 bp and by Agilent 2100 Bioana-

b4444_Ch-04.indd 99 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

100 Cell-Free Circulating DNA: Purification and Analysis Techniques

lyzer.65 The relative concentrations of 135, 490, and 905 bp per 76

bp in plasma were 39%, 18%, and 5.6 % for nonpregnant and 34%,
14%, and 23% for pregnant women, whereas the relative concen-
trations in exosomes were 40%, 18%, and 3.3% for nonpregnant
and 30%, 12%, and 18% for pregnant women. The size distribu-
tion in plasma showed a majority of peaks between 150 and 200 bp
and also peaks between 300 and 500 bp and 1000 and 10,000 bp
for pregnant donors, and a very small peak between 300 and 500
bp for nonpregnant donors. The size distribution of exosomes was
150–550 bp for nonpregnant donors and between 300 and 500 bp
and between 1000 and 10,000 bp for pregnant donors. Further-
more, the analysis of nonpregnant plasma pellet showed 87%, 75%,
and 68% of relative concentrations of 135, 490, and 905 bp per 76
bp, and the majority of peaks were ≥1000 bp with a very small pro-
portion between 150 and 200 bp and 600 and 1000 bp.

Comparison of Fetal cirDNA Extraction Methods

Several studies compared different extraction methods to identify
fetal cirDNA from maternal plasma (Tables 2 and 3). They used
different methods to estimate the yield of fetal cirDNA and total
cirDNA starting with different sample volumes and from various
weeks of gestation (Table 3). Seven matrix-based extractions and
one phenol–chloroform protocol have been evaluated for cirDNA
extraction but only the QIAamp DBM, the QIAamp can, and the
QIAamp DSP virus kits have been compared by ≥ 2 studies.
The NucliSens Magnetic Extraction system, the QIAamp
DSP Virus, and the QIAamp DBM kits were compared using
500 µL (for QIAamp DSP Virus) or 800 µL (for QIAamp DBM and
NucliSens) of maternal plasma samples (13th to 19th or 34th to
37th weeks of gestation) from women who were RhD negative

b4444_Ch-04.indd 100 08-Jul-22 5:58:12 PM

b4444_Ch-04.indd 101

Table 3. Workflow of Fetal cirDNA Extraction Method Comparison from Maternal Plasma

9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Quantification of Fetal Quantification of
Extraction method Maternal plasma ccfDNA Total ccfDNA Reference
NucliSens, QIAamp 500 or 800 µL plasma qPCR assay: qPCR assay of (66)
DSP Virus 13th-19th or 34th-37th 93 bp of RHD gene GAPDH gene
and QIAamp RhD-negative woman
DBM kits and RhD-positive fetus
QIAamp DSP Virus 200 to 1000 µL pooled plasma qPCR assay: qPCR assay of (20)
QIAamp DBM 8th-21th 84 bp of DYS14 gene specific androgen
and QIAamp Male fetus receptor region
CNA kits on X chromosome

Maximizing Yield from Plasma cirDNA Extraction 101

QIAamp CNA 5 and 7 mL plasma Duplex qPCR assay: Duplex qPCR assay: (70)
and Akonni Pregnant with male fetus: 5.3th-13.3th 84 bp of DYS14 gene 81 bp of EIF2C1
TruTip kits Nonpregnant with spike-in male and Digital PCR gene on 1
female fragmented DNA chromosome
Digital PCR
QIAamp DBM kit 200 and 500 µL plasma qPCR assay: qPCR assay: (72)
and THP Pregnant 17th-28th 90 bp of RHD gene 102 bp of b-actin
method RhD positive and husband RhD negative
Nonpregnant
MagnaPure 500 µL plasma Single base extension of UV-spectroscopy (74)
automated 7th–39th exon 10 of RHD gene
system and BCSI Multiplex qPCR assay
SNAP method of exon 3, 4, 5 and 7
08-Jul-22 5:58:12 PM

of RHD gene
  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

102 Cell-Free Circulating DNA: Purification and Analysis Techniques

while their foetus was RhD positive.66 The yields of fetal and total
cirDNA were estimated using qPCR assays of 93 bp of the RHD
exon 767 and of the GAPDH gene,68 respectively. Quantities of fetal
and total cirDNA were higher using NucliSens (1.7- and 2.3-folds,
respectively) and DSP Virus (1.5- and 1.3-folds, respectively) kits
compared to the control QIAamp DBM kit.66
Another study compared three QIAamp kits (DBM, DSP
Virus, and CNA) for isolation of cirDNA from pooled maternal
plasma (8th to 21th week of gestation with a male foetus) using two
qPCR assays of 84 bp of DYS14 on the Y chromosome,69 and the
X chromosome-specific androgen receptor region to quantify fetal
and total cirDNA, respectively.20 Different inputs of plasma and
elution volumes of the extracted cirDNA were used: 1000/100 µL
for the QIAamp CNA kit, 200/200 µL or 500/50 µL for the QIAamp
DBM kit, and 500/50 µL for the QIAamp DSP kit. Consistent with
the data obtained using plasma from nonpregnant donors, the
average yield of fetal cirDNA was considerably higher for QIAamp
CNA (231.45 pg/mL) and QIAamp DSP (188.13 pg/mL) kits com-
pared to the QIAamp DBM kit (34.26 pg/mL). For the same input
of plasma and elution, the yield of fetal cirDNA was 210.22 pg/
mL and 97.59 pg/mL and the average quantity of total cirDNA was
7876 and 4296 pg/mL for QIAamp DSP Virus and QIAamp DBM
kits, respectively. In summary, the QIAamp CNA kit was more suit-
able for higher volumes of plasma and the QIAamp DSP Virus kit
gave a higher yield of cirDNA for low input of plasma.
The QIAamp CNA kit was also compared to the Akonni
TruTip (Akonni Biosystems) kit for 5 mL (for both kits) or 7 mL
(for Akonni TruTip) of nonpregnant and pregnant (5.3 to 13.3
weeks of gestation) plasma samples.70 The amount of fetal and total
cirDNA was quantified using qPCR of a duplex of 84 bp of DYS14
on the Y chromosome69 and 81 bp of EIF2C1 on chromosome 1,

b4444_Ch-04.indd 102 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 103

respectively, and ddPCR.71 Non-pregnant plasma samples were

spiked with a dilution series of male fragmented DNA between 50
and 400 bp and female fragmented DNA between 100 and 1600
bp. The recovery of fetal cirDNA was 60% and 75% using male
fragmented DNA and the recovery of total cirDNA was 40.6% and
93.6% using fragmented female DNA for the Akonni TruTip and
QIAamp CNA kits, respectively. Both kits were very efficient in
extracting small fragments, but the Akonni TruTip preferentially
enriched fetal cirDNA using a step that reduced fragments above
600 bp from total cirDNA. Consequently, the quantity of cirDNA
was higher for the QIAamp CNA kit but the fetal cirDNA/total
cirDNA recovery ratio was higher for the Akonni TruTip.
The QIAamp DBM kit and the THP25 method were com-
pared using 200 mL (QIAamp DBM kit) or 500 mL (THP method)
of maternal plasma from RhD-negative women carrying an
RhD-positive fetus (17th to 28th weeks of gestation), and non-
pregnant women (RhD positive).72 Fetal cirDNA and total cirDNA
were quantified using standard curves using two qPCR assays, 90
bp of the RHD gene exon 7, and 102 bp of the b-globin gene,
respectively.73 The yield, estimated using the Ct value of the fetal
cirDNA was improved using the THP protocol (33.8 ± 1.6) com-
pared to the QIAamp DBM kit (36.1 ± 2.47). In addition, the yield
of total cirDNA was not significantly different for both methods:
32.2 and 30.9 for the QIAamp DBM kit and the THP method,
respectively.
Finally, the MagnaPure automated system was compared to
the binding on a glass surface BCSI SNAPTM card protocol (Blood
Cell Storage Inc.) using 500 µL of maternal plasma (7th to 39th
weeks of gestation).74 The total cirDNA was quantified using
UV-spectroscopy and fetal cirDNA was measured using RHD
gene single base extension (52 SNPs) of a multiplex PCR of exons

b4444_Ch-04.indd 103 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

104 Cell-Free Circulating DNA: Purification and Analysis Techniques

3, 4, 5, and 7 and a qPCR of exon 10. The mean yield of total

cirDNA extracted by BCSI SNAPTM card (33.8 ng/µL) was 2-fold
higher compared to MagnaPure (15.7 ng/µL) and the mean frac-
tion of fetal cirDNA was 116.2 for all samples. The fetal cirDNA
extracted using BCSI SNAPTM card allowed the detection of more
SNPs, especially in earlier pregnancy.
In summary, a number of circulating DNA kits have been
used to extract fetal cirDNA from maternal plasma. Two of the
five comparisons of cirDNA extraction methods (Tables 2 and 3)
showed that the QIAamp CNA kit gave higher yields of fetal and
total cirDNA from 1 mL or 5 mL of maternal plasma.20,70

Conclusions
Numerous extraction methods, either “in-house” or commercially
available, have been developed to isolate cirDNA specifically from
plasma and serum. The dual challenge for these methods is to both
efficiently extract low-abundance cirDNA from the samples and to
capture the variety of fragment lengths, especially when the input
volume is limited. This is particularly the case for tumor-derived
cirDNA and fetal cirDNA, which only represent a fraction of the
total cirDNA, and are more fragmented in plasma.
Comparison of extraction methods has shown that protocols
optimized for cirDNA give a higher yield and better recovery of
short fragments (Tables 1, 2 and 3). The choice of method is also
dependent on the type of cirDNA to be analyzed, the scientific
question asked, and sample limitations. There is not a universal
extraction method to isolate all fragment lengths with the same
recovery efficiency from all types of plasma samples (Table 2).
Consequently, it is important to choose the most appropriate
extraction method for each application and to optimize extraction

b4444_Ch-04.indd 104 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 105

steps (volume of plasma, volume of elution, standard to control

the extraction recovery, etc.) to maximize the recovery of all frag-
ment sizes and the yield of isolated cirDNA. The QIAamp CNA
kit, which is the most commonly used method, appears to be the
most effective and versatile method to maximize the yield of total,
tumor-derived or fetal cirDNAs from a few mL of plasma (Table
2). However, all extraction methods have not been directly com-
pared, they have not been optimized for all applications, for exam-
ple, small versus large sample volumes and they have not been
compared using the same sample, the same quantification method
(fluorescent assay, qPCR, or digital PCR assays), or the same frag-
ment analysis method (microfluidic electrophoresis, qPCR, and
digital PCR assays) (Tables 1 and 3). It is likely that with ongoing
research new techniques will emerge.
In addition, the analysis of cirDNA can also be enhanced
using other approaches. Fetal cirDNA or methylated cirDNA
can be enriched from isolated total cirDNA to improve the anal-
ysis. For a limited amount of plasma, the extraction step can be
replaced by direct analysis of unpurified plasma by digital PCR or
an enrichment step for cirDNA.
Moreover, it is crucial to standardize and optimize the process
for each application especially for the analysis of clinical samples.
To ensure traceability, standardization of processing, and analytical
reproducibility, standard operating procedure, a Laboratory Infor-
mation Management System, and ISO International Standards
(ISO 20186-3: 2019) should be used for plasma sample analysis.
A complete specific liquid biopsy workflow has to be developed
incorporating collection, transport, biobanking, pre-analytical pro-
cess, sample volumes, extraction method, quantification method,
and analysis method for each kind of cirDNA and application
(Tables 1 and 3). Each step of the process can impact the yield

b4444_Ch-04.indd 105 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

106 Cell-Free Circulating DNA: Purification and Analysis Techniques

and fragment sizes recovery of isolated cirDNA and should be

optimized, standardized, and automated to improve target detec-
tion especially in the case of low-abundance tumor cirDNA.
The dramatic results obtained in cancer diagnosis, prediction,
and monitoring have created interest in using cirDNA in a broad
variety of applications. Among these is the promise of cirDNA
becoming a universal biomarker of disease through epigenetic
analysis or the potential of cirDNA to more adequately capture
tumor heterogeneity. This interest will undoubtedly lead to the
development and/or improvement of new extraction for various
types of liquid biopsies (e.g., urine, saliva plasma, cerebrospinal
fluid) and NGS analysis methods in the very near future.

References
1. Thierry AR, El Messaoudi S, Gahan PB, et al. (2016) Origins, structures,
and functions of circulating DNA in oncology. Cancer Metastasis Rev 35:
347–376.
2. Marko MA, Chipperfield R, Birnboim HC. (1982) A procedure for the large-
scale isolation of highly purified plasmid DNA using alkaline extraction and
binding to glass powder. Anal Biochem 121: 382–387.
3. Boom R, Sol CJ, Salimans MM, et al. (1990) Rapid and simple method for
purification of nucleic acids. J Clin Microbiol 28:495–503.
4. Shi B, Shin YK, Hassanali AA, Singer SJ. (2015) DNA binding to the silica
surface. J Phys Chem B 119: 11030–11040.
5. Melzak KA, Sherwood CS, Turner RFB, Haynes CA. (1996) Driving forces
for DNA adsorption to silica in perchlorate solutions. J Colloid Interf Sci
181: 635–644.
6. Romanowski G, Lorenz MG, Wackernagel W. (1991) Adsorption of plasmid
DNA to mineral surfaces and protection against DNase I. Appl Environ
Microbiol 57: 1057–1061.

b4444_Ch-04.indd 106 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 107

7. Tian H, Huhmer AF, Landers JP. (2000) Evaluation of silica resins for direct
and efficient extraction of DNA from complex biological matrices in a min-
iaturized format. Anal Biochem 283: 175–191.
8. Liu L, Guo Z, Huang Z, et al. (2016) Size-selective separation of DNA frag-
ments by using lysine-functionalized silica particles. Sci Rep;6: 22029.
9. Feng J, Gang F, Li X, et al. (2013) Plasma cell-free DNA and its DNA integ-
rity as biomarker to distinguish prostate cancer from benign prostatic hyper-
plasia in patients with increased serum prostate-specific antigen. Int Urol
Nephrol 45: 1023–1028.
10. Vinayanuwattikun C, Winayanuwattikun P, Chantranuwat P, et al. (2013) The
impact of non-tumor-derived circulating nucleic acids implicates the progno-
sis of non-small cell lung cancer. J Cancer Res Clin Oncol 139: 67–76.
11. Page K, Guttery DS, Zahra N, et al. (2013) Influence of plasma processing
on recovery and analysis of circulating nucleic acids. PloS One 8: e77963.
12. Devonshire AS, Whale AS, Gutteridge A, et al. (2014) Towards standard-
isation of cell-free DNA measurement in plasma: Controls for extraction
efficiency, fragment size bias and quantification. Anal Bioanal Chem 406:
6499–64512.
13. Kirsch C, Weickmann S, Schmidt B, Fleischhacker M. (2008) An improved
method for the isolation of free-circulating plasma DNA and cell-free DNA
from other body fluids. Ann N Y Acad Sci 1137: 135–139.
14. Schmidt B, Weickmann S, Witt C, Fleischhacker M. (2005) Improved
method for isolating cell-free DNA. Clin Chem 51: 1561–1563.
15. Jung M, Klotzek S, Lewandowski M, et al. (2003) Changes in concentration
of DNA in serum and plasma during storage of blood samples. Clin Chem
49: 1028–1029.
16. Fong SL, Zhang JT, Lim CK, et al. (2009) Comparison of 7 methods for
extracting cell-free DNA from serum samples of colorectal cancer patients.
Clin Chem 55: 587–589.
17. Fleischhacker M, Schmidt B, Weickmann S, et al. (2011) Methods for isola-
tion of cell-free plasma DNA strongly affect DNA yield. Clinica Chim Acta
412: 2085–2088.
18. Birch L, English CA, O’Donoghue K, et al. (2005) Accurate and robust
quantification of circulating fetal and total DNA in maternal plasma from 5
to 41 weeks of gestation. Clin Chem 51: 312–320.

b4444_Ch-04.indd 107 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

108 Cell-Free Circulating DNA: Purification and Analysis Techniques

19. Lo YM, Tein MS, Lau TK, et al. (1998) Quantitative analysis of fetal DNA in
maternal plasma and serum: implications for noninvasive prenatal diagnosis.
Am J Hum Genet 62: 768–775.
20. Repiska G, Sedlackova T, Szemes T, et al. (2013) Selection of the optimal
manual method of cell free fetal DNA isolation from maternal plasma. Clin
Chem Lab Med 51: 1185–1189.
21. Warton K, Graham LJ, Yuwono N, Samimi G. (2018) Comparison of 4 com-
mercial kits for the extraction of circulating DNA from plasma. Cancer Gen
228: 143–150.
22. Sherwood JL, Corcoran C, Brown H, et al. (2016) Optimised pre-analytical
methods improve KRAS mutation detection in circulating tumour DNA
(ctDNA) from patients with non-small cell lung cancer (NSCLC). PloS One
11:e0150197.
23. Markus H, Contente-Cuomo T, Farooq M, et al. (2018) Evaluation of
pre-analytical factors affecting plasma DNA analysis. Sci Rep 8: 7375.
24. Sambrook MRGaJ. Molecular Cloning: A Laboratory Manual, Cold Spring
Harbor Laboratory Press; 4th edition (June 15, 2012) New York.
25. Xue X, Teare MD, Holen I, et al. (2009) Optimizing the yield and utility of circu-
lating cell-free DNA from plasma and serum. Clinica Chim Acta 404: 100–104.
26. Breitbach S, Tug S, Helmig S, et al. (2014) Direct quantification of cell-free,
circulating DNA from unpurified plasma. PloS One 9: e87838.
27. Hufnagl C, Stöcher M, Moik M, et al. (2013) A modified Phenol-chloroform
extraction method for isolating circulating cell free DNA of tumor patients.
J Nucleic Acids Invest 4: e1.
28. Mauger F, Dulary C, Daviaud C, et al. (2015) Comprehensive evaluation
of methods to isolate, quantify, and characterize circulating cell-free DNA
from small volumes of plasma. Anal Bioanal Chem 407: 6873–6878.
29. Yuan H, Zhu ZZ, Lu Y, et al. (2012) A modified extraction method of cir-
culating free DNA for epidermal growth factor receptor mutation analysis.
Yonsei Med J 53: 132–137.
30. van Dongen JJ, Langerak AW, Bruggemann M, et al. (2003) Design and
standardization of PCR primers and protocols for detection of clonal
immunoglobulin and T-cell receptor gene recombinations in suspect
lymphoproliferations: Report of the BIOMED-2 Concerted Action
BMH4-CT98-3936. Leukemia 17: 2257–2317.

b4444_Ch-04.indd 108 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 109

31. Ponchel F, Toomes C, Bransfield K, et al. (2003) Real-time PCR based

on SYBR-Green I fluorescence: An alternative to the TaqMan assay for a
relative quantification of gene rearrangements, gene amplifications and
micro gene deletions. BMC Biotechnol 3: 18.
32. Board RE, Williams VS, Knight L, et al. (2008) Isolation and extraction of
circulating tumor DNA from patients with small cell lung cancer. Ann N Y
Acad Sci 1137: 98–107.
33. Pinzani P, Salvianti F, Zaccara S, et al. (2011) Circulating cell-free DNA in
plasma of melanoma patients: qualitative and quantitative considerations.
Clinica Chim Acta 412: 2141–2145.
34. Salvianti F, Pinzani P, Verderio P, et al. (2012) Multiparametric analysis of
cell-free DNA in melanoma patients. PloS One 7: e49843.
35. Ebtsam R. Zahera MMA, Hanaa M, et al. (2012) Value of circulating DNA
concentration and integrity as a screening test for detection of cancer in an
Egyptian cohort. Alexandria J Med 48: 187–196.
36. Spurgin J, Ford A, Athanasuleas J, Yeh C. (2015) Next-generation targeted
sequencing of circulating cell-free DNA from droplet volumes of blood.
IJLSR 3: 23–29.
37. Sorber L, Zwaenepoel K, Deschoolmeester V, et al. (2017) A comparison of
cell-free DNA isolation kits: Isolation and quantification of cell-free DNA in
plasma. J Mol Diagn 19: 162–168.
38. Warton K, Lin V, Navin T, et al. (2014) Methylation-capture and next-
generation sequencing of free circulating DNA from human plasma. BMC
Genomics 15: 476.
39. Lee H, Na W, Park C, et al. (2018) Centrifugation-free extraction of circulating
nucleic acids using immiscible liquid under vacuum pressure. Sci Rep 8: 5467.
40. Perez-Barrios C, Nieto-Alcolado I, Torrente M, et al. (2016) Compari-
son of methods for circulating cell-free DNA isolation using blood from
cancer patients: Impact on biomarker testing. Transl Lung Cancer Res 5:
665–672.
41. Sefrioui D, Beaussire L, Perdrix A, et al. (2017) Direct circulating tumor
DNA detection from unpurified plasma using a digital PCR platform. Clin
Biochem 50: 963–966.
42. Mouliere F, Rosenfeld N. (2015) Circulating tumor-derived DNA is shorter
than somatic DNA in plasma. Proc Natl Acad Sci U S A 112: 3178–3179.

b4444_Ch-04.indd 109 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

110 Cell-Free Circulating DNA: Purification and Analysis Techniques

43. Jiang P, Chan CW, Chan KC, et al. (2015) Lengthening and shortening of
plasma DNA in hepatocellular carcinoma patients. Proc Natl Acad Sci U S
A 112: E1317–E1325.
44. Burnham P, Kim MS, Agbor-Enoh S, et al. (2016) Single-stranded DNA
library preparation uncovers the origin and diversity of ultrashort cell-free
DNA in plasma. Sci Rep 6: 27859.
45. Diehl F, Li M, Dressman D, et al. (2005) Detection and quantification of
mutations in the plasma of patients with colorectal tumors. Proc Natl Acad
Sci U S A 102: 16368–16373.
46. Diehl F, Schmidt K, Choti MA, et al. (2008) Circulating mutant DNA to
assess tumor dynamics. Nat Med 14: 985–990.
47. Mouliere F, Robert B, Arnau Peyrotte E, et al. (2011) High fragmentation
characterizes tumour-derived circulating DNA. PloS One 6: e23418.
48. Mouliere F, El Messaoudi S, Pang D, et al. Multi-marker analysis of circu-
lating cell-free DNA toward personalized medicine for colorectal cancer.
Mol oncol 8: 927–941.
49. Andersen RF, Spindler KL, Brandslund I, et al. (2015) Improved sensitivity
of circulating tumor DNA measurement using short PCR amplicons. Clin-
ica Chim Acta 439: 97–101.
50. Sefrioui D, Mauger F, Leclere L, et al. (2016) Comparison of the quantifica-
tion of KRAS mutations by digital PCR and E-ice-COLD-PCR in circulat-
ing-cell-free DNA from metastatic colorectal cancer patients. Clinica Chim
Acta 465: 1–4.
51. Ulz P, Thallinger GG, Auer M, et al. (2016) Inferring expressed genes
by whole-genome sequencing of plasma DNA. Nat Genet 48(10):
1273–1278.
52. Legendre C, Gooden GC, Johnson K, et al. (2015) Whole-genome bisulfite
sequencing of cell-free DNA identifies signature associated with metastatic
breast cancer. Clin Epigenetics 7: 100.
53. Wen L, Li J, Guo H, et al. (2015) Genome-scale detection of hypermethyl-
ated CpG islands in circulating cell-free DNA of hepatocellular carcinoma
patients. Cell Res 25: 1376.
54. Lun FM, Chiu RW, Sun K, et al. (2013) Noninvasive prenatal methylomic
analysis by genomewide bisulfite sequencing of maternal plasma DNA. Clin
Chem 59: 1583–1594.

b4444_Ch-04.indd 110 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Maximizing Yield from Plasma cirDNA Extraction 111

55. Sun K, Jiang P, Chan KC, et al. (2015) Plasma DNA tissue mapping by
genome-wide methylation sequencing for noninvasive prenatal, cancer, and
transplantation assessments. Proc Natl Acad Sci U S A 112: E5503–E5512.
56. Chan RW, Jiang P, Peng X, et al. (2014) Plasma DNA aberrations in sys-
temic lupus erythematosus revealed by genomic and methylomic sequenc-
ing. Proc Natl Acad Sci U S A 111: E5302–E5311.
57. Maggi EC, Gravina S, Cheng H, et al. (2018) Development of a method
to implement whole-genome bisulfite sequencing of cfDNA from cancer
patients and a mouse tumor model. Front Genet 9: 6.
58. Keeley B, Stark A, Pisanic TR, et al. (2013) Extraction and processing of
circulating DNA from large sample volumes using methylation on beads for
the detection of rare epigenetic events. Clinica Chim Acta 425: 169–175.
59. Jensen TJ, Kim SK, Zhu Z, et al. (2015) Whole genome bisulfite sequencing
of cell-free DNA and its cellular contributors uncovers placenta hypometh-
ylated domains. Genome Biol 16: 78.
60. Chan KC, Zhang J, Hui AB, et al. (2004) Size distributions of maternal and
fetal DNA in maternal plasma. Clin Chem 50: 88–92.
61. Lo YM, Chan KC, Sun H, et al. (2010) Maternal plasma DNA sequencing
reveals the genome-wide genetic and mutational profile of the fetus. Sci
Transl Med 2: 61ra91.
62. Fan HC, Blumenfeld YJ, Chitkara U, et al. (2010) Analysis of the size distri-
butions of fetal and maternal cell-free DNA by paired-end sequencing. Clin
Chem 56: 1279–1286.
63. Kimura M, Hara M, Itakura A, et al. (2011) Fragment size analysis of free
fetal DNA in maternal plasma using Y-STR loci and SRY gene amplification.
Nagoya J Med Sci 73: 129–135.
64. Ramezanzadeh M, Salehi M, Farajzadegan Z, et al. (2016) Detection of
paternally inherited fetal point mutations for beta-thalassemia in maternal
plasma using simple fetal DNA enrichment protocol with or without whole
genome amplification: An accuracy assessment. J Matern Neonatal Med 29:
2645–2649.
65. Fernando MR, Jiang C, Krzyzanowski GD, Ryan WL. (2018) Analysis of
human blood plasma cell-free DNA fragment size distribution using Eva-
Green chemistry based droplet digital PCR assays. Clinica Chim Acta 483:
39–47.

b4444_Ch-04.indd 111 08-Jul-22 5:58:12 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

112 Cell-Free Circulating DNA: Purification and Analysis Techniques

66. Clausen FB, Krog GR, Rieneck K, Dziegiel MH. (2007) Improvement in
fetal DNA extraction from maternal plasma. Evaluation of the NucliSens
magnetic extraction system and the QIAamp DSP virus kit in comparison
with the QIAamp DNA blood mini kit. Prenat Diagn 27: 6–10.
67. Clausen FB, Krog GR, Rieneck K, et al. (2005) Reliable test for prena-
tal prediction of fetal RhD type using maternal plasma from RhD negative
women. Prenat Diagn 25: 1040–1044.
68. Johnson KL, Dukes KA, Vidaver J, et al. (2004) Interlaboratory comparison
of fetal male DNA detection from common maternal plasma samples by
real-time PCR. Clin Chem 50: 516–521.
69. Zimmermann BG, Maddocks DG, Avent ND. (2008) Quantification of cir-
culatory fetal DNA in the plasma of pregnant women. Methods Mol Biol
444: 219–229.
70. Holmberg RC, Gindlesperger A, Stokes T, et al. (2013) Akonni TruTip((R))
and Qiagen((R)) methods for extraction of fetal circulating DNA—Evalua-
tion by real-time and digital PCR. PloS One 8: e73068.
71. Fan HC, Blumenfeld YJ, Chitkara U, et al. (2008) Noninvasive diagnosis of
fetal aneuploidy by shotgun sequencing DNA from maternal blood. Proc
Natl Acad Sci U S A 105: 16266–16271.
72. Keshavarz Z, Moezzi L, Ranjbaran R, et al. (2015) Evaluation of a modified
DNA extraction method for isolation of cell-free fetal DNA from maternal
serum. Avicenna J Med Biotechnol 7: 85–88.
73. Wang XD, Wang BL, Ye SL, et al. (2009) Non-invasive foetal RHD genotyp-
ing via real-time PCR of foetal DNA from Chinese RhD-negative maternal
plasma. Europ J Clin Invest 39: 607–617.
74. Adamczyk T, Doescher A, Haydock PV, et al. (2015) The glass slide extraction
system snap card improves non-invasive prenatal genotyping in pregnancies
with antibodies. Transfus Med Hemother 42: 379–384.

b4444_Ch-04.indd 112 08-Jul-22 5:58:12 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0005

Chapter 5
Optimal Design of PCR Assays
for Circulating DNA
Rikke F. Andersen

Abstract

Carefully considered assay design is essential for PCR-based methods

used in circulating DNA (cirDNA) studies. Broadly speaking, there are
three main contexts in which cirDNA is subjected to PCR amplification
and analyses: (1) in order to quantitate overall cirDNA, or a particu-
lar cirDNA sequence; (2) in order to detect mutations or methylation;
and (3) as part of sequencing library preparation. In addition to gen-
eral considerations about primer/probe design, it is equally important to
understand the nature and composition of cirDNA in order to optimize
analyses. Throughout the research literature, many different approaches
have been taken to amplify and quantify cirDNA. Differences in analyt-
ical methods influence results, such that studies become incomparable
and this hampers the progress of promising clinical biomarkers.

Circulating Cell-Free DNA

The origin of circulating DNA (cirDNA) in blood and the mech-
anisms of release have been investigated but are not yet fully

113

b4444_Ch-05.indd 113 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

114 Cell-Free Circulating DNA: Purification and Analysis Techniques

understood. In healthy individuals, the majority of the cirDNA is

of hematopoietic origin.1 Apoptosis appears to be the major release
mechanism for cirDNA, but other mechanisms such as active
release through particular or macromolecular structures also seem
to play an important role.2,3 The genome is not uniformly repre-
sented in circulation; there is an over-representation of Alu ele-
ments and an under-representation of long interspersed nuclear
elements L1 and L2,4–6 confirming that cirDNA is not randomly
released but that active release is involved.

cirDNA Size
Nuclear DNA is wrapped around nucleosomes with linker regions
between the nucleosome cores. The size of the DNA wrapped
around histones and the linker DNA is 146 base pairs (bp) and
20 bp, respectively. Consistent with its proposed origin from apop-
totic cells, most cirDNA is made up of short double-stranded frag-
ments that are 166 bp in length, corresponding to the length of
DNA cleaved between nucleosomes.7 Although the size profile
of cirDNA in healthy individuals is dominated by the peak at
166 bp, other fragment sizes are also seen. Smaller peaks are
found at approximately 143 bp and at 10 bp intervals below that,
corresponding to one turn of the DNA helix wrapped around the
core histone.8,9 This indicates that part of the nucleosomal DNA
sequence is exposed to nucleases. Longer fragments correspond-
ing to the length of DNA wrapped around two or three nucleo-
somes are also found but at much lower levels.7
Recent studies have demonstrated that the amount of short
(<100 bp) fragments of cirDNA is probably more significant than
previously thought.10,11 There are a number of reasons why these
very short fragments do not readily lend themselves to detection
and analysis. First, they cannot be amplified in a PCR reaction if

b4444_Ch-05.indd 114 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 115

the amplicon is larger than the cirDNA fragment; even amplicons

that are the same size or slightly smaller than the cirDNA frag-
ments will not detect them, unless the position of the DNA breaks
aligns closely with ends of the PCR target sequence. Second, they
are difficult to observe on a gel or BioAnalyzer chip, as they seem
to occur in a range of sizes, and hence do not produce a discrete,
readily visible band. Finally, DNA fragments smaller than 100 bp
are not captured by standard NGS library preparation protocols.
Single-stranded NGS library preparation methods, which do not
include a lower limit on the size of fragments sequenced, have
revealed significantly shorter cirDNA fragments than traditional
NGS methods. The periodicity of 10 bp peaks remained down
to 50 bp.10,12 Hence, PCR assays should consider the fragmented
nature of cirDNA, and the population of cirDNA assayed will
depend on the size of the target PCR products.

Nucleosomal Positioning and cirDNA Fragmentation

Patterns
Nucleosomes are not evenly distributed along chromosomes. Posi-
tioning of nucleosomes is important for gene transcription, with
active transcription start sites and upstream regions being without
nucleosomes in order to provide access for transcription factors,
and nucleosomes immediately adjacent to the transcription start
site being phased in a regular pattern.13–15 Nucleosome positioning
needs to be considered in the design of PCR assays for cirDNA.
Assays spanning the transcription start sites of active genes or the
region between phased nucleosomes may not yield a product, as
intact target DNA may be depleted from the cirDNA pool due to
inter-nucleosomal fragmentation.
The number of nucleosomes is increased in exon regions
compared to intron regions and especially at exon–intron and

b4444_Ch-05.indd 115 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

116 Cell-Free Circulating DNA: Purification and Analysis Techniques

intron–exon boundaries.16–18 Interestingly, the 146 bp length of

DNA wrapped around a nucleosome is approximately equiv-
alent to the length of an average exon.19 Longer exons contain
a higher number of nucleosomes phased within the exon.17 The
fragmentation of cirDNA is nonrandom and reflects the nucle-
osomal pattern. DNA within the nucleosomal core is protected
from nucleases by the histones, whereas the linker regions are
more exposed. In regions with fewer nucleosomes, DNA is more
fragmented.20,21 DNA cleavage is sequence-dependent with 10
positions on either side of the DNA cleavage site showing a con-
sistent pattern of preference for specific nucleotides. In particular,
cytosine is over-represented at the cleavage site and at positions 1
and −2.20 The copy number of specific fragments of cirDNA may
consequently depend on the nucleosomal positioning and specific
sequence at a given DNA locus, which has implications for opti-
mal assay design.

cirDNA Methylation
Methylation patterns are correlated with nucleosome positioning.
DNA associated with nucleosomes is more highly methylated than
flanking DNA and a 10-base periodicity in DNA methylation sta-
tus is found in nucleosome-bound DNA. Exon regions are more
methylated than intron regions, which is consistent with the higher
number of nucleosomes in exons.17

Mitochondrial cirDNA
Mitochondrial DNA is also present in the circulation. The size pro-
file of mitochondrial DNA is different from nuclear DNA, since
mitochondrial DNA is not associated with histones and hence not
degraded in the same way as nuclear DNA. According to a study

b4444_Ch-05.indd 116 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 117

by Jiang et al.8 the amount of mitochondrial DNA is only 0.00045%

of the total DNA in circulation even though several copies of the
mitochondrial genome are present in cells. Mitochondrial DNA in
circulation is very short (<100 bp) with a peak at 42 bp,22 indicating
that it is not protected from nucleases in the same way as nuclear
DNA, and not degraded in a systematic way during apoptosis.
By using a single-stranded library preparation method, Burnham
et al. showed that a large proportion of the mitochondrial DNA
is undetected by widely used methods. In this study, circulating
mitochondrial DNA was found at a level of 0.002% of the total
cirDNA.10 By optimizing purification and NGS specifically for
mitochondrial cirDNA, Zhang et al. found 0.14% mitochondrial
DNA in circulation.22

Circulating Cell-Free DNA in Cancer

Early studies on the clinical utility of cirDNA focused on quantita-
tive differences between healthy individuals and cancer patients.
To obtain reliable and comparable results, it is important to under-
stand the differences in cirDNA between these groups. In can-
cer patients, the level of cirDNA is often elevated compared to
healthy individuals, but this is not always observed in patients with
early-stage cancer, and hence the usage of quantitative differences
to identify cancer patients is limited to patients with advanced
disease. To expand the utility of liquid biopsies, qualitative dif-
ferences such as tumor-specific somatic mutations, copy number
variations, and methylation patterns can be analyzed. This invari-
ably complicates the analyses. No universal tumor-specific marker
has been identified and a panel of analyses is needed to cover var-
ious patient groups. Depending on the tissue, tumors display vari-
ous degrees of molecular similarity. In some cancers, tumors harbor

b4444_Ch-05.indd 117 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

118 Cell-Free Circulating DNA: Purification and Analysis Techniques

somatic mutations in hotspot regions and tumor-specific markers

for many patients can be covered by only a few analyses. In other
cancers, hardly any similarities are seen in mutation profiles and
customized analyses would have to be developed for individual
patients. Methylation analyses are based on the fact that various
cancer-related genes are methylated in the tumor but not in nor-
mal tissue. These genes are specific for certain types of cancer and
hence one analysis for all patients with the same type (or sub-type)
of cancer can in theory be used.23
The level of cirDNA in healthy individuals is approximately
1000–10,000 genome equivalents per mL plasma (3.3–33 ng/mL).24,25
In patients with localized disease, levels are often higher and in
metastatic disease levels are significantly elevated in many can-
cers.26 The fraction of tumor-derived DNA in circulation varies
greatly between patients and can reach extremely high levels in
metastatic disease (up to >90% DNA from tumor).27 In localized
disease, levels are lower with undetectable levels in 40%–50% of
patients.27
Circulating tumor DNA is being investigated for use in diag-
nosis of cancer, monitoring of treatment response, detection of
minimal residual disease, and prediction of recurrence and treat-
ment resistance.28 For all of these purposes, an extremely high
level of sensitivity is needed in order for the assay to be clinically
relevant. The current level of detection for the most sensitive
methods is ~0.01% mutated DNA and is set by methodological
limitations. This is equal to 1 mutated molecule in a background
of 10,000 wild-type molecules. To reliably quantify this level, at
least 3 mutated molecules in a background of 30,000 wild-type
molecules must be analyzed. In many cases, however, the amount
of DNA available from plasma samples is less than that and, there-
fore, does not allow for analyses of sufficient sensitivity.

