Jens Brange

Galenics of Insulin
The Physico-chemical and
Pharmaceutical Aspects of Insulin
and Insulin Preparations

With the Collaboration of

B. Skelbaek-Pedersen L. Langkjaer U. Damgaard
H. Ege S. Havelund L.G. Heding K.H. J0rgensen
1. Lykkeberg 1. Markussen M. Pingel
E. Rasmussen

With 34 Figures and 20 Tables

Springer-Verlag Berlin Heidelberg GmbH

JENS BRANGE, M. Se.
Novo Research Institute
Novo Alle
2880 Bagsvaerd, Denmark

This work is subject to copyright. AII rights are reserved, whether the whole or part of the
material is concerned, specifically the rights of translation, reprinting, reuse of iIIustrations,
recitation, broadcasting, reproduction on microfilms or in other ways, and storage in data
banks. Duplication of this publication or parts thereof is only permitted under the provisions
of the German Copyright Law of September 9, 1965, in its version of June 24, 1985, and a
copyright fee must always be paid. Violations fali under the prosecution act of the German
Copyright Law.

ISBN 978-3-662-02528-4 ISBN 978-3-662-02526-0 (eBook)

DOI 10.1007/978-3-662-02526-0

© Springer-Verlag Berlin Heidelberg 1987

Originally published by Springer-Verlag Berlin Heidelberg New York in 1987.

The use of registered names, trademarks, etc. in this publication does not imply, even in the
absence of a specific statement, that such names are exempt from the relevant protective laws
and regulations and therefore free for general use.
Actrapid Lente Neusulin Semilente
Humulin Mixtard NovoPen Semitard
lnitard Monotard Optisulin Ultralente
Insulatard Neulente Protaphane Ultratard
Lentard Neuphane Rapitard Velosulin
Product liability: The publisher can give no guarantee for information about drug dosage
and application thereof contained in this book. In every individual case the respective user must
check its accuracy by consulting other pharmaceuticalliterature.

2127/3130-543210
Foreword

Galenical pharmacy or galenics is the science dealing with the pro-

duction of drug substances from raw materials, the purity of such
substances, their formulation into pharmaceutical preparations with
the desired effects and safety in use, and the quality control, stability
and storage of the preparations. The field has taken its name from the
Greek physician Galen (131-201 A.D.), who had a profound influence
on medicine for many centuries because he collected and systematized
the medicinal knowledge of his time.
The discovery of insulin is attributed to Banting and Best who, in
1921, prepared an extract of the pancreas of the fetal calf and showed
that the extract was capable of reducing the blood sugar level of a
diabetic dog. This outstanding discovery gave rise to the rapid develop-
ment of the manufacture of insulin of bovine and porcine origin.
By 1925, two Danish manufacturers of insulin preparations were
established; both have since been in the forefront ofthe development of
insulin preparations, the latest achievement being the marketing of
human insulin by Novo in 1982. The development of highly purified
human insulin produced semisynthetically from porcine insulin or by
DNA recombinant methods are significant contributions to safe and
efficient insulin therapy.
Insulin is a protein which is destroyed in the gastrointestinal tract.
It is therefore administered by subcutaneous injection using one of the
many available preparations, which have different timing with respect
to onset and duration of effect. The fact that the therapeutic index,
Le., the ratio between a lethal and an effective therapeutic dosis, is
very low for insulin, together with the need for meal-related doses as
well as a continuous basal supply, makes insulin an extremely
challenging drug for pharmaceutical formulation. Improvements in
bioavailability and investigation of alternative routes of administration
of insulin, as well as a number of other proteins, represent major
challenges to the pharmaceutical science today.
The present book covers aH aspects of the galenics of insulin and
will be most valuable to everyone aiming to know more about insulin
and insulin preparations. This review will serve as a comprehensive
reference source as to the properties of insulin and insulin preparations.
In addition, it is a stimulating presentation of the many implications of
insulin preparations and therapy.

Henning Gjelstrup Kristensen

Professor of Pharmacy
Royal Danish School of Pharmacy
Acknowledgements

The present monograph was originally intended for incorporation into

a publication entitled Subcutaneous Insulin Therapy, which the editor
had hoped to publish in 1983. The main aim was to compile a critical
review of the subcutaneous insulin replacement therapy for patients
with diabetes mellitus. The publication of the complete volume has
been delayed considerably, and in late 1986 it was decided to update the
present work and have it published as a single monograph. We are
thankful to Professor Michael Berger and Springer-Verlag, who agreed
to this solution.
The authors wish to thank E. S0rensen, M.Sc., for carrying out the
immunogenicity studies, the prolongation test and biological potency
estimations; A. R. S0rensen, M.Sc., for performing biological potency
estimations; and Torsten Lauritzen, M.D., who was in charge of the
absorption studies performed at Hvid0re Hospital, Klampenborg.
We also wish to acknowledge the skillful technical assistance of
Lene Gmnlund Andersen, Lene Bramsen, Lise Frank, Lone J0rgensen,
Anne-Marie Kolstrup and Birgit Drreby Spon, the proficient assistance
ofInger Jansson, who helped to prepare and assemble the manuscript
and references, and the secretari al assistance of Mariann B. Andreasen
and Dorte Schultz.
We are greatly indebted to Kathleen Larsen for her invaluable
assistance in preparing the manuscript and to Dr. J. Schlichtkrull for
his constructive criticism.
Table of Contents

A. Production of Bovine and Porcine Insulin . . . . . 1

1. Introduction; -. . . . . . . . . . . . . . . 1
II. Physico-chemical Properties of Insulin Relevant
to Production . . . . . . . . . 1
III. Production of Crystalline Insulin . 3
IV. Purification of Crystalline Insulin 5

B. Characterization of Insulin 7
1. Purity . . . . . 7
1. Introduction . . . 7
2. Gel Filtration. . . 7
3. Disc Electrophoresis 8
4. Radioimmunoassay (RIA) 9
5. High Performance Liquid Chromatography (HPLC) 11
II. Standardization . . . 12
1. Bioassay Methods . . . 12
2. Chemical Methods. . . 13
III. Immunogenicity of Insulin. 14