b4444_Ch-05.indd 118 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 119

The literature regarding the size profile of cirDNA in cancer

patients has been inconsistent, although recent studies have been
more in agreement. Several studies have reported that cirDNA
fragments are longer in cancer patients,29–32 while other studies
have reported that cirDNA fragments are shorter.24,33 With the
advancement of NGS techniques, where the exact lengths of indi-
vidual fragments can be determined, it is now clear that the major-
ity of the cirDNA in both healthy individuals and cancer patients
is 166-bp long, but that in cancer patients, a significant proportion
of the tumor-derived cirDNA is shorter. It has been demonstrated
that the size profile of cirDNA is shifted toward shorter fragment
lengths with increasing fraction of tumor DNA in circulation.8
Longer fragments are also found in the circulation but mainly in
patients with lower fractions of tumor DNA.8 These longer frag-
ments have been speculated to be released during necrosis.7 Long
fragments have been observed by gel electrophoresis but have not
been studied by NGS, as the most widely used sequencing meth-
ods are not designed for sequencing very long fragments. Hence,
this fraction of cirDNA is awaiting investigation in more detail by
alternative methods.
cirDNA in cancer patients appears to be released into circu-
lation mainly by apoptosis, as in healthy individuals, but necrosis
and active release may also play important roles.34 The fraction
of cirDNA released by different mechanisms may vary between
patients depending on tumor characteristics (stage, location, size,
tumor type, etc.) perhaps leading to different size profiles of the
cirDNA in patients with different characteristics.
The integrity of cirDNA has been proposed as a marker of
disease. It has been shown that fragments generally are shorter
in patients with metastatic disease than patients with early-stage
disease35 and that variation in integrity index reflects disease

b4444_Ch-05.indd 119 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

120 Cell-Free Circulating DNA: Purification and Analysis Techniques

dynamics.30,36 However, the integrity of cirDNA may vary in

different types of cancer, due to differences in the biology of DNA
release from various tissues.
Circulating mitochondrial DNA has been suggested as a
marker for diagnosis and prognosis in cancer. Mitochondrial DNA
in cancer patients is found at a higher level, and the size profile in
cancer patients shows that the DNA is even shorter than in healthy
subjects.8 It has been found that elevated levels of mitochondrial
cirDNA in cancer patients are associated with a poor prognosis.37–39
As described earlier, nucleosomes are involved in determin-
ing gene expression and hence are positioned differently in differ-
ent cell types/tissues, reflecting different patterns of gene activity.
Because fragmentation occurs more frequently in regions with-
out nucleosomes (i.e., in regions where gene expression is active),
this opens up the possibility of determining the tissue of origin
of cirDNA. The main portion of cirDNA is of hematopoietic and
myeloid origin, but it has been shown that DNA from other tissues
can be identified in circulation, corresponding to cancer tissue in
a specific individual.12 This could potentially be used to classify
cancers of unknown origin, but it also offers potential for more
exact and specific quantification of tumor DNA in patients with
known cancers. If PCR assays are positioned at regions that are
less fragmented in tumor-specific DNA than in normal DNA, a
tumor-specific signal change in cirDNA concentration would be
measured. This, however, requires very detailed knowledge of the
fragmentation patterns of specific tissues and of nucleosome pat-
terns in healthy and diseased individuals.
A similar concept has been demonstrated with DNA meth-
ylation, which also differs between tissues. DNA methylation is
detectable in cirDNA, making it possible to identify its origin by
methylation patterns.40,41

b4444_Ch-05.indd 120 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 121

PCR-based Methods for Quantifying Circulating

Cell-Free DNA
Two PCR-based methods are generally used for cirDNA analyses —
real-time or quantitative PCR (qPCR) and digital PCR (dPCR).
Several platforms exist for both methods. In qPCR, two types
of chemistry are commonly used to detect the PCR product:
probe-based chemistries or dye-based chemistries. Probe-based
methods utilize a fluorescently labelled probe that emits light of
a certain wavelength upon cleavage during PCR amplification.
Hence, an increase in fluorescence is only seen if a specific tar-
get, determined by primer sequences, is amplified and the probe
matches a specific sequence between the primers. The dye-based
methods rely on the specificity of the primers, and increases in
fluorescence are observed when double-stranded DNA is pro-
duced during amplification and the dye intercalates the two
strands. One disadvantage of the dye-based methods is that
once an incorrectly primed amplification cycle has taken place,
the incorrect product incorporates a perfectly matched primer
sequence and forms an efficient template for further amplifica-
tion rounds. Probe-based methods provide an additional level
of specificity since, in addition to the primers, the probe also
requires a complementary sequence to be present within the
product in order for a fluorescence signal to be generated.
dPCR is now widely used for cirDNA analyses. The principle
of dPCR is that samples are partitioned into a large number of
reactions (104 to 107), only some of which include a template mol-
ecule. Each partition contains mastermix, primers, and probes or
dye exactly as in qPCR reactions. If a template molecule is present
in a partition, PCR amplification takes place and fluorescence is
detected as an endpoint measurement, with individual partitions

b4444_Ch-05.indd 121 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

122 Cell-Free Circulating DNA: Purification and Analysis Techniques

designated as positive or negative. Based on the number of positive

and negative partitions, the number of template molecules in the
original sample can be determined with high accuracy. The advan-
tage of dPCR over qPCR lies mainly in the sensitivity for detecting
rare targets (e.g., mutations or methylations) in a high background
(e.g., wild-type or unmethylated DNA). dPCR does not require
the use of standard curves or reference materials for quantification
purposes, as it provides an absolute quantification of the template.
It is, however, a more expensive method that is lower through-
put and requires more hands-on time, so for quantification of total
cirDNA, qPCR would be the preferred method for analyses of
samples, with dPCR used for quantifying standards for standard
curves.
For dPCR, the common approach to analyzing a point muta-
tion is a duplex reaction with two primers and two probes, with each
probe labelled with a different fluorophore. The two probes detect
the wild type and mutated sequence, respectively. Because of the
partitioning of the template molecules, the background is greatly
reduced and assay specificity becomes less vital. The wild-type
probe blocks unspecific detection of wild-type DNA by the muta-
tion-specific probe and quantifies the number of wild-type alleles.
For analyzing point mutations or methylated sites by qPCR,
specific optimizations should be employed to increase the specific-
ity. One common approach is Amplification Refractory Mutation
System qPCR (ARMS-qPCR), in which a mismatch is incorporated
into one of the primer sequences next to the mutated position
to improve discrimination.42 This approach exploits the fact that
while a single mismatched residue within a primer can often be
tolerated, two adjacent mismatched residues render the reaction
too inefficient to proceed. Wild-type blocking oligos that hybrid-
ize to the wild-type sequence and prevent DNA extension can

b4444_Ch-05.indd 122 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 123

also be added to the mutation-detecting reaction to minimize

nonspecific amplification of wild-type sequences, such as the Int-
plex method.43
Co-Amplification at Lower Denaturation Temperature PCR
(COLD-PCR) is a method to enrich for templates containing a
rare mutated sequence. COLD-PCR relies on a two-step dena-
turation protocol. The first denaturation at 94°C is followed by
DNA reannealing at around 70°C. When the mutant sequence
is rare and wild-type sequence is present at a high excess, the
majority of mutation-containing DNA fragments will form het-
eroduplexes with the wild-type sequence. A second denaturation
step is then carried out at a temperature selected to denature
the heteroduplexes but not the perfectly matched homodu-
plexes containing the wild-type sequence only. Single-nucleotide
mismatches alter the melting temperature of double-stranded
DNA slightly (0.2°C–1.5°C), and two fragments differing by
only single-nucleotide mismatch will have different amplifica-
tion efficiencies at the critical temperature. With this approach,
amplification of the mutated sequence is favored and more eas-
ily detected in a downstream application as pyrosequencing or
NGS.44 It is important to note that COLD-PCR improves the
amplification of the minor DNA sequence, so if a sample con-
tained an excess of mutant over wild-type DNA, it is the amplifi-
cation of the wild-type DNA that would be enhanced.
For methylated DNA analysis, methylation is measured by
bisulfite conversion of DNA followed by quantification of meth-
ylated and unmethylated CpG residues by methylation-specific
qPCR (e.g., the MethyLight method).45 Developments of the
method specifically for cirDNA include analyses by dPCR in order
to detect smaller fractions of tumor-related DNA that exhibit a
different methylation pattern.46,47 One of the technical difficulties

b4444_Ch-05.indd 123 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

124 Cell-Free Circulating DNA: Purification and Analysis Techniques

related to analyzing methylated cirDNA is that the bisulphite

conversion process further fragments the DNA strands, leading to
some loss of amplifiable template.48

Target Region
There are a number of considerations when selecting a target
region for PCR analysis of cirDNA. If quantitation of total can-
cer cirDNA is required, it is important to choose a region that is
known not to be amplified or deleted in the cancer of interest. It is
advisable to test several different targets and compare levels.
Some genes are also present in the genome as pseudogenes
and should be used with caution. PCR reactions targeting
pseudogenes with multiple copies in the genome may give dif-
ferent results depending on whether dPCR or qPCR with a
standard curve is used. This is because dPCR will directly quan-
titate the number of pseudogene copies, which will then need to
be compared against the number of known pseudogenes in the
genome in order to calculate how many genome equivalents are
present in the sample. qPCR against a standard curve will directly
quantitate the amount of DNA present since the genomic DNA used
in the standard curve contains the same number of pseudogenes
as the unknown sample. Pseudogenes only contain exons, so if
one of the primers is positioned in an intron region of the gene,
this should not cause problems.
When positioning primers, sequences with known poly-
morphisms should be avoided, in order to maintain consistent
PCR efficiency in samples from different individuals. Primer
dimers and hairpins should be eliminated by investigating
primers in dedicated software, and a BLAST search should be
performed to ensure specific primer binding. The secondary

b4444_Ch-05.indd 124 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 125

structure of the amplicon should also be checked to ensure that

the primer-binding sites are without secondary structures at the
annealing temperature.
When quantifying total cirDNA, there are often several
options for positioning the assay along a gene or DNA region.
Because the fragmentation of DNA is nonrandom, this may have
implications for the quantification of cirDNA. Nucleosome distri-
bution, and hence DNA fragmentation pattern, should be a factor
in guiding positioning, if these are known or can be predicted.
Assays that overlap the positions where DNA is frequently cleaved
(e.g., transcription start sites) may result in lower quantifications
than assays that are completely within the fragments. Gene activ-
ity, and therefore DNA fragmentation pattern, varies between
tissues and the same assay could potentially perform differently
in patients with different cancers. The assay could also perform
differently in healthy individuals making it difficult to compare
DNA quantifications to “normal” levels. DNA fragmentation may
differ between patients depending on type and stage of disease.
Patients with a low fraction of circulating tumor DNA have longer
fragments in circulation while patients with higher fractions have
shorter fragments than healthy individuals, indicating that frag-
mentation pattern develops with the disease.8 This could poten-
tially influence quantifications in patients over time and produce
measured drops in the level of cirDNA that are not real. Again, it
would be advisable to measure more than one target.
When analyzing tumor-specific DNA, positioning is deter-
mined by the genetic abnormality under investigation. When
quantifying point mutations, the assay must cover the specific base
position; when analyzing methylation, CpG residues determine
the positioning. Nucleosome distribution is particularly relevant
to the analysis of CpG sites in promoter-associated CpG islands.

b4444_Ch-05.indd 125 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

126 Cell-Free Circulating DNA: Purification and Analysis Techniques

These regions are, by their nature, adjacent to or overlapping with

transcription start sites, and as such, care must be taken to avoid
positioning the PCR assay over the transcription start site of active
genes where nucleosomes are absent.
Analyzing copy number variations is often done by comparing
quantitative values of two PCR assays from unaffected and ampli-
fied genomic regions. The two assays have to be carefully charac-
terized and compared experimentally in order to avoid artefacts.
Positioning of assays could in this case potentially influence results
if the assays are not in similar positions with regard to fragmenta-
tion patterns. The same is the case with integrity analyses where
two assays of different lengths are compared.

Purification
Because tumor-specific cirDNA is generally shorter than DNA
derived from normal cells, it is important to use isolation and
purification techniques that include short fragments. It has been
shown that spike-in fragments of different lengths are not purified
equally well with different purification kits49 so the appropriate
DNA capture kit needs to be selected. Kits dedicated to purifica-
tion of cirDNA are often optimized for isolating short fragments
to minimize contaminating genomic DNA, but longer fragments
released by necrosis in some patients are also lost.
It is possible to analyze cirDNA directly from plasma without
purification.50 The amount of DNA and the integrity of the DNA
change after purification and depend on the purification method.
This demonstrates that it is also important to consider the puri-
fication method when comparing results from different studies.
Underhill et al.51 demonstrated that by size-selecting for short frag-
ments by polyacrylamide gel electrophoresis, increases in fraction

b4444_Ch-05.indd 126 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 127

of mutated DNA could be observed. Hence, a better sensitivity for

detecting genetic changes could be obtained by excluding larger
fragments from the analyses.

Assay Length
According to the literature, shorter PCR assays lead to higher
amounts of detectable DNA.24,33,52 When designing assays for
PCR-based analyses of cirDNA, it is evident that amplicons
should be less than 166 bp long since the majority of fragments
in circulation are of that size or smaller, and when optimizing
analyses to be tumor specific, very short amplicons are neces-
sary. Decreasing the size of the PCR amplicon increases the
likelihood that it will fall entirely within a cirDNA fragment and
not span a break point.
Quantification by qPCR is dependent on amplification effi-
ciency, which generally increases when amplicons are shorter. This
is in part because amplicon length influences how susceptible the
reaction is to PCR inhibitors, with longer amplicons being more
prone to inhibition.53 When comparing quantifications by different
assays, it is necessary to take this into account. In a direct compar-
ison of quantifications with long (120 bp) and short (85 bp) assays,
the short assays amplified between 2.5 and 5 times more efficiently
than longer assays on the same positive control material.54
Because tumor-derived cirDNA is shorter than normal
cirDNA, the benefit of using short assays is highest with tumor
DNA.24,33 Shortening qPCR assays specific for KRAS-mutated
DNA from 120 to 85 bp increased the amount of DNA detected in
cancer patient samples 3 times.54 This was after taking into account
that short assays performed more efficiently, so the increase is
directly associated with the sample material being shorter.

b4444_Ch-05.indd 127 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

128 Cell-Free Circulating DNA: Purification and Analysis Techniques

The difference in amplification efficiency between assays

of different lengths is of course especially important when using
qPCR for calculating integrity indexes of cirDNA. Assays to ana-
lyze integrity indices are by definition of different lengths, so to
determine the relative quantities of the long and short fragments,
assays have to be carefully compared and variations in efficiencies
taken into account. Otherwise, the result of the integrity index will
depend on the total level of DNA in the sample. For example,
cancer patients often have increased amounts of cirDNA, and with
a longer assay being less efficient, the integrity of the DNA would
appear to be increased compared to healthy controls.
If assays cannot be designed to be of identical length, standard
curves can be generated to identify the amplification efficiency
of each assay and quantifications can be adjusted accordingly. It
would, however, be important to calculate the efficiency on stan-
dard curve material that is comparable to sample material. This
means using fragmented DNA for the standard curve, but since
DNA from plasma is not fragmented in a random manner, it is dif-
ficult to generate a perfectly comparable standard curve material.
Ideally, DNA purified from plasma from a source with high levels
of cirDNA should be used in the standard curve, but this also does
not take into account individual differences in cirDNA fragmen-
tation patterns.
Amplicons of 50–150 bp are recommended for efficient qPCR
assays. For most targets, this is long enough to design primers and
probes of adequate length. For probe-based qPCR methods, the
minimal length of an amplicon is determined by the length of two
primers and the probe. One primer and probe on one strand can
be next to each other and should ideally be as close as possible
without overlapping to ensure rapid cleavage by the polymerase.
For dye-based assays, amplicons will often have to be slightly

b4444_Ch-05.indd 128 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 129

longer than probe-based assays in order to differentiate the PCR

product from primer–dimers in the melt curve analysis. In prac-
tice, this means designing amplicons of 90–150 bp. The quantifi-
cation of a template using intercalating dye depends on the length
of the amplicon and the AT-content of the amplicon sequence.55
The copy number is overestimated with long amplicons so caution
should be taken when comparing different assays or comparing
dye-based with probe-based methods.

Primer/Probe Design
Primers and probes have to be of a certain length to work opti-
mally. The specificity of a primer increases with length. To avoid
unspecific binding, which can significantly lower the efficiency of
an assay, primers generally should be 15–25 bases. The GC con-
tent should be around 50%, and stretches of identical nucleotides
should be avoided. The melting temperature (Tm) of the prim-
ers should ideally be 50°C–65°C and should not differ more than
2°C–5°C between the primers. If the Tm between primers differs
more, there may not be an annealing temperature at which both
primers will work optimally. The Tm is primarily affected by the
primer length and GC content but also primer and salt concentra-
tion in the reaction.
The 5’ position of the probe should not be a G, as this quenches
the fluorescence, even after probe hydrolysis, and the GC content
should be around 50%. The Tm of the probe should be 5°C–10°C
higher than that of the primers. If the melting temperature of the
probe is close to the Tm of the primers, the percentage of probe
bound to the target will be low when primers anneal and exten-
sion begins. The primers would start to amplify a product, but since
many targets would be without probe, the correct amount of target

b4444_Ch-05.indd 129 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

130 Cell-Free Circulating DNA: Purification and Analysis Techniques

would not be measured. The Tm of a probe can be increased by

using modifications such as minor groove binders (MGB), modified
nucleic acids (e.g., peptide nucleic acids (PNA) or locked nucleic
acids (LNA)). When designing assays for mutation detection, the
mismatch should be positioned in the central third or toward the
3’end of the probe but not in the last two nucleotides.

Optimization of Short Amplicon PCR

Short amplicons are more likely to denature during the melting
step of the PCR reaction, and primers and probes are more likely
to bind their target. It may be possible and beneficial to shorten
the length of the denaturation step after the initial cycles. Stan-
dard Taq DNA polymerase has a half-life of approximately 40 min
at 95°C and hence its activity decreases slightly with every PCR
cycle. Depurination events may also occur at elevated tempera-
tures leading to mutations in the PCR product. This is, however,
mainly a problem with long PCR amplicons.
It has also been demonstrated that lowering the denaturation
temperature after a few initial cycles improves the yield of short
PCR products when using Taq polymerase. One study showed
that by lowering the temperature from 94°C to 87°C after 5–10
cycles, the yield of a 110-bp product could be increased 4–6-fold.56
Below 90°C the half-life of Taq polymerase is significantly lon-
ger. After a few cycles, the predominant template for amplification
is PCR amplicons, which denature much more readily than long
genomic DNA, and consequently the denaturation temperature
can be decreased. The extension step can in theory also be short,
as the length of DNA to be amplified is short, and polymerases can
extend DNA at a fast rate.

b4444_Ch-05.indd 130 08-Jul-22 5:59:22 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 131

When quantifying cirDNA, duplicate or triplicate PCR

reactions should be performed and appropriate controls should
be included in each analysis. For tumor cirDNA analyses, this
includes two negative controls (no template control and wild-type
only DNA template) and a positive control consisting of DNA con-
taining the tumor-specific marker diluted in wild-type DNA.
Standard curves should be generated for qPCR analyses to
determine amplification efficiencies. It is difficult to generate
standard curves that are completely comparable to real samples,
especially for analyses for circulating tumor DNA. Samples from a
patient with high levels of cirDNA may be used, but these may be
extremely difficult to obtain — especially in the quantities needed
to analyze standard curves a number of times and would only
reflect fragmentation pattern in one patient. As an alternative,
short synthetic fragments could be used. If possible, the synthetic
fragments should be spiked into a plasma matrix but, if this isn’t
possible, they can be spiked into fragmented genomic DNA.
Primer quality and batch variation should be monitored over
time by plotting Cq values or quantitative measurements for con-
trol samples. It is important to use high-quality primers and probes
and store them correctly, as this can dramatically influence ampli-
fication efficiency and lead to very different Cq values. In dPCR,
variation in primer/probe batches is less of a challenge because the
quantification is performed as an end-point measurement. Fluores-
cence levels could be different for different batches of primers and
probes, but the overall quantification would generally be identical.
Other optimizations can be investigated to improve discrim-
ination between tumor and wild-type DNA. Lower dNTP levels,
higher annealing temperatures, lower primer, MgCl2, and enzyme
concentrations all increase the stringency of the amplification and

b4444_Ch-05.indd 131 08-Jul-22 5:59:22 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

132 Cell-Free Circulating DNA: Purification and Analysis Techniques

favor amplification of the sequence that correctly matches the

primers.

Conclusions
Many factors influence analyses of cirDNA by PCR-based methods.
The fragmented nature of the DNA is particularly important to take
into account when designing and validating assays. The positioning
and length of amplicon can greatly influence the analysis and care
must be taken when comparing results from different assays.
The current knowledge about biological properties of cirDNA
is difficult to implement in designing PCR assays for cirDNA anal-
yses, but in future, more applicable knowledge about fragmenta-
tion patterns and release mechanisms with regard to disease type
and severity may become available.
Many resources are put into finding clinical applications for
cirDNA analyses. Research into the origin and biology of cirDNA
is equally important and necessary in order to develop the most
relevant and clinically useful analyses with regard to specific
patient groups.

References
1. Lui YY, Chik KW, Chiu RW, Ho CY, et al. (2002) Predominant hemato-
poietic origin of cell-free DNA in plasma and serum after sex-mismatched
bone marrow transplantation. Clin Chem 48: 421–427.
2. Peters DL, Pretorius PJ. (2011) Origin, translocation and destination of
extracellular occurring DNA — a new paradigm in genetic behaviour. Clin
Chim Acta 412: 806–811.
3. Thierry AR, El Messaoudi S, Gahan PB, et al. (2016) Origins, structures,
and functions of circulating DNA in oncology. Cancer Metastasis Rev 35:
347–376.

b4444_Ch-05.indd 132 08-Jul-22 5:59:23 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 133

4. Beck J, Urnovitz HB, Riggert J, et al. (2009) Profile of the circulating DNA
in apparently healthy individuals. Clin Chem 55: 730–738.
5. Morozkin ES, Loseva EM, Morozov IV, et al. (2012) A comparative study
of cell-free apoptotic and genomic DNA using FISH and massive parallel
sequencing. Expert Opin Biol Ther 12: S11–S17.
6. Stroun M, Lyautey J, Lederrey C, et al. (2001) Alu repeat sequences are
present in increased proportions compared to a unique gene in plasma/
serum DNA: Evidence for a preferential release from viable cells? Ann N Y
Acad Sci 945: 258–264.
7. Jahr S, Hentze H, Englisch S, et al. (2001) DNA fragments in the blood
plasma of cancer patients: Quantitations and evidence for their origin from
apoptotic and necrotic cells. Cancer Res 61: 1659–1665.
8. Jiang P, Chan CW, Chan KC, et al. (2015) Lengthening and shortening of
plasma DNA in hepatocellular carcinoma patients. Proc Natl Acad Sci U S A
112: E1317–E1325.
9. Zheng YW, Chan KC, Sun H, et al. (2012) Nonhematopoietically derived
DNA is shorter than hematopoietically derived DNA in plasma: A trans-
plantation model. Clin Chem 58: 549–558.
10. Burnham P, Kim MS, Agbor-Enoh S, et al. (2016) Single-stranded DNA
library preparation uncovers the origin and diversity of ultrashort cell-free
DNA in plasma. Sci Rep 6: 27859.
11. Mouliere F, El MS, Pang D, et al. (2014) Multi-marker analysis of circulat-
ing cell-free DNA toward personalized medicine for colorectal cancer. Mol
Oncol 8: 927–941.
12. Snyder MW, Kircher M, Hill AJ, et al. (2016) Cell-free DNA comprises
an in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164:
57–68.
13. Schones DE, Cui K, Cuddapah S, et al. (2008) Dynamic regulation of nucle-
osome positioning in the human genome. Cell 132: 887–898.
14. Struhl K, Segal E. (2013) Determinants of nucleosome positioning. Nat
Struct Mol Biol 20: 267–273.
15. Ma X, Zhu L, Wu X, et al. (2017) Cell-free DNA provides a good represen-
tation of the tumor genome despite its biased fragmentation patterns. PLoS
One 12: e0169231.

b4444_Ch-05.indd 133 08-Jul-22 5:59:23 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

134 Cell-Free Circulating DNA: Purification and Analysis Techniques

16. Andersson R, Enroth S, Rada-Iglesias A, et al. (2009) Nucleosomes are well

positioned in exons and carry characteristic histone modifications. Genome
Res 19: 1732–1741.
17. Chodavarapu RK, Feng S, Bernatavichute YV, et al. (2010) Relationship
between nucleosome positioning and DNA methylation. Nature 466:
388–392.
18. Schwartz S, Meshorer E, Ast G. (2009) Chromatin organization marks
exon-intron structure. Nat Struct Mol Biol 16: 990–995.
19. Lander ES, Linton LM, Birren B, et al. (2001) Initial sequencing and anal-
ysis of the human genome. Nature 409: 860–921.
20. Chandrananda D, Thorne NP, Bahlo M. (2015) High-resolution charac-
terization of sequence signatures due to non-random cleavage of cell-free
DNA. BMC Med Genomics 8: 29.
21. Ivanov M, Baranova A, Butler T, et al. (2015) Non-random fragmentation
patterns in circulating cell-free DNA reflect epigenetic regulation. BMC
Genomics 16: S1.
22. Zhang R, Nakahira K, Guo X, et al. (2016) Very short mitochondrial DNA
fragments and heteroplasmy in human plasma. Sci Rep 6: 36097.
23. Warton K, Mahon KL, Samimi G. (2016) Methylated circulating tumor
DNA in blood: Power in cancer prognosis and response. Endocr Relat
Cancer 23: R157–R171.
24. Diehl F, Li M, Dressman D, et al. (2005) Detection and quantification of
mutations in the plasma of patients with colorectal tumors. Proc Natl Acad
Sci USA 102: 16368–16373.
25. Spindler KL, Appelt AL, Pallisgaard N, et al. (2014) Cell-free DNA in
healthy individuals, non-cancerous disease and strong prognostic value in
colorectal cancer. Int J Cancer 135: 2984–2991.
26. Fleischhacker M, Schmidt B. (2007) Circulating Nucleic Acids (CNAs) and
cancer — a survey. Biochim Biophys Acta 1775: 181–232.
27. Bettegowda C, Sausen M, Leary RJ, et al. (2014) Detection of circulating
tumor DNA in early- and late-stage human malignancies. Sci Transl Med 6:
224ra24.
28. Siravegna G, Bardelli A. (2014) Genotyping cell-free tumor DNA in the
blood to detect residual disease and drug resistance. Genome Biol 15: 449.

b4444_Ch-05.indd 134 08-Jul-22 5:59:23 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 135

29. Hanley R, Rieger-Christ KM, Canes D, et al. (2006) DNA integrity assay:
A plasma-based screening tool for the detection of prostate cancer. Clin
Cancer Res 12: 4569–4574.
30. Jiang WW, Zahurak M, Goldenberg D, et al. (2006) Increased plasma DNA
integrity index in head and neck cancer patients. Int J Cancer 119: 2673–
2676.
31. Umetani N, Kim J, Hiramatsu S, et al. (2006) Increased integrity of free
circulating DNA in sera of patients with colorectal or periampullary cancer:
Direct quantitative PCR for ALU repeats. Clin Chem 52: 1062–1069.
32. Wang BG, Huang HY, Chen YC, et al. (2003) Increased plasma DNA integ-
rity in cancer patients. Cancer Res 63: 3966–3968.
33. Mouliere F, Robert B, Arnau PE, et al. (2011) High fragmentation charac-
terizes tumour-derived circulating DNA. PLoSOne 6: e23418.
34. Diaz LA, Jr., Bardelli A. (2014) Liquid biopsies: Genotyping circulating
tumor DNA. J Clin Oncol 32: 579–586.
35. Madhavan D, Wallwiener M, Bents K, et al. (2014) Plasma DNA integrity as
a biomarker for primary and metastatic breast cancer and potential marker
for early diagnosis. Breast Cancer Res Treat 146: 163–174.
36. Umetani N, Giuliano AE, Hiramatsu SH, et al. (2006) Prediction of breast
tumor progression by integrity of free circulating DNA in serum. J Clin
Oncol 24: 4270–4276.
37. Ellinger J, Muller SC, Wernert N, et al. (2008) Mitochondrial DNA in serum
of patients with prostate cancer: A predictor of biochemical recurrence after
prostatectomy. BJU Int 102: 628–632.
38. Mahmoud EH, Fawzy A, Ahmad OK, Ali AM. (2015) Plasma circulating
cell-free nuclear and mitochondrial DNA as potential biomarkers in the
peripheral blood of breast cancer patients. Asian Pac J Cancer Prev 16:
8299–8305.
39. Mehra N, Penning M, Maas J, et al. (2007) Circulating mitochondrial
nucleic acids have prognostic value for survival in patients with advanced
prostate cancer. Clin Cancer Res 13: 421–426.
40. Lehmann-Werman R, Neiman D, Zemmour H, et al. (2016) Identification
of tissue-specific cell death using methylation patterns of circulating DNA.
Proc Natl Acad Sci U S A 113: E1826–E1834.

b4444_Ch-05.indd 135 08-Jul-22 5:59:23 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

136 Cell-Free Circulating DNA: Purification and Analysis Techniques

41. Sun K, Jiang P, Chan KC, et al. (2015) Plasma DNA tissue mapping by
genome-wide methylation sequencing for noninvasive prenatal, cancer, and
transplantation assessments. Proc Natl Acad Sci U S A 112: E5503–E5512.
42. Newton CR, Graham A, Heptinstall LE, et al. (1989) Analysis of any point
mutation in DNA. The Amplification Refractory Mutation System (ARMS).
Nucleic Acids Res 17: 2503–2516.
43. Mouliere F, El Messaoudi S, Gongora C, et al. (2013) Circulating cell-free
DNA from colorectal cancer patients may reveal high KRAS or BRAF
mutation load. Transl Oncol 6: 319–328.
44. Tost J. (2016) The clinical potential of enhanced-ice-COLD-PCR. Expert
Rev Mol Diagn 16: 265–268.
45. Eads CA, Danenberg KD, Kawakami K, et al. (2000) MethyLight: A
high-throughput assay to measure DNA methylation. Nucleic Acids Res 28:
E32.
46. Garrigou S, Perkins G, Garlan F, et al. (2016) A study of hypermethylated
circulating tumor DNA as a universal colorectal cancer biomarker. Clin
Chem 62: 1129–1139.
47. Li M, Chen WD, Papadopoulos N, et al. (2009) Sensitive digital quantifica-
tion of DNA methylation in clinical samples. Nat Biotechnol 27: 858–863.
48. Worm Orntoft MB, Jensen SO, Hansen TB, et al. (2017) Comparative anal-
ysis of 12 different kits for bisulfite conversion of circulating cell-free DNA.
Epigenetics 12: 626–636.
49. Devonshire AS, Whale AS, Gutteridge A, et al. (2014) Towards standard-
isation of cell-free DNA measurement in plasma: Controls for extraction
efficiency, fragment size bias and quantification. Anal Bioanal Chem 406:
6499–6512.
50. Breitbach S, Tug S, Helmig S, et al. (2014) Direct quantification of cell-free,
circulating DNA from unpurified plasma. PLoS One 9: e87838.
51. Underhill HR, Kitzman JO, Hellwig S, et al. (2016) Fragment length of
circulating tumor DNA. PLoS Genet 12: e1006162.
52. Suzawa K, Yamamoto H, Ohashi K, et al. (2017) Optimal method for quan-
titative detection of plasma EGFR T790M mutation using droplet digital
PCR system. Oncol Rep 37: 3100–3106.
53. Pionzio AM, McCord BR. (2014) The effect of internal control sequence
and length on the response to PCR inhibition in real-time PCR quantita-
tion. Forensic Sci Int Genet 9: 55–60.

b4444_Ch-05.indd 136 08-Jul-22 5:59:23 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Optimal Design of PCR Assays for cirDNA 137

54. Andersen RF, Spindler KL, Brandslund I, et al. (2015) Improved sensitivity
of circulating tumor DNA measurement using short PCR amplicons. Clin
Chim Acta 439: 97–101.
55. Colborn JM, Byrd BD, Koita OA, Krogstad DJ. (2008) Estimation of copy
number using SYBR Green: Confounding by AT-rich DNA and by variation
in amplicon length. Am J Trop Med Hyg 79: 887–892.
56. Yap EP, McGee JO. (1991) Short PCR product yields improved by lower
denaturation temperatures. Nucleic Acids Res 19: 1713.

b4444_Ch-05.indd 137 08-Jul-22 5:59:23 PM

 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0006

Chapter 6
Preparation of Next-Generation
Sequencing Libraries for
Sequencing Circulating DNA
Katrin Heider, Florent Mouliere and Christopher G. Smith

Abstract

Next-generation sequencing of circulating DNA brings a set of chal-

lenges distinct to sequencing genomic DNA, stemming largely from its
low-concentration and heavily fragmented nature. This chapter con-
siders different approaches to sequencing library preparation, and how
these are tailored to circulating DNA analysis.

Challenges Associated with Circulating Tumor DNA

Analysis
Circulating cell-free DNA (cirDNA) that is found in the circula-
tion has emerged as a biomarker of choice to monitor and track
cancer in a minimally invasive manner. The portion of cirDNA
that originates from tumor cells is termed circulating tumor DNA
(ctDNA).1,2 ctDNA can capture the global representation of the
cancer that might be missed using a standard biopsy procedure.3–5

139

b4444_Ch-06.indd 139 08-Jul-22 6:01:24 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

140 Cell-Free Circulating DNA: Purification and Analysis Techniques

The minimally invasive nature of the approach also allows for lon-
gitudinal sampling, which is less harmful than using CT scans or
biopsies.1,2,6,7
Due to the nature, structure, and concentration of cirDNA in
the blood, sequencing of cirDNA requires optimized preparation
methods. Sensitivity for ctDNA detection is hampered by the low
amounts of DNA released in the bloodstream by tumor cells, as well as
by the technological and conceptual limitations.2 The main challenge
in ctDNA detection is to distinguish mutant ctDNA released by the
tumor from the generally overwhelming quantities of cirDNA that
originates from healthy cells in the body. This distinction becomes
easier with a higher concentration of ctDNA, which is related to can-
cer type and stage of disease.8–10 Bettegowda and colleagues showed
that the detectability of ctDNA with a PCR-based assay or Safe-
SeqS, a massively parallel sequencing method, was close to 100% in
the plasma of advanced bladder, colorectal, gastroesophageal, and
ovarian cancer patients.9 In these cases, the concentration of ctDNA
was high enough to allow for detection of mutant fragments. On the
other hand, for advanced glioma or for kidney cancer patients, the
average rate of detection was only around 10%, and the concentra-
tion of ctDNA detected was also minimal (less than 10 copies per 5
mL of plasma in glioblastoma patients).9 The lower levels of ctDNA
in these cancers have since been confirmed in other studies.11,12
ctDNA is thought to be released via apoptosis, necrosis,
proliferation, or active secretion.2,5,13 With an increase in tumor
volume, common for later stage disease, there is a concomitant
increase in the number of cancerous cells with the potential to
release ctDNA. It could also be hypothesized that the surface of
contact of these cells with blood vessels will be increased, favoring
the release of DNA fragments in the blood. For high-grade serous
ovarian carcinoma, it was demonstrated that the overall volume

b4444_Ch-06.indd 140 08-Jul-22 6:01:24 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 141

of the tumor is strongly correlated with the levels of ctDNA mea-

sured in plasma.14 A similar trend was later confirmed for lung
cancer patients at earlier disease stages with lower tumor burden.5
Analysis of ctDNA samples from certain cancer types or at earlier
stage of disease may require more sensitive methods of analysis.
Another challenge in the analysis of ctDNA is that it is highly
fragmented in plasma and other body fluids. The length of plasma
cirDNA fragments is organized around a mode of 166 bp, repre-
senting the length of DNA wrapped around the nucleosome and a
linker histone.15 This mode of distribution is usually complemented
by additional peaks at multiples of 166 bp, representing di- and
trinucleosomes, typical of an apoptotic ladder.16,17 As compared
to cirDNA, ctDNA has been shown to be even more fragmented,
exhibiting a mode of distribution between 133 and 145 bp (or
smaller) in length.11,18,19 Due to the cleavage of DNA by nucleases
during cell death, the composition of bases at the ends of cirDNA
fragments is also altered in cancer compared to healthy controls.20
There are multiple sequencing-based methods that can be
used to analyze cirDNA. The techniques differ in the way the
library is prepared, the size of the genome that will be analyzed,
and the mean sequencing depth per sample. In the following
sections, we will compare different methods for cirDNA library
preparation and sequencing, highlighting the advantages and dis-
advantages associated with each.

Preparing Next-Generation Sequencing Libraries from

Cell-Free DNA
Multiple strategies are currently available to prepare a DNA sam-
ple for next-generation sequencing. Both digital polymerase chain
reaction (dPCR) and library preparation methods are commonly

b4444_Ch-06.indd 141 08-Jul-22 6:01:25 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

142 Cell-Free Circulating DNA: Purification and Analysis Techniques

used in ctDNA analysis; however, this chapter will mostly focus

on the latter. Library preparation usually entails sample frag-
mentation of genomic DNA followed by end repair, appending
sequencer-specific “adapters” to fragments, and amplification.21
Due to the fragmented nature of cirDNA and the generally low
levels of ctDNA, library preparation protocols differ from those
used in analysis of genomic DNA. For this chapter, we will focus
on preparation of libraries for sequencing on Illumina platforms.
However, there are alternative sequencers currently, or soon to
become, available with their own sample preparation require-
ments/protocols.22

Quantification of Cell-Free DNA Prior to Library

Preparation
Before initiating cirDNA library preparation, it is important to
accurately determine the number of cirDNA molecules present
in a sample. The most commonly used methods in the ctDNA
field involve quantification through the use of fluorescent dyes
(using equipment like Thermo Fisher’s Qubit), PCR (quantitative
PCR — qPCR or digital PCR — dPCR), or microfluidic DNA
gel electrophoresis (using equipment like Agilent’s Bioanalyzer).23
The main differences in these methods are with respect to cost,
ease and time of handling, and reproducibility. PCR approaches
are more reproducible than fluorescent dye approaches but come
at a higher cost and more complex setup.24 Especially for sam-
ples at lower cirDNA concentration qPCR is the more appropriate
method since it is more robust and reproducible.23,24
While fluorescent dye or gel separation techniques quan-
tify the total amounts of genomic material, dPCR quantifies the

b4444_Ch-06.indd 142 08-Jul-22 6:01:25 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 143

number of DNA copies in a given sample by quantifying a spe-

cific region in a gene (usually a housekeeping gene is chosen).25
As cancer is a disease that often results in changes in copy num-
ber, this may present a problem for dPCR-based quantification, as
the total levels of cirDNA will not necessarily correspond to the
number of genome equivalents present in the plasma. As such, it
may be more appropriate to average the inferred concentration
as determined by dPCR of multiple genomic loci.26,27 Despite this
concern, dPCR generally represents a more accurate measure of
the concentration of a given sample. Therefore, it is advisable to
quantify the sample using dPCR, even if the fragmented nature of
cirDNA prevents a direct conversion of the detected number of
copies by dPCR into nanograms of DNA.

Library Preparation from DNA

In general, library preparation involves end repair to create mol-
ecules with blunt ends, which is followed by adapter ligation and
PCR amplification.28,29 This final step is generally necessary due to
the low concentrations of input DNA that are commonly available
from cirDNA.28,30 One of the main challenges when sequencing
DNA libraries is to mitigate the bias introduced by PCR amplifica-
tion.29 This challenge is particularly important in cancer genomics
when attempting to sensitively detect single-nucleotide variants
(SNVs) targetable by precision medicine.31 Another consider-
ation is the maintenance of molecular complexity during library
preparation to increase the chances of finding rare mutations with
low allelic frequencies. Current methods typically vary in the effi-
ciency with which the final output represents the molecules that
went into the reaction.29,30

b4444_Ch-06.indd 143 08-Jul-22 6:01:25 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

144 Cell-Free Circulating DNA: Purification and Analysis Techniques

Double-Stranded DNA Library Preparation

The most commonly used library preparation protocols in the
cirDNA field are aimed at sequencing double-stranded DNA.
Indeed, a multitude of kits and methods are available that follow
a similar workflow. While tissue or germline DNA library prepara-
tion kits require a fragmentation step to generate short fragments
(~200 bp) compatible with the sequencer (a potential source of
severe fragment loss),32 cirDNA samples do not have to undergo
DNA shearing due to its already fragmented nature. The cirDNA
sample first undergoes end repair, followed by phosphorylation of
the 5′ prime ends. The 3′ ends undergo an A-tailing step to allow
for ligation of the sequencer-specific adapters. Upon adapter liga-
tion, samples undergo several cycles of PCR to enrich the frag-
ments that are compatible with the sequencer. This is followed by
a final “cleanup” of the library to remove excess adapter dimers
and heteroduplexes.29,32,33 Individual DNA samples can be tagged
with uniquely barcoded adapters. These barcodes consist of a short
and unique stretch of bases and allow for multiplexed sequencing.
Molecules originating from the same sample will obtain the same
string of unique bases; therefore, each barcoded DNA fragment
in the sequencing pool can be attributed to its sample of origin.31
Competition in the market of library preparation kits has
helped to lower the prices while also improving their quality. The
required input for most kits starts at less than a nanogram of DNA,
allowing for sample preparation even when very little material is
available.30
The most commonly used double-stranded library prepa-
ration protocols tend to capture fragments greater than 100
bp in length.34 It is still unclear why shorter fragments are not
observed but there are different hypotheses. Shorter molecules

b4444_Ch-06.indd 144 08-Jul-22 6:01:25 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 145

could be lost during the extraction or library clean up steps, or

be damaged or of single-stranded nature and thereby not cap-
tured by the double-stranded protocol. This results in the loss
of fragments shorter than 100 bp, which may be enriched for
ctDNA.18,19 Single-stranded library preparation methods have
emerged as an alternative library preparation approach that may
overcome this problem.34,35

Single-Stranded DNA Library Preparation

Developed for the analysis of ancient, degraded DNA, current
single-stranded DNA library preparation methods allow for the
capture of single-stranded, double-stranded, and damaged DNA
fragments.35 Therefore, as compared to double-stranded library
preparation protocols that capture only double-stranded DNA,
single-stranded DNA library preparation allows for the analysis of
a broader range of DNA fragments. Indeed, in a study that used
a single-stranded protocol on DNA from a cohort of transplant
patients, a greater portion of 50–100 bp fragments was recovered
using this approach as compared to a double-stranded equivalent.34
Also, when applied to plasma samples from healthy individuals, a
greater representation of shorter DNA fragments was recovered as
compared to the standard double-stranded library preparation.36
An updated version of the Gansauge and Meyer ancient DNA pro-
tocol was published in 2017, showing greater recovery of DNA
and an improved turn-around time.37 Previous research suggested
an enrichment of ctDNA in shorter fragments.18,19,38,39 Therefore,
the use of single-stranded DNA library preparation protocols that
improve the coverage of shorter fragments might be expected to
improve the recovery, and detection, of ctDNA. However, initial
reports seem to indicate that there is no improvement in ctDNA

b4444_Ch-06.indd 145 08-Jul-22 6:01:25 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

146 Cell-Free Circulating DNA: Purification and Analysis Techniques

detection using a single-stranded DNA library,40 raising questions

as to the origin of the additional DNA fragments that are captured
by this protocol.
Other single-stranded protocols include the commercially
available SMART ChIP-Seq kit that is based on template switch-
ing. This protocol was initially developed for RNA approaches.41,42
Compared to the ancient DNA protocols described previously, the
SMART ChIP–Seq protocol involves a simpler, and quicker, work-
flow. However, a recent publication indicates that this method
preferentially captures fragments with poly dA/dT tracts, which
could induce a bias in cirDNA recovery.43 Similar comparisons
should be performed on the other library preparation methods to
obtain a clearer picture of the sequencing biases specific to each
method.

Enrichment of ctDNA
While single-stranded library preparation protocols aid in cap-
turing shorter DNA fragments, which might be biased toward
a higher proportion of ctDNA, they do not actually enrich for
ctDNA. One means by which one could potentially do so is by
carrying out size selection, either through in vitro or in silico
means. This is based on the observation that ctDNA fragments
are predominantly shorter than 150 bp, while cirDNA of noncan-
cerous origin is predominantly 166 bp in length.11,15,18,19,39,44,45 Such
approaches have indeed shown an improvement in ctDNA detec-
tion.11,19,40,45 In addition to size selection, methods leveraging the
thermodynamic properties of primers for sequencing also enable
enrichment in tumor signal.46
While selecting for certain fragments may lead to more con-
fident mutation detection, the general level of background noise

b4444_Ch-06.indd 146 08-Jul-22 6:01:25 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 147

remains a challenge. Indeed, with an error rate of just below 1%,

current sequencing yields data with high background noise.32,47
Additionally, given the low concentration of cirDNA, PCR ampli-
fication is usually warranted during library preparation. This fur-
ther leads to the introduction of PCR errors that may reach allelic
fractions as high or higher than true mutations.48,49 Methods have
evolved to aid in the downstream analysis and differentiation of
true mutations from noise.

Sequencing Error Correction Using Unique Molecular

Identifiers
The concept of uniquely tagging individual molecules was first
described in 2003 in the context of determining the unique num-
ber of mRNA molecules in a given sample.50 In 2011, unique
molecular identifiers (UMIs) were first used to reduce PCR back-
ground noise and obtain a cleaner sequencing result.51,52 UMIs are
composed of a string of entirely random nucleotides that tag each
molecule of the initial sample in a unique way.51 After sequencing,
the UMIs are used to group fragments of the same origin into fam-
ilies, which are then combined into a single consensus sequence.53
As shown in Fig. 1, this consensus sequence aids in identifying
PCR and sequencing errors, which should only be present in some,
but not all, members of a given family. Conversely, true mutations
should be represented in most, if not all, of the members of that
family. Thus, the use of UMIs allows for error suppression and
more stringent and confident calling, which improves downstream
analyses.
UMIs have made an impact within the field of ctDNA
research. Examples of their application include the Safe–SeqS
approach with an error rate of only 3.5 × 10−6 mutations/bp (70-fold

b4444_Ch-06.indd 147 08-Jul-22 6:01:25 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

148 Cell-Free Circulating DNA: Purification and Analysis Techniques

Fig. 1: Schematic representation of background noise reduction through

UMIs. Unique molecules are tagged with different adapters before undergoing
PCR amplification and sequencing. Utilizing read position and UMI sequence
information, PCR and sequencing errors are removed and molecules are col-
lapsed into a consensus sequence.

less than without using UMIs), which was used for a multicancer
study led by Bettegowda and colleagues.9,48 UMIs were also incor-
porated into the iDES CAPP–Seq approach used by Newman and
colleagues to study early-stage lung cancers. Using this approach,
the authors showed that, together with in-silico background “pol-
ishing,” a sensitivity of 0.0025% could be achieved.54
Despite their promise, there are some limitations when using
UMIs. For example, it is possible that PCR errors that occur early
in the amplification process may not be filtered out. Furthermore,
sequencing errors in the sequence of the UMI itself may lead to
assignment of fragments to the wrong family.47
The widespread development of high-depth sequencing for
ctDNA mutation analysis, as well as the availability of standard-
ized kits, leads to an easier implementation of the approach with-
out the necessity of prior experience.55 Tools such as CONNOR,
MAGERI, UMI-tools, and Agilent’s SureCall further aid with the
analysis of UMI data.53,56–58
Building upon the background noise suppression achieved
through the use of UMIs, Schmitt and colleagues developed a

b4444_Ch-06.indd 148 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 149

technique called duplex sequencing. They not only incorporate

UMIs into their library preparation but also retain strand direc-
tionality and origin to further reduce background noise.59 After
generating a strand-based consensus sequence, they identify which
two strands came from the same original double-stranded mole-
cule and combine those two consensus strands into a final duplex
consensus sequence. They estimate their error rate to be at 3.8 ×
10−10 or even lower, allowing for very sensitive mutation calling.
However, this degree of read combining warrants a great initial
sequencing depth, resulting in a much increased sequencing cost.59
More recently, Cohen and colleagues developed an approach (Saf-
erSeqS,60 a spin-off of the SafeSeq approach described earlier) that
applies identical barcodes to both strands of the input DNA frag-
ment before strand-specific hemi-nested PCR enrichment. Muta-
tions can be assigned to the Watson and Crick strands — those
called in both strands are considered bonafide mutations. As with
the approach developed by Schmitt and colleagues,59 SaferSeqS
achieves impressive sensitivity and specificity, detecting variants at
frequencies below 1 in 100,000.

PCR-Free Library Preparation

In 2015, Karlsson and colleagues described a library preparation
method for cirDNA and noninvasive prenatal diagnosis that does
not require a PCR step.61 Their method not only avoids amplifica-
tion bias induced through PCR but also removes any errors that
PCR would have produced. Based on the single-stranded library
preparation introduced by Gansauge and colleagues, the authors
altered the protocol to remove any PCR amplification steps.35,61
Unlike other kits that employ PCR-free library preparation and
that require more than 500 ng input material, the method pro-
posed by Karlsson and colleagues requires only 50 ng input making

b4444_Ch-06.indd 149 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

150 Cell-Free Circulating DNA: Purification and Analysis Techniques

it a feasible method for the analysis of cirDNA,61 although the

amount is still fairly large in the context of typical cirDNA yields.
Comparing their method to a normal double-stranded library
preparation protocol, as well as a UMI approach, the authors
concluded that the amplification-free library preparation process
increases coverage and decreases GC bias. Finally, it also retains
strand information, which is otherwise lost during PCR-based
library preparation.61 The authors highlight that their method is
most suitable for degraded samples and samples requiring suffi-
cient coverage. When accurate sequencing is needed, for example,
rare mutation detection in samples with low levels of ctDNA, they
would still recommend a UMI-based approach.61
Other library preparation methods also aid in reducing PCR
and sequencing errors but have limited applications in the field of
cirDNA so far. Examples are multiple displacement amplification
(MDA) and rolling circle amplification (RCA).47,62
Having highlighted the main options and new developments
available for cirDNA library preparation, ctDNA enrichment,
and error rate reduction, we will now turn toward the different
sequencing methods available for cirDNA analysis.

Tailoring Approaches for Cell-Free DNA Sequencing

Upon generation of a library, different sequencing methods are
available depending on the required analysis. One can either pro-
ceed directly with whole genome sequencing (WGS) or enrich the
sample for some part of the genome. In the following sections,
different sequencing approaches will be explained in more detail
and their advantages and disadvantages, as well as their applica-
tion to the study of plasma DNA, will be discussed. We will group
the different methods based on the size of the genome they target
and provide an overview of the main differences between them.

b4444_Ch-06.indd 150 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 151

Whole Genome Sequencing

WGS of plasma DNA provides researchers with a comprehensive
view of the entire genome of the patient,63 facilitating the analy-
sis of somatic copy number variations (CNVs), structural variants
(SVs), and SNVs. With the development of digital karyotyping,
it has become possible to analyze copy number changes of the
genome in greater detail, even with sufficient sensitivity to observe
focal events involving small genomic regions (<1M bp).64 Digital
karyotyping laid the foundation for later copy number analysis and
its implementation to plasma analysis. WGS of cancerous plasma
samples was first employed in 2012 when CNVs were identified,
even in the absence of matched tumor specimens.65 Using CNVs
to detect cancer represents a very useful tool since amplifications
and deletions are often an early event in cancer development.66
Additionally, it is possible to infer copy number status from low-
depth data, even with a sequencing depth of <1x, in turn allowing
simple and cost-effective interrogation of ctDNA.67,68 The ability
to detect CNVs at low depth has led to the use of terms such as
shallow whole genome sequencing (sWGS) or low-pass WGS.
For sWGS-based CNV analysis, the genome is apportioned
into “bins” of an equally sized length (ranging from kb to Mb). The
sequencing reads that map within each of these bins are counted
and this value is compared to the average bin read count across
the entire genome. Any bins spanning regions that are amplified
or deleted will contain greater or fewer reads than the respective
average bin count. Such a ratio is obtained for every bin across the
length of the genome, generating a CNV profile. This graphic rep-
resentation of the genome allows one to quickly visualize ampli-
fications and deletions, as well as compares the overall pattern
between different samples from the same patient, or between dif-
ferent patients. The chosen bin size is dependent on the sequencing

b4444_Ch-06.indd 151 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

152 Cell-Free Circulating DNA: Purification and Analysis Techniques

depth, where greater sequencing depth allows for a smaller bin

size and, therefore, a more detailed CNV profile.
One can convert the above CNV profiles into quantifiable
metrics, for example, using a z-score as described by Heitzer and
colleagues.69 This method determines a score for each sample
based on the deviation in read distribution from a cohort of control
plasma samples.69 A sequencing depth coverage of 0.1× to 0.2× is
sufficient to determine the z-score of a patient.69 Other tools regu-
larly used for quantifying ctDNA, based on CNV profiles, include
tMAD11 and ichorCNA.67 In addition to CNV, the cirDNA frag-
mentation patterns could also be converted as a quantitative met-
ric to assess tumor fraction.70
One can also call SNVs from WGS data; however, greater
sequencing depth is needed to ensure sufficient confidence in
mutation calling. Indeed, the detection limit of SNVs is directly
correlated with the coverage; the more a sample is sequenced, the
lower the mutation frequency detection threshold (i.e., greater
sensitivity). At a sequencing depth of 17×, Chan and colleagues
were able to identify both CNVs and SNVs in plasma samples from
four patients with hepatocellular carcinoma and one patient with
both ovarian and breast cancers.3 Comparing the copy number
data from plasma and matched tumor tissue, the authors observed
a better correlation with increasing tumor size and ctDNA frac-
tion in the plasma.3 For the patient with breast and ovarian can-
cers, CNV analysis suggested that the copy number alterations of
both the breast and ovarian lesions could be detected in plasma,
demonstrating that plasma analysis has the potential to overcome
spatial tumor heterogeneity.3
Furthermore, the concordance between SNV calls of the
plasma and tumor samples of the hepatocellular carcinoma
patients was between 15% and 94%, respectively, and was proba-

b4444_Ch-06.indd 152 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 153

bly related to tumor size and ctDNA concentration in the plasma.3

For the patient with ovarian and breast cancers, the authors
showed that mutations shared between different tumor regions
were better represented in the plasma, as reflected by their higher
allele fraction.3 Conversely, sub-clonal tumor mutations were less
well-represented and had a greater chance of being missed.3,67
While depth of sequencing can prohibit SNV calling due
to low depth, Landau and colleagues developed an approach
(MRDetect)71 that relies on integration of patient-specific somatic
SNVs (as identified by prior sequencing of matched tumour tis-
sue) across the entire breadth of sequencing for sensitive detec-
tion of ctDNA to fractions as low as parts per 10−5. The sensitivity
of MRDetect is related to the overall mutation rate in the sample
of interest and will be better for cancers with high mutation rates
such as melanoma and lung.
As outlined previously, WGS is a powerful tool to obtain
genome-wide information for a sample. However, use of this
tool becomes increasingly cost prohibitive as the required depth
of sequencing increases. Indeed, even a modest overall depth of
~30× (generally used to genotype polymorphisms) typically has a
high enough sequencing cost such as to preclude use in most clin-
ical and research settings.

Capture-Based Sequencing Approaches

While WGS of plasma DNA can provide a comprehensive over-
view of a patient’s genome, it lacks sensitivity for the detection
of low-frequency mutations. By restricting the size of the region
to be analyzed, one can reach a greater depth per sample while
maintaining or decreasing the cost of sequencing. This greater
depth results in a greater sensitivity and the ability to reliably

b4444_Ch-06.indd 153 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

154 Cell-Free Circulating DNA: Purification and Analysis Techniques

detect mutations at lower frequencies.63 After library preparation,

samples can undergo a “hybrid capture” process. In this method,
regions of interest are selected by the annealing of complementary
DNA or RNA “baits” or primer pairs. Captured regions are ampli-
fied while uncaptured regions are washed away. Depending on
the design of the capture baits, the captured regions may include
the entire exome (whole exome sequencing, WES), or some other
custom-designed region (custom capture sequencing). In the
case of the latter, current approaches typically target commonly
mutated genes, or already identified mutations.
A major limitation of capture-based approaches can be
uneven coverage, due to the fact that some regions lack complexity
and are difficult to target. As a result, they will not be captured and
enriched as well during the hybridization, making their analysis
more difficult.72 WGS does not include a capture step, resulting in
a more even and less biased coverage of the genome.

Whole Exome Sequencing

Assuming that most driver mutations will occur in the actively tran-
scribed part of the genome, targeting the exome allows research-
ers to focus their sequencing efforts on this ~1% of the genome.
Many companies offer exome capture reagents, varying in the cap-
ture approach, complexity of the protocol, and covered regions.73
WES has been demonstrated to be a useful tool for moni-
toring tumor evolution. Murtaza and colleagues applied WES to
serial plasma samples with high (>10%) levels of ctDNA. An input
of 2.3 ng DNA (690 haploid genome equivalents) and a sequenc-
ing depth of 31–160× was sufficient to call mutations from patient
plasma.4 Using longitudinally obtained samples, it was possible to
track tumor evolution throughout treatment.4

b4444_Ch-06.indd 154 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 155

Furthermore, Girotti and colleagues applied WES to plasma

samples from stage III melanoma patients. They were able to suc-
cessfully identify resistance mutations, highlighting the potential for
ctDNA as a disease monitoring tool.74 Dietz and colleagues showed
the utility of WES on low-volume serum samples from stage III
nonsmall cell lung cancer patients (NSCLC). However, due to lim-
ited sample input and sequencing depth (68×) they validated only
17% of mutations identified in matched tumor specimens.75 Never-
theless, they also identified a potential resistance mutation that was
absent in the corresponding tumor sample, which could explain why
treatment for that patient became ineffective.75 WES was applied
to plasma samples from metastatic cancer patients in a study by
Butler and colleagues.76 They showed good correlation between
tumor and plasma mutations and again identified additional plasma
mutations that were absent in the tumor sample.76 However, their
study focused on only two patients, utilizing 15 mL and 25 mL of
plasma at sequencing depths of 309× and 561×, respectively.76
While there are a growing number of studies showing the fea-
sibility of WES on plasma and serum samples, it is important to
consider the current limitations. All studies thus far have focused
on later-stage cancer patients, in whom levels of ctDNA are gener-
ally higher.9,10 Additionally, the sequencing depth in these studies
is quite high which, while allowing for greater sensitivity in muta-
tion calling, increases the overall cost of this approach. Applied to
a larger cohort or in a clinical setting, WES would generally not be
currently feasible.

Custom Capture Sequencing

As an alternative to “off the shelf” exome capture kits, custom cap-
ture methods have emerged, allowing one to focus sequencing efforts

b4444_Ch-06.indd 155 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

156 Cell-Free Circulating DNA: Purification and Analysis Techniques

on particular genomic loci of interest. This, in turn, means that

either the same number of samples can be sequenced to a greater
depth, allowing for more sensitive mutation calling, or more sam-
ples can be sequenced to the same depth, enabling application to
larger cohorts.
Cancer Personalized Profiling by Deep Sequencing (CAPP–
Seq) is one example of a custom capture sequencing approach
that relies on the presence of recurring mutations across a can-
cer cohort. Using publicly available WES, exon, and structural
rearrangement data, common mutations and rearrangements are
identified and used to design a targeted gene panel.77 Focusing
on NSCLC, a 125-kb panel was designed and applied to plasma
samples from stage I–IV patients. The small panel size allowed for
deep sequencing (~10,000× in this study) at a reasonable cost per
sample. For later-stage patients, CAPP–Seq showed 100% sensi-
tivity for ctDNA detection and proved advantageous over imag-
ing technology for disease detection, emphasizing the potential of
capture-based methods for cancer diagnosis and monitoring.77 The
same authors further developed the iDES CAPP–Seq described
previously, a technique that implements molecular barcodes and
polishing of sequencing “noise.” These developments allow for a
sensitivity of 0.0025%, sufficient for the detection of ctDNA at
very low levels.9,54
The recent TRACERx study used patient-specific ampli-
con sequencing panels for the enrichment of libraries to improve
detection and sensitivity.5 Based on WES tumor data, a median
of 18 SNVs were identified for each of 96 patients. These regions
were enriched in libraries prepared from plasma samples. While
not as sensitive as iDES CAPP–Seq, the approach used by Abbosh
and colleagues was able to detect ctDNA in 94% of stage I squa-
mous cell carcinoma NSCLC patients. On the other hand, the

b4444_Ch-06.indd 156 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 157

ctDNA detection rate in patients with adenocarcinoma NSCLC

was only 12.8% at stage I, highlighting the interaction between
ctDNA detection and cancer (sub-)type.5,9
Phallen and colleagues also used a targeted capture approach
to detect ctDNA in patients with different cancer types and stages
of disease. Their targeted error correction sequencing (TEC–Seq)
method utilizes an 81-kb gene panel containing 58 cancer-related
genes. Error suppression of their data is based on the start and end
position of the reads, as well as a small number of dual-index bar-
code adapters.10 Collapsing reads purely based on position is not
advised since the size distribution of cirDNA is too tight and differ-
ent unique molecules could have the same start and end position
by chance.10 Using their general gene panel deep sequencing with
30,000× coverage, Phallen and colleagues were able to detect ctDNA
in 62% of all stage I and II patients and 77% of all stage III and IV
patients.10 Similar to the analysis of Bettegowda and colleagues, they
also found a difference in detection between different cancer types
and an increased detectability with later-stage disease.9,10
Recently a method reliant on custom capture sequencing,
combined with custom approaches for signal enrichment and
sequencing noise suppression, has been used for sensitive detec-
tion of ctDNA.78 This approach, described by Wan and Heider and
referred to as INVAR (INtegration of VAriant Reads), targets and
integrates signal from hundreds to thousands of patient-specific
mutations (as identified by prior sequencing of matched tumour
tissue) and has been demonstrated to reliably quantify ctDNA to
levels as low as parts per hundred thousand in patients with early-
stage lung, kidney,12 breast, and other cancers.78 In the case of
the former, ctDNA was detected in 12 of 19 patients with stage
I–III NSCLC. INVAR also detected stage IV melanoma in 50 of
52 (96%) baseline and follow-up plasma samples.

b4444_Ch-06.indd 157 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

158 Cell-Free Circulating DNA: Purification and Analysis Techniques

Alternative Sequencing-Based Methods

Beyond hybrid capture-based approaches, other methods allow for
sensitive mutation detection. For example, tagged amplicon deep
sequencing (TAm–Seq) is an amplification-based approach that
combines singleplex and multiplex amplification steps.79 Regions
of interest are first enriched by amplification in a multiplexed
PCR, thereby reducing later sampling bias. The sample is then
split and a singleplex PCR of each of the regions of interest is car-
ried out.79 Amplicons are selected based on publicly available data
or cohort-specific knowledge of mutations.79 The size of the ampli-
cons is around 100 bp, accounting for the majority of cirDNA frag-
ments based on the known distribution of fragment lengths.15,44
TAm–Seq has a similar sensitivity to dPCR and reduces sampling
bias through the first multiplexed amplification step.80
Another sequencing-based method is the anchored multi-
plex PCR (AMP).81 Initially, designed for the detection of gene
rearrangements, this method uses a region-specific forward
primer with a universal reverse primer. AMP also detects SNVs,
CNVs, insertions, and deletions. Any fragment that contains the
sequence corresponding to the forward primer will be amplified
and sequenced.81 This may prove advantageous in the context of
ctDNA analysis where the great fragmentation of DNA could lead
to a given fragment not covering both primer-binding sites. Using
the AMP method, only one primer has to be complementary to a
given fragment, thereby increasing the chances of capturing the
region of interest.
A recent method, described by Douville and colleagues,82 uses
a single primer to amplify ~350,000 amplicons containing repeti-
tive elements that are evenly distributed throughout the genome.
The approach, termed RealSeqS (Repetitive Element Aneuploidy

b4444_Ch-06.indd 158 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 159

Sequencing System), was shown to detect aneuploidy in 49% of

883 nonmetastatic cancer patients. Critically, RealSeqS was found
to generate meaningful results from as little as 3 pg of input DNA.
McDonald and colleagues developed another sensitive
detection method relying on linear pre-amplification followed by
single-stranded DNA ligation and multiplex PCR termed targeted
digital sequencing (TARDIS).83 Fragment size information and
UMIs are used to suppress errors, resulting in sensitivities of 91%
and 53% at allelic fractions of 3 in 104 and 3 in 105 with a specificity
of 96%. TARDIS was applied to stage I–III breast cancer sam-
ples and detected ctDNA in all pre-treatment samples and 77% of
post-treatment samples.83
Lastly, tools like Personalized Analysis of Rearranged Ends
(PARE) utilize SVs to detect ctDNA via a PCR-based method.84
SVs are a common and patient-specific feature among human can-
cers and are identified through WGS of the tumor sample. Prim-
ers covering these unique rearrangement sites can be applied to
plasma samples in order to detect ctDNA. PARE is very sensitive
since it does not rely on identification of single nucleotide varia-
tions (prone to sequencing and PCR errors, as discussed above),
but rather on the presence or absence of a PCR product.84 The
major drawback of this method is the need for a tumor sample for
initial SV detection, which is not always available.

Summary
Here we have introduced different methods to prepare sequenc-
ing libraries from cirDNA samples. One of the most commonly
used approaches in the field is double-stranded DNA library
preparation. There are a variety of high-quality and easy-to-use
kits available, making it a robust method for the preparation of

b4444_Ch-06.indd 159 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

160 Cell-Free Circulating DNA: Purification and Analysis Techniques

cirDNA samples for sequencing. However, double-stranded

library preparations will capture mostly nondegraded fragments
longer than 100 bp, thereby not representing all cirDNA popu-
lations. Single-stranded DNA library preparation methods have
partly addressed this problem by allowing the capture of both
shorter and degraded DNA fragments, as well as single- and
double-stranded DNAs. However, some of these methods are
complex and time consuming while others have shown a bias in
fragment incorporation. Both double-stranded and single-stranded
DNA library preparation methods contain a PCR step to amplify
the cirDNA. Indexing with unique molecular barcodes has been
proposed to counteract potential PCR and sequencing biases that
arise during the standard preparation process. While indexing will
remove a great number of PCR and sequencing errors, they might
still fail to correct for early PCR errors or lose data due to PCR
and sequencing errors in the barcodes themselves. Recently, PCR-
free amplification methods with lower input requirements have
been developed, notably for single-cell genomics, making it a fea-
sible alternative to cirDNA sample preparation. The removal of
the PCR step results in less-biased sample preparation and elimi-
nates mutations that would otherwise have been induced through
PCR errors. Unfortunately, this method requires an input of
50 ng, which is not easily achievable in all cancer settings. We have
highlighted the most important advantages and disadvantages of
the different library preparation methods in Table 1.
Moreover, we have presented in this chapter various methods
for sequencing cirDNA from plasma samples. Figure 2 provides
a schematic overview of the preparation process involved in the
described methods.
The main difference between the methods is the size of the
genome that will be sequenced and the resulting sensitivity, spec-
ificity, and sequencing cost per sample. As shown in Fig. 3, the

b4444_Ch-06.indd 160 08-Jul-22 6:01:26 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 161

Table 1. Comparison of Library Preparation Methods

Method PCR? Main advantage Main disadvantage
Double-stranded Yes Quick, optimized Loss of shorter and degraded
library protocols and fragments
preparation4,48 low input
Single- Yes Low input, Bias in fragment recovery,
strandedlibrary capturing short more complex protocols
preparation37,42 and degraded
fragments
Unique molecular Yes Reduced noise by Early PCR errors are missed;
identifiers48,54 read collapsing sequencing error in UMI
into consensus sequence incorrectly assigns
sequences family members. Greater
sequencing depth is required
to generate a family of
sufficient size for consensus
calling
PCR-freelibrary No Reduces GC Greater amount of input
preparation61 bias, better material needed
coverage, no
PCR errors

methods also differ in the type of somatic event they can detect.
While WGS provides the broadest and least biased overview of
a given sample, it is also the most expensive if an appreciable
depth of sequencing is required. Using capture-based methods,
one decreases the window of the genome that is being targeted,
thereby reducing the total sequencing cost per sample. Off-the-
shelf capture approaches exist that target the whole exome or a
cancer specific gene sets. Alternatively, one can design custom
capture panels to improve the sequencing of bespoke regions of
interest. Lastly, alternative sequencing approaches such as TAm–
Seq, RealSeqS, TARDIS, and AMP that can still cover multiple
genomic regions were discussed. AMP in particular is an intrigu-
ing method since it only requires one site-specific primer, allowing

b4444_Ch-06.indd 161 08-Jul-22 6:01:26 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

162 Cell-Free Circulating DNA: Purification and Analysis Techniques

Fig. 2: Overview of preparation process of different cirDNA sequencing

methods described in this document.

(a) (b)

Fig. 3: (a) Overview of described methods for cirDNA analysis. Methods

are grouped by variants they can optimally detect (structural variants (SV), sin-
gle-nucleotide variants (SNV), and copy number variants (CNV)). (b) Methods
are compared with respect to cost, sensitivity, and area of the genome covered.

b4444_Ch-06.indd 162 08-Jul-22 6:01:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 163

the user to capture fragments of any length as long as they contain

the one primer-binding site. Semi-ligated PCR methods, such as
AMP, can also be adapted for sensitively detecting new rearrange-
ments in cirDNA.
When choosing the most appropriate sequencing method,
it is important to have a good estimation of the expected levels
of ctDNA in a given sample. Depending on the ctDNA fraction,
more or less sensitive methods will be required for analysis. One
potentially useful practice could be to implement sWGS of a
given plasma sample prior to choosing the analysis approach. This
is because sWGS is a comparatively cheap and rapid technique
that can assess likely ctDNA levels. Indeed, the mFast-SeqS and
ichorCNA approaches have been demonstrated to aid in deter-
mining ctDNA concentration in the plasma, in-turn identifying
the best analysis platform.67,68 These approaches allow the user
to group the samples into low and high tumor burden ctDNA
samples and, based on this classification, one can then choose an
approach for the analysis of the specific alterations in the sample.
The only caveat to this would be the reliance on somatic CNVs —
this approach would not be as effective in cancers known to carry
few somatic CNVs.26
Our understanding of cirDNA in the cancer setting and ways to
analyze it have improved greatly since it was first described.85,86 The
first diagnostic test using plasma ctDNA to identify NSCLC patients
with an EGFR exon19 deletion has already been FDA approved
for the clinic. This alteration is important as patients carrying this
mutation can be effectively treated with the tyrosine kinase inhibi-
tor gefitinib (Iressa).87 More recently, the COBAS® EGFR mutation
test v2 was FDA approved for using plasma samples from NSCLC
patients to detect alterations that allow treatment with the tyrosine
kinase inhibitors erlotinib (Tarceva) and osimertinib (Tagrisso).88

b4444_Ch-06.indd 163 08-Jul-22 6:01:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

164 Cell-Free Circulating DNA: Purification and Analysis Techniques

Multiple studies have also shown the utility of ctDNA as a means

of identifying recurrence sooner than conventional methods, with
important clinical implications.5,89
ctDNA has been demonstrated to represent most of the cel-
lular populations that make up a given tumor. As such, it allows
a global “snap-shot” of the disease at a given time. Conversely,
conventional biopsy techniques might not capture this global rep-
resentation.3–5 Additionally, the ease of sampling and minimally
invasive nature of cirDNA collection allow for serial monitoring of
patients, which would otherwise be harmful using biopsies or CT
scans.6,7 Finally, with the development of easy-to-use kits and the
continuous reduction in sequencing costs, the analysis of ctDNA
in the clinic represents an increasingly viable and attractive option.
Recent developments that tailor sequencing to the biology of
ctDNA, innovative sequencing solutions, as well as integration of
machine learning analysis of large scale cohorts will also lead to
intriguing new avenues for implementation of liquid biopsy in the
clinic.

Acknowledgments
The authors would like to thank Nitzan Rosenfeld and all the
members of the Rosenfeld group, especially I. Hudecova and
D. Gale for proof reading the manuscript. This research was
funded by Cancer Research UK (grant numbers A20240), the
European Research Council (ERC) under the European Union’s
Seventh Framework Programme (FP/2007-2013) ERC Grant
Agreement number 337905 and supported by the NIHR Cam-
bridge BRC. The views expressed are those of the author(s) and
not necessarily those of the NIHR or the Department of Health
and Social Care.

b4444_Ch-06.indd 164 08-Jul-22 6:01:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 165

References
1. Siravegna G, Marsoni S, Siena S, Bardelli A. (2017) Integrating liquid biop-
sies into the management of cancer. Nat Rev Clin Oncol 14: 531–548 [http://
www.nature.com/doifinder/10.1038/nrclinonc.2017.14]
2. Wan JCM, Massie C, Garcia-Corbacho J, et al. (2017) Liquid biopsies
come of age: Towards implementation of circulating tumour DNA. Nat Rev
Cancer 17: 223–238.
3. Chan KCA, Jiang P, Zheng YWL, et al. Cancer genome scanning in plasma:
Detection of tumor-associated copy number aberrations, single-nucleotide
variants, and tumoral heterogeneity by massively parallel sequencing. Clin
Chem 59: 211–224 [http://www.ncbi.nlm.nih.gov/pubmed/23065472]
4. Murtaza M, Dawson S-J, Tsui DWY, et al. (2013) Non-invasive analysis
of acquired resistance to cancer therapy by sequencing of plasma DNA.
Nature 497: 108–112.
5. Abbosh C, Birkbak NJ, Wilson GA, et al. (2017) Phylogenetic ctDNA anal-
ysis depicts early-stage lung cancer evolution. Nature 545: 446–451.
6. Fazel R, Krumholz HM, Wang Y, et al. (2009) Exposure to low-dose ionizing
radiation from medical imaging procedures. N Engl J Med 361: 849–857
[http://www.ncbi.nlm.nih.gov/pubmed/19710483]
7. Overman MJ, Modak J, Kopetz S, et al. (2013) Use of research biopsies in
clinical trials: Are risks and benefits adequately discussed? J Clin Oncol 31:
17–22.
8. Schwarzenbach H, Hoon DSB, Pantel K. (2011) Cell-free nucleic acids as
biomarkers in cancer patients. Nat Rev Cancer 11: 426–437 [http://dx.doi.
org/10.1038/nrc3066]
9. Bettegowda C, Sausen M, Leary RJ, et al. (2014) Detection of circulating
tumor DNA in early-and late-stage human malignancies. Sci Transl Med 6:
224ra24–224ra24 [http://stm.sciencemag.org/cgi/doi/10.1126/scitranslmed.
3007094]
10. Phallen J, Sausen M, Adleff V, et al. (2017) Direct detection of early-stage
cancers using circulating tumor DNA. Sci Transl Med 9: eaan2415.
11. Mouliere F, Chandrananda D, Piskorz AM, et al. (2018) Enhanced detec-
tion of circulating tumor DNA by fragment size analysis. Sci Transl Med 10:
eaat4921.

b4444_Ch-06.indd 165 08-Jul-22 6:01:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

166 Cell-Free Circulating DNA: Purification and Analysis Techniques

12. Smith CG, Moser T, Mouliere F, et al. (2020) Comprehensive characteri-

zation of cell-free tumor DNA in plasma and urine of patients with renal
tumors. Genome Med 12(1): 23.
13. Jahr S, Hentze H, Englisch S, et al. (2001) DNA fragments in the blood
plasma of cancer patients: Quantitations and evidence for their origin from
apoptotic and necrotic cells. Cancer Res 61: 1659–1665.
14. Parkinson CA, Gale D, Piskorz AM, et al. (2016) Exploratory analysis of
TP53 mutations in circulating tumour DNA as biomarkers of treatment
response for patients with relapsed high-grade serous ovarian carcinoma: A
retrospective study. Mardis ER, editor. PLOS Med 13: e1002198.
15. Lo YMD, Chan KCA, Sun H, et al. (2010) Maternal plasma DNA sequenc-
ing reveals the genome-wide genetic and mutational profile of the fetus.
Sci Transl Med 2010 Dec 8; 2(61): 61ra91 [http://stm.sciencemag.org/con-
tent/2/61/61ra91.full]
16. Chandrananda D, Thorne NP, Bahlo M. (2015) High-resolution charac-
terization of sequence signatures due to non-random cleavage of cell-free
DNA. BMC Med Genomics 8: 29 [http://bmcmedgenomics.biomedcentral.
com/articles/10.1186/s12920-015-0107-z]
17. Lichtenstein AV, Melkonyan HS, Tomei LD, Umansky SR. (2001) Circu-
lating nucleic acids and apoptosis. Ann N Y Acad Sci 945: 239–249 [http://
www.ncbi.nlm.nih.gov/pubmed/11708486]
18. Mouliere F, Robert B, Arnau Peyrotte E, et al. (2011) High fragmentation
characterizes tumour-derived circulating DNA. Lee T, editor. PLOS One 6:
e23418 [http://dx.plos.org/10.1371/journal.pone.0023418]
19. Underhill HR, Kitzman JO, Hellwig S, et al. (2016) Fragment length of
circulating tumor DNA. PLOS Genet 2016 Jul 18;12(7): e1006162
20. Jiang P, Sun K, Peng W, et al. (2020) Plasma DNA end-motif profiling as
a fragmentomic marker in cancer, pregnancy, and transplantation. Cancer
Discov 10: 664–673 [https://pubmed.ncbi.nlm.nih.gov/32111602/]
21. Shendure J, Ji H. (2008) Next-generation DNA sequencing. Nat Biotechnol
26: 1135–1145 [http://www.nature.com/doifinder/10.1038/nbt1486]
22. Valle-Inclan JE, Stangl C, de Jong AC, et al. (2021) Optimizing nanopore
sequencing-based detection of structural variants enables individualized cir-
culating tumor DNA-based disease monitoring in cancer patients. Genome

b4444_Ch-06.indd 166 08-Jul-22 6:01:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 167

Med 13: 86 [https://genomemedicine.biomedcentral.com/articles/10.1186/

s13073-021-00899-7]
23. Urosevic N, Inglis TJJ, Grasko J, Lim EM. (2013) Evaluation of clinical
laboratory methods for plasma cell-Free DNA analysis in suspected sep-
ticaemia. J Med Diagn Meth 2(3) [https://www.omicsgroup.org/journals/
evaluation-of-clinical-laboratory-methods-for-plasma-cell-free-dna-
analysis-in-suspected-septicaemia-2168-9784.1000123.php?aid=17035]
24. Ponti G, Maccaferri M, Manfredini M, et al. The value of fluorimetry
(Qubit) and spectrophotometry (NanoDrop) in the quantification of cell-
free DNA (cfDNA) in malignant melanoma and prostate cancer patients.
Clin Chim Acta 479: 14–19 [http://linkinghub.elsevier.com/retrieve/pii/
S000989811830007X]
25. Vogelstein B, Kinzler KW. (1999) Digital PCR. Proc Natl Acad Sci USA 96:
9236–9241 [http://www.ncbi.nlm.nih.gov/pubmed/10430926]
26. Ciriello G, Miller ML, Aksoy BA, Senbabaoglu Y, Schultz N, Sander C.
(2013) Emerging landscape of oncogenic signatures across human can-
cers. Nat Genet 45: 1127–1133 [http://www.nature.com/doifinder/10.1038/
ng.2762]
27. Devonshire AS, Whale AS, Gutteridge A, et al. (2014) Towards standard-
isation of cell-free DNA measurement in plasma: Controls for extraction
efficiency, fragment size bias and quantification. Anal Bioanal Chem 406:
6499–6512 [http://www.ncbi.nlm.nih.gov/pubmed/24853859]
28. Linnarsson S. (2010) Recent advances in DNA sequencing methods —
general principles of sample preparation. Exp Cell Res 316: 1339–1343
[http://www.sciencedirect.com/science/article/pii/S0014482710000984]
29. Head SR, Komori HK, LaMere SA, et al. (2014) Library construction for
next-generation sequencing: Overviews and challenges. Biotechniques 56:
61–64, 66, 68, passim [http://www.ncbi.nlm.nih.gov/pubmed/24502796]
30. Van Dijk EL, Jaszczszyn Y, Thermes C. (2014) Library preparation methods for
next-generation sequencing: Tone down the bias. Exp Cell Res 322: 12–20.
31. Malapelle U, Pisapia P, Rocco D, et al. (2016) Next generation sequenc-
ing techniques in liquid biopsy: Focus on non-small cell lung cancer
patients. Transl Lung Cancer Res 5: 505–510 [http://www.ncbi.nlm.nih.gov/
pubmed/27826531]

b4444_Ch-06.indd 167 08-Jul-22 6:01:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

168 Cell-Free Circulating DNA: Purification and Analysis Techniques

32. Quail MA, Kozarewa I, Smith F, et al. (2008) A large genome center’s
improvements to the Illumina sequencing system. Nat Methods 5: 1005–
1010 [http://www.nature.com/doifinder/10.1038/nmeth.1270]
33. van Dijk EL, Auger H, Jaszczyszyn Y, Thermes C. (2014) Ten years of next-gen-
eration sequencing technology. Trends Genet 30: 418–426 [http://www.sci-
encedirect.com/science/article/pii/S0168952514001127?via%3Dihub]
34. Burnham P, Kim MS, Agbor-Enoh S, et al. (2016) Single-stranded DNA
library preparation uncovers the origin and diversity of ultrashort cell-
free DNA in plasma. Sci Rep 6: 27859 [http://www.nature.com/articles/
srep27859]
35. Gansauge M-T, Meyer M. (2013) Single-stranded DNA library preparation
for the sequencing of ancient or damaged DNA. Nat Protoc 8: 737–748
[http://www.nature.com/doifinder/10.1038/nprot.2013.038]
36. Snyder MW, Kircher M, Hill AJ, Daza RM, Shendure J. (2016) Cell-
free DNA comprises an in vivo nucleosome footprint that informs its tis-
sues-of-origin. Cell 164: 57–68.
37. Gansauge M-T, Gerber T, Glocke I, et al. (2017) Single-stranded DNA library
preparation from highly degraded DNA using T4 DNA ligase. Nucleic Acids
Res 45: e79 [http://www.ncbi.nlm.nih.gov/pubmed/28119419]
38. Mouliere F, Rosenfeld N. (2015) Circulating tumor-derived DNA is shorter
than somatic DNA in plasma. Proc Natl Acad Sci 112: 3178–3179.
39. Jiang P, Chan CWM, Chan KCA, et al. (2015) Lengthening and shortening
of plasma DNA in hepatocellular carcinoma patients. Proc Natl Acad Sci
112: E1317–E1325.
40. Moser T, Ulz P, Zhou Q, et al. (2017) Single-stranded DNA library prepa-
ration does not preferentially enrich circulating tumor DNA. Clin Chem
63: 1656–1659 [http://clinchem.aaccjnls.org/content/early/2017/08/07/clinchem.
2017.277988.long]
41. Zhu YY, Machleder EM, Chenchik A, Li R, Siebert PD. (2001) Reverse
transcriptase template switching: A SMART approach for full-length cDNA
library construction. Biotechniques 30: 892–897 [http://www.ncbi.nlm.nih.
gov/pubmed/11314272]
42. Turchinovich A, Surowy H, Serva A, Zapatka M, Lichter P, Burwinkel B.
(2014) Capture and amplification by tailing and switching (CATS). An
ultrasensitive ligation-independent method for generation of DNA librar-

b4444_Ch-06.indd 168 08-Jul-22 6:01:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 169

ies for deep sequencing from picogram amounts of DNA and RNA. RNA
Biol 11: 817–828 [http://www.ncbi.nlm.nih.gov/pubmed/24922482]
43. Vardi O, Shamir I, Javasky E, Goren A, Simon I. (2017) Biases in the
SMART-DNA library preparation method associated with genomic poly
dA/dT sequences. Bandapalli OR, editor. PLOS ONE 12: e0172769 [http://
dx.plos.org/10.1371/journal.pone.0172769]
44. Thierry AR, Mouliere F, Gongora C, et al. (2010) Origin and quantifi-
cation of circulating DNA in mice with human colorectal cancer xeno-
grafts. Nucleic Acids Res 38: 6159–6175 [https://academic.oup.com/nar/
article-lookup/doi/10.1093/nar/gkq421]
45. Mouliere F, Piskorz AM, Chandrananda D, et al. (2017) Selecting short
DNA fragments in plasma improves detection of circulating tumour DNA.
bioRxiv: 134437.
46. Song P, Chen SX, Yan YH, et al. (2021) Selective multiplexed enrichment
for the detection and quantitation of low-fraction DNA variants via low-
depth sequencing. Nat Biomed Eng 5(7): 690–701 [https://pubmed.ncbi.
nlm.nih.gov/33941896/]
47. Lou DI, Hussmann JA, McBee RM, et al. (2013) High-throughput DNA
sequencing errors are reduced by orders of magnitude using circle sequenc-
ing. Proc Natl Acad Sci 110: 19872–19877 [http://www.ncbi.nlm.nih.gov/
pubmed/24243955]
48. Kinde I, Wu J, Papadopoulos N, Kinzler KW, Vogelstein B. (2011) Detec-
tion and quantification of rare mutations with massively parallel sequencing.
Proc Natl Acad Sci U S A 108: 9530–9535 [http://www.ncbi.nlm.nih.gov/
pubmed/21586637]
49. Robin JD, Ludlow AT, LaRanger R, Wright WE, Shay JW. (2016) Compari-
son of DNA quantification methods for next generation sequencing. Sci Rep
6: 24067 [http://www.nature.com/articles/srep24067]
50. Hug H, Schuler R. (2003) Measurement of the number of molecules of a
single mRNA species in a complex mRNA preparation. J Theor Biol 221:
615–624 [http://www.ncbi.nlm.nih.gov/pubmed/12713944]
51. Casbon JA, Osborne RJ, Brenner S, Lichtenstein CP. (2011) A method
for counting PCR template molecules with application to next-generation
sequencing. Nucleic Acids Res 39: e81–e81 [https://academic.oup.com/nar/
article-lookup/doi/10.1093/nar/gkr217]

b4444_Ch-06.indd 169 08-Jul-22 6:01:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