C. Insulin Preparations. . . . . 17
1. Introduction. . . . . . . 17
II. Pharmaceutical Chemistry . 18
1. Insulin in Solution. . 18
2. Association of Insulin . 20
3. Metal Ion Binding. . . 22
4. Crystals and Crystallization . 23
5. Chemical Reactions . . . . 27
6. Mechanisms of Prolongation 27
III. Rapid-Acting Preparations 29
1. Introduction . . . . . . . 29
2. Acid and Neutral Formulations 30
3. Manufacture . . . . 31
4. Absorption . . . . . 31
IV. Protracted Preparations . 32
1. Introduction . . . . 32
2. Protamine Insulins. . 34
3. Insulin Zinc Suspensions (Lente Insulins) 36
x Table of Contents

4. Biphasic Preparations . . . . . . . . 38
5. Other Types of Protracted Preparations . 39
V. Auxiliary Substances 40
1. Introduction . 40
2. Preservatives . . 40
3. Isotonic Agents . 42
4. Buffering Substances . 42
VI. Characterization and Testing 43
1. Introduction . . . . . . 43
2. Analytical Tests Applied to aU Preparations . 43
3. Analytical Control Specific for Special Preparations 43
4. Prolongation Tests . 45
VII. Insulin Strengths. . . . . . . . . 46
1. Introduction . . . . . . . . . 46
2. Strengths of Insulin Preparations 46
VIII. Mixing Insulin. . . 48
1. Insulin Mixtures. . . . . . . . 48
2. Dilutions. . . . . . . . . . . 50
3. Addition to Intravenous Infusion Fluids 51
IX. Stability . . . . . 52
1. Physical Stability . 52
2. Chemical Stability . 54
3. Biological Stability . 58
4. Immunogenicity Studies 60
X. Storage and U se of Insulin 60
1. Storage . . . . . . . 60
2. Withdrawal of Insulin from the Vial 62
3. Mixing Techniques. . . . 63
4. Pre10ading of Syringes . . 63
XI. Insulin for Delivery Systems . 65
1. Introduction . . . . . . 65
2. Use of Present Insulin Solutions in Infusion Pumps 65
3. Physical Stabilization of Insulin Solutions . 67
XII. Insulin Derivatives . . . . . . . . 69
XIII. Alternative Administration of Insulin 70.

D. Human Insulin . 75
1. Manufacture 75
1. Sources 75
2. Conversion of Porcine Insulin . 75
3. Biosynthesis in E. coli . . . . 76
4. Biosynthesis in Saccharomyces cerevisiae 79
II. Preparations. 80

References . 83

Subject Index 101

 A. Production of Bovine and Porcine Insulin

1. Introduction

Soon after the discovery of insulin by Banting and Best in 1921 processes were
established for the manufacture of bovine and porcine insulin. Although these
first insulins were very impure and caused undesirable side-effects they drastically
improved the existence and the prospects oflife for hundreds ofthousands of dia-
betics. Later developments in insulin production leading to today's highly puri-
fied products have further improved the efficacy and safety ofinsulin therapy.

II. Physico-chemical Properties of Insulin Relevant to Production

That insulin is a protein was early recognised (Wintersteiner et al. 1928). The elu-
cidation ofits primary structure, however, was not established unti11955 after the
extensive work of Sanger and coworkers, as reviewed by Sanger (1959). Fig. 1
shows the structure of porcine insulin. The bovine insulin molecule contains Ala
(instead of Thr) and Val (instead ofIle) in positions 8 and 10 ofthe A-chain, re-
spectively. Due to the size and structure of the molecule insulin may be called a
polypeptide as well as a protein. The molecular weights are 5734 and 5778 for
bovine and porcine insulin, respectively.
The insulin molecule contains many ionizable groups, due to 6 amino acid res-
idues capable of attaining a positive charge and 10 amino acid residues capable
of attaining a negative charge. Figure 2 shows the net charge of the insulin mol-

A-Chain

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 26 29 30

B-Chain

Fig. 1. The primary structure of porcine insulin

2 Production of Bovine and Porcine Insulin

Z -2 Fig. 2. The net charge Z of the insulin

-4
monomer as a function of pH. The following
values of intrinsic pKa were used for the
-6 calculation: 7.2 (oc-NH 2 ), 9.6 (e-NH 2), 6.0
(imidazole group), 11.9 (guanidine group),
-8
3.6 (oc-COOH), 4.7 (l'-COOH) and 9.6
-10 (phenolic groups)
O 2 4 6 8 10 12

pH

ecule as a function of pH calculated on the basis of intrinsic pK values found by

Tanford and Epstein (1954). The net charge is zero at pH 5.5, in good agreement
with the electrophoretically determined isoelectric pH of 5.3-5.4 (Li 1954). One
or more of the 6 amide groups in the insulin molecule may be lost by hydrolysis
in solution (in particular at low pH, e.g. in 0.1 N HCl at 37 0c) leading to the for-
mation of up to 6 carboxyl groups (Sundby 1962). Insulin degrades in alkaline
medium (pH > 10) primarily due to its content of cystine residues (Cavallini et al.
1970). Insulin can be degraded by proteolytic enzymes, including those present
in the pancreas, like trypsin, chymotrypsins and carboxypeptidases.
The solubility of insulin depends on a number of factors, such as the nature
of the solvent, pH, temperature and the concentration of divalent metal ions and
salts. In aqueous medium insulin can be precipitated at a pH interval around the
isoelectric pH of 5.3-5.4 (Tanford and Epstein 1954). Bovine insulin precipitates
easier than porcine insulin. At pH below 4 and above 7, both insulins are fairly
soluble in the absence of Zn2+. The precipitation zone is broadened towards
higher pH values with increasing Zn 2 + concentration (SchlichtkrullI958). Insu-
lin can be salted out at high concentrations of salts. Even at as Iowa concentra-
tion of insulin as 0.2 g/l a nearly complete precipitation takes place in 2 M NaCl
at pH 2-3. Insulin is very soluble in homogenous mixtures ofwater and organic
solvents, e.g. in 50-70% (v/v) ethanol. This uncommon property of a protein can
be ascribed to the relatively low molecular weight of insulin and its relatively high
content ofhydrophobic amino acid residues (Val, Ile, Leu, Tyr, Phe). When the
content of organic solvent in the mixture is very high the solubility of insulin de-
creases, depending on the kind of solvent, tempera ture, pH and salt content.
In aqueous solution insulin forms many kinds of association compounds due
to non-covalent bondings; for a survey of these phenomena cf. C. II. 2. Addition
of urea or organic solvents miscible with water, such as ethanol and acetic acid,
counteracts such association. This effect is utilized in the chromatographic puri-
fication of insulin as described in A. IV.
Several crystal forms of insulin are known. Most important are the rhombo-
hedral crystals composed of unit cells containing hexamers of insulin with 2 or
Production of Crystalline Insulin 3