170 Cell-Free Circulating DNA: Purification and Analysis Techniques

52. Kou R, Lam H, Duan H, et al. (2016) Benefits and challenges with apply-
ing unique molecular identifiers in next generation sequencing to detect
low frequency mutations. Wang J, editor. PLOS ONE 11: e0146638 [http://
dx.plos.org/10.1371/journal.pone.0146638]
53. Smith T, Heger A, Sudbery I. (2017) UMI-tools: Modeling sequencing
errors in unique molecular Identifiers to improve quantification accuracy.
Genome Res 27: 491–499 [http://www.ncbi.nlm.nih.gov/pubmed/28100584]
54. Newman AM, Lovejoy AF, Klass DM, et al. (2016) Integrated digital error
suppression for improved detection of circulating tumor DNA. Nat Biotech-
nol 34: 547–555.
55. Perakis S, Speicher MR. (2017) Emerging concepts in liquid biopsies. BMC
Med 15: 75 [http://www.ncbi.nlm.nih.gov/pubmed/28381299]
56. UM BRCF Bioinformatics Core. (2017) CONNOR [https://github.com/
umich-brcf-bioinf/Connor]
57. Shugay M, Zaretsky AR, Shagin DA, et al. (2017) MAGERI: Computational
pipeline for molecular-barcoded targeted resequencing. Gardner PP, edi-
tor. PLOS Comput Biol 13: e1005480 [http://dx.plos.org/10.1371/journal.
pcbi.1005480]
58. Agilent. (2017) SureCall [https://www.genomics.agilent.com/en/NGS-
Data-Analysis-Software/SureCall/?cid=AG-PT-154&tabId=AG-PR-
1196]
59. Schmitt MW, Kennedy SR, Salk JJ, Fox EJ, Hiatt JB, Loeb LA. (2012)
Detection of ultra-rare mutations by next-generation sequencing. Proc
Natl Acad Sci U S A 109: 14508–14513 [http://www.ncbi.nlm.nih.gov/
pubmed/22853953]
60. Cohen JD, Douville C, Dudley JC, et al. (2021) Detection of low-frequency
DNA variants by targeted sequencing of the Watson and Crick strands. Nat
Biotechnol 39(10):1220–1227.
61. Karlsson K, Sahlin E, Iwarsson E, Westgren M, Nordenskjöld M, Linnars-
son S. (2015) Amplification-free sequencing of cell-free DNA for prenatal
non-invasive diagnosis of chromosomal aberrations. Genomics 105: 150–
158 [http://www.ncbi.nlm.nih.gov/pubmed/25543032]
62. Dean FB, Hosono S, Fang L, et al. (2002) Comprehensive human genome
amplification using multiple displacement amplification. Proc Natl Acad Sci
U S A 99: 5261–5266 [http://www.ncbi.nlm.nih.gov/pubmed/11959976]

b4444_Ch-06.indd 170 08-Jul-22 6:01:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 171

63. Goodwin S, McPherson JD, McCombie WR. (2016) Coming of age: Ten
years of next-generation sequencing technologies. Nat Rev Genet 17: 333–
351 [schatz]
64. Wang T-L, Maierhofer C, Speicher MR, et al. (2002) Digital karyotyping.
Proc Natl Acad Sci U S A 99: 16156–16161 [http://www.ncbi.nlm.nih.gov/
pubmed/12461184]
65. Leary RJ, Sausen M, Kinde I, et al. (2012) Detection of chromosomal
alterations in the circulation of cancer patients with whole-genome
sequencing. Sci Transl Med 4: 162ra154 [http://www.ncbi.nlm.nih.gov/
pubmed/23197571]
66. Hanahan D, Weinberg RA. (2000) The hallmarks of cancer. Cell 100: 57–70
[http://linkinghub.elsevier.com/retrieve/pii/S0092867400816839]
67. Adalsteinsson VA, Ha G, Freeman SS, et al. (2017) Scalable whole-exome
sequencing of cell-free DNA reveals high concordance with metastatic
tumors. Nat Commun 8: 1324.
68. Belic J, Koch M, Ulz P, et al. (2015) Rapid identification of plasma DNA
samples with increased ctDNA levels by a modified FAST-SeqS approach.
Clin Chem 61: 838–849.
69. Heitzer E, Ulz P, Belic J, et al. (2013) Tumor-associated copy number
changes in the circulation of patients with prostate cancer identified through
whole-genome sequencing. Genome Med 5: 30 [http://genomemedicine.
biomedcentral.com/articles/10.1186/gm434]
70. Cristiano S, Leal A, Phallen J, et al. (2019) Genome-wide cell-free DNA
fragmentation in patients with cancer. Nature 570(7761): 385–389 [http://
www.nature.com/articles/s41586-019-1272-6]
71. Zviran A, Schulman RC, Shah M, et al. (2020) Genome-wide cell-free DNA
mutational integration enables ultra-sensitive cancer monitoring. Nat Med
26: 1114–1124.
72. Genohub. (2015) Whole Genome Sequencing (WGS) vs. Whole Exome
Sequencing (WES) [https://blog.genohub.com/2015/02/21/whole-genome-
sequencing-wgs-vs-whole-exome-sequencing-wes/]
73. Samorodnitsky E, Jewell BM, Hagopian R, et al. (2015) Evaluation of
hybridization capture versus amplicon-based methods for whole-ex-
ome sequencing. Hum Mutat 36: 903–914 [http://www.ncbi.nlm.nih.gov/
pubmed/26110913]

b4444_Ch-06.indd 171 08-Jul-22 6:01:28 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

172 Cell-Free Circulating DNA: Purification and Analysis Techniques

74. Girotti MR, Gremel G, Lee R, et al. (2016) Application of sequencing, liq-
uid biopsies, and patient-derived xenografts for personalized medicine in
melanoma. Cancer Discov 6: 286–299.
75. Dietz S, Schirmer U, Mercé C, et al. (2016) Low input whole-exome
sequencing to determine the representation of the tumor exome in circulat-
ing DNA of non-small cell lung cancer patients. PLOS ONE 11: e0161012
[http://www.ncbi.nlm.nih.gov/pubmed/27529345]
76. Butler TM, Johnson-Camacho K, Peto M, et al. (2015) Exome sequencing
of cell-free DNA from metastatic cancer patients identifies clinically action-
able mutations distinct from primary disease. Richards KL, editor. PLOS
ONE 10: e0136407 [http://dx.plos.org/10.1371/journal.pone.0136407]
77. Newman AM, Bratman S V, To J, et al. (2014) An ultrasensitive method for
quantitating circulating tumor DNA with broad patient coverage. Nat Med
20: 548–554 [http://www.ncbi.nlm.nih.gov/pubmed/24705333]
78. Wan JCM, Heider K, Gale D, et al. (2020) ctDNA monitoring using
patient-specific sequencing and integration of variant reads. Sci Transl Med
12: eaaz8084.
79. Forshew T, Murtaza M, Parkinson C, et al. (2012) Noninvasive identifica-
tion and monitoring of cancer mutations by targeted deep sequencing of
plasma DNA. Sci Transl Med 4: 136ra68–136ra68.
80. Dawson S-J, Tsui DWY, Murtaza M, et al. (2013) Analysis of circulating
tumor DNA to monitor metastatic breast cancer. N Engl J Med 368: 1199–
1209.
81. Zheng Z, Liebers M, Zhelyazkova B, et al. (2014) Anchored multiplex PCR
for targeted next-generation sequencing. Nat Med 20: 1479–1484 [http://
www.nature.com/doifinder/10.1038/nm.3729]
82. Douville C, Cohen JD, Ptak J, et al. Assessing aneuploidy with repeti-
tive element sequencing. Proc Natl Acad Sci U S A 2020 Mar 3; 117(9):
4858–4863. doi: 10.1073/pnas.1910041117. Epub 2020 Feb 19.
83. McDonald BR, Contente-Cuomo T, Sammut S-J, et al. (2019) Personalized
circulating tumor DNA analysis to detect residual disease after neoadjuvant
therapy in breast cancer. Sci Transl Med 11: eaax7392 [http://stm.sciencemag.
org/lookup/doi/10.1126/scitranslmed.aax7392]
84. Leary RJ, Kinde I, Diehl F, et al. (2010) Development of personalized
tumor biomarkers using massively parallel sequencing. Sci Transl Med 2:
20ra14 [http://www.ncbi.nlm.nih.gov/pubmed/20371490]

b4444_Ch-06.indd 172 08-Jul-22 6:01:28 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Next-Generation Sequencing Libraries for Sequencing cirDNA 173

85. Leon SA, Shapiro B, Sklaroff DM, Yaros MJ. (1977) Free DNA in the serum
of cancer patients and the effect of therapy. Cancer Res 37: 646–650 [http://
www.ncbi.nlm.nih.gov/pubmed/837366]
86. Stroun M, Anker P, Lyautey J, Lederrey C, Maurice PA. (1987) Isolation
and characterization of DNA from the plasma of cancer patients. Eur J Can-
cer Clin Oncol 23: 707–712 [https://www.sciencedirect.com/science/article/
pii/0277537987902665]
87. European Medicines Agency. (2014) Iressa: EPAR-Product Information
[http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_-_
Product_Information/human/001016/WC500036358.pdf]
88. FDA. (2016) cobas® EGFR Mutation Test v2 approval [https://www.access-
data.fda.gov/cdrh_docs/pdf15/P150044A.pdf]
89. Pereira E, Camacho-Vanegas O, Anand S, et al. (2015) Personalized circu-
lating tumor DNA biomarkers dynamically predict treatment response and
survival in gynecologic cancers. PLOS ONE 10: e0145754 [http://dx.plos.
org/10.1371/journal.pone.0145754]

b4444_Ch-06.indd 173 08-Jul-22 6:01:28 PM

 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0007

Chapter 7
Blood Nucleases Affecting
Circulating DNA in Serum
and Plasma
Gustavo Barra

Abstract

Circulating cell-free DNA (cirDNA) is extracellular DNA occurring in

the blood.1 It is a biomarker of growing interest in various clinical fields,
especially in prenatal diagnosis and oncology, because it allows sampling
of fetal or tumor DNA while avoiding the risks associated with inva-
sive tissue biopsies. The existence of a DNA-degrading activity in the
blood is also well established; however, the impact of these enzymes as
a preanalytical variable in cirDNA assays and studies have been long
neglected. The significance of blood DNAses is 2-fold. Firstly, they are
intimately involved in the physiological mechanisms of cirDNA gen-
eration and clearance and as such are responsible for the size profile,
cleavage patterns, and concentration of cirDNA in blood. Secondly, the
activity of blood DNAses in plasma and serum after sample collection
contributes to experimental artifacts including decreasing DNA concen-
tration, changing size distribution, and DNA sequences of interest falling

175

b4444_Ch-07.indd 175 08-Jul-22 6:02:29 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

176 Cell-Free Circulating DNA: Purification and Analysis Techniques

below detectable levels. A detailed understanding of DNases biological

properties is essential to making the best use of cirDNA in different
clinical settings. Thus, this chapter is a review of the types, functions,
biology, and clinical correlates for the DNases found in the circulation.
Moreover, it contains a summary of the main results observed by our
research group about the blood DNase ex vivo effect over the cirDNA
in the most commonly used cell-free samples.

Biological Role of Blood DNases

Unprotected exogenous single-stranded oligonucleotides and
double-stranded DNA are rapidly degraded in vitro2–4 and in
vivo5,6 suggesting that protection against foreign nucleic acids is
one of blood DNases’ biological roles.7 A role in the prevention
of horizontal transfer of gene sequences from one cell to another
can also be speculated.8 Similarly, these enzymes degrade endoge-
nous nucleic acids that appear in the circulation as consequence of
DNA released by living, apoptotic and necrotic cells, as well as by
other mechanisms (e.g., NETosis), maintaining the physiological
level of DNA in the circulation.9
An inverse correlation between the cirDNA yield and blood
DNase activity has been reported10,11 and failure of the DNA clear-
ance mechanism has been involved in pathogenesis of autoimmune
diseases (e.g., systemic lupus erythematosus).12–14 For example,
serum DNAse I, C1q15 and factor VII-activating protease16 coop-
erate in the degradation of chromatin and removal of dying cells
by macrophages, which is essential for tissue homeostasis and res-
olution of inflammation.17
Howsoever, DNases do not seem to be the major mecha-
nism for cirDNA removal from the bloodstream.18,19 In addition
to its breakdown by DNases, intake by cells,8 liver,8 and kidney20
metabolism can clear cirDNA. The association of the DNA with

b4444_Ch-07.indd 176 08-Jul-22 6:02:29 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 177

Fig. 1: Sources of cirDNA and physiological clearance mechanism.

nucleosomes seems to have an important effect on the uptake and

breakdown processes.8
In spite of the clearance mechanism, fetal cirDNA has a mean
half-life of only 16.3 min in maternal plasma19 and tumor cirDNA
in the blood has a half-life of ~2 hr.21 This evidence suggests that
cirDNA elimination from blood is fast, and a continuous supply
is necessary to maintain a specific cirDNA sequence at a constant
level in the circulation (Fig. 1).

Role of DNases in cirDNA Origins

cirDNA is fragmented, occurring mainly as short, but also as long,
double-stranded molecules. Thus, it is generated by mechanisms
that involve the action of DNases. Understanding which enzymes
are involved in its origins would contribute to making the best use
of these molecules. In this section, a connection between cirDNA
fragment size and the DNAses that generate such fragment pat-
terns will be discussed.

b4444_Ch-07.indd 177 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

178 Cell-Free Circulating DNA: Purification and Analysis Techniques

cirDNA Origin
Recent studies have shown that majority of cirDNA in healthy
individuals originates from hematopoietic cell death and it is a het-
erogeneous mixture of sequences from lymphoid and myeloid cell
types.22 CirDNA yields ladder-like banding patterns equivalent to
whole-number multiples of the ~170 bp nucleosomal unit DNA
in gel electrophoresis,23,24 which is similar to the pattern observed
in degraded DNA from apoptotic cells25 suggesting that cirDNA
originates from cell dying by this process. Indeed, apoptosis is a
common, critical, and actively regulated process during hema-
topoiesis and in mature hematopoietic cell types.26 Additionally,
nondividing cells, such as lymphocytes, and cultured cell lines
including HL-60 spontaneously release a nucleoprotein complex
within a homeostatic system in which newly synthetized DNA is
preferentially released.27 This spontaneously released DNA also
has a ladder-like banding pattern in gel electrophoresis. Larger
DNA molecules (>10,000 bp) have also been observed in cirDNA
and such molecules are speculated to come from cells dying via
necrosis.28 DNA released by necrosis is incompletely and nonspe-
cifically digested and thus smears on electrophoretic separation
due to its larger fragment sizes.27
In nonphysiological conditions such as cancer or acute/
chronic tissue damage there is additional sequence contribution
from one or more nonhematopoietic cell types to the cirDNA
pool. These additional sequences, as well as the aberrant contribu-
tion from hematopoietic cell lineages, have tremendous potential
as a “liquid biopsy.”22 However, the relative contribution of each
above-cited processes to the cirDNA levels in pathological con-
ditions (e.g., myocardial infarction, stroke, autoimmune disorders
and cancer) is still under investigation.

b4444_Ch-07.indd 178 08-Jul-22 6:02:31 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 179

Finally, NETosis is another mechanism that may produce

cirDNA in pathological conditions. Upon activation, neutrophils
(the most abundant population of white blood cells) can release
extracellular nucleic acids decorated with histones and granular
proteins capable of entrapping pathogens.29 These DNA struc-
tures, named neutrophil extracellular traps (NETs), can be formed
within blood vessels30 (Fig. 1).

cirDNA Fragment Size

Advanced technologies of massively parallel sequencing have pro-
vided unprecedented opportunities to investigate the size profile
of cirDNA.31 The distribution of fragment lengths of cirDNA is
multimodal. The most prominent peak was around ~167 bp, with
next occurring at ~340 bp. A much wider mode is observed around
~510 bp. These three peaks appear to correspond to the lengths of
DNA associated with a mono-, di-, and tri-nucleosome structure,
respectively.32 While the first cell-free DNA mode peaks at ~167 bp, it
is preceded by a series of smaller peaks occurring at approximately
10 bp periodicity below ~145 bp (~145, ~134, ~123, ~113, ~102,
~92, and ~82 bp).32,33 Moreover, minute quantities (0.06%-–0.3%)
of sequences larger than 1000 bp have been observed.34 This adds
evidence to the hypothesis that the nucleosome packaging and the
approximately 10 bp 360° turn of the double helix are key deter-
minants for the fragmentation of the cirDNA.32

DNA Degradation During Apoptosis

Firstly the chromosomal DNA is cleaved into large fragments of
50–300 kB (a size consistent with chromatin loop domains8) by
topoisomerase II,35 apoptosis-inducing factor,36 and/or caspase

b4444_Ch-07.indd 179 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

180 Cell-Free Circulating DNA: Purification and Analysis Techniques

activated DNase (CAD).37 Subsequently, CAD (caspase-dependent

apoptotic pathway)38,39 or endonuclease G40 DNase g and DNase I41
(caspase-independent apoptotic pathway) cleave these large DNA
fragments in oligonucleosomes and/or mononucleosomes.42 Then,
after the apoptotic cells are engulfed by macrophages, DNase II
further degrades the fragmented DNA in their lysosomes.43 Given
the high cellular turnover and/or saturation of the mechanism, it is
not surprising that some DNAs escape final cleavage/degradation
and thus appear in the circulation.44

DNA Protection by Nucleosome Structure

The core particle of a nucleosome consists of an octamer of two
copies of the four histones H2A, H2B, H3, and H4 — around
which ~146 bp of helical DNA is wrapped.8 An additional ~20
bp should be added if the DNA connected to the peripheral
histone H1 is considered (~165 bp).32 Negatively charged DNA
is electrostatically bound to the positively charged histones and
individual mononucleosomes are connected by a stretch of linker
DNA.45 During apoptosis, DNases cleave linker DNA more eas-
ily than nucleosome-bound DNA, thus resulting in the classic
ladder pattern of fragment sizes associated with programmed
cell death.46

DNases that Cleave the DNA within the Nucleosome Core

As detailed below, DNase I can cut within and outside nucleoso-
mal DNA generating a cleavage signature of 10.3 base oscillation
that corresponds to the accessibility of the minor groove as DNA
winds around the nucleosome. A similar nick pattern is observed
for Endonuclease G but not for CAD.

b4444_Ch-07.indd 180 08-Jul-22 6:02:31 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 181

Blood DNases and the cirDNA Cleavage Signature

The fragment size profile described above suggests that cirDNA
fragments are derived from the enzymatic processing of DNA from
apoptotic cells.33 However, DNase I (the main nuclease found in
blood) generates similar DNA cleavage patterns and cannot be
excluded as a player in the process that generates the observed
cleavage signature. Thus, it is possible that, independent of the
mechanisms that may produce cirDNA (apoptosis, necrosis, active
release, and netosis), blood DNases shape its size pattern. In con-
clusion, cellular DNases and/or blood DNases together with the
protection conferred by nucleosome packing are likely to be the
key determinants for cirDNA fragment size distribution observed
in the blood.

Types of DNAses
DNases are enzymes capable of hydrolyzing the most stable
chemical bond found in biological molecules, the phosphodiester
bond.47 Enzymes capable of degrading DNA in the human blood
are deoxyribonuclease I, deoxyribonuclease II, phosphodiester-
ase I, DNA-hydrolyzing antibodies, and lactoferrin. Some of their
characteristics will be detailed in the following and a summary can
be found in Table 1. The DNases involved in apoptosis will also be
addressed in this section and are summarized in Table 2.

DNAses in Blood
Deoxyribonuclease I (DNAse I)
Deoxyribonuclease I is responsible for 90% of the DNase activity
of blood.48 It is characterized by neutral pH optimum, bivalent

b4444_Ch-07.indd 181 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

182 Cell-Free Circulating DNA: Purification and Analysis Techniques

Table 1. Characteristics of the DNAses Found in Blood

DNases Concentration Preferred substrate Bivalent ions EDTA
DNAse I High dsDNA>>ssDNA Dependent Inhibited
DNAse II Low dsDNA>>ssDNA Independent Not inhibited
Phosphodiesterase I Low ssDNA, Dependent Inhibited
RNA>>dsDNA
DNA-hydrolyzing Low dsDNA>>ssDNA Dependent Inhibited
antibodies
Lactoferrin Low dsDNA>>ssDNA Stimulated Inhibited

Table 2. Characteristics of the DNAses Involved in Apoptosis

DNases Location Preferred substrate Bivalent ions EDTA
Caspase-activated Nuclei/ dsDNA Dependent Inhibited
DNase cytoplasm
Endonuclease G Mitochondria ssDNA, Dependent Inhibited
RNA>>dsDNA
DNAse g Nuclei ssDNA>>dsDNA Dependent Inhibited
L-DNase II Cytoplasm dsDNA>>ssDNA Independent Not inhibited

metal ion (Ca2+ and Mg2+) requirement for catalytic activity,49 and
formation of oligonucleotides and nucleotides with a hydroxyl
group at the 3′-end and a phosphate group at the 5′-end.50 DNase
I is 100–500 times more active in hydrolysis of double-stranded
DNA than of single-stranded.51 The enzyme binds only with the
DNA minor groove,52,53 interacts with both DNA strands,51 and
introduces one to several single-strand breaks into both strands,
which are shifted by several nucleotides.54 DNase I digests with a
~10 bp periodicity around nucleosomes matching the exposure of
the DNA minor groove as it wraps around histones.55,56
DNAse I is a secretory protein.7 Pancreas and parotid glands
are the enzyme’s major sources, consistent with its role in digesting
nucleic acids in the gastrointestinal tract.57 It is also found in kidney,

b4444_Ch-07.indd 182 08-Jul-22 6:02:31 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 183

urine, blood, and seminal fluid, suggesting additional functions.48

A significant portion of blood DNAse I has pancreatic origin; other
sources include the pituitary.7 The concentration of active DNAse
I level in blood measured by single radial enzyme-diffusion assay
is 65 ± 27 units/mg protein, a value corresponding to 4.4 units/L.48
For comparison, its level in urine is 6000 ± 2000 × 103 units/mg
protein.48 In blood, the analysis of catalytic activity is more appro-
priate than the analysis of enzyme concentration,58 because the
proportion of the active form is influenced by the presence of the
enzyme’s natural inhibitor, actin.59

Deoxyribonuclease II (DNAse II)

Human DNase II is characterized by acidic pH optimum, lack
of bivalent ion (Ca2+ and Mg2+) requirement, and formation of
DNA reaction products with a hydroxyl group at the 5′-end and
a phosphate group at the 3′-end.60 It is 5–10 times more active
on double-stranded DNA than on single-stranded and cannot
hydrolyze RNA.61 Similarly to DNase I, DNAse II hydrolyses
the phosphodiester backbone by single-strand nicks rather than
double-strand cuts.60,62 The enzyme is found in the lysosomes of
almost all human cells and is believed to be involved in intracellu-
lar breakdown of DNA.7,60 Its activity is also found in saliva, blood,
urine, and seminal liquid.7 Human blood is characterized by low
content of DNAse II activity, 0.11 ± 0.009 units/mg protein.58
Finally, since DNase II has no divalent cation requirement, it is
insensitive to the chelating agent EDTA.62

Phosphodiesterase I
Members of the phosphodiesterase I family have alkaline pH opti-
mum, require bivalent metal ions (Ca2+, Zn2+ and Mg2+) for catalytic

b4444_Ch-07.indd 183 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

184 Cell-Free Circulating DNA: Purification and Analysis Techniques

activity, and remove 5′-mononucleotides successively from the

3′-hydroxy termini of both DNA and RNA.63,64 There is evidence
that these enzymes show preference for single-stranded or dena-
tured substrates.65,66 The membrane form of phosphodiesterase I is
one of the key enzymes responsible for degradation of nucleic acid
fragments to nucleosides.67 Hydrolyzing foreign DNA and RNA,
this enzyme plays a protective role.7 Family members have been
found in the kidney, pancreas, uterus, liver, and heart, and in the
following body fluids: seminal liquid, serum, bile, urine, milk, and
cerebrospinal liquid.7,68 The normal values in human blood serum
were determined to be 33 ± 6.4 units/L by a phosphodiesterase
activity assay68 and 36 ng/mL by ELISA.69 EDTA totally inhibits
the catalytic activity of phosphodiesterase I.68,70

DNA-Hydrolyzing Antibodies
Autoantibodies with DNA-nicking activity have been detected in
the serum of patients with autoimmune diseases (e.g., systemic
lupus erythematous,71 multiple sclerosis,72 and others7). Their
effectiveness of hydrolysis of double-stranded DNA is 3–5 times
higher than single-stranded.73 The DNA-nicking activity is depen-
dent on metal ions (Mg2+, Mn2+, and Ca2+) and is inhibited by
EDTA.71 The cleavage patterns of DNA-hydrolyzing antibodies do
not have any cleavage-site specificity and differ from those pro-
duced by DNAse from human serum and DNAse I.71 Their activ-
ity in patient serum is two orders of magnitude less than that of
DNAse I71 and is not detected or negligible in healthy subjects.71,72

Lactoferrin
Lactoferrin is a unique polyfunctional protein and possesses
five different catalytic activities: RNase, DNase, phosphatase,

b4444_Ch-07.indd 184 08-Jul-22 6:02:31 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 185

ATPase, and amylase.7 The enzyme is more effective in cleavage

of double-strand DNA than that of oligonucleotides, has maximal
activity at neutral pH, and is stimulated by bivalent ions (Cu2+,
Zn2+, Ca2+, Mg2+, and Mn2+). EDTA at concentrations of 10 mM or
higher inhibited its activity.74 Lactoferrin is the major nuclease of
human milk74 and has been found in a wide range of other human
external secretion (saliva, tears, urine, and others)75 as well as in
the specific granules of neutrophilic leukocytes.76 The enzyme is
present in blood at a very low concentration and its level appears
correlated to the neutrophil turnover.77 In healthy subjects, the lac-
toferrin level in plasma is 168 ± 100 µg/L (imunoenzymatic assay)
and its concentration is slightly higher in serum (237 ± 155 µg/L),78
probably because of neutrophil lysis during coagulation.79

Intracellular DNases Involved in the Apoptosis

Two classes of nucleases degrade the cellular DNA during apop-
tosis: (a) the cell autonomous nucleases (e.g., CAD and Endo
G), which cleave DNA within the dying cell and are responsible
for DNA laddering; and (b) the cell nonautonomous nucleases
(e.g., lysosomal DNase II), which derives from the cells that have
phagocytized the apoptotic remnants or destroy the DNA that is
released into the extracellular compartment.
Under physiological conditions, the intracellular nucleases
involved in apoptosis are confined to the cell. However, if cells lyse
during blood collection, storage, or handling, they will be released
into the specimen (e.g., serum or plasma) and could contribute
to the ex vivo DNA-degrading activity of the sample. While the
appearance of genomic DNA in cirDNA isolated from plasma,
concomitant with an unexpectedly high measurement of DNA
concentration, can be indicative of cell lysis and contamination with

b4444_Ch-07.indd 185 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

186 Cell-Free Circulating DNA: Purification and Analysis Techniques

cellular components, absence of genomic DNA does not necessarily

exclude that cytoplasmic cell contents haven’t been released.
The nucleus, supported by the nuclear lamina, is relatively
robust and lysis of the plasma membrane is likely to precede lysis
of the nucleus.80 Isolated cell nuclei are pelleted by relatively
low centrifugation speeds (e.g., 800 × g),81 such as are typically
used to separate plasma from cellular blood components. Hence,
nuclei from partially lysed cells, if present, will be separated from
plasma during blood centrifugation along with red blood cells
and leukocytes, and an absence of genomic DNA in the sample
does not necessarily indicate that the plasma hasn’t been contam-
inated with cytoplasmic contents. The same considerations apply
to serum, which is well known to have a higher cirDNA concen-
tration than plasma, and this increased concentration has been
attributed to cell lysis during the coagulation.79 Hence, intracellu-
lar DNAses involved in apoptosis are described in the following,
and their activity profile and characteristics are summarized in
Table 2.

Caspase-Activated DNase
CAD, also called DFF40, is characterized by neutral pH opti-
mum, Mg2+ requirement, and formation of DNA reaction prod-
ucts with a hydroxyl group at the 3′-end and a phosphate group at
the 5′-end. It is specific for double-stranded DNA not for single
stranded DNA or RNA and exhibits an extraordinary preference
for cleaving the internucleosomal linker regions in chromatin (the
enzyme is unable to cut DNA bound to the histone octamer).82
The enzyme cuts both DNA strands and generates blunt ends or
ends with 1-base 5′-overhangs.83 Consequently, its action on chro-
matin results in the classical DNA ladders observed upon apop-
totic cell death (~180 pb periodicity).84 CAD is found in the cell

b4444_Ch-07.indd 186 08-Jul-22 6:02:31 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 187

nuclei and cytoplasm heterodimerized with its inhibitor, which is

released upon apoptotic activation.85

Endonuclease G
Endonuclease G is a mitochondrial nuclease that translocates to
the nucleus during apoptosis.86 The enzyme is characterized by
neutral pH optimum, Mg2+ or Mn2+ ion requirement, and forma-
tion of DNA reaction products with a hydroxyl group at the 3′-end
and a phosphate group at the 5′-end. Endonuclease G “prefers”
single-stranded DNA or RNA over double-stranded DNA.87 Endo-
nuclease G chromatin cleavage often results in single-stranded
nicks between nucleosomes in the DNA linkers (~190 bases peri-
odicity) and within the nucleosome core (~10 base periodicity).88
These cleavage patterns are similar to those of DNase I digestion
products.89 However, unlike DNase I, Endonuclease G prefer-
entially attacks single-stranded regions, allowing for targeting by
single-stranded nicks for adjacent strand cleavage to generate
double-stranded nucleosomal length fragments.88 DNase I and
exonuclease III stimulate its ability to generate dsDNA cleavage
products at physiological ionic strength in vitro.88

DNAse g
DNase g is characterized by neutral pH optimum, Ca2+ and Mg2+
requirement, and formation of single-stranded DNA breaks with a
hydroxyl group at the 3′-end and a phosphate group at the 5′-end.
The enzyme is stored in the nuclear envelope lumen and released
into the nucleus in the late apoptotic phase and accelerates DNA
fragmentation. Thus, DNA fragmentation is initiated by CAD/
DFF40 and DNase g completes the digestion of the genomic DNA
in dying cells as apoptosis’ final caretaker.90 The enzyme action on

b4444_Ch-07.indd 187 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

188 Cell-Free Circulating DNA: Purification and Analysis Techniques

naked DNA generates the typical DNA ladder as seen in apoptosis

but only at low ionic strength buffer conditions. However, in the
presence of histone H1, the DNase g coactivator, naked DNA, and
chromatin are effectively cleaved at physiological ionic strength.91

Leukocyte Elastase Inhibitor-Derived DNAse II (L-DNase II)

Leukocyte Elastase Inhibitor (LEI)-derived DNase II (L-DNase
II) is characterized by acidic pH optimum, bivalent ions indepen-
dency, and formation DNA breaks with a hydroxyl group at the
5′-end and a phosphate group at the 3′-end.87 The enzyme induces
the cleavage of DNA into an oligonucleosomal ladder in vitro.92
Curiously, L-DNAse II originates from the LEI, a member of the
serine protease inhibitors superfamily. The acidic treatment or
action of some proteases (e.g., elastase, cathepsin D, and other
proteases activated during apoptosis) over LEI changes its enzy-
matic activity; the antiprotease activity is lost and the endonuclease
activity is gained. The modification also unveils a nuclear localiza-
tion signal. Hence, L-DNAse II enters into the nucleus, cleaves the
DNA, and finally leads to apoptosis.93 This is a caspase-indepen-
dent pathway and an interesting example in which an anti-apop-
totic protein acquires pro-apoptotic function.87

Other Apoptosis DNases

There are several other less characterized apoptosis DNases — for
review see.86,87

Clinical Correlates of Blood DNAse I Activity

Serum DNase I activity has been measured in several diseases/
conditions (Table 3). If compared to the levels found in healthy

b4444_Ch-07.indd 188 08-Jul-22 6:02:31 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 189

Table 3. Diseases or Conditions with High/Low Serum DNase I Activity

Compared to Controls
High serum DNase I activity Low serum DNase I activity
Oral cancer94 Malignant lymphomas95
Breast cancer96 Stomach and colon cancer11
Transient myocardial ischaemia97,98 Pancreatic cancer and pancreatitis99
Acute myocardial infarction100 Overnight fast101
Antineutrophil cytoplasmic Surgical trauma101
antibody-associated vasculitis102
Type 2 diabetes103 Systemic lupus erythematosus104,105
Exercise106 Prostate tumors10

individuals, blood DNA-degrading activity can be classified into

two categories: high serum DNase activity compared to controls
or low serum DNase activity compared to controls (see below).
Although the focus of intensive research, the diagnostic utility of
serum DNAse activity has not been translated to clinical practice
yet. Differences in the methods used, low-level of enzymatic activ-
ity, and large inter-individual variations are some of the factors that
contribute to this scenario.58

Assays to Monitor the DNase Activity

Fluorometric, electrochemical, and immunological assays have
been developed to monitor the DNase activity in clinical samples;
further details can be found in the review by Sato and Takenaka.107
The fluorometric assays can be divided into three classes (Fig. 2).

Assay Based on Decrease in Fluorescence of

Noncovalent DNA Dyes
These assays rely on the fact that noncovalent DNA dyes (e.g.,
Pico Green, SYBR Green, or ethidium bromide) fluoresce when

b4444_Ch-07.indd 189 08-Jul-22 6:02:31 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

190 Cell-Free Circulating DNA: Purification and Analysis Techniques

(a)

(b)

(c)

Fig. 2: The principles of fluorometric assays to monitor the DNAse activity

in clinical samples: (a) assays based on decrease in fluorescence of DNA inter-
calating dyes; (b) assays based on increase in fluorescence of attenuated DNA
dyes; (c) assays based on cleavage of dual-labeled probes. Attenuated dyes (grey
ellipses), deattenuated dyes (blue ellipses), and quencher (black ellipse).

bound to double-stranded DNA molecules. The degradation of

the double-stranded DNA molecules by DNases decreases the
fluorescence, which is proportional to the enzyme activity (Fig.
2a).108 These assays can be performed in solution or in a matrix
(e.g., agarose gels — the single radial enzyme-diffusion assay).48,109

Assays Based on Increase in Fluorescence Self-attenuated

Covalent DNA Dyes
These assays use DNA molecules covalently conjugated to a flu-
orescent dye. The proximity with the DNA attenuates its fluores-
cence. With DNA cleavage by Dnases, the fluorescence increases
proportionally to the enzyme activity (Fig. 2b).110

b4444_Ch-07.indd 190 08-Jul-22 6:02:32 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 191

Assays Based on Fluorescence Deattenuation of Dual-labeled

Probes (e.g., qPCR Hydrolysis Probes)
These assays use single-stranded oligonucleotides labeled with a
fluorophore and a quencher. The oligo cleavage by DNAses phys-
ically separates both labels, and the fluorescence increase reports
the enzyme activity (Fig. 2c).111,112

DNAse Activity and the Initial Circulation Cell-Free DNA

Concentration in Plasmas and Serum
Blood nucleases degrade cirDNA ex vivo and are a preanalyti-
cal source of variation that decreases the sensitivity of molecular
assays. For many years, their impact on cirDNA was neglected.113
Our group conceived independently an assay based on a qPCR
hydrolysis probe degradation to monitor the DNase activity in clin-
ical samples. However, previous similar protocols were identified
after a literature search. By using this method, we measured blood
DNase activity in the major cell-free specimens: plasma-EDTA,
plasma-heparin, plasma-citrate 3.2%, and serum.113,114

Endogenous DNase Activity Assay

Endogenous DNase activity assay consisted of the admixture of
the crude sample with a qPCR hydrolysis probe in a qPCR mas-
ter mix. The reactions were then incubated isothermally at 37°C
for 24 hr on a qPCR system. The fluorescence is measured every
30 min. The qPCR master mix contains ROX (passive reference dye)
and probe labeled with FAM (active dye). When the hydrolysis
probe is intact, the quencher is close to FAM and its fluorescence
is mitigated. The FAM fluorescence increases upon hydrolysis
probe cleavage by DNase activity of the sample. The unit of the

b4444_Ch-07.indd 191 08-Jul-22 6:02:32 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

192 Cell-Free Circulating DNA: Purification and Analysis Techniques

(a) (b)

Fig. 3: qPCR hydrolysis probe kinetics observed in the endogenous DNase

activity assay for (a) plasma-EDTA (magenta circles) and serum (red circles)
and (b) 1U of DNase I (red squares) and H2O (magenta squares).

reaction is the deltaRn, which is the ratio of fluorescence values

between active and passive dyes in each measurement subtracted
by the ratio of the first measurement. A typical graphical rep-
resentation of the endogenous DNase activity assay in paired
plasma-EDTA and serum samples is shown in Fig. 3 (a plot of
deltaRn versus time).113

Plasma-EDTA
Endogenous DNase activity is highly inhibited in plasma-EDTA
(Fig. 2a). It makes this matrix the specimen of choice to avoid
the DNases’ ex vivo impact on cirDNA. Indeed, EDTA works
as an indirect inhibitor. This anticoagulant chelates divalent ions
(Ca2+, Mg2+ and Mn2+), which are essential for the activity of many
DNAses, especially DNAse I (the major DNase found in blood)
providing higher stability to the analyte.113
Although the inhibition is high, it is not complete. A resid-
ual DNase activity is observed in plasma-EDTA and it could be

b4444_Ch-07.indd 192 08-Jul-22 6:02:33 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 193

(a) (b)

Fig. 4: DNase activity (a) and initial cirDNA (b) of the main plasma tube
types and serum.

attributed to the following: (a) the concentration of EDTA in

regular tubes is not enough for full inhibition113 and (b) the small
concentration of DNAse II found in blood, an enzyme that is inde-
pendent of bivalent ions (Fig. 4a).
Another advantage of the plasma-EDTA is that the original
admixture of DNA sequences found in the bloodstream tends
to be preserved in this specimen, because of the low contamina-
tion with leukocyte genomic DNA in promptly processed blood
samples.115 Such contamination could decrease the cirDNA assay
sensitivity for rare DNA sequences. Consequently, plasma-EDTA
better represents the physiological status of the cirDNA in vivo114
(Fig. 4b).

Plasma-Citrate 3.2%
Similar to EDTA, sodium citrate is an anticoagulant that chelates
bivalent ions.116,117 However, blood DNase activity of this speci-
men is only partially inhibited in the endogenous DNase activity
assay.114 This may be due to the sodium citrate concentration in the

b4444_Ch-07.indd 193 08-Jul-22 6:02:34 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

194 Cell-Free Circulating DNA: Purification and Analysis Techniques

collection tube being lower than the optimum for full enzymatic
inhibition (Fig. 4a).
Moreover, just after the blood draw, the initial amount of
cirDNA in this specimen is similar to plasma-EDTA. The above-
cited lower contamination with leukocyte genomic DNA is also
observed. These results make the plasma-citrate the best alterna-
tive to plasma-EDTA and suggest that chelating of bivalent ions is
a mechanism to both inhibit blood DNAses and avoid the introduc-
tion of unnecessary DNA sequences to the specimen114 (Fig. 4b).