4 structural zinc atoms. The properties of these crystals and the kinetics of crys-
tallization are summarized in C. II. 4. Two different crystallization methods are
important for present day insulin production. The first method is based on crys-
tallization at pH 6 in citrate buffer containing Zn2+ and 15% (v/v) acetone
(Petersen 1945). The crystals formed contain two structural zinc atoms per
hexamer ofinsulin (Schlichtkru1l1956a). The second method is based on crystal-
lization at pH 5.5 in an acetate buffer containing 7% NaCI and 8-9 mg Zn2+ per
gram of insulin. U sing this method the crystals contain four structural zinc atoms
per hexamer ofinsulin (Schlichtkru1l1958).
Detailed descriptions of the physico-chemical properties of insulin from a
more general point of view, including insulins from other species than beef and
pork, have been published in review form (Klostermeyer and Zahn 1971, Humbel
et al. 1972, Schlichtkrull et al. 1975b).

ITI. Production of Crystalline Insutin

Until recently insulin for therapeutic use has been produced almost exclusively
from bovine and porcine pancreas. According to a report from WHO (1980) there
will be sufficient supplies of pancreas to meet requirements of insulin for the next
decades.
The principles for the production of crude insulin have remained essentially
unchanged throughout the past 40-50 years (Romans et al. 1940, Schlichtkrull et
al. 1975b). Scott (1934) realized that insulin crystallizes in the presence of Zn 2 +,
which formed the basis for the subsequent introduction of crystallization as a step
for purification of crude (amorphous) insulin. Figure 3 shows the main steps as-
sumed to be representative for present day production of non-chromatographed
insulin. The pancreata are deepfrozen (1) as soon as possible after slaughter in
order to avoid enzymatic degradation of insulin. If kept below - 20 °c during
transport and storage (2) the processing ofthe glands can wait for several months
without detecta bIe loss of insulin. The content of insul in in pancreas is usually in
the order of 0.02%. The use of acid ethanol/water for extraction (3,4) ensures
protection against enzymatic degradation, a high solubility of insulin and a low
solubility of proteolytic enzymes, their precursors and many other proteins. The
neutralization of the extract (5) renders possible the removal of substances which
would otherwise create difficulties later on in the process. The evaporation step
(6) serves the triple purpose of concentrating the extract, recovering ethanol and
rendering possible the removal of fatty substances (mostly fatty acids and phos-
pholipids) dissolved in the extract. In the salting-out step (7) insulin is separated
from nonproteinaceous compounds and also from some foreign proteins and
polypeptides. The salt-cake contains 15-20% insulin on the basis of total protein,
depending on the kind of pancreas. The salting-out process may be repeated at
a lower concentration of NaCI in order to obtain a further purification. The con-
tent of insulin in the precipitate formed at a pH -value close to the isoelectric point
(8) is 50-60%. The first crystallization (9) is advantageously carried out in an ace-
tone-Zn 2 +-containing citrate buffer (cf.A.II). Insulin crystals with a very low
content of glucagon are obtained by this method (Weitzel et al. 1953). The second
4 Production of Bovine and Porcine Insulin

( 1) Collection and deep-freezing of pancreata at

slaughterhouses.

(2) Transport to the factory via cold store.

....
Comminution of the frozen glands and addition of
(3) acid ethanol/water.

....
(4) Extraction at pH 1-3 and at a final ethanol
concentration of approx 60% (v/v).

(5) Neutralization of the extractand removal of the

precipitate formed.

(6) Acidification of the extract to pH 3-4, evaporation

of ethanol in vacuo and removal of fatty material.

(7) Salting·out in the concentrate by addition of 2-3

molii NaC), isolation of the salt·cake and
dissolution in water.

(8) Adjustment .of pH to 5.0-5.5, isolation of the

precipitate and dissolution in acid.

(9) Crystallization, isolation of the crystals, washing

and dissolution in acid.

Fig. 3. Flow sheet of the production

(1O) washmg
Rec~stallization! is~lation of the crystals,
and drymg In vacuo. of porcine and bovine insulin from
pancreas

crystallization (10) may be carried out by the acetate-salt method mentioned

under A II. The recrystallization results in crystals with lower content of non-in-
sulin-like contaminants. The purity of the twice crystallized insulin is 80-90%
(content ofinsulin with the structure found by Sanger). Ifthe insulin is not going
to be purified by chromatographical methods, it is often subjected to further crys-
tallizations before it is formulated into pharmaceutical preparations.
During the processing some insulin is inevitably lost. Other factors, however,
also influence the yields, viz. species and race, age when slaughtered, the feeding
and treatment of the animals in general, as well as the handling of the pancreata.
The yields of crystalline insulin from one kilogram pancreas originating from beef
cattle or from pigs are in the range of 100-200 mg. In dairy cattle the yields are
somewhat below this range; whereas calf pancreas may yield as much as 400 mg
of crystalline insulin per kg pancreas.
Reviews from a more general point of view on isolation of insulin (also con-
cerning laboratory methods) have been published by Burgermeister et al. (1975)
and Schlichtkrull (1975b).
Purification of Crystalline Insulin 5

IV. Purification of Crystalline Insulin

After it became possible to crystallize insulin on a large scale, it was believed for
some years that essentially pure insulin could at last be produced. Contributory
to this assumption was the observation by Jorpes (1949) that insulin recrystallized
up to 7 times was better tolerated by patients suffering from allergic reactions
after treatment with conventional insulin. However, due to developments in ana-
lytical methods from the beginning of the 1950's until the late 1960's, it became
increasingly c1ear that it was not justified to consider insulin purified solely by
crystallization as a pure compound.
Heterogeneity of once crystallized or recrystallized insulin was revealed by a
variety of methods, such as counter-current distribution (Harfenist and Craig
1952), partition chromatography (Carpenter 1958), anion- or cation-exchange
chromatography (Thompson and O'Donne1l1960, Cole 1960), disc electrophore-
sis (Mirsky and Kawamura 1966) and gel filtration (Steiner 1967).
Identification of the detected non-insulin components, beyond the finding
that some of them arose by deamidation of insulin, was not commenced untiI the
work of Steiner and Oyer (1967), c10sely followed by the work of Steiner (1967),
Chance et al. (1968), Schmidt and Arens (1968) and Steiner et al. (1968). Through
the work of these researchers, it became c1ear that the b-component, revealed by
gel filtration of crystalline insulin (cf. Fig.4, B. 1. 2), was mainly composed of
proinsulin, intermediates thereof and a non-convertible dimer of insulin, whereas
two of the bands seen in disc electrophoresis of crystalline insulin (cf. Fig. 5,
B. 1. 3), were due to proinsulin and its intermediates, respectively.
The a-component seen by gel filtration of once crystallized insulin (Fig. 4,
cf. B. 1. 2) is composed of a series of high molecular weight compounds with a
small content of insulin-related material (Schlichtkrull et al. 1974). The content
of a-component can be considerably reduced by recrystallization, whereas the b-
component is only slightly reduced in this way. Insulin itself is present in the c-
component, which contains, in addition, insulin-like components, revealed by
disc electrophoresis (Schlichtkrull et al. 1972), such as monodesamido insulin and
a mixture of monoarginine- and monoethyl insulins formed by hydrolysis, con-
version of proinsulin and esterification of insulin during extraction, respectively.
The results ofimmunogenicity studies in rabbits provoked the hypothesis that
insulin impurities, not the insulin itself, were responsible for the immunogenicity
of recrystallized insulin in patients (Schlichtkrull et al. 1970). As a consequence
it was found desirable to purify insulin on a large scale, so that it displayed only
a single component when analysed by both disc electrophoresis and gel filtration.
In order for insulin to fulfill these criteria a method was developed based on an-
ion-exchange column chromatography in ethanolic solution (J0rgensen et al.
1970, Schlichtkrull et al. 1972). It was later found that the purified insulin con-
tained approx. 500 ppm homologous Proinsulin-Like Immunoreactivity (PLI),
which could be reduced to less than 1 ppm by an additional chromatography
(Schlichtkrull 1974). A survey of the development is given in the paper by
J0rgensen et al. (1982). The twice chromatographed insulin has been named
monocomponent (MC) insulin (Novo). Another example of twice chromato-
graphed insulin in production is single component (SC) insulin (Lilly), made by
6 Production of Bovine and Porcine Insulin

a combination of gel filtration in 1 M acetic acid and anion-exchange chromatog-

raphy in 7 Murea (Chance et al. 1976). The porcine RI insulin (Nordisk Insulin-
laboratorium) is a third example of twice chromatographed insulin.
Agreat part of the insulin produced today is only purified by a single chro-
matography aiming at removal of the a- and b-component of crystalline insulin.
Examples are SP-insulin (Lilly), CR- and CS-insulins (Hoechst) and mono-pic in-
sulin (Organon). Most insulins used in current therapy have been subjected to one
or another form of chromatographical purification, although a number of manu-
facturers stiH produce insulin purified only by crystallization.
B. Characterization of Insulin

1. Purity

1. Introduction
Proper characterization of the purity of insulin requires a combination of analyti-
cal methods based on various principles (Jergensen et al. 1982). The four most
important analytical principles for the characterization of insulin with respect to
purity are the now c1assical biochemical methods of gel filtration (fractionation
by molecular size) and disc electrophoresis (fractionation mainly by charge) to
which have been added the highly sensitive methods of radioimmunoassay (RIA)
and, recently, the high performance liquid chromatography (HPLC) methods. Im-
proved modifications of HPLC techniques are expected to replace gel filtration
and disc electrophoresis in the future.

2. Gel Filtration
In 1967 it was shown by means of gel filtration that commercial insulin, purified
solely by crystallization, contained impurities with a higher molecular weight
than that of insulin (Steiner 1967). These impurities were later identified to be
mainly proinsulin, proinsulin intermediates and a covalent insulin dimer (Steiner
et al. 1968).
Since then several gel filtration systems for the analysis of insulin have been
described (Rolando and Torroba 1972, Schlichtkrull et al. 1974, Fisher and
Porter 1981). Common to nearly alI methods is the use of Bio-Gel P-30 (Bio-Rad
Laboratories) or Sephadex G-50 (Pharmacia Fine Chemicals) and 1-3 M acetic
acid as eluent. In this medium insulin is fulIy dissociated. Figure 4 shows three
examples of gel filtration chromatograms of porcine insulin. In once crystallized
insulin three distinct peaks are seen (a, b and c).
The a-component comprises high molecular weight substances
(MW > 25,000) derived from pancreatic tissue proteins. Its concentration is re-
duced by crystallizations, but small amounts are still present in 5 times crystal-
lized insulin, since antibodies against a-component are detectable in nearly alI pa-
tients treated with insulin ofthis purity. (Heding et al. 1980).
The b-component contains proinsulin and related substances, which are re-
moved to only a slight extent by crystalIizations.
FinalIy, the c-component comprises insulin and derivatives of insulin with
practicalIy the same molecular size (insulin ethyl ester, arginine insulin, deami-
date.9 insulin, etc.) (Schlichtkrull et al. 1974).
 Care should be taken when analysing insulin preparations by gel filtration to
avoid possible misinterpretations due to the preservatives of the preparations
(phenol, m-cresol or methylparaben). These substances absorb UV light strongly
at the wavelength used for detection of insulin (275-280 nm) resulting in the lar-
gest peak of the chromatogram. An example of such a misinterpretation has been
described by Schlichtkrull (1977 a). Insulin isolated from preparations still con-
tains traces of preservatives which willlead to peaks in the chromatograms. As
the preservatives are of low molecular weight and tend to adsorb to the gel, they
are eluted after the insulin peaks.
Gel filtration of insulin using HPLC technique (cf. B.I.5)has been described
by Welinder (1980).