Plasma-Heparin
Among the tested specimens, plasma-heparin showed the highest
DNase activity in the endogenous DNase activity assay (Fig. 4a).
It is 17-fold higher compared to plasma-EDTA.114 The mecha-
nism by which heparin prevents blood coagulation is not based on
divalent cation chelation. The anticoagulant action of heparin lies
in its ability to bind to and enhance the inhibitory activity of the
plasma protein antithrombin against several serine proteases of
the coagulation system, most importantly factors IIa (thrombin),
Xa and IXa.118 Because DNase activity is a temperature-triggered
mechanism,113 strictly low-temperature condition control should
be applied to heparin samples to neutralize the DNases’ ex vivo
impact on cirDNA (it can be more flexible for plasma-EDTA).
On the other hand, the initial amount of cirDNA in this speci-
men is slightly higher compared to the paired plasma-EDTA and
plasma-citrate (Fig. 4b). It suggests a small contamination with
leukocyte’s genomic DNA after the blood draw, which could be
explained by direct effect of heparin over these cells.119 Again, ex
vivo introduction of unnecessary nucleic acids could decrease the

b4444_Ch-07.indd 194 08-Jul-22 6:02:34 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 195

cirDNA assay sensitivity for rare DNA sequences. Taken together,

these observations indicate that plasma-heparin should be avoided
in cirDNA studies and analyses.

Serum
Serum has second highest DNase activity among the tested spec-
imens, which is 15-fold higher than plasma-EDTA (Fig. 4a). To
neutralize the DNases’ ex vivo impact on cirDNA, the strictly
low-temperature condition control should also be applied to
serum samples.113 Moreover, due to leukocyte lysis during coag-
ulation, a phenomenon not fully understood yet, serum is highly
contaminated with genomic DNA (Fig. 4b).114 Thus, as with hep-
arin, serum’s cell-free DNA admixture doesn’t reflect the in vivo
condition, and consequently should be avoided in cirDNA stud-
ies and analyses. Conversely, serum is an acceptable alternative to
whole blood for patient genomic DNA analysis.120

Nonanticoagulated Plasma
The nonanticoagulated plasma is obtained by blood draw in a plain
tube followed by immediate centrifugation. Probably, this is the
specimen that best represents the in vivo physiological status of
blood DNases and cell-free DNA, but is difficult to obtain and
handle. We observed similar DNA-degrading activity in nonan-
ticoagulated plasma as in paired serum, although they were dif-
ferent between individuals,113 suggesting that lysis of leukocytes
and release of cellular contents do not alter the nuclease profile
of serum compared to nonanticoagulated plasma. However, fur-
ther experiments are required to reach the final conclusion on this
question (Fig. 4).

b4444_Ch-07.indd 195 08-Jul-22 6:02:34 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

196 Cell-Free Circulating DNA: Purification and Analysis Techniques

Influence of Temperature
Blood DNase activity is a temperature-dependent mechanism113
and temperature is an important preanalytical factor for specimens
that lack additives with bivalent ion chelation capacity (e.g., serum
and plasma-heparin).114 In serum, no difference is observed in the
cirDNA yield after incubation at −20°C and 4°C for 24 hr, oth-
erwise, a decrease occurs at room temperature and 37°C. Because
EDTA indirectly inhibits blood DNases, the cirDNA is protected
from degradation in plasma-EDTA at 37°C (at least for 24 hr).
Despite the EDTA inhibition, the small but still detectable nucle-
ase activity leads to a significant reduction of the cirDNA yield
after 48 hr at 37°C.113

EDTA Concentration
The addition of a 10-fold serial dilution of EDTA to nonantico-
agulated plasma (and also to serum) resulted in a stepwise reduc-
tion of the blood DNase activity, evidencing that EDTA indirectly
inhibits blood’s DNases.113 However, the EDTA concentration in
regular EDTA tubes ranges from 1.5 to 2.0 mg per mL of blood,121
corresponding to approximately 5 × 10−3 M. We observed a sig-
nificant but not complete inhibition of the DNase activity at this
concentration. These results suggest that the cirDNA assays would
benefit from a 10-fold increase of EDTA concentration in the col-
lection tube since no or negligible DNAse activity was observed at
5 × 10−2 M (Fig. 4).113

Other Known Factors

Considering the remnant blood DNase activity ex vivo in regu-
lar EDTA tubes, cirDNA analysis could benefit from a 10-fold
increase of EDTA concentration in the collection tubes (Fig. 5).

b4444_Ch-07.indd 196 08-Jul-22 6:02:34 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 197

Fig. 5: A serial dilution of EDTA was added to nonanticoagulated plasma

before the endogenous DNase activity assay in order to investigate the
EDTA-mediated inhibition of hydrolysis probe degradation. A stepwise inhibi-
tion of the hydrolysis probe degradation was observed, consistent with EDTA
indirectly inhibiting the sample’s DNase activity.

Tube manufacturers should consider producing such a tube, spe-

cifically for the liquid biopsy field. Moreover, in the endogenous
DNase activity assay, the positive control showed hydrolysis probe
degradation kinetics that was not as continuous as the serum sam-
ples. We cannot distinguish if this effect is secondary to a higher
DNase amount in the serum together with the presence of inhibi-
tors, or secondary to the complex composition of DNA-degrading
enzymes in serum (Fig. 3).113

Approaches to Minimizing Preanalytical Artifacts in

cirDNA Analysis
The aim of cirDNA assays and studies is frequently the detection of
rare tumoral or fetal sequences in a huge background of wild-type

b4444_Ch-07.indd 197 08-Jul-22 6:02:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

198 Cell-Free Circulating DNA: Purification and Analysis Techniques

or maternal DNA, respectively. With the blood drawn, the supply

of these rare DNA sequences terminates and they become exposed
to the blood DNA-degrading activity, which could decrease the
assay sensitivity. Moreover, the ex vivo introduction of nucleic
acids from leukocytes further dilutes the signal and should also be
avoided. In summary, to minimize these preanalytical effects over
circulation cell-free DNA the following recommendations should
be considered:

(1) Use a collection tube with a DNase inhibitor (e.g., EDTA) to

minimize the enzyme action during blood processing, trans-
portation, and storage.
(2) Avoid coagulation and leukocyte lysis.
(3) Process the blood into its cell-free subproducts and/or
physically separate the cellular content as soon as possible
avoiding unwanted DNA contamination (e.g., centrifugation
and aliquoting or blood draw in tubes with gel barrier).
(4) Transport the sample at refrigerator temperature (2°C to 8°C)
or less, as DNAse activity is a temperature-triggered enzymat-
ic activity.
(5) Plan to extract the DNA and detect the rare DNA as soon as
possible, as it is difficult to neutralize all DNAse activity.
6) If necessary, store the crude sample at freezer temperature
(−12°C to −20°C) or less.

Limitations of Current Knowledge

The limitations of the current knowledge are:

a) Relative contribution of apoptosis, necrosis, netosis, and other

mechanisms to the cirDNA pool under pathological conditions.

b4444_Ch-07.indd 198 08-Jul-22 6:02:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 199

b) Complete understanding of the cirDNA clearance from circu-

lation, including the relative contribution of DNases, intake by
cells, and liver and kidney metabolism.
c) The diagnostic utility and clinical application of blood DNAse
activity measurement.
d) Why DNase activity of the serum has a different kinetics when
compared to the DNase I positive control in endogenous
DNase activity assay.
e) How to inhibit both bivalent metal ion dependent and inde-
pendent DNases ex vivo in the collection tube for cirDNA
analysis.
f) Exact mechanism by which the blood clot formation introduc-
es genomic DNA into serum.

References
   1. Jung K, Fleischhacker M, Rabien A. (2010) Cell-free DNA in the blood as
a solid tumor biomarker — a critical appraisal of the literature. Clin Chim
Acta 411: 1611–1624.
   2. Chiou HC, Tangco MV, Levine SM, et al. (1994) Enhanced resistance
to nuclease degradation of nucleic acids complexed to asialoglycopro-
tein-polylysine carriers. Nucleic Acids Res 22: 5439–5446.
   3. Cox RA, Gokcen M. (1976) Comparison of serum DNA, native DNA-bind-
ing and deoxyribonuclease levels in ten animal species and man. Life Sci
19: 1609–1614.
   4. Wickstrom E. (1986) Oligodeoxynucleotide stability in subcellular extracts
and culture media. J Biochem Biophys Methods 13: 97–102.
   5. Agrawal S, Temsamani J, Galbraith W, Tang J. (1995) Pharmacokinetics of
antisense oligonucleotides. Clin Pharmacokinet 28: 7–16.
   6. Geary RS, Norris D, Yu R, Bennett CF. (2015) Pharmacokinetics, biodis-
tribution and cell uptake of antisense oligonucleotides. Adv Drug Deliv-
ery Rev 87: 46–51.

b4444_Ch-07.indd 199 08-Jul-22 6:02:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

200 Cell-Free Circulating DNA: Purification and Analysis Techniques

   7. Baranovskii AG, Buneva VN, Nevinsky GA. (2004) Human deoxyribonu-

cleases. Biochem Biokhimiia 69: 587–601.
   8. Peters DL, Pretorius PJ. (2011) Origin, translocation and destination of
extracellular occurring DNA — a new paradigm in genetic behaviour.
Clin Chim Acta 412: 806–811.
   9. van der Vaart M, Pretorius PJ. (2007) The origin of circulating free DNA.
Clin Chem 53: 2215.
10. Cherepanova AV, Tamkovich SN, Bryzgunova OE, et al. (2008) Deoxyri-
bonuclease activity and circulating DNA concentration in blood plasma of
patients with prostate tumors. Ann N Y Acad Sci 1137: 218–221.
11. Tamkovich SN, Cherepanova AV, Kolesnikova EV, et al. (2006) Circulat-
ing DNA and DNase activity in human blood. Ann N Y Acad Sci 1075:
191–196.
12. Napirei M, Karsunky H, Zevnik B, et al. (2000) Features of systemic lupus
erythematosus in Dnase1-deficient mice. Nat Genet 25: 177–181.
13. Yasutomo K, Horiuchi T, Kagami S, et al. (2001) Mutation of DNASE1 in
people with systemic lupus erythematosus. Nat Genet 28: 313–314.
14. Martinez Valle F, Balada E, Ordi-Ros J, Vilardell-Tarres M. (2008) DNase
1 and systemic lupus erythematosus. Autoimmun Rev 7: 359–363.
15. Gaipl US, Beyer TD, Heyder P, et al. (2004) Cooperation between C1q
and DNase I in the clearance of necrotic cell-derived chromatin. Arthritis
Rheum 50: 640–649.
16. Stephan F, Marsman G, Bakker LM, et al. (2014) Cooperation of factor
VII-activating protease and serum DNase I in the release of nucleosomes
from necrotic cells. Arthritis Rheum 66: 686–693.
17. Liang YY, Rainprecht D, Eichmair E, et al. (2015) Serum-dependent pro-
cessing of late apoptotic cells and their immunogenicity. Apoptosis Int J
Program Cell Death 20: 1444–1456.
18. Yu SC, Lee SW, Jiang P, et al. (2013) High-resolution profiling of fetal
DNA clearance from maternal plasma by massively parallel sequencing.
Clin Chem 59: 1228–1237.
19. Lo YM, Zhang J, Leung TN, et al. (1999) Rapid clearance of fetal DNA
from maternal plasma. Am J Hum Genet 64: 218–224.
20. Botezatu I, Serdyuk O, Potapova G, et al. (2000) Genetic analysis of DNA
excreted in urine: a new approach for detecting specific genomic DNA
sequences from cells dying in an organism. Clin Chem 46: 1078–1084.

b4444_Ch-07.indd 200 08-Jul-22 6:02:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 201

21. Diehl F, Schmidt K, Choti MA, et al. (2008) Circulating mutant DNA to
assess tumor dynamics. Nat Med 14: 985–990.
22. Snyder MW, Kircher M, Hill AJ, et al. (2016) Cell-free DNA comprises an
in vivo nucleosome footprint that informs its tissues-of-origin. Cell 164:
57–68.
23. Fleischhacker M, Schmidt B. (2007) Circulating nucleic acids (CNAs) and
cancer — a survey. Biochimica et Biophysica Acta 1775: 181–232.
24. Giacona MB, Ruben GC, Iczkowski KA, et al. (1998) Cell-free DNA in
human blood plasma: length measurements in patients with pancreatic
cancer and healthy controls. Pancreas 17: 89–97.
25. Wyllie AH, Morris RG, Smith AL, Dunlop D. (1984) Chromatin cleav-
age in apoptosis: association with condensed chromatin morphology and
dependence on macromolecular synthesis. J Pathol 142: 67–77.
26. Opferman JT. (2007) Life and death during hematopoietic differentiation.
Curr Opin Immunol 19: 497–502.
27. Stroun M, Lyautey J, Lederrey C, et al. (2001) About the possible origin
and mechanism of circulating DNA apoptosis and active DNA release.
Clin Chim Acta 313: 139–142.
28. Jahr S, Hentze H, Englisch S, et al. (2001) DNA fragments in the blood
plasma of cancer patients: quantitations and evidence for their origin from
apoptotic and necrotic cells. Cancer Res 61: 1659–1665.
29. Yipp BG, Kubes P. (2013) NETosis: how vital is it? Blood 122: 2784–2794.
30. Manda A, Pruchniak MP, Arazna M, Demkow UA. (2014) Neutrophil
extracellular traps in physiology and pathology. Cent Eur J Immunol 39:
116–121.
31. Jiang P, Lo YM. (2016) The long and short of circulating cell-free DNA and
the ins and outs of molecular diagnostics. Trends Genet 32(6): 360–371.
32. Chandrananda D, Thorne NP, Bahlo M. (2015) High-resolution charac-
terization of sequence signatures due to non-random cleavage of cell-free
DNA. BMC Med Genom 8: 29.
33. Lo YM, Chan KC, Sun H, et al. (2010) Maternal plasma DNA sequencing
reveals the genome-wide genetic and mutational profile of the fetus. Sci
Transl Med 2: 61ra91.
34. Cheng SH, Jiang P, Sun K, et al. (2015) Noninvasive prenatal testing by
nanopore sequencing of maternal plasma DNA: feasibility assessment.
Clin Chem 61: 1305–1306.

b4444_Ch-07.indd 201 08-Jul-22 6:02:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

202 Cell-Free Circulating DNA: Purification and Analysis Techniques

35. Solovyan VT, Bezvenyuk ZA, Salminen A, et al. (2002) The role of
topoisomerase II in the excision of DNA loop domains during apoptosis.
J Biol Chem 277: 21458–21467.
36. Susin SA, Lorenzo HK, Zamzami N, et al. (1999) Molecular characteriza-
tion of mitochondrial apoptosis-inducing factor. Nature 397: 441–446.
37. Widlak P. (2000) The DFF40/CAD endonuclease and its role in apoptosis.
Acta Biochim Pol 47: 1037–1044.
38. Liu X, Zou H, Slaughter C, Wang X. (1997) DFF, a heterodimeric protein
that functions downstream of caspase-3 to trigger DNA fragmentation
during apoptosis. Cell 89: 175–184.
39. Enari M, Sakahira H, Yokoyama H, et al. (1998) A caspase-activated
DNase that degrades DNA during apoptosis, and its inhibitor ICAD.
Nature 391: 43–50.
40. Li LY, Luo X, Wang X. (2001) Endonuclease G is an apoptotic DNase
when released from mitochondria. Nature 412: 95–99.
41. Oliveri M, Daga A, Cantoni C, et al. (2001) DNase I mediates internucle-
osomal DNA degradation in human cells undergoing drug-induced apop-
tosis. Eur J Immunol 31: 743–751.
42. Elmore S. (2007) Apoptosis: a review of programmed cell death. Toxicol
Pathol 35: 495–516.
43. Kawane K, Nagata S. (2008) Nucleases in programmed cell death. Meth-
ods Enzymol 442: 271–287.
44. Bischoff FZ, Lewis DE, Simpson JL. (2005) Cell-free fetal DNA in mater-
nal blood: kinetics, source and structure. Hum Reprod Update 11: 59–67.
45. Hariton-Gazal E, Rosenbluh J, Graessmann A, et al. (2003) Direct trans-
location of histone molecules across cell membranes. J Cell Sci 116:
4577–4586.
46. Luger K. (2003) Structure and dynamic behavior of nucleosomes. Curr
Opin Genet Dev 13: 127–135.
47. Westheimer FH. (1987) Why nature chose phosphates. Science 235:
1173–1178.
48. Nadano D, Yasuda T, Kishi K. (1993) Measurement of deoxyribonuclease
I activity in human tissues and body fluids by a single radial enzyme-diffu-
sion method. Clin Chem 39: 448–452.

b4444_Ch-07.indd 202 08-Jul-22 6:02:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 203

49. Gueroult M, Picot D, Abi-Ghanem J, et al. (2010) How cations can assist
DNase I in DNA binding and hydrolysis. PLoS Comput Biol 6: e1001000.
50. Moore S. (1981) Pancreatic DNase. In: PD Boyer (eds), The Enzymes,
Academic Press, New York, pp. 281–296.
51. Drew HR. (1984) Structural specificities of five commonly used DNA
nucleases. J Mol Biol 176: 535–557.
52. Weston SA, Lahm A, Suck D. (1992) X-ray structure of the DNase I-d(G-
GTATACC)2 complex at 2.3 A resolution. J Mol Biol 226: 1237–1256.
53. Lahm A, Suck D. (1991) DNase I-induced DNA conformation. 2 A struc-
ture of a DNase I-octamer complex. J Mol Biol 222: 645–667.
54. Riley DE. (1980) Deoxyribonuclease I generates single-stranded gaps in
chromatin deoxyribonucleic acid. Biochemistry 19: 2977–2992.
55. Noll M. (1974) Internal structure of the chromatin subunit. Nucleic Acids
Res 1: 1573–1578.
56. Cousins DJ, Islam SA, Sanderson MR, et al. (2004) Redefinition of the
cleavage sites of DNase I on the nucleosome core particle. J Mol Biol 335:
1199–1211.
57. Lazarus RA, Wagener JA. (2013) Recombinant Human Deoxyribonuclease
I. Pharmaceutical Biotechnology. Springer, New York, pp. 321–336.
58. Cherepanova A, Tamkovich S, Pyshnyi D, et al. (2007) Immunochemical
assay for deoxyribonuclease activity in body fluids. J Immunol Methods
325: 96–103.
59. Sierakowska H, Shugar D. (1977) Mammalian nucleolytic enzymes. Prog
Nucl Acid Res Mol Biol 20: 59–130.
60. Evans CJ, Aguilera RJ. (2003) DNase II: genes, enzymes and function.
Gene 322: 1–15.
61. Murai K, Yamanaka M, Akagi K, Anai M. (1980) Purification and
properties of deoxyribonuclease II from human urine. J Biochem 87:
1097–1103.
62. Baker KP, Baron WF, Henzel WJ, Spencer SA. (1998) Molecular cloning
and characterization of human and murine DNase II. Gene 215: 281–289.
63. Gross SC, Watabe M, Goodarzi G, et al. (1990) Organ-specific distribu-
tion of isozymes of 5´-nucleotide phosphodiesterase in mouse. Comp Bio-
chem Physiol B Comp Biochem 95: 821–824.

b4444_Ch-07.indd 203 08-Jul-22 6:02:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

204 Cell-Free Circulating DNA: Purification and Analysis Techniques

64. Luthje J, Ogilvie A. (1987) 5′-Nucleotide phosphodiesterase isoenzymes

in human serum: quantitative measurement and some biochemical prop-
erties. Clin Chim Acta 164: 275–284.
65. Eder PS, DeVine RJ, Dagle JM, Walder JA. (1991) Substrate specificity
and kinetics of degradation of antisense oligonucleotides by a 3´ exonucle-
ase in plasma. Antisense Res Dev 1: 141–151.
66. Ito K, Yamamoto T, Minamiura N. (1987) Phosphodiesterase I in human
urine: purification and characterization of the enzyme. J Biochem 102:
359–367.
67. Goding JW, Terkeltaub R, Maurice M, et al. (1998) Ecto-phosphodies-
terase/pyrophosphatase of lymphocytes and non-lymphoid cells: structure
and function of the PC-1 family. Immunol Rev 161: 11–26.
68. Hynie I, Meuffels M, Poznanski WJ. (1975) Determination of phosphodi-
esterase I activity in human blood serum. Clin Chem 21: 1383–1387.
69. Frittitta L, Camastra S, Baratta R, et al. (1999) A soluble PC-1 circu-
lates in human plasma: relationship with insulin resistance and associated
abnormalities. J Clin Endocrinol Metabol 84: 3620–3625.
70. Belli SI, Goding JW. (1994) Biochemical characterization of human PC-1,
an enzyme possessing alkaline phosphodiesterase I and nucleotide pyro-
phosphatase activities. Eur J Biochem / FEBS 226: 433–443.
71. Shuster AM, Gololobov GV, Kvashuk OA, et al. (1992) DNA hydrolyzing
autoantibodies. Science 256: 665–667.
72. Baranovskii AG, Ershova NA, Buneva VN, et al. (2001) Catalytic het-
erogeneity of polyclonal DNA-hydrolyzing antibodies from the sera of
patients with multiple sclerosis. Immunol Lett 76: 163–167.
73. Cherepanova AV, Tamkovich SN, Vlasov VV, Laktionov PP. (2007) Blood
deoxyribonuclease activity in health and diseases. Biomeditsinskaia
khimiia 53: 488–496.
74. Babina SE, Kanyshkova TG, Buneva VN, Nevinsky GA. (2004) Lactofer-
rin is the major deoxyribonuclease of human milk. Biochem Biokhimiia
69: 1006–1015.
75. Masson PL, Heremans JF, Dive CH. (1966) An iron-binding protein com-
mon to many external secretions. Clin Chim Acta 14: 5.
76. Masson PL, Heremans JF, Schonne E. (1969) Lactoferrin, an iron-bind-
ing protein in neutrophilic leukocytes. J Exper Med 130: 643–658.

b4444_Ch-07.indd 204 08-Jul-22 6:02:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 205

77. Bennett RM, Kokocinski T. (1979) Lactoferrin turnover in man. Clin Sci
57: 453–460.
78. Barthe C, Galabert C, Guy-Crotte O, Figarella C. (1989) Plasma and
serum lactoferrin levels in cystic fibrosis. Relationship with the presence
of cystic fibrosis protein. Clin Chim Acta 181: 183–188.
79. Lee TH, Montalvo L, Chrebtow V, Busch MP. (2001) Quantitation of
genomic DNA in plasma and serum samples: higher concentrations
of genomic DNA found in serum than in plasma. Transfusion 41:
276–282.
80. Grandinetti G, Smith AE, Reineke TM. (2012) Membrane and nuclear
permeabilization by polymeric pDNA vehicles: efficient method for gene
delivery or mechanism of cytotoxicity? Mol Pharm 9: 523–538.
81. Valenzuela SM, Martin DK, Por SB, et al. (1997) Molecular cloning
and expression of a chloride ion channel of cell nuclei. J Biol Chem 272:
12575–12582.
82. Widlak P, Garrard WT. (2006) Unique features of the apoptotic endonu-
clease DFF40/CAD relative to micrococcal nuclease as a structural probe
for chromatin. Biochem Cell Biol 84: 405–410.
83. Widlak P, Li P, Wang X, Garrard WT. (2000) Cleavage preferences of the
apoptotic endonuclease DFF40 (caspase-activated DNase or nuclease) on
naked DNA and chromatin substrates. J Biol Chem 275: 8226–8232.
84. Allan J, Fraser RM, Owen-Hughes T, Keszenman-Pereyra D. (2012)
Micrococcal nuclease does not substantially bias nucleosome mapping.
J Mol Biol 417: 152–164.
85. Woo EJ, Kim YG, Kim MS, et al. (2004) Structural mechanism for inacti-
vation and activation of CAD/DFF40 in the apoptotic pathway. Mol Cell
14: 531–539.
86. Counis MF, Torriglia A. (2006) Acid DNases and their interest among
apoptotic endonucleases. Biochimie 88: 1851–1858.
87. Samejima K, Earnshaw WC. (2005) Trashing the genome: the role of
nucleases during apoptosis. Nat Rev Mol Cell Biol 6: 677–688.
88. Widlak P, Li LY, Wang X, Garrard WT. (2001) Action of recombi-
nant human apoptotic endonuclease G on naked DNA and chromatin
substrates: cooperation with exonuclease and DNase I. J Biol Chem 276:
48404–48409.

b4444_Ch-07.indd 205 08-Jul-22 6:02:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

206 Cell-Free Circulating DNA: Purification and Analysis Techniques

89. Lutter LC. (1979) Precise location of DNase I cutting sites in the
nucleosome core determined by high resolution gel electrophoresis.
Nucleic Acids Res 6: 41–56.
90. Mizuta R, Mizuta M, Araki S, et al. (2009) DNase gamma-dependent and
-independent apoptotic DNA fragmentations in Ramos Burkitt’s lym-
phoma cell line. Biomed Res 30: 165–170.
91. Mizuta R, Mizuta M, Araki S, et al. (2006) Action of apoptotic endonu-
clease DNase gamma on naked DNA and chromatin substrates. Biochem
Biophy Res Commun 345: 560–567.
92. Torriglia A, Perani P, Brossas JY, et al. (1998) L-DNase II, a molecule that
links proteases and endonucleases in apoptosis, derives from the ubiqui-
tous serpin leukocyte elastase inhibitor. Mol Cell Biol 18: 3612–3619.
93. Torriglia A, Martin E, Jaadane I. (2017) The hidden side of SERPINB1/
Leukocyte Elastase Inhibitor. Semin Cell Develop Biol 62: 178–186.
94. Scully C, Spandidos DA, Ward Booth P, et al. (1981) Serum alkaline
deoxyribonuclease in oral cancer and premalignant lesions. Biomedicine /
[publiee pour l’AAICIG] 35: 179–180.
95. Economidou-Karaoglou A, Lans M, Taper HS, et al. (1988) Variations in
serum alkaline DNase activity. A new means for therapeutic monitoring of
malignant lymphomas. Cancer 61: 1838–1843.
96. Spandidos DA, Ramandanis G, Garas J, Kottaridis SD. (1980) Serum
deoxyribonucleases in patients with breast cancer. Eur J Cancer 16:
1615–1619.
97. Arakawa K, Kawai Y, Kumamoto T, et al. (2005) Serum deoxyribonuclease
I activity can be used as a sensitive marker for detection of transient myo-
cardial ischaemia induced by percutaneous coronary intervention. Eur
Heart J 26: 2375–2380.
98. Morikawa N, Kawai Y, Arakawa K, et al. (2007) Serum deoxyribonuclease
I activity can be used as a novel marker of transient myocardial ischaemia:
results in vasospastic angina pectoris induced by provocation test. Eur
Heart J 28: 2992–2997.
99. Funakoshi A, Wakasugi H, Ibayashi H. (1979) Clinical investigation of
serum deoxyribonuclease: II. Clinical studies of serum deoxyribonuclease
activity in pancreatic disease. Gastroenterol Jpn 14: 436–440.

b4444_Ch-07.indd 206 08-Jul-22 6:02:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Blood Nucleases Affecting cirDNA in Serum and Plasma 207

100. Kuribara J, Tada H, Kawai Y, et al. (2009) Levels of serum deoxyribonuclease

I activity on admission in patients with acute myocardial infarction can
be useful in predicting left ventricular enlargement due to remodeling.
J Cardiol 53: 196–203.
101. Lundqvist EH, Sjovall K, Eneroth PH. (1992) Influences of diet and surgical
trauma on serum alkaline DNase activity levels. Clin Chim Acta 205: 43–49.
102. Wang H, Sha LL, Ma TT, et al. (2016) Circulating level of neutrophil
extracellular traps is not a useful biomarker for assessing disease activity
in antineutrophil cytoplasmic antibody-associated vasculitis. PloS One 11:
e0148197.
103. Zhu B, Gong Y, Chen P, et al. (2014) Increased DNase I activity in dia-
betes might be associated with injury of pancreas. Mol Cell Biochem 393:
23–32.
104. Vancevska A, Nikolic A, Bonaci-Nikolic B, et al. (2016) Assessment of
deoxyribonuclease activity in serum samples of patients with systemic
lupus erythematosus: fluorescence-based method versus ELISA. J Clin
Lab Anal 30(6): 797–803.
105. Skiljevic D, Jeremic I, Nikolic M, et al. (2013) Serum DNase I activity in
systemic lupus erythematosus: correlation with immunoserological mark-
ers, the disease activity and organ involvement. Clin Chem Lab Med 51:
1083–1091.
106. Velders M, Treff G, Machus K, et al. (2014) Exercise is a potent stimulus
for enhancing circulating DNase activity. Clin Biochem 47: 471–474.
107. Sato S, Takenaka S. (2014) Highly sensitive nuclease assays based on
chemically modified DNA or RNA. Sensors 14: 12437–12450.
108. Choi SJ, Szoka FC. (2000) Fluorometric determination of deoxyribonu-
clease I activity with PicoGreen. Anal Biochem 281: 95–97.
109. Macanovic M, Lachmann PJ. (1997) Measurement of deoxyribonuclease I
(DNase) in the serum and urine of systemic lupus erythematosus (SLE)-
prone NZB/NZW mice by a new radial enzyme diffusion assay. Clin Exper
Immunol 108: 220–226.
110. Trubetskoy VS, Hagstrom JE, Budker VG. (2002) Self-quenched covalent
fluorescent dye-nucleic acid conjugates as polymeric substrates for enzy-
matic nuclease assays. Anal Biochem 300: 22–26.

b4444_Ch-07.indd 207 08-Jul-22 6:02:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

208 Cell-Free Circulating DNA: Purification and Analysis Techniques

111. Sonawane S, Khanolkar V, Namavari A, et al. (2012) Ocular surface

extracellular DNA and nuclease activity imbalance: a new paradigm
for inflammation in dry eye disease. Invest Ophthalmol Visual Sci 53:
8253–8263.
112. Kiedrowski MR, Kavanaugh JS, Malone CL, et al. (2011) Nuclease mod-
ulates biofilm formation in community-associated methicillin-resistant
Staphylococcus aureus. PloS One 6: e26714.
113. Barra GB, Santa Rita TH, de Almeida Vasques J, et al. (2015) EDTA-me-
diated inhibition of DNases protects circulating cell-free DNA from ex
vivo degradation in blood samples. Clin Biochem 48: 976–981.
114. Barra GB, Silva VM, Jácomo RH, et al. (2014) Sodium citrate at 8% is
equivalent to EDTA as anticoagulant of choice for circulating cell-free
DNA analysis: low contamination by blood cells genomic DNA and inhi-
bition of blood nuclease activity. Clin Chem 60: S190.
115. El Messaoudi S, Rolet F, Mouliere F, Thierry AR. (2013) Circulating cell
free DNA: preanalytical considerations. Clin Chim Acta 424: 222–230.
116. Mann KG, Whelihan MF, Butenas S, Orfeo T. (2007) Citrate anticoag-
ulation and the dynamics of thrombin generation. J Thromb Haemost 5:
2055–2061.
117. Lee G, Arepally GM. (2012) Anticoagulation techniques in apheresis:
from heparin to citrate and beyond. J Clin Apheresis 27: 117–125.
118. Gray E, Hogwood J, Mulloy B. (2012) The anticoagulant and antithrom-
botic mechanisms of heparin. Handb Exper Pharmacol (207): 43–61.
119. Adachi I, Iwaki H, Adachi H, et al. (1986) Heparin-induced leukocyte
lysis in vitro. J Pharmacobio Dyn 9: 207–210.
120. Santa Rita TH, Jácomo RH, Abdalla LF, et al. (2015) Enrichment of
genomic DNA amount in serum by transport and storage at ambient tem-
perature makes it an alternative matrix for molecular assays: high-yield of
DNA, automated DNA extraction- friendly and direct use as template in
qPCR. Clin Chem 61: S176.
121. Clinical and Laboratory Standards Institute. Tubes and Additives for
Venous and Capillary Blood Specimen Collection: Approved Standard,
6th ed. Clinical and Laboratory Standards Institute, 2010.

b4444_Ch-07.indd 208 08-Jul-22 6:02:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0008

Chapter 8
Isolating Circulating Exosomes
as Biomarkers: Challenges and
Opportunities
Alexander Semaan and Anirban Maitra

Abstract

In addition to free extracellular DNA, blood plasma also contains nucleic

acids encased in vesicles with lipid bilayer membranes. These can be
divided into exosomes and extracellular vesicles, and although the two
types overlap in some markers and physical properties, they are gener-
ated by different mechanisms. Exosomes can be actively secreted from
cells, including cancer cells, and the stability of their cargo makes them
a promising source of biomarker targets. In this chapter, we describe the
origins, isolation strategies, nucleic acid content, and biomarker poten-
tial of exosomes.

Extracellular Vesicle Properties and Nomenclature

Extracellular vesicles (EV) are a heterogeneous class of cell-
derived membranous particles with a storied history. Initially,
these membranous particles were described as “garbage bins” of

209

b4444_Ch-08.indd 209 08-Jul-22 6:05:49 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

210 Cell-Free Circulating DNA: Purification and Analysis Techniques

living or dying cells and received limited attention.1 In the early

1980s two groups independently reported the discovery of what
today is termed “exosomes”, and since then, the interest in these
vesicles has increased significantly.2,3
There is much confusion and even contradictory definitions
of exosomes and other vesicles within the literature, which is due
to the heterogeneity of the EV population.4 It is important to stress
the presence of two distinct EV types that differ by origin, biolog-
ical and pathological function, and size. Although the presented
classification is an oversimplification, it reflects the current state of
knowledge and is based on the recommendations of the Interna-
tional Society for Extracellular Vesicles (ISEV). With that said, it
is possible that the current stated classification system will change
or expand in the future based on ongoing studies.

1. Exosomes (size: 50–150 nm, buoyant density: 1.11–1.19 g/mL),

also known as nanovesicles or dexosomes. The term “exo-
some” was first introduced by Johnstone et al. in 1987 with the
description of intraluminal vesicles (ILV).1 Biogenesis of these
particles is through inward budding of the endomembrane
system into the multivesicular endosome (MVE) by fission of
cytosol. After additional cargo loading and maturation, they are
released into the extracellular space upon fusion of MVEs with
the cellular lipid membrane and termed exosomes.2,3 Exosomes
have crucial functions in multiple physiological and pathologi-
cal processes including cell–cell communication, inflammation,
stem cell expansion, and tumorigenesis,5 and are detectable
in nearly all body fluids including blood, saliva, urine, breast
milk, amniotic fluid, and ascites. Importantly, the exosomal
cargo composition varies greatly and includes proteins, liquids,
and nucleic acids. The diversity of exosomal content and its

b4444_Ch-08.indd 210 08-Jul-22 6:05:49 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 211

modification based on the conditions of the cell of origin points

to a highly selective loading of exosomal cargo depending on
the physiological or pathological state.6
2. EV (size: 50 nm up to 1000 nm, buoyant density: undefined),
also known as apoptotic bodies, microparticles, microvesicles,
blebbing vesicles, and oncosomes. In contrast to the produc-
tion of exosomes within the endosomal system, microvesicles
are formed by ectocytosis. This process involves the outward
budding and fission of the cellular membrane to form secret-
ed particles.7 Currently there is no clear subclassification of
EV, which is mainly due to the fact that nobody knows how
many functionally distinct subtypes exist. There are substantial
efforts ongoing to close this gap in knowledge and characterize
the heterogenous nature of EVs.8,9

Although of distinct origin, both vesicle types share com-

mon intracellular mechanisms and sorting machineries, lead-
ing to significant overlap in membrane bound components and
cargo. As has been confirmed by reports investigating overlapping
subgroups of vesicles after isolation, the two populations are not
completely distinguishable by ultracentrifugation and density gra-
dient isolation techniques based on their physical properties (size,
density).10 Additionally, many details in key biogenesis functions
are still waiting to be revealed. This chapter will focus on the role,
cargo, and isolation methods of exosomes.

Exosomal Function and Biomarker Potential

Once discounted as garbage bins of cells, exosomes have experi-
enced a renaissance over the last decades. Research into exosomes
and other vesicles has gained traction as we begin to under-
stand how they play a crucial role in cell–cell communication to

b4444_Ch-08.indd 211 08-Jul-22 6:05:49 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

212 Cell-Free Circulating DNA: Purification and Analysis Techniques

orchestrate multiple physiological functions, including immune

signaling, angiogenesis, aging, differentiation, and proliferation.11–13
In particular, their ability to transfer genetically encoded messages
has garnered attention, because for decades, intercellular commu-
nication was thought to be only possible in a paracrine, endocrine,
and neuronal manner or by direct contact between cells.
Given their role in cell–cell communication, the presence of
an “exosomal fingerprint” as related to disease-specific cargo may
be used as biomarker for detection, surveillance, prognosis, and
therapy guidance in multiple cancer types.14 Exosomes, together
with circulating tumor cells (CTC) and circulating cell-free DNA
(cirDNA), are part of a new approach to tumor characterization
known as “liquid biopsies.”15 In comparison to invasive tissue biop-
sies, blood draws can be used to help monitor treatment response,
drug resistance, predict recurrence, and also serve as auxiliary stag-
ing and diagnostic markers.16,17 EV and exosomes contain a myriad
of tumor-derived molecules that reflect, in real time, the current
state of the releasing cell. In contrast to fragmented cirDNA and
CTCs, accumulating evidence suggests exosomal cargo is superior
as a clinical biomarker in a variety of diseases.18–20 This may be due
to the cargo’s protection by the lipid bilayer, as well as a higher
concentration of nucleic acid copies due to selective exosomal
loading.17
In the context of clinical utility, a urine exosome gene expres-
sion assay together with standard-of-care (SOC) tests including
PSA levels, age, race, and family history was able to reliably distin-
guish ≥7 Gleason from low-grade prostate tumors or benign disease.
These results suggest the potential for auxiliary exosomal disease
staging to help guide personalized therapies and protect patients
from harmful over- or undertreatment.21 Another example of the
use of exosomes in treatment monitoring was shown recently by

b4444_Ch-08.indd 212 08-Jul-22 6:05:49 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 213

Bernard et al. The authors demonstrated in a longitudinal cohort

of over 36 patients with pancreatic ductal adenocarcinoma that
an increase in mutational KRAS burden within exosomes, but not
ctDNA KRAS level, during therapy is associated with disease pro-
gression.20 Especially promising is the potential identification of
targetable mutations within liquid biopsies for patient stratifica-
tion in the context of personalized medicine. Herein, research-
ers have been able to detect clinically actionable BRAF V600E
mutations in patients,22 as well as changes in the exosomal cargo
from melanoma cell lines and patient-derived xenografts after
treatment with the BRAF inhibitor Vemurafenib.23 Additionally,
EGFR L858R and T790M mutations from lung cancer–derived
exosomes, which represent susceptible genotypes to tyrosine
kinase inhibitors, have been found with higher sensitivity within
exosomes compared to cirDNA.24
Exosomes have also been shown to provide a scaffold for
enrichment of tumor-specific material in circulation by exploiting
the protein surfaceome of these vesicles. Enrichment of exosomes
leads to greater sensitivity of mutant detection, facilitating molec-
ular profiling of emerging mechanisms of resistance.25 The abil-
ity to modify and engineer exosomal cargo is also currently being
evaluated as an immune-inert treatment vehicle for targeted drug
delivery and immunotherapeutic approach.15,16
Following Peter Paget’s “soil and seed” theory, exosomes have
also been described as mediators of the so-called “pre-metastatic
niche” formation. Herein, these vesicles have been shown to trans-
port essential tumor-secreted factors to prime stromal tissue at
distant organ sites for the outgrowth of CTC.26
It is also important to note that exosomes have major roles
outside of tumorigenesis, including neurodegenerative and cardio-
vascular diseases.27 Exosomes secreted physiologically by certain

b4444_Ch-08.indd 213 08-Jul-22 6:05:49 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

214 Cell-Free Circulating DNA: Purification and Analysis Techniques

cell populations might harbor important immunomodulatory

and cytoprotective properties derived from their releasing cells.
For example, microvesicles released by mesenchymal stem cells
(MSCs) have been shown to have a function in tissue repair and
are currently being evaluated for their role in regenerative medi-
cine.28 Particularly in myocardial infarction (MI), there is increas-
ing evidence of a possible protective role of purified exosomes, for
example, embryonic stem cell–derived MSC exosomes in animal
models of MI.29,30 Many other pathophysiological mechanisms,
like deposition of Tau protein in early Alzheimer disease, have
been thought to occur via exosomal release, although no effective
treatment strategy targeting this deposition has been proposed.31,32
In contrast, memory and educational effects of exosomes in
immune cells have been already used to prime organisms for
future infections,32 which is a promising strategy to protect organ-
isms from multidrug-resistant bacteria infections.