3. Disc Electrophoresis
Since the introduction of disc electrophoresis in polyacrylamide gels by Ornstein
(1964) and Davis (1964), and Mirsky and Kawamura's (1966) subsequent appli-
cation of this method for the analysis of insulin, it has been widely used for the
characterization of insulin purity (Tjioe and Wacker 1972, Schlichtkrull et al.
1974, Krause and Beyer 1975, Kasama et al. 1980, Fisher and Porter 1981).
Although various electrophoretic systems are used by the different authors,
they are all modifications ofthe original system by Davis and Ornstein operating
at a slightly alkaline pH (8-9). Variations in acrylamide concentration, load per
tube, content of dissociating agent (urea), and dye used for staining have been de-
scribed. One of the most reproducible methods allows application of 100-200 /lg
ofinsulin per 5 x 50 mm tube containing 1 mI of gel (Schlichtkrull et al. 1974). At
higher loads the insulin and monodesamido insulin are no longer distinguishable
as two separate bands.
Radioimmunoassay (RIA) 9

Cathode

Proinsulin

Arg-insulins and ethyl insulins

Fig. 5. Approximate positions

of protein components by disc
Insulin and insulin dimers electrophoresis of recrystallized
insulin. As indicated the
Monodesamido insulin positions of proinsulin
intermediates of pork and beef
Anode insulin are a little different
Pork Beet

Staining with Amidoschwarz in combination with diffusional destaining in

3% acetic acid is among the most reproducible methods for visualizing protein
bands in polyacrylamide gels. Destaining by electrophoresis cannot be recom-
mended, as weak bands may be removed by this method. A slightly more sensitive
stain is Coomassie Brilliant Blue G 250, as used by Diezel et al. (1972), but me-
ticulously standardized conditions during the staining procedure are required to
obtain reproducible results. Therefore staining with Amidoschwarz is more con-
venient for routine purposes.
Figure 5 shows a schematic presentation of the position of the different pro-
tein bands resulting from disc electrophoresis ofinsulin purified solely by crystal-
lizations. A rough quantification of the bands can be made by comparing with
gels containing known amounts of the impurities added to monocomponent in-
sulin. Another method is to compare with a series of gels to which varying
amounts of insulin have been applied as a protein standard, as prescribed in the
British Pharmacopoeia (1980).
Because ofvariation in ability to bind the dye and variation in band sharpness
between the different impurities the latter method is less accurate than the
former.

4. Radioimmunoassay (RIA)
Radioimmunochemical methods are by far the most sensitive used in the evalu-
ation ofinsulin purity with a detection limit in the range of 1-500 ppb by weight
(1 ppb ~ 1 part impurity in 109 parts of insulin). RIAs are two thousand to one
million times more sensitive than the other methods described in this chapter (de-
tection limits about 0.1 %). RIAs have proved their great potential in detecting
trace amounts in insulin of other polypeptide hormones originating from the ex-
tracţion of the pancreatic glands.
10 Characterization of Insulin

Shortly after the discovery of proinsulin, the precursor of insulin, in 1968, it

was found that commercial insulin preparations contained about 2% ofproinsu-
lin, and during the following years R1As were developed for bovine and porcine
proinsulin and C-peptide (Yip and Logothetopoulus 1969, Heding et al. 1974).
The proinsulin RIA was the first of its kind used to evaluate insulin purity.
Proinsulin is composed of the insulin moiety and the connecting peptide. The
sequence ofthe latter is highly variable between species. This allows the establish-
ment of species specific proinsulin RIAs. Two methods have been used to elim-
inate interference from insulin:
a) The anti-proinsulin serum is passed through a column ofimmobilized insul in
which removes insulin antibodies. The possibility of insulin interference is thus
eliminated, and labelled proinsulin can be used as tracer (Yip and Logothetopou-
los 1969, Aaby 1979).
b) The anti-proinsulin serum is used without pretreatment and insulin interfer-
ence overcome by using a tracer that cannot be displaced by insulin, such as 125 1_
Tyr-C-peptide (Heding et al. 1974). C-peptide contains no site for labelling with
1251. Tyrosine is therefore coupled to the N-terminal amino acid ofthe C-peptide
to yield Tyr-C-peptide, which is then labelled with 125 1 (Markussen et al. 1970).

In addition to molecules containing the intact C-peptide moiety, the proinsu':

lin RIAs may detect molecules containing fragments of the C-peptide. Thus, a
better description of the material measured is given by the designation PLI
(Proinsulin-Like Immunoreactivity), indicating that the trace contaminants in in-
sulin may not only be proinsulin, but also other compounds with sequences par-
tially identical to that of proinsulin.
The PLI in insulin is attributable to several impurities, consequently, standar-
dization of reagents and methodology is necessary in order to obtain reproducible
estimates of the PLI content in insulin (Damgaard and Kruse 1982, Kruse et al.
1984).
Shortly after a RIA for Glucagon-Like Immunoreactivity (GLI) became
available analysis for GLI in insulin samples was introduced (Unger et al. 1970,
Heding 1971). One ofthe major problems ofthis R1A is caused by the sensitivity
of glucagon to proteolytic enzymes. Great care must be taken to avoid false es-
timates stemming from enzyme contamination leading to degradation of stan-
dard, tracer or the sample GLI. To minimize the problem a proteinase inhibitor,
such as aprotinin, is added to the buffer to inhibit proteolytic degradation.
During the last decade new polypeptide hormones have been discovered to be
present in the pancreas. This led to the application of additional R1As for the
characterization of insulin purity, namely for Pancreatic Polypeptide (PP)
(Chance et al. 1979), somatostatin (Patel and Reichlin 1979, Tronier and Larsen
1982) and Vasoactive Intestinal Peptide (VIP) (Fahrenkrug and Schaffalizky de
Muckadell 1977, Bloom 1979). Thus at least five different R1As have been used
for the characterization of insulin.
The level of some of the hormonal contaminants in insulin has been measured
and found to vary greatly from brand to brand (Bloom et al. 1978, Fitz-Patrick
and Pate11981, Mizuno et al. 1980). Sutcliffe and Bristow (1984) reported that the
High Performance Liquid Chromatography (HPLC) 11