Exosomal Biogenesis and Regulation

Regulation of exosomal biogenesis and cargo loading is a complex
and not yet fully understood process that involves multiple distinct
pathways.33 It should be noted that though EV and exosomes orig-
inate from different parts of the cell, they share some common
intracellular mechanisms.34 Exosomal biogenesis may be split into
three separate steps: (1) formation of ILV within MVE; (2) further
maturation and transport to the cell membrane, and (3) fusion of
MVEs and release of exosomes (Fig. 1). It remains unclear if cells
form a single, or multiple types, MVE. Some research points to the
fact that only certain MVE subpopulations have the ability to fuse
with the cell membrane, while others are directed to lysosomes.35
An important step in the formation of ILV comprises the clus-
tering of lipids and membrane-associated proteins in membrane

b4444_Ch-08.indd 214 08-Jul-22 6:05:49 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 215

Fig. 1: Pathways of exosomal biogenesis. ER — endoplasmic reticulum;

MVB — multivesicular endosome.

microdomains. This process can happen in an Endosomal

Sorting Complex Required for Transport (ESCRT)-dependent or
ESCRT-independent manner.36,37 The former requires a cascade
of ESCRT complexes to aggregate transmembrane cargo followed
by inward budding and fission of early exosomes or ILVs. Detailed
knockout studies targeting 23 ESCRT-related proteins revealed
a complex synergy within the ESCRT machinery, which consists
of four major complexes associated with AAA ATPase Vps4 com-
plex.38 Apart from ILV formation, the ESCRT has also been shown
to control sorting of ubiquitinated proteins into ILVs. Details on
the regulation of exosome loading, ESCRT machinery, and their
related proteins, such syntenin and the syndecans, have been pre-
viously published.39–42

b4444_Ch-08.indd 215 08-Jul-22 6:05:51 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

216 Cell-Free Circulating DNA: Purification and Analysis Techniques

ESCRT-independent exosome formation on the other hand

requires either the formation of ceramides by hydrolyzation of
sphingomyelin43 or a process that involves proteins of the tetra-
spanin family (e.g., CD63, CD9, and CD81) that are abundantly
expressed on exosomes.44 Exosome formation may also result from
co-sorting with certain protein classes such as chaperones includ-
ing heat shock 70 kDa protein (HSP70 and HSP71)45 or lipid rafts
harboring glycosylphosphatidylinositol (GPI)-anchored proteins.46
In summary, exosomal biogenesis is a complex, cell, and
cargo-specific process influenced by multiple physiological, devel-
opmental, and pathological processes.

Exosomal Cargo
Protected by a characteristic bilipid layer, exosomal cargo is
shielded from degradation in the extracellular space and blood
stream. Exosomes may contain a large variety of different mole-
cules, often mirroring the current state of their releasing cell.35,36
The cargo is either processed from within the cells in the Golgi
apparatus or upon internalization from the plasma membrane,
finally leading to the formation of ILV. The exosomal content
comprises membrane bound and soluble proteins, lipids, carbohy-
drates, and nucleic acids47 (Fig. 2). The following section focuses
on exosomal nucleic acids; details on exosomal protein content
have been published previously.48

Nucleic Acids in Exosomes

Nucleic acids in exosomes include transcriptomic and intact
mRNAs, mRNA fragments, noncoding RNA (lnRNA, snRNA,
siRNA, piRNA, ncRNA, and miRNA) as well as single- or
double-stranded DNA.11,39 In contrast to cellular RNA, exosomal

b4444_Ch-08.indd 216 08-Jul-22 6:05:51 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 217

Fig. 2: Exosomal cargo includes protein content, nucleic acid content, and
lipid content.

RNA (exoRNA) is mostly devoid of the large and small ribosomal

RNA subunits, which are often used for validation in cellular RNA
assays.49 In general, RNA cargo differs based on the cells of ori-
gin and their homeostatic state. In particular, cell-specific exoso-
mal microRNA has been characterized as “fingerprints,” where
shifts in composition can occur in cases of disease, infection, and
cancer.26,50–53 Besides the aforementioned pathologic shifts, there
have also been reports of gender-specific expression patterns for
microRNA in urine-derived exosomes.54 In general, exoRNA is
more fragmented (typically <700 nucleotides) than intracellular

b4444_Ch-08.indd 217 08-Jul-22 6:05:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

218 Cell-Free Circulating DNA: Purification and Analysis Techniques

RNA (400–12,000 nucleotides).55 The exact cause of this is not

known, but it may be related to the RNA interference machin-
ery within or after the formation of exosomes, RNA degenera-
tion within the exosomes, or artificial changes upon isolation and
downstream analysis.
In comparison to other extracellular nucleic acids, such as
cirDNA, that is present as fragmented DNA in circulation (~170
base pairs), exosomal DNA is preserved as significantly longer
lengths. Even high–molecular weight DNA has been found in
plasma-derived exosomes, representing the entire genome, and
mirrors the mutational status of parental cells.22,56 There have also
been reports of high-quality, double-stranded genomic DNA in
exosomes isolated from cell lines and human samples by ultracen-
trifugation.57 Based on these findings, exosomal DNA and RNA
harbor immense potential as biomarkers in many diseases.

Selective RNA Loading into Exosomes

The specific exosomal cargo content unique to certain cell types
suggests an active RNA loading and selection process based on
physiologic conditions.58,59 Several mechanisms have been pro-
posed for this selective loading. Exosomal microRNA expresses a
distinct EXOmotif (GGAG tetranucleotide) and harbors a consen-
sus sequence within the 3’UTR end (25nt sequence, which con-
tains a short CTGCC core domain on a stem-loop structure and
carries a miR-1289-binding site). Both of these motifs may act as
“zip codes” for intracellular loading into exosomes.60,61 Addition-
ally, different machineries have been implicated in exosomal RNA
sorting including the ESCRT-II complex, tetraspanin-enriched
microdomains, ceramides, protein argonaute-2 (AGO2) and
microRNA-induced silencing complex (mRISC).62,63 Finally, it is

b4444_Ch-08.indd 218 08-Jul-22 6:05:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 219

also important to note that cellular RNA levels and their availability
will directly affect exosomal RNA loading.64

Measurement of Exosomal RNA Quantity and Quality

As RNA is intrinsically unstable, particularly in the presence of
external factors (e.g., heat, mechanical forces, UV light, and
RNases), and since exosomal RNA in itself is already largely frag-
mented, it is important to follow stringent laboratory techniques
when conducting RNA-based research. This includes use of
RNase-free working areas and reagents (chemical or mechanical
inactivation of RNase prior to use), and working on ice and in a
time-efficient manner for sample processing. To effectively quan-
tify the low concentration of exosomal nucleic acids, it is necessary
to rule out extravesicular contamination. This involves treating
isolated exosomes with RNases before exosomal lysis in order to
degrade extra-exosomal soluble or protein-bound nucleic acids.
While extracellular RNA will be degraded, intravesicle cargo is
protected by the lipid bilayer. In addition to the removal of any
extravesicular RNAs, RNase treatment also helps to resolve exo-
somal aggregation and increase yield.65 In the case of potential
protein contamination (e.g., cell lines cultured with protein-rich
media), addition of proteinase-K before exosomal lysis releases
any protein-bound RNA and reduces protein and protein-bound
RNA contamination.66
Quantification of exosomal nucleic acid levels remains chal-
lenging, but with increasing sensitivity of commercial detection
systems, it is becoming more manageable. Traditional spectropho-
tometer devices like Nanodrop are typically insufficiently sensitive
to reliably quantify exoRNA concentrations. Fluorometer devices
like Qubit have a lower limit of detection for nucleic acids (down

b4444_Ch-08.indd 219 08-Jul-22 6:05:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

220 Cell-Free Circulating DNA: Purification and Analysis Techniques

to 0.25 ng RNA/μl) and are able to distinguish between RNA and

DNA, thus also allowing digital elimination of DNA contamina-
tion. Another option for exoRNA quantification involves utilizing
a Bioanalyzer Pico Chip, which is able to detect concentrations
down to 50 pg/μL. Importantly, the Bioanalyzer provides an
electrophoresis-like virtual RNA size profile, which also helps to
visually confirm the presence of different-sized nucleic acids.67
It has to be noted that quality control of RNA on the Bioanalyzer
is often based on the large ribosomal subunits (18S and 28S) that
are often nonexistent in exoRNA samples. As the quality of RNA
cannot be measured, some reports suggest that the system may
be error-prone based on vulnerability to salt concentration in
the sample, ladder variability, and co-detection of contaminating
DNA.68 Among the most sensitive methods, real-time PCR has the
ability to detect quantities of RNA down to 1 fg. Though incredi-
bly precise, one has to bear in mind that this technique is sensitive
to DNA contamination. Detailed comparison of different RNA
quantification methods has been previously described.69
Depending on the intended downstream analysis, several
issues must be kept in mind to minimize potential bias. Sources
of bias that especially affect high throughput techniques are sum-
marized in the following. First, the extraction methodology may
impact the final RNA composition. For example, TRIzol tends
to concentrate strong anti-GC content especially in small RNA
amounts, as highly structured small RNAs or those with low GC
content are less efficient in interacting with carrier.70 Also, differ-
ent NGS library preparation kits might affect exoRNA sequencing
results by variation in adaptor ligation or hybridization,71 although
novel strategies have been developed to tackle this obstacle. Addi-
tionally, some kits favor either smaller or larger RNA fragments
that might lead to an underrepresentation of “mid-size” RNA

b4444_Ch-08.indd 220 08-Jul-22 6:05:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 221

fragments in the final library pool. The sequencing platform itself,

and subsequent bioinformatics processing of the sequencing data
(e.g., mapping order to multiple databases), also has an impact on
RNA sequencing results through various parameters (for details
see72). None of the current library processing or analysis methods
are completely unbiased, which renders it necessary to perform an
orthogonal method of validation, e.g., by RT-qPCR and normaliza-
tion of data.
In comparison to mRNA, small RNAs, such as microRNAs,
are relatively stable. Their abundance within exosomes and their
function as regulators of posttranscriptional gene expression73
make microRNAs attractive candidates as cancer biomarkers.
There is also an increasing body of evidence for functional cell-
to-cell transfer of miRNAs via exosomes, adding to their roles in
tumorigenesis.64,74 There are currently various commercial kits
available for isolation of exosomal microRNA from primary cul-
tures as well as patient biofluids.

Exosomal Identification
Based on the increasing significance of exosomes in health and
disease, multiple principles and guidelines for exosome detec-
tion have been established, for example, the position statement
from the ISEV.75 It is always recommended to run all experiments
with appropriate positive and negative control samples to iden-
tify systemic errors and sources of contamination. Additionally,
guidelines recommend the use of different methods of exosome
verification. The gold standard for exosomal imaging is transmis-
sion electron microscopy (TEM). Preparations require fixation,
embedding, cutting, and staining, which results in a collapsed arti-
factual cup-shaped structure. TEM also allows for reliable size

b4444_Ch-08.indd 221 08-Jul-22 6:05:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

222 Cell-Free Circulating DNA: Purification and Analysis Techniques

measurements of EV, and specific subpopulations can be iden-

tified through immuno-gold-labelling. Standard views should
comprise wide field view as well as close-ups of single vesicles.75
Another method for exosome identification is flow-cytometry. As
exosomes are too small in size to be reliably detected through flow
cytometry, vesicles are typically captured onto large particles or
beads and then fluorescently tagged with antibodies of interest
including CD63 or CD9.76
Another popular method of exosome detection is through
nanoparticle tracking analysis (NTA).77 This technique measures
size distribution and vesicle concentration based on their Brown-
ian motion with a dark field microscope, although size distribu-
tions should be validated with values calculated through TEM.78,79
NTA is not able to differentiate nonvesicular particles such as
protein aggregates with overlapping size distributions of vesicles
but is superior in size measurements compared to dynamic light
scattering.80 Another method for characterization of single vesicles
is known as resistive pulse sensing (RPS), which is based on the
Coulter principle and tracks size and concentration of nanoparti-
cles passing through pores of a stationary phase.81 Similar to NTA,
RPS is not capable of distinguishing between nonvesicular and
vesicular particles.
Besides identification based on their physical properties,
isolation of exosomes after enrichment is often verified by West-
ern blot, ELISA, or proteomic analysis in order to demonstrate
the presence of abundantly expressed exosomal-specific proteins.
Thanks to evolving isolation and analytical technologies, the list
of new exosomal protein content is constantly growing. As of
July 2018, a total of around 7500 different proteins have been
described, but only 500 of them account for approximately 90% of
exosomal proteins.82 The general characterization of exosomes is

b4444_Ch-08.indd 222 08-Jul-22 6:05:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 223

typically done by using at least three exosomal markers (described

in the following).74,75 Additionally, the ISEV recommends confirm-
ing the under-representation or absence of intracellular proteins
associated with compartments other than the plasma membrane
or endosomes, and comparing protein composition of the iso-
lated exosomal fraction with supernatant and negative controls.
Protein analysis is ideally performed by Western blot, FACS, or
global proteomic analysis using mass spectrometry techniques in a
semi-quantitative assay.75
The most commonly used markers for exosome identifica-
tion are the tetraspanin family members CD9, CD63, and CD81.
More recent reports have also identified them to be abundantly
expressed in apoptotic bodies and microvesicles.83 Based on the
cargo sorting machinery involved in their biogenesis, ESCRT com-
ponents such as ALG-2-interacting protein X (ALIX) and tumor
susceptibility gene 101 protein (TSG101) can also be identified,
but these are not specific to exosomes.84 Some papers indicate a
lack of commonly used markers in certain subpopulations of exo-
some such as small vesicles. For example, Kowal et al. identified
certain subcategories of small vesicles/exosomes that lack bona
fide exosomal markers CD9 and CD63, which suggests that their
biogenesis does not involve the endosomal pathway.8 In contrast,
all small vesicles with the size of exosomes have shown abundant
expression in Annexin XI, ADAM10, ACE, and EHD4. These
results are consistent with other recent proteomic characteriza-
tions of EV and exosomes from platelets and tumor cells85–87 and
might point to the fact that these markers represent higher accu-
racy for exosomal isolation. In summary, after exclusion of larger
vesicles by isolation techniques, CD63 might be one of the best
markers to pull down or identify endosome-derived exosomes,
although some cell types do secrete CD63-devoid vesicles.87–89

b4444_Ch-08.indd 223 08-Jul-22 6:05:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

224 Cell-Free Circulating DNA: Purification and Analysis Techniques

Other markers have been proposed, which are described as

more cancer specific. For example, similar to the field of CTC,
the use of EPCAM antibodies has been reported as a specific
method to isolate tumor exosomes from epithelial carcinomas.90
More recently, we have shown that CLDN4, EPCAM, CD151,
LGALS3BP, HIST2H2BE, and HIST2H2BF can specifically iso-
late exosomes from pancreatic ductal adenocarcinoma. Biofluid
specificity has also been reported with regard to exosomal surface
marker repertoire, whereby lectin-abundant exosomes have been
identified in patient-derived urine samples.91

Exosomal Isolation Strategies

The first step in the successful isolation of EV is to optimize the
collection conditions of the starting fluid. If working with exo-
somes from cell culture supernatant, contamination of the super-
natant with other sources of extracellular exosomes such as fetal
bovine serum (FBS) has to be minimized or at the very least taken
into consideration in downstream analysis.92
If the starting material comprises biofluids such as mamma-
lian whole blood, precautions should be taken to prevent platelet
activation.93,94 This is typically done through the use of anticoag-
ulants in blood collection tubes; however, it is important to know
that anticoagulants may impair downstream analysis, specifically
PCR inhibition by heparin.95 As blood is one of the most commonly
used sources of exosomes, standardization of blood collection and
processing is of upmost importance. For example, prolonged stasis
using tourniquets should be avoided to prevent hemolysis,96 patient
volume status should be taken into account, and even circadian
rhythms have been shown to have an impact in the context of liq-
uid biopsy analysis. Other pre-analytical variables to consider are

b4444_Ch-08.indd 224 08-Jul-22 6:05:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 225

standardization of needle diameter as well as the need to discard

the first 2 mL of blood collected to avoid contamination with non-
blood cells.97 Few reports exist comparing the exosomal fraction
derived from human serum versus plasma, but as serum is highly
influenced by the coagulation system, plasma is typically consid-
ered the preferred source for exosomal isolation. Among blood
collection tubes, acid citrate dextrose (ACD) tubes have been
reported to be superior because of their impairment of platelet
exosome formation.98 Handling of the blood vials is also important
for exosomal quality. Exosomal yields are maximized with gentle
handling and fast processing after collection, usually within 4 hr.99
To minimize the number of white and red blood cells, as well as
platelets, that might release exosomes upon activation, plasma is
separated from whole blood. Room temperature, two-step centrif-
ugation (2 × 15 min, 2500g) with lowest deceleration settings is
recommended for plasma isolation. Plasma collection should also
stop approximately 5 mm above the buffy coat meniscus.100 It is
also important to consider the short- and long-term storage of col-
lected samples, where snap freezing of aliquots has been recom-
mended for exosome storage. There is a limited amount of studies
investigating the potential impact of long-term freezing on the
recovery, morphology, and impact on downstream analysis. In gen-
eral, fresh processing of isolated exosomal pellet provides the best
yield, whereas extended freezer storage of exosomes may increase
their size and change their physical properties, such as their rep-
ertoire of surface proteins. These morphological changes are due
to an expansion of ice crystals in the spherical lipid bilayer and
consequent membrane disruption, which may ultimately lead to
membrane fusion of neighboring vesicles.101 Nonetheless, storage
temperatures of –70°C to –80°C are recommended over higher
temperature (–20°C to –60°C ) for long-term storage.102

b4444_Ch-08.indd 225 08-Jul-22 6:05:54 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

226 Cell-Free Circulating DNA: Purification and Analysis Techniques

Herein, we will describe six current strategies for exosome

isolation. These are either based on different physical proper-
ties of the vesicles (e.g., size, density, shape) or on the capture of
exosomes by specific molecules that are abundantly expressed on
their surface. All have their own advantages and disadvantages and
have to be tailored and optimized based on specific needs (Fig. 3).

Ultracentrifugation
Ultracentrifugation (Fig. 3a) uses centrifugal forces to separate
particles in suspension based on their density, size, and shape, as
well as the density and viscosity of the solvent. Ultracentrifuga-
tion applies exceptionally high centrifugal forces of up to 100,000g
to isolate nanosized vesicles of interest and is still considered the
gold standard for exosome purification. Due to its long-term cost
efficiency and relative purity of extraction, this technique is used
by many of the laboratories studying exosome biology,103 although
the initial capital expense is high. Nonetheless, it is important to
highlight that different equipment (e.g., rotor type and diameter)
and isolation parameters like centrifugation time and speed pro-
duce significantly different populations of exosomes with specific
protein and RNA content.104,105
Usually sequential ultracentrifugation steps with increasing
gravitational (“g”) speeds are carried out in order to eliminate cel-
lular debris and isolate EV of a desired size. Exosomal purity can be
increased by combining density gradient separation with ultracen-
trifugation. This also allows the separation of low-density exosomes
from high-density protein aggregates, which are often a source of
contamination from differential ultracentrifugation.106 There are
two major subtypes of density-gradient centrifugation (Fig. 3b).
Isopycnic centrifugation sediments particles based on their density
profile to a defined equilibrium position (equilibrium of density)

b4444_Ch-08.indd 226 08-Jul-22 6:05:54 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 227

(a) (b)

(c) (d)

(e) (f)

Fig. 3: Strategies for exosome purification: (a) differential centrifugation,

(b) density gradient centrifugation, (c) ultrafiltration, (d) precipitation, (e) size
exclusion chromatography, (f) affinity capture.

but needs high centrifugation speed and a long running time. The
isopycnic position is 1.10–1.21 g/mL in case of the commonly used
solvent cesium chloride.107 In contrast, moving-zone ultracentrifu-
gation uses differences in the molecular weight (size and mass) for

b4444_Ch-08.indd 227 08-Jul-22 6:05:59 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

228 Cell-Free Circulating DNA: Purification and Analysis Techniques

separation. Although moving-zone ultracentrifugation allows for

the separation of vesicles with different sizes but similar densities,
it is important to mention that centrifugation time directly affects
the position of the exosomal fraction within the density gradient,
which therefore needs to be controlled rigorously. Insufficient cen-
trifugation time will result in an incomplete separation, whereas
excess centrifugation time will co-precipitate all analytes. Addi-
tionally, loading volumes are very limited, which restricts a high
through put.

Ultrafiltration
This technique uses exosomal size, deformation capability, and
molecular weight to selectively pass them through a membrane
filter with defined exclusion limits (Fig. 3c). The biggest advan-
tage of ultrafiltration is the ease of use, the short processing time,
and the low cost per sample. On the other hand, the filtration
process may result in possible deformation of vesicles and the
loss of specificity in isolating exosomal fractions. This technique
is of particular use in cell-free samples like urine where research-
ers have shown similar efficiency of vesicle isolation compared
to much more time-consuming protocols like ultracentrifuga-
tion.108 For suspensions with a high cellular content, removal of
large cells and particles has to be performed prior to ultrafiltra-
tion. Yield may be affected by clogging or vesicle trapping on the
membranes.109

Polyethylene Glycol (PEG) Precipitation

Polyethylene glycol (PEG) precipitation (Fig. 3d) uses exosomal
solubility and dispersal properties to isolate them from a suspen-
sion. The technique requires a hydrophilic polymer (e.g., PEG),

b4444_Ch-08.indd 228 08-Jul-22 6:05:59 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 229

which binds water molecules and forces hydrophobic contents to

settle. The precipitate itself is usually generated from pre-processed
samples (removal of cells, cell fragments, and debris by centrifu-
gation) upon an incubation step at 4°C followed by centrifugation
or filtration for harvesting the precipitate. The workflow is easy
to handle, cost efficient, does not require specialized equipment
and is easily scalable to sample size.110 Nonetheless, exosomes are
not exclusively precipitated, and the background of hydrophobic
molecules and particles — especially co-precipitating proteins —
can be high.111 Also, precipitating conditions must be optimized to
different viscosity levels of samples, leading to the requirement for
pre- and post-precipitation cleaning steps.

Size Exclusion Chromatography (SEC)

Size exclusion chromatography (SEC) (Fig. 3e) uses the hydro-
dynamic volume of particles for their separation and can be used
for low-volume samples. Despite its name, SEC uses not only the
size of the particle for separation but also their varying diffusion
volumes over a stationary porous phase. This is one of its main
differences from techniques like ultrafiltration that separate par-
ticles based solely on their physical size. Upon entering the sta-
tionary polymer layer with micron-scale polymer beads containing
pores of different sizes, smaller particles experience a leap in their
available diffusion volume leading to a prolonged retention time.
In contrast, larger particles with a smaller available diffusion vol-
ume are quickly forced out of the porous layer in the isolation
column. Thus, larger particles elute faster through the stationary
polymer compared to smaller molecules. A big advantage of SEC
is its resulting high purity of exosomes and data reproducibility.
Unfortunately, it’s a time-consuming technique that requires spe-
cialized technical expertise and equipment. A possible source of

b4444_Ch-08.indd 229 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

230 Cell-Free Circulating DNA: Purification and Analysis Techniques

contamination to consider in SEC is the co-isolation of lipoproteins

(high- and low-density lipoproteins and chylomicrons) that show
similar diffusion volumes to exosomes.

Immunoprecipitation/Microfluidics Devices
Immunoaffinity-based separation assays (Fig. 3f) use exoso-
mal antigens to capture vesicles. A basic requirement for the
use of any immunoprecipitation is the presence of a unique,
membrane-bound antigen that is expressed at a significant concen-
tration on the outer lipid layer of exosomes. Herein, one might rely
on “general” exosomal identification markers (e.g., CD9, TSG101,
or CD69) or target cancer–specific molecules. Preferably, no sol-
uble counterpart of these molecules should exist, in order to pre-
vent unintended and uncontrolled binding. The most commonly
used separation technique is magnetic pull-down, where magnetic
beads are coated with specific antibodies with a high affinity for
the desired exosomal antigen. The bound vesicles can then be
either directly lysed to discharge their content or released from
the beads for further downstream analysis (see section “Exosomal
identification” for details on the exosomal antigens).
Other methods use microfluidics devices that exploit
membrane-coated immunoprecipitation on a chip.112 These
devices provide an efficient method for exosomal isolation from
low-volume samples. Some systems also allow the on-chip inte-
gration of exosomal characterization and nucleic acid extraction.
For example, ExoChip uses anti-CD63 IgG antibodies to capture
exosomes, quantifies the number of captured exosomes after fluo-
rescent carbocyanine dye (DiO, 3,3´-Dioctadecyloxacarbocyanine
Perchlorate) staining with a plate reader, and enables downstream
isolation of intact exosomal RNA.113 With advancing production

b4444_Ch-08.indd 230 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 231

capacity and more precise manufacturing, microfluidics devices

promise a more efficient, time-saving, and cost-effective approach
to isolate exosomes that is consolidated on a single chip, although
it has to be mentioned that to date, no system meets the stringent
criteria required in a clinical trial setting. Immunoaffinity capture
methods result in highly purified exosomal fractions but at the
cost of lower exosomal yields. This is due to the fact that one is
limited in isolating the specific exosomal phenotype of interest.
Additionally, it is important to note that exosomes may aggregate
or be shielded by protein or platelet coating, resulting in a reduced
binding to antibodies.

Field Flow Fractionation

In contrast to classical chromatography, field flow fractionation
(FFF) uses differences in particle diffusion capabilities to sepa-
rate particles in a long narrow channel. Exosomes in solution are
pumped through this chamber with a continuous parabolic flow.
In a separation zone this continuous flow is exposed to a crossfield
comprised of a perpendicular force, for example, electric charges,
magnetism, light or gravity. This interaction between the para-
bolic flow and the perpendicular force separates exosomes based
on the Brownian motion of particles. In contrast to classical chro-
matographs, smaller particles like exosomes are separated early
creating a reversed eluate profile.114 In addition to classical FFF,
other separation forces such as acoustic nanofilters using ultra-
sonic waves have been described as a strategy in exosome isola-
tion, which allows for versatile size selection of vesicles.115
Various commercial kits for the direct isolation of stable exo-
somal microRNAs have been released. In principle, exosomes
are isolated in a first step on a membrane filter. Following lysis,

b4444_Ch-08.indd 231 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

232 Cell-Free Circulating DNA: Purification and Analysis Techniques

exosomal microRNA content is purified in a column through a

consecutive protocol of washing solutions. Many of these isolation
kits are popular because they are time efficient and user friendly,
but it is important to stress that most of the kits produce a high
exosomal yield but also suffer from a relatively high albumin con-
tamination. For detailed comparison of different performance of
kits, please refer to the technique comparison in the following sec-
tion, as well as to other studies.67,116,117
All of these methods carry advantages and disadvantages, and
ultimately it is the user’s decision as to which strategy provides
the most effective means of answering the biological question of
interest. Combining different isolation principles may be a pow-
erful tool to minimize contamination and achieve a high purity,
but is time consuming and leads to a significant decrease in yield.
Several reports have been published providing valuable guidelines
in identifying common pitfalls and outlining consensus findings for
researchers in this field.59,93,100 Another important source of com-
parison is the crowdsourcing knowledgebase that centralizes EV
studies with the aim of a more standardized approach.118

Comparison Between Isolation Techniques

Although the choice of exosomal isolation technique has a signif-
icant impact on the findings and reproducibility of a study, there
are a limited number of high-quality studies comparing the yield
and quality of different isolation methods using the same input
material.119,120 In 2014, van Deun compared two commercial iso-
lation kits ExoQuick (System Biosciences) and Total Exosome
Isolation Reagent (Invitrogen) to ultracentrifugation and density
gradient centrifugation (OptiPrep, Sigma Aldrich) using superna-
tant from a breast cancer cell line.119 The authors showed that,

b4444_Ch-08.indd 232 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 233

based on electron microscopy and Western blot, OptiPrep density

gradient centrifugation outperformed all other techniques in
terms of purity, although with significantly lower yields. Similarly,
Ting-Tang et al. compared a variety of commercial isolation meth-
ods with ultracentrifugation using cell culture media and found
the highest purity, but the lowest yield, of exosomes was obtained
with ultracentrifugation.67 In 2017, Helwa et al. compared differ-
ent volumes of noncancerous human serum processed with four
different exosome isolation methodologies: miRCURY (Exiqon),
ExoQuick (System Biosciences), Total Exosome Isolation Reagent
(Invitrogen), and ultracentrifugation.117 The authors used NTA,
Western blot analysis (for CD9 and CD63), TEM, and droplet
digital PCR to confirm exosomal isolation and compare yields.
They concluded that the commercial kits provide higher yields for
most input volumes, although no evaluation of contamination in
the exosome isolates was carried out. In general, commercial kits
have shown up to 100-fold higher RNA yield but with accompany-
ing higher contamination of soluble proteins. Other studies have
shown superior purification when using immunoaffinity-based
approaches compared to density gradient ultracentrifugation.121,122
In summary, there exists a tradeoff between exosomal purity and
yield, which must be adapted to the needs of the study.

Artifacts
Though exciting new results have been presented over the last
decade in the field of EV, exosomal research is still heavily prone to
bias and artifacts. This stems from the fact that there is no uniform
definition of exosomes and other EV subtypes. This nonstringent
terminology across published articles leads to overlapping defini-
tions of vesicle subpopulations.33 It is also important to keep in

b4444_Ch-08.indd 233 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

234 Cell-Free Circulating DNA: Purification and Analysis Techniques

mind that most vesicles isolated from human blood originate from
healthy cells. The mean EV concentration in healthy individuals
reaches approximately 107–109/mL blood and it needs significant
technical abilities to analyze changes over this background123 and
distinguish exosomes originating from, for example, cancer cells.
Herein, the problem with most published protocols for isolation
of EV is their nonselective isolation of all circulating exosomes, in
contrast to a cancer-specific exosomal pull-down.
Due to their small diameter and cargo load, validation of exo-
somal isolation is still challenging. In this context, several former
proposed “exosomal markers” have been shown to be present in a
wide variety of EV subtypes (e.g., major histocompatibility com-
plex molecules, flotillins and heat shock proteins), meaning that
studies that have used these likely had a heterogeneous popula-
tion of vesicle types in their analysis.8,124 Additionally, the primary
isolation fluid (e.g., saliva, urine, blood, cell culture supernatant)
significantly affects the possible contaminants that may be pres-
ent. For example, bacteria-derived material is a source of contam-
ination in human nasal fluid and saliva, whereas Tamm–Horsefall
glycoprotein has been reported as major contaminant in human
urine.59
Co-isolation of other soluble particles is a common source of
artifacts. Physical characteristics overlapping with large proteins,
protein aggregates, and lipoproteins, especially in complex human
suspensions like blood, saliva, or urine, make the isolation of
exosomes based on their physical properties prone to contamina-
tion.109,125 Protein complexes and single proteins may also co-bind
microRNA and therefore significantly affect downstream analy-
sis. Another source of contamination are nonexosomal–associated
forms of extracellular RNA bound to lipoprotein (HDL and LDL)
complexes, ribonucleoprotein complexes (RNPs) or viral particles

b4444_Ch-08.indd 234 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 235

that might be co-isolated with exosomes.125 Both lipoproteins, as

well as postprandially increased chylomicrons, have been shown
to carry nucleic acids, and their co-purification might lead to
compromised results.126,127 Specifically, each lipoprotein fraction
(HDL, LDL, VLDL, and chylomicrons) has a size or buoyant den-
sity profile that overlaps with exosomes. In this context, although
VLDL and chylomicrons have overlapping sizes to exosomes (≥60
nm), they may be separated by their different density buoyance of
<1.06 g/cm3 versus exosomal 1.11–1.19 g/cm3 density buoyance,
whereas HDL can be separated from exosomes based solely on
their smaller size (10 nm vs. 50–150 nm), although they show an
overlapping density buoyance.
To reliably distinguish between extravesicular and intravesic-
ular RNA, exosomal samples should be treated with RNase and
proteinase prior to exosome lysis. Herein, some reports question
the presence of high–molecular weight DNA and ribosomal RNA
in exosomes and ascribe their presence in exosomes to co-isolation
with other molecules.59 Like human samples, cell culture super-
natant comprises its own challenges for exosomal isolation. There
is, for example, concern in the exosomal community regarding the
contamination of RNA results from cell culture–isolated exosomes
by carryover from FBS used in culture media.92

Conclusions
EV, including exosomes, have experienced a renaissance over
the last decade. Multiple new discoveries about their role in
crucial (patho-)physiological mechanisms, as well as their highly
protected cargo that shows immense potential as a source of bio-
markers, resulted in a spike of interest. Despite their importance,
EV isolation and characterization harbors multiple challenges

b4444_Ch-08.indd 235 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

236 Cell-Free Circulating DNA: Purification and Analysis Techniques

and pitfalls. Following now established guidelines together with

careful experimental design offer the opportunity to fully unlock
their potential and integrate EV biomarkers into routine clinical
workup.

References
   1. Johnstone RM, Adam M, Hammond JR, Orr L, Turbide C. (1987) Ves-
icle formation during reticulocyte maturation. Association of plasma
membrane activities with released vesicles (exosomes). J Biol Chem 262:
9412–9420.
   2. Pan BT, Teng K, Wu C, Adam M, Johnstone RM. (1985) Electron micro-
scopic evidence for externalization of the transferrin receptor in vesicular
form in sheep reticulocytes. J Cell Biol 101: 942–948.
   3. Harding C, Heuser J, Stahl P. (1984) Endocytosis and intracellular pro-
cessing of transferrin and colloidal gold-transferrin in rat reticulocytes:
demonstration of a pathway for receptor shedding. Eur J Cell Biol 35:
256–263.
   4. Gould SJ, Raposo G. (2013) As we wait: coping with an imperfect nomen-
clature for extracellular vesicles. J Extracell Vesicles 2: 20389.
   5. Cicero Lo A, Stahl PD, Raposo G. (2015) Extracellular vesicles shuffling
intercellular messages: for good or for bad. Curr Opin Cell Biol 35: 69–77.
   6. Raposo G, Stoorvogel W. (2013) Extracellular vesicles: exosomes, microve-
sicles, and friends. J Cell Biol 200: 373–383.
   7. Tricarico C, Clancy J, D’Souza-Schorey C. (2017) Biology and biogenesis
of shed microvesicles. Small GTPases 8: 220–232.
   8. Kowal J, Arras G, Colombo M, et al. (2016) Proteomic comparison defines
novel markers to characterize heterogeneous populations of extracellular
vesicle subtypes. Proc Natl Acad Sci USA 113: E968–E977.
   9. Willms E, Johansson HJ, Mäger I, et al. (2016) Cells release subpopula-
tions of exosomes with distinct molecular and biological properties. Sci
Rep 6: 22519.
10. Shen B, Wu N, Yang J-M, Gould SJ. (2011) Protein targeting to exosomes/
microvesicles by plasma membrane anchors. J Biol Chem 286: 14383–14395.

b4444_Ch-08.indd 236 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 237

11. Valadi H, Ekstr K, Bossios A, Margareta S., Lee JJ, Lötvall JO. (2007)
Exosome-mediated transfer of mRNAs and microRNAs is a novel mecha-
nism of genetic exchange between cells. Nat Cell Biol 9: 654–659.
12. Robbins PD, Dorronsoro A, Booker CN. (2016) Regulation of chronic
inflammatory and immune processes by extracellular vesicles. J Clin Invest
126: 1173–1180.
13. Ibrahim A, Marbán E. (2016) Exosomes: fundamental biology and roles in
cardiovascular physiology. Annu Rev Physiol 78: 67–83.
14. Hoshino A, Costa-Silva B, Shen T-L, et al. (2015) Tumour exosome integ-
rins determine organotropic metastasis. Nature 527: 329–335.
15. Crowley E, Di Nicolantonio F, Loupakis F, Bardelli A. (2013) Liquid
biopsy: monitoring cancer-genetics in the blood. Nat Rev Clin Oncol 10:
472–484.
16. Wan JCM, Massie C, Garcia-Corbacho J, et al. (2017) Liquid biopsies
come of age: towards implementation of circulating tumour DNA. Nat
Rev Cancer 17: 223–238.
17. Siravegna G, Marsoni S, Siena S, Bardelli A. (2017) Integrating liq-
uid biopsies into the management of cancer. Nat Rev Clin Oncol 14:
531–548.
18. Halvaei S, Daryani S, Eslami-S Z, et al. (2018) Exosomes in cancer liquid
biopsy: a focus on breast cancer. Mol Ther Nucleic Acids 10: 131–141.
19. Nedaeinia R, Manian M, Jazayeri MH, et al. (2017) Circulating exosomes
and exosomal microRNAs as biomarkers in gastrointestinal cancer. Cancer
Gene Ther 24: 48–56.
20. Bernard V, Kim DU, San Lucas FA, et al. (2018) Circulating nucleic acids
associate with outcomes of patients with pancreatic cancer. Gastroenterol-
ogy 156(1): 108–118.e4.
21. McKiernan J, Donovan MJ, O’Neill V, et al. (2016) A novel urine exo-
some gene expression assay to predict high-grade prostate cancer at initial
biopsy. JAMA Oncol 2: 882–889.
22. Thakur BK, Zhang H, Becker A, et al. (2014) Double-stranded DNA in
exosomes: a novel biomarker in cancer detection. Cell Res 24: 766–769.
23. Lunavat TR, Cheng L, Einarsdottir BO, et al. (2017) BRAFV600 inhibi-
tion alters the microRNA cargo in the vesicular secretome of malignant
melanoma cells. Proc Natl Acad Sci USA 114: E5930–E5939.

b4444_Ch-08.indd 237 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

238 Cell-Free Circulating DNA: Purification and Analysis Techniques

24. Krug AK, Enderle D, Karlovich C, et al. (2018) Improved EGFR mutation
detection using combined exosomal RNA and circulating tumor DNA in
NSCLC patient plasma. Ann Oncol 29: 700–706.
25. Castillo J, Bernard V, San Lucas FA, et al. (2017) Surfaceome profiling
enables isolation of cancer-specific exosomal cargo in liquid biopsies from
pancreatic cancer patients. Ann Oncol 29: 223–239.
26. Peinado H, Zhang H, Matei IR, et al. (2017) Pre-metastatic niches:
organ-specific homes for metastases. Nat Rev Cancer 17: 302–317.
27. Xu J-Y, Chen G-H, Yang Y-J. (2017) Exosomes: a rising star in falling
hearts. Front Physiol 8: 494.
28. Baglio SR, Pegtel DM, Baldini N. (2012) Mesenchymal stem cell secreted
vesicles provide novel opportunities in (stem) cell-free therapy. Front
Physiol 3: 359.
29. Ailawadi S, Wang X, Gu H, Fan G-C. (2015) Pathologic function and ther-
apeutic potential of exosomes in cardiovascular disease. Biochim Biophys
Acta 1852: 1–11.
30. Lai RC, Arslan F, Lee MM, et al. (2010) Exosome secreted by MSC
reduces myocardial ischemia/reperfusion injury. Stem Cell Res 4:
214–222.
31. Saman S, Kim W, Raya M, et al. (2012) Exosome-associated tau is secreted
in tauopathy models and is selectively phosphorylated in cerebrospinal
fluid in early Alzheimer disease. J Biol Chem 287: 3842–3849.
32. Fais S, O’Driscoll L, Borras FE, et al. (2016) Evidence-based clinical use
of Nanoscale extracellular vesicles in Nanomedicine 10: 3886–3899.
33. van Niel G, D’Angelo G, Raposo G. (2018) Shedding light on the cell biol-
ogy of extracellular vesicles. Nat Rev Mol Cell Biol 19: 213–228.
34. Stoorvogel W. (2015) Resolving sorting mechanisms into exosomes. Cell
Res 25: 531–532.
35. Möbius W, Ohno-Iwashita Y, van Donselaar EG, et al. (2002) Immuno-
electron microscopic localization of cholesterol using biotinylated and
non-cytolytic perfringolysin O. J Histochem Cytochem 50: 43–55.
36. Wollert T, Wunder C, Lippincott-Schwartz J, Hurley JH. (2009) Mem-
brane scission by the ESCRT-III complex. Nature 458: 172–177.
37. Stuffers S, Wegner CS, Stenmark H, Brech A. (2009) Multivesicular endo-
some biogenesis in the absence of ESCRTs. Traffic 10: 925–937.