Table 1. Typical examples of contaminant contents

(ppm) in twice crystallized and mon6component insulin
MC Novo

Contaminant Twice crystallized Monocomponent

insulin insulin

PLI 20,000 ~l
GLI 45 ~0.1
PP 43 ~0.01
Somatostatin 1.4 ~0.D1
VIP 0.3 ~0.D1

PLI content in commercial bovine insulin formulations ranged from 0.23 to 4.8%
for conventionally purified insulins, whereas the content in highly (chromato-
graphically) purified bovine insulin varied between less than 1 ppm and 1160 ppm
(0.1 %). The contents of different hormone contaminants need not be correlated,
as different purification methods may yield completely different compositions of
the hormonal contamination (J0rgensen et al. 1982). Therefore it is unjustified to
use a single contaminant as a common denominator of purity. A more com-
prehensive radioimmunochemical characterization should be used to evaluate
purity.
Typical examples of the contaminant content of twice crystallized and mono-
component (MC) insulin Novo are shown in Table 1. It shows that careful appli-
cation of modern purification technology can reduce the content of contaminat-
ing polypeptides in insulin to a level at or below the detection limit ofthe sensitive
RIAs.

5. High Performance Liquid Chromatography (HPLC)

The most recent method for the analysis of insulin is HPLC. In particular the re-
verse phase mode ofHPLC, with a nonpolar matrix as bed and a buffered mixture
ofwater and organic solvent as eluent, has revolutionized polypeptid,e analysisby
combining high selectivity and short time of analysis. Separation of bovine and
porcine insulin was the first important application of HPLC in insulin chemistry
(Biemond et al. J979, Damgaard and Markussen 1979, Dinner and Lorenz 1979),
followed by the separation of human insulin from the two other insulin species
(Terabe et al. 1979). HPLC methods for separating several species inc1uding
chicken, rabbit, sheep and horse have since been published (Ohta et al. 1983,
Rivier and McClintock 1983). The introduction of HPLC means that high sensi-
tivity species analysis of insulin preparations can now be performed in less than
one hour.
The determination of monodesamido insulin formed by acid hydrolysis dur-
ing extraction of insulin has traditionally been performed in a semiquantitative
manner by disc gel electrophoresis (ef. B. I. 3). A number of HPLC methods for
the quantitation of monodesamido insulin have been published, and in most cases
the species composition and monodesamido content can be determined in the
same analysis (Szepesi and Gazdag 1981, Lloyd 1982). Lloyd and Corran (1982)
12 Characterization of Insulin

have described an HPLC method capable of separating bovine, porcine and hu-
man insulin and their respective monodesamido derivatives in the same analysis.
In contrast to the number of methods described for species analysis and determi-
nation of monodesamido insulin, few publications have described the use of
HPLC for quantitation of impurities and insulin derivatives in insulin prepara-
tions. HPLC profiles of porcine insulin (J0rgensen et al. 1982, Welinder 1984),
recombinant as well as semisynthetic human insulin (Chance et al. 1981 a,
Markussen et al. 1982) have appeared.
HPLC of conventional crystallized insulin has shown that insulin derivatives
with virtually the same molecular weight as insulin, e.g. monoarginyl-insulin,
monodesamido insulin and insulin ethylester, can be determined under isocratic
conditions, i.e. using a constant eluent composition throughout the e1ution.
Higher molecular weight compounds, such as proinsulin, proinsulin intermedi-
ates and covalent insulin dimers, need gradient elution with increasing elution
strength in order to be analyzed with reasonable sensitivity in samples of insulin
(J0rgensen et al. 1982). Recent1y, Welinder et al. (1986) have reviewed the
litera ture concerning reversed-phase HPLC of insulin and investigated the
influence of column geometry and eluent composition on resolution and recovery.
Considering the increasing use of HPLC in insulin analysis during the last few
years, little doubt remains that it will be one of the fundamental methods for
species analysis and characterization of insulin purity in the future.

II. Standardization

1. Bioassay Methods
Several in vivo and in vitro bioassay methods for insulin have been developed.
Only the in vivo bioassays based on the hypoglycemic effect of the insulin after
injection into an experimental animal will be mentioned, since only these are rec-
ognized by health authorities for control of the biological potency of insulin for
therapeutic use. The methods differ primarily in the species of animal employed,
in the experimental design, and the procedure for handling the data.
During 1922-1926 bioassay methods for insulin were developed, based on the
measurement of the hypoglycemic response in rabbits and the convulsive re-
sponse in mice, in order to assess the potency of the hormone relative to an inter-
national standard. The 4th International Standard is soon to be replaced by a 5th
generation ofinsulin standards ofpure porcine, bovine and human insulin (WHO
1982). The classical bioassay methods, which contributed to the establishment of
biometry, are still the basis for the pharmacopoeial methods for determination
of insulin potency.
The rabbit blood sugar assay using a cross-over design was first suggested by
Marks (1925). Since then there have been many publications recommending
minor modifications to the original rabbit assay (Smith 1969).
Simultaneously with the development of the rabbit blood sugar assay, work
was undertaken to utilize the insulin induced convulsions in mice first observed
by Fraser (1923). Hemmingsen and Krogh (1926) and Trevan and Bock (1926)
found that the percentage of mice responding increased with increasing dosage,
Chemi cal Methods 13