b4444_Ch-08.indd 238 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 239

38. Colombo M, Moita C, van Niel G, et al. (2013) Analysis of ESCRT

functions in exosome biogenesis, composition and secretion highlights the
heterogeneity of extracellular vesicles. J Cell Sci 126: 5553–5565.
39. Hurley JH, Hanson PI. (2010) Membrane budding and scission by the
ESCRT machinery: it’s all in the neck. Nat Rev Mol Cell Biol 11: 556–566.
40. Christ L, Raiborg C, Wenzel EM, Campsteijn C, Stenmark H. (2017)
Cellular functions and molecular mechanisms of the ESCRT mem-
brane-scission machinery. Trends Biochem Sci 42: 42–56.
41. Baietti MF, Zhang Z, Mortier E, et al. (2012) Syndecan-syntenin-ALIX
regulates the biogenesis of exosomes. Nat Cell Biol 14: 677–685.
42. Villarroya-Beltri C, Baixauli F, Gutiérrez-Vázquez C, Sánchez-Madrid
F, Mittelbrunn M. (2014) Sorting it out: regulation of exosome loading.
Semin Cancer Biol 28: 3–13.
43. Trajkovic K, Hsu C, Chiantia S, et al. (2008) Ceramide triggers budding of
exosome vesicles into multivesicular endosomes. Science 319: 1244–1247.
44. van Niel G, Charrin S, Simoes S, et al. (2011) The tetraspanin CD63 reg-
ulates ESCRT-independent and -dependent Endosomal sorting during
melanogenesis. Dev Cell 21: 708–721.
45. Geminard C, de Gassart A, Blanc L, Vidal M. (2004) Degradation of AP2
during reticulocyte maturation enhances binding of hsc70 and Alix to a
common site on TFR for sorting into exosomes. Traffic 5: 181–193.
46. de Gassart A, Geminard C, Fevrier B, Raposo G, Vidal M. (2003) Lipid
raft-associated protein sorting in exosomes. Blood 102: 4336–4344.
47. Colombo M, Raposo G, Thery C. (2014) Biogenesis, secretion, and inter-
cellular interactions of exosomes and other extracellular vesicles. Annu
Rev Cell Dev Biol 30: 255–289.
48. Haraszti RA, Didiot M-C, Sapp E, et al. (2016) High-resolution proteomic
and lipidomic analysis of exosomes and microvesicles from different cell
sources. J Extracell Vesicles 5: 32570.
49. Jenjaroenpun P, Kremenska Y, Nair VM, Kremenskoy M, Joseph B, Kuroch-
kin IV. (2013) Characterization of RNA in exosomes secreted by human
breast cancer cell lines using next-generation sequencing. PeerJ 1: e201.
50. Wang K, Zhang S, Weber J, Baxter D, Galas DJ. (2010) Export of microR-
NAs and microRNA-protective protein by mammalian cells. Nucleic Acids
Res 38: 7248–7259.

b4444_Ch-08.indd 239 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

240 Cell-Free Circulating DNA: Purification and Analysis Techniques

51. Li L, Li C, Wang S, et al. (2016) Exosomes derived from hypoxic oral

squamous cell carcinoma cells deliver miR-21 to normoxic cells to elicit a
prometastatic phenotype. Cancer Res 76: 1770–1780.
52. Ramakrishnaiah V, Thumann C, Fofana I, et al. (2013) Exosome-mediated
transmission of hepatitis C virus between human hepatoma Huh7.5 cells.
Proc Natl Acad Sci USA 110: 13109–13113.
53. Fong MY, Zhou W, Liu L, et al. (2015) Breast-cancer-secreted miR-122
reprograms glucose metabolism in premetastatic niche to promote metas-
tasis. Nat Cell Biol 17: 183–194.
54. Ben-Dov IZ, Whalen VM, Goilav B, Max KEA, Tuschl T. (2016) Cell and
Microvesicle urine microRNA deep sequencing profiles from healthy indi-
viduals: observations with potential impact on biomarker studies. PLoS
ONE 11: e0147249.
55. Batagov AO, Kurochkin IV. (2013) Exosomes secreted by human cells
transport largely mRNA fragments that are enriched in the 3′-untranslated
regions. Biol Direct 8: 12.
56. San Lucas FA, Allenson K, Bernard V, et al. (2015) Minimally invasive
genomic and transcriptomic profiling of visceral cancers by next-genera-
tion sequencing of circulating exosomes. Ann Oncol 27(4): 635–641.
57. Kahlert C, Melo SA, Melo SA, et al. (2014) Identification of double-
stranded genomic DNA spanning all chromosomes with mutated KRAS
and p53 DNA in the serum exosomes of patients with pancreatic cancer. J
Biol Chem 289: 3869–3875.
58. Mittelbrunn M, Gutiérrez-Vázquez C, Villarroya-Beltri C, et al. (2011)
Unidirectional transfer of microRNA-loaded exosomes from T cells to
antigen-presenting cells. Nat Commun 2: 282.
59. Mateescu B, Kowal EJK, van Balkom BWM, et al. (2017) Obstacles and
opportunities in the functional analysis of extracellular vesicle RNA-an
ISEV position paper. J Extracell Vesicles 6: 1286095.
60. Villarroya-Beltri C, Gutiérrez-Vázquez C, Sánchez-Cabo F, et al. (2013)
Sumoylated hnRNPA2B1 controls the sorting of miRNAs into exosomes
through binding to specific motifs. Nat Commun 4: 2980.
61. Koppers-Lalic D, Hackenberg M, Bijnsdorp IV, et al. (2014) Nontem-
plated nucleotide additions distinguish the small RNA composition in cells
from exosomes. Cell Rep 8: 1649–1658.

b4444_Ch-08.indd 240 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 241

62. Gibbings DJ, Ciaudo C, Erhardt M, Voinnet O. (2009) Multivesicular

bodies associate with components of miRNA effector complexes and mod-
ulate miRNA activity. Nat Cell Biol 11: 1143–1149.
63. Kosaka N, Iguchi H, Yoshioka Y, Takeshita F, Matsuki Y, Ochiya T. (2010)
Secretory mechanisms and intercellular transfer of microRNAs in living
cells. J Biol Chem 285: 17442–17452.
64. Squadrito ML, Baer C, Burdet F, et al. (2014) Endogenous RNAs modu-
late microRNA sorting to exosomes and transfer to acceptor cells. Cell Rep
8: 1432–1446.
65. Cheng L, Sharples RA, Scicluna BJ, Hill AF. (2014) Exosomes provide a
protective and enriched source of miRNA for biomarker profiling com-
pared to intracellular and cell-free blood. J Extracell Vesicles 3: 23743.
66. Shelke GV, Lasser C, Gho YS, Lotvall J. (2014) Importance of exosome
depletion protocols to eliminate functional and RNA-containing extracel-
lular vesicles from fetal bovine serum. J Extracell Vesicles 3: 24783.
67. Tang Y-T, Huang Y-Y, Zheng L, et al. (2017) Comparison of isolation meth-
ods of exosomes and exosomal RNA from cell culture medium and serum.
Int J Mol Med 40: 834–844.
68. Becker C, Hammerle-Fickinger A, Riedmaier I, Pfaffl MW. (2010) mRNA
and microRNA quality control for RT-qPCR analysis. Methods 50: 237–243.
69. Aranda R, Dineen SM, Craig RL, Guerrieri RA, Robertson JM. (2009)
Comparison and evaluation of RNA quantification methods using viral,
prokaryotic, and eukaryotic RNA over a 10(4) concentration range. Anal
Biochem 387: 122–127.
70. Kim Y-K, Yeo J, Kim B, Ha M, Kim VN. (2012) Short structured RNAs
with low GC content are selectively lost during extraction from a small
number of cells. Mol Cell 46: 893–895.
71. Huang X, Yuan T, Tschannen M, et al. (2013) Characterization of human
plasma-derived exosomal RNAs by deep sequencing. BMC Genomics 14:
319.
72. Hill AF, Pegtel DM, Lambertz U, et al. (2013) ISEV position paper: extra-
cellular vesicle RNA analysis and bioinformatics. J Extracell Vesicles 2:
22859.
73. Iwakawa H-O, Tomari Y. (2015) The functions of MicroRNAs: mRNA
decay and translational repression. Trends Cell Biol. 25: 651–665.

b4444_Ch-08.indd 241 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

242 Cell-Free Circulating DNA: Purification and Analysis Techniques

74. Bhome R, Del Vecchio F, Lee G-H, et al. (2018) Exosomal microRNAs
(exomiRs): small molecules with a big role in cancer. Cancer Lett 420:
228–235.
75. Lotvall J, Hill AF, Hochberg F, et al. (2014) Minimal experimental require-
ments for definition of extracellular vesicles and their functions: a posi-
tion statement from the International Society for Extracellular Vesicles.
J Extracell Vesicles 3: 26913.
76. Cointe S, Judicone C, Robert S, et al. (2017) Standardization of micropar-
ticle enumeration across different flow cytometry platforms: results of a
multicenter collaborative workshop. J Thromb Haemost 15: 187–193.
77. van der Pol E, Hoekstra AG, Sturk A, Otto C, van Leeuwen TG, Nieu-
wland R. (2010) Optical and non-optical methods for detection and
characterization of microparticles and exosomes. J Thromb Haemost 8:
2596–2607.
78. Dragovic RA, Gardiner C, Brooks AS, et al. (2011) Sizing and phenotyping
of cellular vesicles using nanoparticle tracking analysis. Nanomedicine 7:
780–788.
79. Soo CY, Song Y, Zheng Y, et al. (2012) Nanoparticle tracking analysis mon-
itors microvesicle and exosome secretion from immune cells. Immunology
136: 192–197.
80. van der Pol E, Coumans FAW, Sturk A, Nieuwland R, van Leeuwen TG.
(2014) Refractive index determination of nanoparticles in suspension
using nanoparticle tracking analysis. Nano Lett 14: 6195–6201.
81. Vogel R, Coumans FAW, Maltesen RG, et al. (2016) A standardized
method to determine the concentration of extracellular vesicles using tun-
able resistive pulse sensing. J Extracell Vesicles 5: 31242.
82. Smith ZJ, Lee C, Rojalin T, et al. (2015) Single exosome study reveals sub-
populations distributed among cell lines with variability related to mem-
brane content. J Extracell Vesicles 4: 28533.
83. Crescitelli R, Lasser C, Szabó TG, et al. (2013) Distinct RNA profiles in
subpopulations of extracellular vesicles: apoptotic bodies, microvesicles
and exosomes. J Extracell Vesicles 2: 20677.
84. Raiborg C, Stenmark H. (2009) The ESCRT machinery in endosomal sort-
ing of ubiquitylated membrane proteins. Nature 458: 445–452.

b4444_Ch-08.indd 242 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 243

85. Xu R, Greening DW, Rai A, Ji H, Simpson RJ. (2015) Highly-purified

exosomes and shed microvesicles isolated from the human colon cancer
cell line LIM1863 by sequential centrifugal ultrafiltration are biochemi-
cally and functionally distinct. Methods 87: 11–25.
86. Aatonen MT, Ohman T, Nyman TA, Laitinen S, Grönholm M, Siljander
PRM. (2014) Isolation and characterization of platelet-derived extracellu-
lar vesicles. J Extracell Vesicles 3: 24692.
87. Oksvold MP, Kullmann A, Forfang L, et al. (2014) Expression of B-cell
surface antigens in subpopulations of exosomes released from B-cell lym-
phoma cells. Clin Ther 36: 847–862.e1.
88. Yoshioka Y, Konishi Y, Kosaka N, Katsuda T, Kato T, Ochiya T. (2013)
Comparative marker analysis of extracellular vesicles in different human
cancer types. J Extracell Vesicles 2: 20424.
89. Kim D-K, Nishida H, An SY, Shetty AK, Bartosh TJ, Prockop DJ. (2016)
Chromatographically isolated CD63+CD81+ extracellular vesicles from
mesenchymal stromal cells rescue cognitive impairments after TBI. Proc
Natl Acad Sci USA 113: 170–175.
90. Belov L, Matic KJ, Hallal S, Best OG, Mulligan SP, Christopherson RI.
(2016) Extensive surface protein profiles of extracellular vesicles from can-
cer cells may provide diagnostic signatures from blood samples. J Extracell
Vesicles 5: 25355.
91. Samsonov R, Shtam T, Burdakov V, et al. (2016) Lectin-induced aggluti-
nation method of urinary exosomes isolation followed by mi-RNA analysis:
application for prostate cancer diagnostic. Prostate 76: 68–79.
92. Wei Z, Batagov AO, Carter DRF, Krichevsky AM. (2016) Fetal bovine
serum RNA interferes with the cell culture derived extracellular RNA. Sci
Rep 6: 31175.
93. Witwer KW, Buzás EI, Bemis LT, et al. (2013) Standardization of sample
collection, isolation and analysis methods in extracellular vesicle research.
J Extracell Vesicles 2. doi: 10.3402/jev.v2i0.20360. eCollection 2013
94. Lacroix R, Judicone C, Mooberry M, et al. (2013) Standardization of
pre-analytical variables in plasma microparticle determination: results of
the international society on thrombosis and Haemostasis SSC collabora-
tive workshop. J Thromb Haemost 11: 1190–1193.

b4444_Ch-08.indd 243 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

244 Cell-Free Circulating DNA: Purification and Analysis Techniques

95. Beutler E, Gelbart T, Kuhl W. (1990) Interference of heparin with the

polymerase chain reaction. BioTechniques 9: 166.
96. Lippi G, Salvagno GL, Montagnana M, Franchini M, Guidi GC. (2006)
Venous stasis and routine hematologic testing. Clin Lab Haematol 28:
332–337.
97. Hefler L, Grimm C, Leodolter S, Tempfer C. (2004) To butterfly or to
needle: the pilot phase. Ann Intern Med 140: 935–936.
98. Gyorgy B, Pálóczi K, Kovács A, et al. (2014) Improved circulating micro-
particle analysis in acid-citrate dextrose (ACD) anticoagulant tube. Thromb
Res 133: 285–292.
99. Lacroix R, Judicone C, Poncelet P, et al. (2012) Impact of pre-analytical
parameters on the measurement of circulating microparticles: towards
standardization of protocol. J Thromb Haemost 10: 437–446.
100. Coumans FAW, Brisson AR, Buzas EI, et al. (2017) Methodological guide-
lines to study extracellular vesicles. Circ Res 120: 1632–1648.
101. Maroto R, Zhao Y, Jamaluddin M, et al. (2017) Effects of storage tempera-
ture on airway exosome integrity for diagnostic and functional analyses.
J Extracell Vesicles 6: 1359478.
102. Trummer A, De Rop C, Tiede A, Ganser A, Eisert R. (2009) Recovery
and composition of microparticles after snap-freezing depends on thawing
temperature. Blood Coagul Fibrinolysis 20: 52–56.
103. Zarovni N, Corrado A, Guazzi P, et al. (2015) Integrated isolation and
quantitative analysis of exosome shuttled proteins and nucleic acids using
immunocapture approaches. Methods 87: 46–58.
104. Cvjetkovic A, Lotvall J, Lasser C. (2014) The influence of rotor type
and centrifugation time on the yield and purity of extracellular vesicles.
J Extracell Vesicles 3: 23111.
105. Chevillet JR, Kang Q, Ruf IK, et al. (2014) Quantitative and stoichiometric
analysis of the microRNA content of exosomes. Proc Natl Acad Sci USA
111: 14888–14893.
106. Grapp M, Wrede A, Schweizer M, et al. (2013) Choroid plexus transcytosis
and exosome shuttling deliver folate into brain parenchyma. Nat Commun
4: 2123.
107. Miranda KC, Bond DT, Levin JZ, et al. (2014) Massively parallel sequencing
of human urinary exosome/microvesicle RNA reveals a predominance of
non-coding RNA. PLoS ONE 9: e96094.

b4444_Ch-08.indd 244 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Isolating Circulating Exosomes as Biomarkers 245

108. Cheruvanky A, Zhou H, Pisitkun T, et al. (2007) Rapid isolation of urinary

exosomal biomarkers using a nanomembrane ultrafiltration concentrator.
Am J Physiol Renal Physiol 292: F1657–F1661.
109. Liga A, Vliegenthart ADB, Oosthuyzen W, Dear JW, Kersaudy-Kerhoas
M. (2015) Exosome isolation: a microfluidic road-map. Lab Chip 15:
2388–2394.
110. Rider MA, Hurwitz SN, Meckes DG. (2016) ExtraPEG: a polyethylene
glycol-based method for enrichment of extracellular vesicles. Sci Rep 6:
23978.
111. Gyorgy B, Módos K, Pállinger E, et al. (2011) Detection and isolation of
cell-derived microparticles are compromised by protein complexes result-
ing from shared biophysical parameters. Blood 117: e39–e48.
112. Zhao Z, Yang Y, Zeng Y, He M. (2016) A microfluidic ExoSearch chip for
multiplexed exosome detection towards blood-based ovarian cancer diag-
nosis. Lab Chip 16: 489–496.
113. Kanwar SS, Dunlay CJ, Simeone DM, Nagrath S. (2014) Microfluidic
device (ExoChip) for on-chip isolation, quantification and characterization
of circulating exosomes. Lab Chip 14: 1891–900.
114. Kang D, Oh S, Ahn S-M, Lee B-H, Moon MH. (2008) Proteomic analy-
sis of exosomes from human neural stem cells by flow field-flow fraction-
ation and nanoflow liquid chromatography-tandem mass spectrometry.
J Proteome Res 7: 3475–3480.
115. Lee K, Shao H, Weissleder R, Lee H. (2015) Acoustic purification of extra-
cellular microvesicles. ACS Nano 9: 2321–2327.
116. Ding M, Wang C, Lu X, et al. (2018) Comparison of commercial exosome
isolation kits for circulating exosomal microRNA profiling. Anal Bioanal
Chem 410: 3805–3814.
117. Helwa I, Cai J, Drewry MD, et al. (2017) A comparative study of serum
exosome isolation using differential ultracentrifugation and three com-
mercial reagents. PLoS ONE 12: e0170628.
118. EV-TRACK Consortium, Van Deun J, Mestdagh P, et al. (2017)
EV-TRACK: transparent reporting and centralizing knowledge in extra-
cellular vesicle research. Nat Methods 14: 228–232.
119. Van Deun J, Mestdagh P, Sormunen R, et al. (2014) The impact of dis-
parate isolation methods for extracellular vesicles on downstream RNA
profiling. J Extracell Vesicles 3: 24858.

b4444_Ch-08.indd 245 08-Jul-22 6:06:00 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

246 Cell-Free Circulating DNA: Purification and Analysis Techniques

120. Andreu Z, Rivas E, Sanguino-Pascual A, et al. (2016) Comparative analysis

of EV isolation procedures for miRNAs detection in serum samples.
J Extracell Vesicles 5: 31655.
121. Nakai W, Yoshida T, Diez D, et al. (2016) A novel affinity-based method for
the isolation of highly purified extracellular vesicles. Sci Rep 6: 33935.
122. Tauro BJ, Greening DW, Mathias RA, et al. (2012) Comparison of ultra-
centrifugation, density gradient separation, and immunoaffinity capture
methods for isolating human colon cancer cell line LIM1863-derived exo-
somes. Methods 56: 293–304.
123. Arraud N, Linares R, Tan S, et al. (2014) Extracellular vesicles from blood
plasma: determination of their morphology, size, phenotype and concen-
tration. J Thromb Haemost 12: 614–627.
124. Groot Kormelink T, Arkesteijn GJA, Nauwelaers FA, van den Engh G,
Nolte-‘t Hoen ENM, Wauben MHM. (2016) Prerequisites for the analysis
and sorting of extracellular vesicle subpopulations by high-resolution flow
cytometry. Cytometry A 89: 135–147.
125. Arroyo JD, Chevillet JR, Kroh EM, et al. (2011) Argonaute2 complexes
carry a population of circulating microRNAs independent of vesicles in
human plasma. Proc Natl Acad Sci USA 108: 5003–5008.
126. Vickers KC, Palmisano BT, Shoucri BM, Shamburek RD, Remaley AT.
(2011) MicroRNAs are transported in plasma and delivered to recipient
cells by high-density lipoproteins. Nat Cell Biol 13: 423–433.
127. Sódar BW, Kittel A, Pálóczi K, et al. (2016) Low-density lipoprotein mim-
ics blood plasma-derived exosomes and microvesicles during isolation and
detection. Sci Rep 6: 24316.

b4444_Ch-08.indd 246 08-Jul-22 6:06:00 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_0009

Chapter 9
DNA Methylation Analysis
of Circulating DNA
Biomarkers
Kristina Warton, Clare Stirzaker, Goli Samimi and Susan Clark

Abstract

DNA methylation alterations are often more consistent between indi-

vidual tumors than mutations, and offer an alternative target for circu-
lating DNA assays in cancer. Here we summarize different approaches
to methylation analysis, with a focus on their technical parameters in the
context of circulating DNA research.

DNA Methylation
DNA methylation is the presence of a methyl group covalently
linked to the cytosine base in the context of a CpG dinucleotide
(CpG site)1 and provides a layer of epigenetic information spec-
ifying how the DNA sequence may be interpreted in the con-
text of individual cell and tissue biology.2 Approximately 70% of

247

b4444_Ch-09.indd 247 08-Jul-22 6:11:33 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

248 Cell-Free Circulating DNA: Purification and Analysis Techniques

gene promoters are associated with regions that have maintained

CpG sites through evolution, known as CpG islands.3 Generally,
unmethylated CpG island–associated promoters are more loosely
packaged and therefore accessible to transcription factors, such
that the gene can be activated in response to the appropriate stim-
uli. In contrast, methylated CpG island promoters are commonly
more densely packaged and as such prevent accessibility to tran-
scription factor binding and corresponding gene activation. DNA
methylation patterns are established during cell differentiation and
vary by cell type,4 and this epigenetic variation allows signalling
pathways to elicit different transcriptional responses depending
on the tissue and cell type involved. Cell-type DNA methylation
patterns are preserved when DNA is released from necrotic or
apoptotic cells into the circulating cell-free DNA (cirDNA) pool
and is thus a ready source of biomarkers accessible via a “liquid
biopsy” blood draw. In this chapter, we describe different meth-
ods for analyzing methylation of cirDNA at candidate regions that
may have biomarker potential and highlight the technical consid-
erations when assays are applied to cirDNA samples.

cirDNA Methylation as a Cell-of-Origin Biomarker

Since DNA methylation is involved in cell differentiation, indi-
vidual cell types defining a tissue carry distinct methylation foot-
prints.5–7 For this reason, the different cell types contributing to the
cirDNA pool can be identified by the methylation pattern detected
in the cirDNA. DNA methylation analysis from blood samples has
shown that the bulk of cirDNA in healthy individuals is derived
from leukocytes, erythrocyte progenitors, and endothelial cells,
with a minor contribution from hepatocytes.8–11 Tissues compro-
mised by pathological processes experience higher cell death, with

b4444_Ch-09.indd 248 08-Jul-22 6:11:33 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 249

the appearance of corresponding cell-specific DNA methylation

pattern in plasma. For example, patients with recently diagnosed
diabetes show an increase in pancreatic β-cell cirDNA,12 while
hepatocyte cirDNA is increased in patients experiencing rejection
of a liver transplant and sepsis.13

cirDNA Methylation as a Cancer Biomarker

Aberrant DNA methylation patterns are a frequent event associ-
ated with carcinogenesis and commonly occur early in the tumor
cell formation process.5,14 Promoter CpG island methylation is
one of the mechanisms by which tumor suppressor genes are
inactivated, leading to uncontrolled cell division.15 Alterations
in DNA methylation patterns in gene promoters and intergenic
regions are also associated with different characteristics of the
tumor, including histological and molecular subtypes that can
shape the signalling pathways involved in tumorigenesis.16 Can-
cer-related DNA methylation changes are also commonly shared
within a given type of tumor. For example, the HOXA9 promoter
and the EN1 promoter are methylated in 95% and 80% of high-
grade serous ovarian tumors, respectively, and unmethylated in
healthy ovarian epithelium,17 while the 14-3-3 sigma promoter
is methylated in 96% of breast tumors, and unmethylated in
healthy breast tissue.18 Methylation-specific PCR assays that
detect the methylation status of a given promoter are routinely
applied to cohorts of tissue samples.19 A positive outcome of the
PCR is indicative of DNA methylation, with no DNA sequencing
required.
Detection of circulating tumor–derived DNA (ctDNA) in
blood plasma through tumor-specific methylation patterns also
offers an alternative or complementary approach to the detection

b4444_Ch-09.indd 249 08-Jul-22 6:11:33 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

250 Cell-Free Circulating DNA: Purification and Analysis Techniques

of cancer-derived genetic mutations.20 For example, the two genes

most commonly mutated in breast cancer, TP53 and PK3KCA, are
only altered in 37% and 36% of cases, respectively.21 Even when a
cancer is characterized by mutations in a particular gene, as is the
case with TP53 in ovarian cancer,22,23 those mutations are typically
spread over many thousands of base pairs of DNA. Frequent,
consistent point mutations, such as those in KRAS in pancreatic
cancer,20 are the exception rather than the norm. This diversity
provides a challenge for the design of targeted assays to detect
ctDNA in blood. Prior knowledge of specific mutations, obtained
from sequencing tumor tissue DNA, is a straightforward and effec-
tive approach but requires labor-intensive characterization of each
individual patient tumor before the ctDNA assay can be created24;
it is therefore not a strategy applicable in the context of cancer
diagnostic tests. In cancer types that have a degree of consistency
in their mutations, DNA sequencing panels targeting the regions
of interest can be applied directly to cirDNA.25

Analysis of DNA Methylation

A wide range of techniques have been developed for the study
of DNA methylation. While most of these approaches incor-
porate a bisulfite treatment step,19,26,27 enzymatic cytosine con-
version is also possible28 and several approaches are bisulfite
free.29,30 These protocols have been largely developed for the
analysis of relatively abundant, high–molecular weight (HMW)
genomic DNA, and are now being translated with varying suc-
cess to ctDNA analysis. In contrast to genomic DNA, ctDNA is
typically found at low concentrations and is heavily fragmented.
Here, we review the different methods of cirDNA methylation
analysis and consider their advantages and limitations in the
context of investigating cirDNA.

b4444_Ch-09.indd 250 08-Jul-22 6:11:33 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 251

Fig. 1: Conversion of unmethylated cytosines to uracil by bisulfite treatment.

Methylated cytosine does not undergo the reaction as the deamination step is
blocked by the methyl group.

Bisulfite-Based Methods
Bisulfite conversion of DNA has been the backbone of DNA meth-
ylation analysis for over three decades.31–33 Bisulfite reacts with
unmethylated cytosine by nucleophilic attack of the C6 carbon,
and the sulfonate group can then be removed via hydrolytic deam-
ination with alkali treatment to produce uracil (Fig. 1). Methylated
cytosine does not undergo this reaction to an appreciable extent
and remains unchanged. Hence following bisulfite treatment, all
unmethylated cytosines are converted to uracil, while methylated
cytosines remain as cytosine, resulting in sequence differences
between methylated and unmethylated DNA (Fig. 2). The meth-
ylation sites can then be analyzed by PCR and sequencing.

b4444_Ch-09.indd 251 08-Jul-22 6:11:35 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

252 Cell-Free Circulating DNA: Purification and Analysis Techniques

Fig. 2: Bisulfite treatment creates a sequence difference between methylated

and unmethylated DNA after conversion of unmethylated cytosines to uracil.

A large number of bisulfite conversion kits are commercially

available; however, relatively few studies have evaluated these
specifically in the context of cirDNA.34–37 Techniques that bisul-
fite convert cirDNA face the same challenges as techniques that
purify cirDNA, namely, that the target DNA is in low abundance
and fragmented. However, one advantage of cirDNA is that it
undergoes less fragmentation than HMW genomic DNA during
bisulfite treatment, as the short fragments are resistant to further
degradation.38,39
As short DNA fragments do not have the same matrix binding
and recovery characteristics as large DNA fragments (see Chapter 4),
for the purpose of cirDNA studies it is important that methods are
evaluated using DNA input of the appropriate size. For example,
Worm Orntoft and colleagues used a 131-bp custom PCR product
to evaluate the performance of 11 commercial bisulfite conver-

b4444_Ch-09.indd 252 08-Jul-22 6:11:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 253

sion kits and one in-house method.36 They observed a large vari-
ability in kit performance, which was more pronounced for the
small (131 bp) PCR product than large genomic leukocyte DNA.
The EZ Methylation-Direct kit from Zymo Research was the high-
est rated kit, with the Epijet kit from ThermoFisher Scientific a
close second. Holmes and colleagues compared nine different
bisulfite conversion kits, manufactured by three different biotech
companies34; however, this study only used FFPE DNA, also a
fragmented substrate, in the direct kit comparisons. The kits that
were selected all performed with similar efficiency; however, a
decrease in the recovery of FFPE bisulfite-treated DNA com-
pared to tissue DNA was reported.34 Only one kit, the innuCon-
vert Bisulphite Body Fluids Kit, was tested for performance with
plasma cirDNA samples and yielded ~150 ng cirDNA per mL of
plasma. This unexpectedly high yield may be due to leukocyte
lysis during pre-processing blood storage (Chapter 2), as sample
blood sample handling parameters are not reported in the study.
Finally, a study used qPCR to compare recovery of cirDNA that
was bisulfite converted using either the EpiTect or the EpiTect
Fast kits from Qiagen and found that the EpiTect Fast resulted
in approximately double the yield as well as allowing more rapid
sample processing.38

PCR of Bisulfite-Converted cirDNA

Following bisulfite conversion, DNA can be selectively amplified
and quantitated by PCR. Bisulfite-treated DNA templates are
generally more challenging to amplify; as in the initial PCR cycles,
the DNA polymerase must contend with uracil instead of thymine
in the DNA backbone,40 and runs of consecutive dTs are more
frequent following the conversion of unmethylated cytosines.41

b4444_Ch-09.indd 253 08-Jul-22 6:11:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

254 Cell-Free Circulating DNA: Purification and Analysis Techniques

In the following, we address considerations for PCR of bisulfite-

converted cirDNA. Despite these challenges, PCR is commonly
used to analyze DNA methylation in tissue and cirDNA; indeed, a
PCR assay for the methylated SEPT9 promoter in cirDNA forms
the basis of an FDA-approved blood test for colorectal cancer.42
Blood plasma tests for methylated cirDNA biomarkers for hepa-
tocellular carcinoma and lung cancer are also regulatory agency
approved (reviewed by Ref. (19)).19

(i) Amplicon size

Amplicon size is an important consideration in any cirDNA PCR
design (Chapter 5) but even more so for bisulfite-treated DNA
and cirDNA. cirDNA occurs mainly as fragments ~167 bp in size,
hence PCR assays targeting sequences above this size will have
limited available DNA template. Below 167 bp, assay sensitivity
progressively increases with decreasing amplicon size, reflecting
a higher number of available intact targets.43 Bisulfite treatment
leads to fragmentation and loss of template44; hence small ampli-
cons are important for maximizing available target sequence; how-
ever, a recent study has shown that cirDNA has some resistance
to fragmentation during bisulfite treatment, most likely due to its
shorter length.38 There is a lower limit to PCR product size, which
must accommodate the length of both primers, and often a probe
sequence. Where the PCR product is to be analyzed by gel elec-
trophoresis, the size of the product must be distinguishable from
the size of any potential primer dimer.

(ii) Effect of nucleosome positioning

Regions of differential methylation that regulate gene expres-
sion are typically found in CpG islands associated with gene pro-
moters.45 Nucleosomes at the start sites of active genes are not
randomly distributed; rather, they are regularly and consistently

b4444_Ch-09.indd 254 08-Jul-22 6:11:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 255

positioned with a nucleosome-free region at the transcription start

site (TSS).46 Indeed, the cirDNA fragmentation pattern at TSS
sites has allowed researchers to determine whether a particular
gene is active in the cells that contributed to the cirDNA pool and
the identification of the cells of origin.46
Due to the regular nucleosome spacing at TSS sites, care
must be taken when designing methylation-specific primers
(MSP) at gene promoters. If the PCR assay spans the TSS of an
active gene, a negative result may not reflect the methylation sta-
tus of the gene but rather the fact the DNA is cleaved at the TSS
and thus not available as a template for PCR. For example, before
the fragmentation pattern of cirDNA (that stems from the regu-
lar spacing of nucleosomes at TSS sites) had been described in
the literature, Jahr and colleagues designed a PCR assay to test
whether endothelial cells contributed DNA to the cirDNA pool.47
This assay targeted the unmethylated E-selectin promoter, known
to be expressed specifically in endothelial cells.48 The assay suc-
cessfully amplified genomic endothelial cell DNA but generated
no PCR product when cirDNA was used as the template, leading
to the conclusion that endothelial cell DNA was not present in the
cirDNA pool. However, this negative result was artifactual, and
due to the primers spanning the TSS. Fifteen years later, Moss and
colleagues used Illumina methylation array data to show that endo-
thelial cells contributed approximately 10% of the total cirDNA in
healthy individuals.8

(iii) Primer design

Methylation-specific PCR: Depending on the needs of the
assay, PCR of bisulfite-converted DNA can be designed to spe-
cifically amplify either the methylated (MSP) or unmethylated
sequence (USP), or amplify both unmethylated and methylated
sequences nonselectively. Methylation-specific PCR is depen-

b4444_Ch-09.indd 255 08-Jul-22 6:11:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

256 Cell-Free Circulating DNA: Purification and Analysis Techniques

dent on mismatches where the primers overlap CpG sites. In

theory, and occasionally in practice,49 a single CpG site resulting
in a primer-template mismatch is sufficient to make the primers
specific for a selected methylation state, but this is dependent on
individual sequence characteristics such as annealing temperature
bias. In general, at least three mismatches, at CpG sites, between
each of the primers and the undesired template provide a sufficient
basis for developing a specific PCR reaction, but once again, this is
dependent on individual sequence. A critical consideration is the
positioning of the mismatches within the primer, with mismatches
within the 3′-most bases providing the highest specificity.49,50
Unbiased PCR: Unbiased PCR, that is PCR amplifying both the
methylated and the unmethylated target with equal efficiency, is car-
ried out when the relative proportions of the methylated and unmet-
hylated region are to be determined by a downstream technique,
such as next-generation sequencing or clonal sequencing. Equal
amplification of both targets is difficult as the base pair composition
post-bisulfite treatment is very different. While the primers may be
designed to be unbiased, PCR bias can occur due to the inherent
difference in sequence of methylated and unmethylated strands.51
PCR conditions need to be optimized and primers ideally are
designed such that they do not span sequences containing CpG sites
or have an inosine residue to pair with the C of the CpG site.52
In methylation analysis, the DNA molecules of interest can
occur in a background excess of opposite methylation status DNA.
This is particularly the case when detecting methylated tumor
DNA in plasma or DNA contributions from rare cell types. In this
scenario, the PCR reaction needs to not only efficiently amplify the
desired target but also avoid the opposite methylation status DNA,
which has a very similar sequence. DNA duplex stability decreases
with increasing temperature and decreasing magnesium, so the

b4444_Ch-09.indd 256 08-Jul-22 6:11:36 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 257

unspecific PCR reaction with the primer mismatch can be made

less efficient, or eliminated by raising annealing temperature or
lowering magnesium. However, this can also decrease the effi-
ciency of the perfectly matched primer reaction, so developing
methylation-specific PCR assays involves systematically identify-
ing temperature/magnesium combinations at which priming of the
nonspecific target does not occur while retaining sufficient reac-
tion efficiency for the specific target. Selected PCR additives can
also be empirically tested for their effect on the specificity and
efficiency of the reaction, with dimethyl sulfoxide having proved
useful for aiding the amplification of GC-rich templates.53
“Alternate” strand PCR assay design: Following bisulfite treatment,
there are two distinct DNA sequences available for assays testing
the methylation status of a region of interest. In an unmethylated
C–G pair, only the C is modified by bisulfite treatment, while the
G remains unchanged, forming at T–G mismatch. Thus the two
DNA strands are no longer complementary and offer two different
targets, each of which can be used to determine the methylation
status of the original (Fig. 3). In cases where one DNA strand is
not amenable to PCR, either because a PCR assay with a suitable
annealing temperature and favorable self-complementarity param-
eters cannot be designed, or because the PCR can be designed but
fails in laboratory testing, it is well worth attempting to design an
assay that utilizes the other strand. Thus, each region does offer
two options at designing specific and efficient PCRs.

Evaluation of Methylation Following PCR Amplification of

Bisulfite-converted cirDNA
(i) Visualizing the PCR product
The PCR product can be visualized by traditional gel electrophore-
sis or by one of the more recent capillary electrophoresis methods

b4444_Ch-09.indd 257 08-Jul-22 6:11:36 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

258 Cell-Free Circulating DNA: Purification and Analysis Techniques

(a)

(b)

(c)

Fig. 3: Following bisulfite conversion, the two DNA strands are no longer
complementary.

that also rely on matrix-based DNA separation by size (e.g., Bio-

analyzer, TapeStation, Qiaxel, Advanced Analytical Fragment Sep-
arator). However, while these methods can determine whether the
product is of the expected size, they cannot differentiate between
amplification of methylated and unmethylated sequences, as both
generate a product of the same length, and additional measures to
determine which product was amplified are required.

(ii) Quantitative PCR

Quantitative PCR (qPCR) techniques are equally applicable
to bisulfite-treated DNA and bisulfite-untreated DNA. DNA
intercalating dyes such as SYTO9 or SYBR Green can be used
to quantitate the target of interest, with the additional advantage
that the melt curve will vary with the methylation status of the

b4444_Ch-09.indd 258 08-Jul-22 6:11:38 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 259

amplified product, thus providing some information regarding the

specificity of the reaction. In general, the methylated product will
have a higher cytosine content, hence a higher melt temperature;
however, this does not apply in cases where differentially meth-
ylated cytosines are only found within the primer sequence and
not in the region between the primers. This is because every PCR
product incorporates primer molecules, and all but the few initial
template DNA fragments are copied from primers. Hence, the
sequence of a PCR product within the primer regions will always
match the primers, even if the product was initially generated by
a nonspecific mispriming event. For this reason, only the differen-
tially methylated CpG sites that lie between primers contribute to
a melt temperature difference between methylated and unmeth-
ylated DNA (Fig. 4).
Methylation-specific qPCR assays that incorporate a probe
sequence between the primers (MethylLight) can also be used to
quantitate the target template and monitor reaction specificity, as
the probe can be designed to overlap CpG sites thereby selec-
tively visualizing amplification of the sequence of interest, and not
the related, mismatched DNA. This approach has commonly been
used to detect tumor DNA in the cell-free DNA pool.54,55 Alter-
nately to MethylLight conversion-specific detection of DNA meth-
ylation can be achieved using real-time PCR (ConLight-MSP) to
avoid false positives that commonly occur in DNA derived from
clinical samples.56

(iii) Digital PCR

Digital PCR involves partitioning the PCR reaction into tens of
thousands of individual reactions and counting how many of these
contain amplified product once cycling is finished.57 Amplified
product indicates the presence of a specific target molecule in the

b4444_Ch-09.indd 259 08-Jul-22 6:11:38 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

260 Cell-Free Circulating DNA: Purification and Analysis Techniques

(a)

(b)

Fig. 4: Mispriming in methylation-specific qPCR (a) When CpG sites with dif-
ferent methylation status are present between the primers, the misprimed prod-
uct can be distinguished by a different melt temperature. (b) When there are no
CpG sites with different methylation status between the primers, the misprimed
and the correct PCR product cannot be distinguished by the melt curve.

b4444_Ch-09.indd 260 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 261

original mix, so it is effectively counting individual DNA targets.

The method does not require a standard curve and provides an
increase in sensitivity over traditional qPCR methods. It is particu-
larly suited to cirDNA applications in which low amounts of target
template are often a technical hurdle.
An additional benefit of digital PCR relates to the aforemen-
tioned difficulty in detecting rare molecules in a high background
of DNA with the opposite methylation status. As described, using
a probe can selectively visualize amplification of the correct tar-
get, but it cannot prevent amplification of the incorrect sequence,
which may, if sufficient, decrease the efficiency of the correct
reaction, leading to low sensitivity. The partitioning in digital PCR
means that in those droplets that contain the correct target, the
ratio of correct target to related sequence is much higher; hence
inhibition due to related sequence amplification will be less pro-
nounced. In those droplets that contain only the incorrect, related
sequence, some amplification may take place, but there will be no
fluorescence due to the specificity of the probe.