and this led to the establishment of the theoretical basis for quantal response
bioassays. Many studies have been carried out since of the precision and repro-
ducibility of the mouse convulsion assay (Smith 1969, Stewart 1974).
Later, Eneroth and Aahlund (1968, 1970a, b) have developed a mouse blood
glucose twin cross-over assay.
Several papers have been published showing that the rabbit blood sugar assay,
the mouse convulsion assay and the mouse blood glucose assay give essentially
similar bioassay results (Miles et al. 1952, Bangham and Mussett 1959, Ashford
et al. 1969, Bangham et al. 1978). However, an analysis of more than 500 rabbit
blood sugar assays by new multivariate statistical methods showed that the
potencies based on blood sugar responses Y2, 1 and 2 Y2 hours after the injection of
pure porcine, bovine and human insulins relative to the present mixed species
standard showed significant variation depending on the blood sampling times
(V01und et al. 1982, Pingel et al. 1985). Porcine and human insulin potencies
decreased by 12% and 18%, respectively, from the Y2-h to the 2 Y2-h response,
whereas bovine insulin potencies increased by 9%. Since the standard is approx-
imately an even mixture of porcine and bovine insulin these results could be due to
porcine and human insulin having a quicker onset and shorter duration of
hypoglycemic effect than bovine insulin. This was confirmed in assays of porcine
relative to bovine insulin and by direct comparison of mean blood sugar curves. It
was concluded that the rabbit blood sugar assay is invalid when the test and
standard insulin have a different species composition. Hence, pure species insulin
standards are needed for this assay.
In the mouse convulsion assay no difference in the effect of porcine and bovine
insulin has been demonstrated (Pingel et al. 1982). Thus this bioassay system can
be used whether the test and standard preparation are of the same or of different
insulin species.

2. Chemical Methods
Assessing the potency of insulins by a bioassay implies a certain risk of variation
in the strength of the final insulin preparations. The use of a bioassay was man-
datory in the years following the discovery of insulin, since the composition and
purity oftheinsulins produced could vary considerably from batch to batch. To-
day, when insulins can be produced with a uniform and very high purity, it has
become possible to reduce the potency variation between batches of pharmaceuti-
cal insulin preparations by calculating the potency of the insulin by accurate
chemical methods of analysis, such as Kjeldahl nitrogen determination (Pingel et
al. 1982) or HPLC (Kroeff and Chance 1982). Quantitation by HPLC of insulin
and related substances in bulk insulin and in insulin preparations has been found
to give an estimate of the potency that is in good agreement with that obtained
by the rabbit blood sugar assay (Smith et al. 1985) and the mouse blood sugar
assay (Fisher and Smith 1986). The HPLC method was found to reduce analysis
time and to give more reproducible results. A bioassay may be used for con-
trol, and the potency estimate should be consistent with the estimate obtained by
the chemical methods. The biological potency of MC insulin whether porcine,
bovine or human has been found to be the same on amolar basis: 168.10 6 IV/moI
14 Characterization of Insulin

corresponding to 184 IU jmg N (Pingel et al. 1982) determined by the mouse con-
vulsion assay.

nI. Immunogenicity of Insulin

Awareness of immunological side-effects of insulin therapy became apparent dur-

ing the late forties, and it was later established that virtuaHy aH insulin treated dia-
betics had insulin antibodies (Berson et al. 1956, Berson and Yalow 1959, 1964).
In some patients the presence of high levels of antibodies caused severe insulin
resistance due to the elimination of the insulin as antigen-antibody complexes
(Berson and Yalow 1960). In other patients the antibodies were found to change
the control of glycemia by firstly capturing the injected insulin, thereby reducing
its biological effect, and later releasing it from the complexes, which could lead
to hypoglycemia (Berson and Yalow 1960). These side-effects were thought to be
inevitable until the introduction of the highly purified insulins in the early se-
venties.
Besides inducing insulin antibodies, insulin preparations containing detect-
able amounts of contaminants such as PP, VIP, glucagon and a-component, can
induce antibodies against aH these substances in insulin-treated diabetics (Bloom
et al. 1979, Villalpando and Drash 1979, Heding et al. 1980). The biological sig-
nificance of these antibodies is unknown. Radical elimination of the contami-
nants from the insulin has completely abolished the induction of such antibodies.
When patients are transferred from conventional crystallized insulin to MC insu-
lin a rapid decrease in the level ofantibodies against the contaminants isobserved
(Heding et al. 1980). The mentioned antibodies aH belong to the IgG class ofim-
munoglobulin and bind and neutralize their antigen.
Antibodies ofthe IgE class specific for PP, a-component and insulin may also
be induced by insulin preparations (Falholt et al. 1983). The IgE, against insulin
causes aHergic manifestations. Insulin purified to MC specifications (Table 1) did
not induce IgE" irrespective ofthe insulin species (Falholt 1982).
During the development of MC-insulin, aiming at reducing the formation of
insulin antibodies (SchlichtkruH et al. 1970), an animal model was developed to
control the various batches of insulin. The rabbit was found to be the most sui-
table species, as it develops insulin antibodies with similar capacity and affinity
binding constants as do diabetic patients (SchlichtkruH et al. 1972). There is a
considerable variation between the responses in rabbits to the same immunologi-
cal stimulus but, irrespective of its shortcomings, the method is useful as a test
for immunogenicity whenever new preparations or new technology for the pro-
duction ofinsulin is introduced. In this rabbit model monocomponent porcine as
well as human insulin exhibit non or very low antibody formation and the two
insulin species cannot be distinguished from each other. In clinical trials, however,
the immunogenicity ofhuman insulin (cf. D.II) has, in several studies, been found
to be lower than that ofporcine insulin. After 6 months' treatment of 102 newly
diagnosed diabetics with either porcine or human monocomponent insulin
Schernthaner et al. (1983) found IgG insulin antibodies in only 14% of the
patients treated with human insulin, but in 29% of those treated with porcine
Immunogenicity of Insulin 15

insulin. The antibody titres were also significant1y lower in the patients treated
with human compared to the patients who received porcine insulin. In a multi-
centre trial Heding et al. (1984) compared the two insulin species in 135 newly
diagnosed diabetic children and conc1uded that human monocomponent insulin
has a lower immunogenicity than porcine insulin of the same purity during the
first year ofinsulin treatment. In a long term study Luyckx and coworkers (1986)
found that the percentage of patients who remained antibody-free after 12-21
months of treatment was 67-75% in the group treated with monocomponent
human insulin and only 25-43 % in the one receiving monocomponent porcine
insulin, and the insulin antibody titres, when present, were lower in subjects
treated with human insulin.
Velcovsky and Federlin (1984) also observed lower antibody concentrations in
patients treated solely with human insulin compared to a group treated with
porcine insulin and it was found that four diabetic patients with delayed-type
allergy to animal insulin couldbe treated with human insulin without any
problems. They conc1uded that the introduction ofhuman insulins represents an
important advance from the immunological point of view.
c. Insulin Preparations