(iv) Methyl-BEAMing
Methyl-BEAMing (Beads Emulsion Amplification Magnetics)
is a digital approach to detect and quantitate DNA methylation
in samples, particularly applicable when the methylated mole-
cules are present in a large excess of unmethylated background
DNA.58 The procedure involves methylation-specific amplification
of bisulfite-converted DNA in aqueous compartments suspended
within an oil phase, which also includes beads that bind ampli-
fied PCR product. After amplification is complete, the beads, now
coated with selectively amplified DNA, are labelled with hybrid-
ization probes and counted by flow cytometry.58 This method has
been used to detect methylated vimentin in plasma from colorectal

b4444_Ch-09.indd 261 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

262 Cell-Free Circulating DNA: Purification and Analysis Techniques

cancer patients58 and used to evaluate drug sensitivity in colorectal

cancer.59,60

Additional Methodologies to Increase the Selectivity of

Methylation-Specific PCR
While dye-based and probe-based methods can help deter-
mine whether the amplified product is methylated or unmet-
hylated, neither of them contribute to PCR reaction specificity.
As described, cirDNA molecules with the methylation status of
interest often occur in a background excess of molecules with
the opposite methylation status. For a given region of interest, it
may only be possible to identify a temperature and magnesium
combination in which the amplification of the related sequence is
reduced but not eliminated. Additional strategies can then be used
to suppress the amplification of the related sequence and create a
highly specific PCR assay.

(i) Headloop suppression PCR

The headloop suppression PCR strategy61 can either be used on
its own to selectively amplify DNA with the desired methyla-
tion status, or as an adjunct to increase the specificity of MSP.17
In this method, one or both of the primers contain an additional
sequence at the 5′ end, which is complementary to a sequence
within the undesired PCR product. In the event that mispriming
and DNA extension occurs, the PCR molecule transcribed con-
tains a self-complementary sequence and is able to form a hairpin
loop structure that blocks further priming. Because this intramo-
lecular binding is much more efficient than primer annealing, the
misprimed sequence is effectively removed from further partici-
pating in the PCR reaction. This mechanism is sufficiently efficient

b4444_Ch-09.indd 262 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 263

that it can give selective amplification even in the absence of

methylation-specific sequences within the actual primer, and sen-
sitive down to around 25 pg of target DNA.61 This method has been
successfully used to monitor prostate cancer patients receiving
chemotherapy by tracking cirDNA-methylated GSTP1 levels.62,63

(ii) HeavyMethyl
Another strategy to prevent the amplification of the unmethylated
DNA sequence, known as HeavyMethyl, uses a methylation-specific
oligonucleotide blocker.64 The nonextendable blockers bind within
the primer binding sites and prevent the binding of the primers.
Together with methylation-specific fluorescent probes, this can be
sufficient to selectively amplify methylated DNA even when the
primer does not favor the methylated sequence.64 This approach,
together with a MethylLight probe, was successfully used in the
SEPT9 assay for the detection of methylated colorectal cancer
DNA in patient plasma.65

(iii) Ms-SNuPE
Ms-SNuPE is a variation of SNuPE, or Single-Nucleotide Primer
Extension, which interrogates the nucleotide present at a selected
position to genotype single-nucleotide polymorphisms.66 In the
SNuPE method, the DNA region is amplified by PCR, then
annealed to a primer that terminates one base short of the posi-
tion of interest. The single base incorporated into that position
can then be identified by using either radioactively or fluores-
cently labelled nucleotides. In Ms-SNuPE, the same approach is
followed, but the DNA is first bisulfite treated so that differences
in sequence are not due to allelic variation but due to methylation
status.67 Ms-SNuPE has been combined with HeavyMethyl PCR
to detect colorectal cancer–specific methylation in cirDNA.68

b4444_Ch-09.indd 263 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

264 Cell-Free Circulating DNA: Purification and Analysis Techniques

Illumina Methylation Arrays

Illumina methylation arrays interrogate the methylation sta-
tus of selected individual CpG sites. Unlike the PCR-based
methods described previously, the methylation arrays are con-
sidered a “whole genome” analysis approach. Early versions of
the arrays targeted around 27,000 individual CpG sites (27K
Array), mostly positioned within promoter CpG islands. The
throughput was then increased to approximately 450,000 CpG
sites and most recently, 850,000 CpG sites (HM450K Array and
MethylationEPIC BeadChip, respectively). The EPIC array
has broader coverage and includes a greater range of genomic
elements in which methylation is known to play a regulatory
role.69 The array method is based on the Infinium genotyping
assay but interrogates differences at CpG sites, rather than
single-nucleotide polymorphisms.70 Bisulfite conversion and
whole genome amplification of DNA are followed by binding to
bead-coded complementary DNA probes, which stop one base
short of the residue of interest. A polymerase then extends the
bead bound probe by a single residue, which can be identified
by fluorescent labelling, thus revealing the methylation state of
the CpG site.
Along with technical developments that increase the num-
ber of methylation sites analyzed, there has been a progres-
sive reduction in the amount of input DNA required, with the
MethylationEPIC BeadChip specifying 250 ng input DNA. Moss
and colleagues were able to reach the 250 ng input requirement
by using large blood draws (20 mL) and pooling cirDNA from
multiple donors. This allowed them to characterize genome-wide
methylation in cirDNA and identify the dominant tissue(s) that
contribute to the cirDNA pool in healthy individuals.8

b4444_Ch-09.indd 264 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 265

Bisulfite Sequencing of CirDNA

Analysis of cirDNA methylation following bisulfite treatment
can also be identified directly by next-generation sequencing.
Sequencing can analyze either the whole genome or targeted
regions of interest. As with other methods in cirDNA research,
the challenge in applying this approach is that sequencing
protocols typically require DNA input quantities that are not eas-
ily obtained from biological samples, with whole genome bisul-
fite sequencing (WGBS) generally requiring more input DNA.
WGBS of cirDNA has been successfully applied; for example,
Jensen and colleagues applied WGBS to cirDNA from pregnant
women to study the DNA fraction derived from the placenta,71
while Sun and colleagues used it to identify the tissues that con-
tribute to the cirDNA pool in pregnant women, and in patients
with liver cancer and follicular lymphoma.11 Other studies using
next-generation sequencing of bisulfite-treated DNA have used a
target approach; Lehmann-Werman and colleagues used targeted
parallel sequencing of three hepatocyte markers to observe the
increase in liver-derived cirDNA in patients following liver trans-
plantation and patients with sepsis.13 Similarly, Lam et al. reported
next-generation sequencing of targeted panels to detect DNA
methylation in cirDNA.26

Bisulfite-Free Methods for CirDNA Methylation

Evaluation
Although the majority of published studies of cirDNA methyl-
ation rely on methods incorporating a bisulfite conversion step,
bisulfite-free methods have also been applied. These methods
have the advantage that the degradation of DNA resulting from

b4444_Ch-09.indd 265 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

266 Cell-Free Circulating DNA: Purification and Analysis Techniques

bisulfite treatment is avoided; however, they do generally require

higher DNA inputs than the commonly used methylation-specific
PCR following bisulfite treatment. Nonetheless, they provide an
alternative approach.

Methylation-Sensitive Restriction Enzyme Digest

One bisulfite-free approach is to restrict the DNA included in
the analysis to the methylated target sequence by removing the
unmethylated sequence with a methylation-sensitive restric-
tion enzyme (MSRE). For example, HpaII and Hin6I only cut
unmethylated DNA, effectively removing it from the sample. The
remaining methylated target can then be quantitated by standard
qPCR or analyzed by other methods. As mentioned previously, the
sample does not undergo the additional fragmentation and DNA
loss associated with bisulfite treatment; however, targets are lim-
ited to those DNA regions that contain the appropriate restriction
site. This type of technique has been applied to colorectal cancer,72
breast cancer,73 and lymphoma.74
The MethDet56 assay is based on enrichment of the methyl-
ated target by MSRE digest, followed by PCR amplification and
microarray quantification.75 Variations on the method employing
methylation capture rather than MSRE, and readouts other than
microarray have also been proposed.75 The assay works on as little
as 200 pg of input DNA, making it well suited to cirDNA analy-
ses75; however, as with other MSRE-based techniques, it is limited
by the locations of restriction enzyme sites. The MethDet56 assay
has been applied to measuring pre- and post-treatment cirDNA
changes in ER-positive breast cancer patients,76 differentiating
between pancreatitis and pancreatic cancer77 and between benign
and malignant ovarian tumors.78 One weakness of the MethDet56

b4444_Ch-09.indd 266 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 267

assay is that with 56 methylation targets, large numbers of patient

samples are needed to achieve statistical significance after correc-
tion for multiple comparisons. However, this can be addressed by
follow-up studies once a more limited set of candidate biomarkers
has been identified by the initial exploratory analysis.

Methylation Capture and Sequencing

Selective capture of methylated DNA followed by sequencing can
identify the methylated DNA regions in samples. This approach
has been widely applied to tissues but has been more challenging
in cirDNA samples due to the high DNA inputs required. Shen
and colleagues successfully carried out cell-free methylated DNA
precipitation and high-throughput sequencing (cfMeDIP-seq) to
identify methylated cirDNA regions in cancer patient plasma and
infer the site tumor origin.79 A key step in developing the protocol
was optimizing the method for low DNA inputs. In control exper-
iments, the authors were able to use as little as 1–10 ng of sheared
input DNA and recapitulate results obtained from standard
Me-DIP, reduced representation bisulfite sequencing (RRBS),
and WGBS, which require 100 ng, 1000 ng, and 2000 ng of DNA,
respectively.
A methylation capture-and-sequence approach based on
methyl-binding protein (MethylMiner, Invitrogen) has been
applied to cirDNA, but, once again, the standard protocol used
for genomic DNA from tissue was adjusted to accommodate the
low inputs of available cirDNA.29 Specifically, the protocol was
scaled down to limit nonspecific binding to excess bead surface
area, and the stringency of the wash steps was increased. Success-
ful next-generation sequencing was carried out using only 50 ng
of cirDNA input per sample, with the low DNA input mitigated

b4444_Ch-09.indd 267 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

268 Cell-Free Circulating DNA: Purification and Analysis Techniques

by avoiding the loss of sample that occurs at the gel purification

size restriction step, which has been estimated to be up to 95%.80
Clonal sequencing was used to confirm that the method was spe-
cific in identifying methylated and unmethylated DNA regions.29

Tissue Selection to Identify Methylated cirDNA Cancer

Biomarkers
Cancer-related DNA methylation changes can be identified by
comparing tumor methylation with adjacent normal tissue. Since
the quantity of DNA obtained from tissues is generally high, lab-
oratory techniques for biomarker discovery in tissue are not lim-
ited by the amount of available input material. However, for the
purpose of methylated cirDNA biomarker discovery, the DNA
methylation of the healthy tissue-of-origin is not relevant if that
tissue does not contribute to the cirDNA pool in healthy individ-
uals. For example, 14-3-3 sigma is methylated in 96% of breast
cancers, and unmethylated in healthy breast epithelium; how-
ever, it cannot be used as a ctDNA biomarker because it is also
methylated in leukocytes.18
The appropriate negative controls for excluding DNA meth-
ylation present in cirDNA from healthy individuals are tissues that
normally contribute to cirDNA, but, until recently, it wasn’t known
what those tissues were. Leukocytes were the obvious candidate
for a major cirDNA contributor, and experimentally this has shown
to be the case.47 However, erythrocyte progenitors,8,47 endothelial
cells,8 and, less obviously, hepatocytes,8,11,13 also contribute signif-
icant amounts, with erythrocyte progenitors and endothelial cells
together accounting for around 40% of the total.8 This is an exam-
ple of how a good understanding of cirDNA biology in healthy
individuals can inform and advance biomarker discovery. Further

b4444_Ch-09.indd 268 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 269

work is required to characterize how the cirDNA pool changes

in various healthy physiological states (e.g., following exercise and
during menstruation), and during common minor illnesses (e.g.,
respiratory tract infection), if a high specificity of cancer biomark-
ers is to be achieved from tissue-based discovery strategies.
One benefit of not using the healthy counterpart of the tumor
tissue to exclude biomarkers is that it allows inclusion of regions
that are methylated in the healthy tissue but don’t normally appear
in the cirDNA pool. In that scenario, the biomarker would not be
indicative of a cancer-associated methylation pattern, but rather,
of a particular tissue experiencing abnormal cell death and turn
over, suggestive of carcinogenesis.
Tissue-based discovery strategies generally allow access to
large amounts of DNA that lend themselves to genome-wide anal-
yses.81 Promising biomarker candidates identified in tissue can
then be evaluated in plasma samples by methylation-specific PCR,
which can test a smaller number of candidates with much greater
sensitivity. However, with genome-wide methods being refined to
accommodate ever-smaller amounts of input DNA, biomarker dis-
covery directly from plasma is also feasible.79 The advantage of this
strategy is that tissues contributing to the cirDNA pool in healthy
control subjects do not need to be separately accounted for and
analyzed.

Conclusions
Analysis of cirDNA methylation can generate information about
disease states as diverse as cancer, cardiac infarction, and diabetes.
A colorectal cancer diagnostic test based on circDNA methylation
is already FDA approved and available to patients, and a test for
diagnosing and monitoring other cancers are in the development

b4444_Ch-09.indd 269 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

270 Cell-Free Circulating DNA: Purification and Analysis Techniques

pipeline. As in other areas of cirDNA research, there are gains

in sensitivity and specificity to be made from improved analy-
sis techniques, which will in turn lead to more effective clinical
applications.

References
1. Bestor TH. (1988) Cloning of a mammalian DNA methyltransferase. Gene
74: 9–12.
2. Greenberg MVC, Bourc’his D. (2019) The diverse roles of DNA methyl-
ation in mammalian development and disease. Nat Rev Mol Cell Biol 20:
590–607.
3. Saxonov S, Berg P, Brutlag DL. (2006) A genome-wide analysis of CpG dinu-
cleotides in the human genome distinguishes two distinct classes of promot-
ers. Proc Natl Acad Sci U S A 103: 1412–1417. doi:10.073/pnas.0510310103.
Epub 2006 Jan 23.
4. Bergman Y, Cedar H. (2013) DNA methylation dynamics in health and dis-
ease. Nat Struct Mol Biol 20: 274–281. doi: 10.1038/nsmb.2518.
5. Dor Y, Cedar H. (2018) Principles of DNA methylation and their implica-
tions for biology and medicine. Lancet 392: 777–786. doi:10.1016/S0140-
6736(18)31268-6. Epub 2018 Aug 9.
6. Jones PA. (2012) Functions of DNA methylation: Islands, start sites, gene
bodies and beyond. Nat Rev Genet 13: 484–492. doi:10.1038/nrg3230.
7. Schubeler D. (2015) Function and information content of DNA methyla-
tion. Nature 517: 321–326. doi:10.1038/nature14192.
8. Moss J, Magenheim J, Neiman D, et al. (2018) Comprehensive human cell-
type methylation atlas reveals origins of circulating cell-free DNA in health
and disease. Nat Commun 9: 5068. doi:10.1038/s41467-018-07466-6.
9. Guo S, Diep D, Plongthongkum N, Fung HL, Zhang K, Zhang K. (2017)
Identification of methylation haplotype blocks aids in deconvolution of het-
erogeneous tissue samples and tumor tissue-of-origin mapping from plasma
DNA. Nat Genet 49: 635–642. doi:10.1038/ng.3805. Epub 2017 Mar 6.
10. Lam WKJ, Gai W, Sun K, et al. (2017) DNA of erythroid origin is present in
human plasma and informs the types of Anemia. Clin Chem 63: 1614–1623.
doi:10.373/clinchem.2017.272401. Epub 2017 Aug 7.

b4444_Ch-09.indd 270 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 271

11. Sun K, Jiang P, Chan KC, et al. (2015) Plasma DNA tissue mapping by
genome-wide methylation sequencing for noninvasive prenatal, cancer, and
transplantation assessments. Proc Natl Acad Sci U S A 112: E5503–E5512.
doi: 10.1073/pnas.1508736112. Epub 2015 Sep 21.
12. Lehmann-Werman R, Neiman D, Zemmour H, et al. (2016) Identification
of tissue-specific cell death using methylation patterns of circulating DNA.
Proc Natl Acad Sci U S A 113: E1826–34. doi: 10.073/pnas.1519286113.
Epub 2016 Mar 14.
13. Lehmann-Werman R, Magenheim J, Moss J, et al. (2018). Monitoring liver
damage using hepatocyte-specific methylation markers in cell-free circulat-
ing DNA. JCI Insight 3: 120687. doi:10.1172/jci.insight
14. Jones PA, Baylin SB. (2007) The epigenomics of cancer. Cell 128: 683–692.
doi:10.1016/j.cell.2007.01.029.
15. Clark SJ, Melki J. (2002) DNA methylation and gene silencing in cancer:
Which is the guilty party? Oncogene 21: 5380–5387.
16. Lee ST, Wiemels JL. (2016) Genome-wide CpG island methylation
and intergenic demethylation propensities vary among different tumor
sites. Nucleic Acids Res 44: 1105–1117. doi: 10.093/nar/gkv038. Epub 2015
Oct 12.
17. Montavon C, Gloss BS, Warton K, et al. (2012) Prognostic and diagnostic
significance of DNA methylation patterns in high grade serous ovarian can-
cer. Gynecol Oncol 124: 582–588.
18. Umbricht CB, Evron E, Gabrielson E, Ferguson A, Marks J, Sukumar S.
(2001) Hypermethylation of 14-3-3 sigma (stratifin) is an early event in
breast cancer. Oncogene 20: 3348–3353.
19. Locke WJ, Guanzon D, Ma C, et al. (2019) DNA methylation cancer bio-
markers: Translation to the clinic. Front Genet 10: 1150.
20. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA, Jr.,
Kinzler KW. (2013) Cancer genome landscapes. Science 339: 1546–1558.
21. Network TCGA. (2012) Comprehensive molecular portraits of human
breast tumours. Nature 490: 61–70. doi: 10.1038/nature11412. Epub 2012
Sep 23.
22. Ahmed AA, Etemadmoghadam D, Temple J, et al. (2010) Driver mutations
in TP53 are ubiquitous in high grade serous carcinoma of the ovary. J Pathol
221: 49–56. doi: 10.1002/path.2696.

b4444_Ch-09.indd 271 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

272 Cell-Free Circulating DNA: Purification and Analysis Techniques

23. Kandoth C, Schultz N, Cherniack AD, et al. (2013) Integrated genomic

characterization of endometrial carcinoma. Nature 497: 67–73. doi: 10.1038/
nature12113.
24. Diehl F, Schmidt K, Choti MA, et al. (2008) Circulating mutant DNA to
assess tumor dynamics. Nat Med 14: 985–990.
25. Forshew T, Murtaza M, Parkinson C, et al. (2012) Noninvasive identification
and monitoring of cancer mutations by targeted deep sequencing of plasma
DNA. Sci Transl Med. 4: 136ra68. doi: 10.1126/scitranslmed.3003726.
26. Lam D, Luu PL, Song JZ, et al. (2020) Comprehensive evaluation of
targeted multiplex bisulphite PCR sequencing for validation of DNA meth-
ylation biomarker panels. Clin Epigenetics 12: 90.
27. Li Y, Tollefsbol TO. (2011) DNA methylation detection: Bisulfite genomic
sequencing analysis. Methods Mol Biol 791: 11–21.
28. Feng S, Zhong Z, Wang M, Jacobsen SE. (2020) Efficient and accurate
determination of genome-wide DNA methylation patterns in Arabidopsis
thaliana with enzymatic methyl sequencing. Epigenetics Chromatin 13: 42.
29. Warton K, Lin V, Navin T, et al. (2014) Methylation-capture and
next-generation sequencing of free circulating DNA from human plasma.
BMC Genomics 15: 476. doi: 10.1186/471-2164-15-476
30. Nair SS, Coolen MW, Stirzaker C, et al. (2011) Comparison of methyl-DNA
immunoprecipitation (MeDIP) and methyl-CpG binding domain (MBD)
protein capture for genome-wide DNA methylation analysis reveal CpG
sequence coverage bias. Epigenetics 6: 34–44.
31. Frommer M, McDonald LE, Millar DS, et al. (1992) A genomic sequencing
protocol that yields a positive display of 5-methylcytosine residues in indi-
vidual DNA strands. Proc Natl Acad Sci U S A 89: 1827–1831.
32. Clark SJ, Harrison J, Paul CL, Frommer M. (1994) High sensitivity map-
ping of methylated cytosines. Nucleic Acids Res 22: 2990–2997.
33. Clark S, Frommer M. (1995) Deamination with NaHSO3 in DNA methylation
studies in: DNA and nucleoprotein structure in vivo. Biomedical Publishers. RG
Landes Company, Austin & Springer-Verlag, New York, pp. 123–132.
34. Holmes EE, Jung M, Meller S, et al. (2014) Performance evaluation of kits
for bisulfite-conversion of DNA from tissues, cell lines, FFPE tissues, aspi-
rates, lavages, effusions, plasma, serum, and urine. PLOS ONE 9: e93933.
doi: 10.1371/journal.pone.0093933. eCollection 2014.

b4444_Ch-09.indd 272 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 273

35. Leontiou CA, Hadjidaniel MD, Mina P, Antoniou P, Ioannides M, Patsalis PC.
(2015) Bisulfite conversion of DNA: Performance comparison of differ-
ent kits and methylation quantitation of epigenetic biomarkers that have
the potential to be used in non-invasive prenatal testing. PLOS ONE 10:
e0135058. doi: 10.1371/journal.pone.0135058.
36. Worm Orntoft MB, Jensen SO, Hansen TB, Bramsen JB, Andersen CL. (2017)
Comparative analysis of 12 different kits for bisulfite conversion of circulating cell-
free DNA. Epigenetics 12: 626–636. doi: 10.1080/15592294.2017.1334024.
Epub 2017 May 30.
37. Yi S, Long F, Cheng J, Huang D. (2017) An optimized rapid bisulfite con-
version method with high recovery of cell-free DNA. BMC Mol Biol 18: 24.
doi: 10.1186/s12867-017-0101-4.
38. Werner B, Yuwono NL, Henry C, et al. (2019) Circulating cell-free DNA
from plasma undergoes less fragmentation during bisulfite treatment than
genomic DNA due to low molecular weight. PLOS ONE 14: e0224338. doi:
10.1371/journal.pone.. eCollection 2019.
39. Millar DS, Warnecke PM, Melki JR, Clark SJ. (2002) Methylation
sequencing from limiting DNA: Embryonic, fixed, and microdissected cells.
Methods 27: 108–113.
40. Millar D, Christova Y, Holliger P. (2015) A polymerase engineered for bisul-
fite sequencing. Nucleic Acids Res 43: e155. doi: 10.1093/nar/gkv798. Epub
2015 Aug 13.
41. Warnecke PM, Stirzaker C, Song J, Grunau C, Melki JR, Clark SJ. (2002)
Identification and resolution of artifacts in bisulfite sequencing. Methods
27: 101–107.
42. Church TR, Wandell M, Lofton-Day C, et al. (2002) Prospective evalua-
tion of methylated SEPT9 in plasma for detection of asymptomatic colorec-
tal cancer. Gut 63: 317–325. doi: 10.1136/gutjnl-2012-304149. Epub 2013
Feb 13.
43. Andersen RF, Spindler KL, Brandslund I, Jakobsen A, Pallisgaard N. (2015)
Improved sensitivity of circulating tumor DNA measurement using short
PCR amplicons. Clin Chim Acta 439: 97–101.
44. Grunau C, Clark SJ, Rosenthal A. (2001) Bisulfite genomic sequencing: Sys-
tematic investigation of critical experimental parameters. Nucleic Acids Res
29: E65–E5.

b4444_Ch-09.indd 273 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

274 Cell-Free Circulating DNA: Purification and Analysis Techniques

45. Jones PA, Baylin SB. (2002) The fundamental role of epigenetic events in
cancer. Nat Rev Genet 3: 415–428.
46. Snyder MW, Kircher M, Hill AJ, Daza RM, Shendure J. (2016)
Cell-free DNA comprises an in vivo nucleosome footprint that informs its
tissues-of-origin. Cell 164: 57–68.
47. Jahr S, Hentze H, Englisch S, et al. (2001) DNA fragments in the blood
plasma of cancer patients: Quantitations and evidence for their origin from
apoptotic and necrotic cells. Cancer Res 61: 1659–1665.
48. Smith GM, Whelan J, Pescini R, Ghersa P, DeLamarter JF, Hooft van
Huijsduijnen R. (1993) DNA-methylation of the E-selectin promoter represses
NF-kappa B transactivation. Biochem Biophys Res Commun 194: 215–221.
49. Stadhouders R, Pas SD, Anber J, Voermans J, Mes TH, Schutten M. (2010)
The effect of primer-template mismatches on the detection and quantifica-
tion of nucleic acids using the 5′ nuclease assay. J Mol Diagn 12: 109–117.
50. Ghedira R, Papazova N, Vuylsteke M, Ruttink T, Taverniers I, De Loose M.
(2009) Assessment of primer/template mismatch effects on real-time PCR
amplification of target taxa for GMO quantification. J Agric Food Chem 57:
9370–9377. doi: 10.1021/jf901976a.
51. Warnecke PM, Stirzaker C, Melki JR, Millar DS, Paul CL, Clark SJ. (1997)
Detection and measurement of PCR bias in quantitative methylation analy-
sis of bisulphite-treated DNA. Nucleic Acids Res 25: 4422–4426.
52. Clark SJ, Statham A, Stirzaker C, Molloy PL, Frommer M. (2006) DNA
methylation: Bisulphite modification and analysis. Nat Protoc 1: 2353–2364.
53. Strien J, Sanft J, Mall G. (2013) Enhancement of PCR amplification of mod-
erate GC-containing and highly GC-rich DNA sequences. Mol Biotechnol
54: 1048–1054.
54. Philipp AB, Nagel D, Stieber P, et al. (2014) Circulating cell-free meth-
ylated DNA and lactate dehydrogenase release in colorectal cancer. BMC
Cancer 14: 245. doi: 10.1186/471-2407-14-245.
55. Matuschek C, Bolke E, Lammering G, et al. (2010) Methylated APC and
GSTP1 genes in serum DNA correlate with the presence of circulating
blood tumor cells and are associated with a more aggressive and advanced
breast cancer disease. Eur J Med Res 15: 277–286.
56. Rand K, Qu W, Ho T, Clark SJ, Molloy P. (2002) Conversion-specific
detection of DNA methylation using real-time polymerase chain reaction
(ConLight-MSP) to avoid false positives. Methods 27: 114–120.

b4444_Ch-09.indd 274 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 275

57. Quan PL, Sauzade M, Brouzes E. (2018) dPCR: A technology review.

Sensors (Basel) 18: 1271.
58. Li M, Chen WD, Papadopoulos N, et al. (2009) Sensitive digital quantifica-
tion of DNA methylation in clinical samples. Nat Biotechnol 27: 858–863.
doi: 10.1038/nbt.559
59. Barault L, Amatu A, Bleeker FE, et al. (2015) Digital PCR quantification
of MGMT methylation refines prediction of clinical benefit from alkylat-
ing agents in glioblastoma and metastatic colorectal cancer. Ann Oncol 26:
1994–1999. doi: 10.093/annonc/mdv272. Epub 2015 Jun 25.
60. Overman MJ, Morris V, Moinova H, et al. (2016) Phase I/II study of azacit-
idine and capecitabine/oxaliplatin (CAPOX) in refractory CIMP-high met-
astatic colorectal cancer: Evaluation of circulating methylated vimentin.
Oncotarget 7: 67495–67506. doi: 10.18632/oncotarget.1317.
61. Rand KN, Ho T, Qu W, et al. (2005) Headloop suppression PCR and its
application to selective amplification of methylated DNA sequences. Nucleic
Acids Res 33: e127. doi: 10.1093/nar/gni120.
62. Mahon KL, Qu W, Devaney J, et al. (2014) Methylated glutathione S-transferase
1 (mGSTP1) is a potential plasma free DNA epigenetic marker of prognosis
and response to chemotherapy in castrate-resistant prostate cancer. Br J Can-
cer 111: 1802–1809. doi: 10.038/bjc.2014.463. Epub Aug 21.
63. Mahon KL, Qu W, Lin HM, et al. (2018) Serum free methylated glutathione
S-transferase 1 DNA levels, survival, and response to docetaxel in meta-
static, castration-resistant prostate cancer: Post hoc analyses of data from a
phase 3 trial. Eur Urol 19: 30855–30858.
64. Cottrell SE, Distler J, Goodman NS, et al. (2004) A real-time PCR assay for
DNA-methylation using methylation-specific blockers. Nucleic Acids Res
32: e10. doi: 1093/nar/gnh008.
65. Lofton-Day C, Model F, Devos T, et al. (2008) DNA methylation biomark-
ers for blood-based colorectal cancer screening. Clin Chem 54: 414–423.
doi: 10.1373/clinchem.2007.095992. Epub 2007 Dec 18.
66. Greenwood AD, Burke DT. (1996) Single nucleotide primer extension:
Quantitative range, variability, and multiplex analysis. Genome Res 6:
336–348.
67. Gonzalgo ML, Liang G. (2007) Methylation-sensitive single-nucleotide
primer extension (Ms-SNuPE) for quantitative measurement of DNA
methylation. Nat Protoc 2: 1931–1936. doi: 10.038/nprot.2007.271.

b4444_Ch-09.indd 275 08-Jul-22 6:11:41 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

276 Cell-Free Circulating DNA: Purification and Analysis Techniques

68. Tierling S, Schuster M, Tetzner R, Walter J. (2010) A combined HM-PCR/

SNuPE method for high sensitive detection of rare DNA methylation. Epi-
genetics Chromatin 3: 12. doi: 0.1186/756-8935-3-12.
69. Pidsley R, Zotenko E, Peters TJ, et al. (2016) Critical evaluation of the
illumina MethylationEPIC BeadChip microarray for whole-genome DNA
methylation profiling. Genome Biol 17: 208.
70. Bibikova M, Barnes B, Tsan C, et al. (2011) High density DNA methylation
array with single CpG site resolution. Genomics 98: 288–295. doi: 10.1016/j.
ygeno.2011.07.007. Epub Aug 2.
71. Jensen TJ, Kim SK, Zhu Z, et al. (2015) Whole genome bisulfite sequencing
of cell-free DNA and its cellular contributors uncovers placenta hypometh-
ylated domains. Genome Biol 16: 78. doi: 10.1186/s13059-015-0645-x.
72. Bhangu JS, Beer A, Mittlbock M, et al. (2018) Circulating free methyl-
ated tumor DNA markers for sensitive assessment of tumor burden and
early response monitoring in patients receiving systemic chemotherapy for
colorectal cancer liver metastasis. Ann Surg 268: 894–902. doi: 10.1097/
SLA.0000000000002901.
73. Kristiansen S, Nielsen D, Soletormos G. (2016) Detection and monitor-
ing of hypermethylated RASSF1A in serum from patients with metastatic
breast cancer. 8: 35. doi:10.1186/s13148-016-0199-0. eCollection 2016.
74. Deligezer U, Yaman F, Erten N, Dalay N. (2003) Frequent copresence
of methylated DNA and fragmented nucleosomal DNA in plasma of lym-
phoma patients. Clin Chim Acta 335: 89–94.
75. Levenson VV, Melnikov AA. (2011) The MethDet: A technology for bio-
marker development. Expert Rev Mol Diagn 11: 807–812. doi: 10.1586/
erm.11.74
76. Liggett TE, Melnikov AA, Marks JR, Levenson VV. (2011) Methylation pat-
terns in cell-free plasma DNA reflect removal of the primary tumor and
drug treatment of breast cancer patients. Int J Cancer 128: 492–499. doi:
10.1002/ijc.25363. Epub 2010 Apr 5.
77. Liggett T, Melnikov A, Yi QL, et al. (2010) Differential methylation of cell-
free circulating DNA among patients with pancreatic cancer versus chronic
pancreatitis. Cancer 116: 1674–1680. doi: 10.002/cncr.24893.
78. Liggett TE, Melnikov A, Yi Q, et al. (2011) Distinctive DNA methylation
patterns of cell-free plasma DNA in women with malignant ovarian tumors.

b4444_Ch-09.indd 276 08-Jul-22 6:11:41 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

DNA Methylation Analysis of cirDNA Biomarkers 277

Gynecol Oncol 120: 113–120. doi: 10.1016/j.ygyno.2010.09.019. Epub

Nov 6.
79. Shen SY, Singhania R, Fehringer G, et al. (2018) Sensitive tumour detec-
tion and classification using plasma cell-free DNA methylomes. Nature 563:
579–583. doi: 10.1038/s41586-018-0703-0. Epub 2018 Nov 14.
80. Quail MA, Kozarewa I, Smith F, et al. (2008) A large genome cen-
ter’s improvements to the Illumina sequencing system. Nat Methods 5:
1005–1010. doi: 10.38/nmeth.270.
81. Charlton J, Williams RD, Weeks M, et al. (2014) Methylome analysis identi-
fies a Wilms tumor epigenetic biomarker detectable in blood. Genome Biol
15: 434. doi: 10.1186/s13059-014-0434-y.

b4444_Ch-09.indd 277 08-Jul-22 6:11:41 PM

 B1948 Governing Asia

This page intentionally left blank

B1948_1-Aoki.indd 6 9/22/2014 4:24:57 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

© 2022 World Scientific Publishing Company

https://doi.org/10.1142/9789811244681_bmatter

Index

apoptosis, 10, 27, 114, 119, 140, 93, 95, 104, 115, 117, 126,
141, 176, 178–182, 185–188, 144–146, 160, 252, 254
198, 248 cell autonomous nucleases, 185
apoptotic bodies, 211, 223 cell-free methylated DNA
assays to monitor the DNase precipitation and high-
activity, 189 throughput sequencing
(cfMeDIP-seq), 267
biobanking, 22, 29, 36, 47–55, 59, cell nonautonomous nucleases,
64, 65, 105 185
biomarker discovery, 53, 54, 268, cell-of-origin biomarker, 4, 96,
269 113, 114, 120, 217, 248, 255,
biospecimens, 24, 31, 37, 47–54, 267, 268
58 cirDNA cleavage signature, 116,
bisulfite conversion of DNA, 141, 175, 180, 181
30, 58, 64, 96, 97,123, 124, cirDNA library, 4, 10, 12, 92, 115,
250–258, 261, 263–267 117, 139–154, 159–161, 220,
blood-derived PCR inhibitors, 221
52, 53, 58, 59 88, 127 cirDNA methylation, 3, 22, 27,
blood DNAses, 24, 27, 175–199 30, 58, 65, 34, 35, 97, 98,
blood storage and handling, 116–118, 120, 122–124,
21–41, 47–55, 185, 198, 225, 253 247–270
broad consent, 50 cirDNA methylation as a cancer
biomarker, 3, 22, 34, 117,
capturing shorter DNA 118, 249, 250, 254, 262, 263,
fragments, 10, 78, 79, 88, 92, 266–269

279

b4444_Index.indd 279 08-Jul-22 6:12:52 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

280 Cell-Free Circulating DNA: Purification and Analysis Techniques

cirDNA origin, 10, 27, 114, 119, DNA methylation, 22, 27, 30, 58,
140, 141, 176, 178–182, 96, 97, 116–118, 120, 123, 125,
185–188, 198, 248 247–270
cirDNA storage, 30, 48, 51, 55, DNA Protection by Nucleosome
cirDNA yield, 6, 10, 11, 23–25, Structure, 36, 56, 66, 96, 114,
27–28, 40, 55, 58, 63–106, 150, 115, 116, 120, 125, 126,
176, 196, 253 180–182, 186, 187, 254, 255
clinical data management, double-stranded DNA library
49–51 preparation, 144, 145, 150
clonal evolution, 3, 153
column-based cirDNA early detection and diagnosis of
purification, 63, 66–68, 77, 79, cancer, 3, 7, 49, 52, 53, 157
84, 88 enrichment of ctDNA, 4, 11–13,
concentration of cirDNA in 84, 87, 146, 150, 156, 213
plasma, 5–7, 29, 54–57, 66, enzymatic cytosine conversion,
88–90, 94, 96, 100, 140, 143, 250
163, 185, 186, 191, 250 ethical considerations, 50, 51
CpG dinucleotide, 123, 125, 247, exosomal biogenesis, 210,
248, 256, 259, 260, 264 214–216
CpG islands, 125, 248, 249, 254, exosomal cargo, 210–212,
264 216–219
exosomal fingerprint, 212, 217
degradation of DNA, 9–11, 13, exosomal isolation strategies, 224
27, 30, 50, 176, 179, 180, 184, exosomal loading, 212
190, 196 experimental artifacts, 5, 55, 56,
density gradient isolation 126, 175, 197, 198, 221, 233,
techniques, 211, 226–228, 232, 234, 255,
233, 235
detecting disease recurrence, 3, field flow fractionation (FFF), 231
21, 49, 118, 164, 212 fluorescent dyes, 55, 121, 128,
dexosomes, 210 129, 142, 189–192, 230, 258,
digital PCR (dPCR), 6, 10, 28, formalin-fixed paraffin embedded
37, 38, 40, 41, 64, 121–124, (FFPE), 13, 253
122, 131, 141–143, 158, 233,
259, 261 headloop suppression PCR, 262
DNA degradation, 9–11, 13, 27, healthy controls, 22, 48, 53, 83,
30, 50, 176, 179, 180, 184, 190, 86, 90–92, 94, 125, 128, 141,
196 184, 188, 268, 269

b4444_Index.indd 280 08-Jul-22 6:12:52 PM

 9”x6”  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques

Index 281

HeavyMethyl, 263 microfluidic DNA gel

hematopoietic cell death, 114, electrophoresis, 38, 64, 69,
178 71–73, 89, 105, 142
hemolysis, 37, 224 microvesicles, 211
hemolytic index, 37 minimal residual disease
detection, 3, 118
Illumina methylation arrays, 255, mitochondrial DNA, 65, 92, 93,
264 116, 117, 120
Illumina platforms, 97, 142, 255, monitoring of cancer, 3, 57, 64,
254 106, 118, 139, 155, 156, 164,
intercellular communication, 212, 263, 269
210–212 Ms-SNuPE, 263

library preparation kits, 115, 144, nanoparticle tracking analysis,

220 222
lipid bilayer membranes, 209, nanovesicles, 210
212, 219, 225 Next-generation sequencing
liquid-liquid cirDNA purification, (NGS), 6, 57, 92, 139–164
66, 74, 80–82, 84, 88, 91 NGS library preparation
low-abundance in plasma/serum, protocols, 115, 139–164, 265,
6, 63–65, 98, 104, 106, 252 267
noninvasive tumor genotyping,
magnetic bead, 57, 63, 68, 79, 80, 3
87, 89, 94, 97, 100 nucleases, 4, 9, 11, 12, 114, 116,
matrix binding, 63, 66–68, 77, 79, 117, 141, 175–199
82, 84, 88, 230, 261 nucleosome positioning, 36, 56,
methylation-sensitive restriction 66, 96, 114, 115, 116, 120, 125,
enzyme (MSRE), 266 126, 180–182, 186, 187, 254,
methylation-specific PCR, 249, 255
255, 256, 260
methylation-specific primers, optimal volume of plasma, 53
255, 260
methylation unbiased PCR, 256 phenol–chloroform protocols, 63,
Methyl-BEAMing (Beads 69, 74, 76, 81–85, 97, 100
Emulsion Amplification polyethylene glycol (PEG)
Magnetics), 261 precipitation, 228

b4444_Index.indd 281 08-Jul-22 6:12:52 PM

  b4444   Cell-Free Circulating DNA: Purification and Analysis Techniques 9”x6”

282 Cell-Free Circulating DNA: Purification and Analysis Techniques

preanalytical handling, 5, 10, sequencing-based methods,

21–41, 50, 64, 79, 105, 175, 139–164
191, 196, 197, 224 sequencing library preparation,
predicting and/or monitoring 113, 139–164
response to therapy, 8, 49, 64, single-stranded library
212 preparation, 115, 117, 145, 146
primer binding, 124, 125, 129, size exclusion chromatography
158, 163, 263 (SEC), 229
primer/probe design, 113, 129, SNuPE, or Single-Nucleotide
254, 259 Primer Extension, 263
processing of blood specimens, 5, standardized operating
10, 22, 23, 29, 33, 50, 55, 193, procedures (SOPs) for
198, 224, 225, 253 specimen collection and
protein surfaceome, 213 processing, 49

QIAamp Circulating Nucleic transmission electron microscopy

Acid kit, 63, 68, 71–74, 76–80, (TEM), 221
82, 84, 86–93, 95–97, 99, treatment monitoring, 3, 49, 57,
101–105 64, 118, 139, 156, 164, 212,
263, 269
quality assurance, 31
ultracentrifugation, 211, 226
recurrence monitoring, 3, 57, 64,
106, 118, 139, 155, 156, 164, ultrafiltration, 228
212, 263, 269 unique molecular identifiers
reporting results, 51 (UMIs), 147

b4444_Index.indd 282 08-Jul-22 6:12:52 PM