1. Introduction

Banting and Best revolutionized diabetes therapy 60 years ago and extended the
life expectancy of a young diabetic by decades. Since then the advances have been
less dramatic, but important improvements and many refinements have emerged
in the course of time.
The major achievements contributing to this progress are listed in Table 2.
Improvement in insulin purity - an essential element as regards character and
quality of an insulin preparation - was the first challenge and has been a goal ever
since, stimulated by the great advances in our knowledge of insulin from the work
of Abel, Scott and Fisher, Jorpes, Sanger, Berson and Yalow, Schlichtkrull, Mir-
sky and Kawamura, Steiner and their coworkers, resulting in an ever increasing
purity and the introduction of Actrapid - the first neutral insulin solution, MC-
insulin - the first chromatographically purified insulin and, recently, human in-
sulin.

Table 2. Major scientific achievements influencing the development and use of insulin
preparations

1922 Isolation of insulin Banting and Best

1926 Crystallization of insulin Abel
1928 Insulin is a protein Wintersteiner et al.
1934 Zinc insulin crystallization Scott
1936 Protamine insulin Hagedorn et al.
1936 Protamine zinc insulin Scott and Fisher
1946 Isophane insulin (NPH) Krayenbiihl and Rosenberg
1949 Recrystallization reduces allergy Jorpes
1951-52 Lente series Hallas-MlIlller et al.
1955 Primary structure of insulin Sanger et al.
1956 Radioimmunoassay Berson, Yalow et al.
1958 4 Zn insulin crystallization Schlichtkrull
1959 Biphasic insulin (Rapitard) Schlichtkrull
1961 Neutral insulin solution (Actrapid) Schlichtkrull et al.
1966 Heterogeneity (disc electrophoresis) Mirsky and Kawamura
1967 Proinsulin Steiner
1969 Tertiary structure Hodgkin and co-workers (Adams et al.)
1970 Monocomponent insulin Schlichtkrull et al.
1974 Continuous infusion Pfeiffer et al., Albisser et al.
1979-81 Human insulin Goeddel et al., Chance et al., Markussen
1980-82 Insulin pump implantation Buchwald et al. Irsigler et al., Schade et al.
18 Insulin Preparations

Numerous attempts were made to retard the subcutaneous absorption of in-

sulin to spare the diabetics the chore and discomfort of multiple daily injections.
Major milestones are the protamine- and protamine-zinc insulins, isophane
(NPH) insulin, Lente insulins and Biphasic insulin. Crystallization of insulin and
elucidation of the role of zinc for the association and structure of insulin have
played an important role for this evolution and our understanding of insulin.
Abel, Scott and Fisher, Schlichtkrull, and Hodgkin and their coworkers are the
main contributors in this field.
A multitude of different insulin preparations are today available for the treat-
ment of diabetics (cf. C. III and IV). The important characteristics from a phar-
maceutical point of view, the analytical control, as well as the chemical and
physicochemical properties relevant to their storage and use are discussed in vari-
ous sections, however, it is not within the scope of the present chapter to go into
details about the clinical usefulness of the various existing preparations.
Increasing awareness of the significance of strict metabolic control in the pre-
vention of long-term complications of diabetes has somewhat paradoxically led
to increasing use of the multiple daily injections of the early days of insulin ther-
apy. Another approach is the use of insulin delivery systems initiated by the work
of Pfeiffer et al. and Albisser et al.
The first implantations of such devices were performed in 1980-1981, a form
of application which wiU require more stable insulin preparations. Before non-
physiological additives are introduced forstabilization purposes thorough phar-
macological and toxicological testing is mandatory. The same applies to the use of
insulin analogues. (cf. C.x.II).
Alternative administration of insulin, such as rectal and nas al, as weB as
peroral, with absorption promoting agents or in liposomes, has been tried repeat-
edly with very limited success. New approaches like polymer implants and lectin-
bound maltose insulins are being investigated (cf. C.xIII).
Despite 60 years' use insulin preparations are stiU undergoing development
aiming at improving their therapeutic efficacy.

II. Pharmaceutical Chemistry

1. Insulin in Solution
The solubility of insulin depends on the species and purity of the insulin, the sol-
vent, composition of the solvent, pH, zinc ion content, as well as the content of
certain other divalent metal ions, ion strength and temperature.
In water insulin is practically insoluble at its isoelectric point of pH 5.4
(Fredericq and Neurath 1950), but is easily soluble at pH lower than 4. At neutral
and alkaline reaction the solubility is strongly dependent on the concentration of
zinc ions and on the species of insulin (SchlichtkrullI958). Jeffrey et al. (1976)
found the solubility of zinc-free bovine insulin at pH 7.0 to be 4.5-5 gjl, corre-
sponding to approximately 125 IUjml. Milthorpe et al. (1977) reported solubility
of bovine zinc insulin crystals (2 Znjhexamer) at pH 7.0 to be 0.9 gjl, correspond-
ing to 24 IU jml.
Insulin in Solution 19

L09 ( dissolved insulin • 100 )

total insulin

I
I Fig. 6. Solubility of rhombohedral insulin crystals
~ of different species and purity. Determination
o \ according to Schlichtkrull (1958) after 2 days at
20-25 ec in a medium containing 0.01 M sodium
\
\
\
\ , acetate,O.oI Mveronal (pH 7.4), 0.7% sodium
ch10ride and 40 IU insulin/ml. Monocomponent
-1
x'-X porcine insulin (x --- x) (Skelb!ek-Pedersen,
''''--X---X--X unpublished). Porcine (o-o) and bovine (e-e)
insulin of conventional purity (Schlichtkrull1958)
o 2 3
\19 Zn 2 + (total)/IU

U sing insulin purified only by crystallization, and having a total Zn 2 + content

of 0.37 Jlg/IU, the concentration ofinsulin in solution at pH 7.4 was found to be
2.4 IU/ml and 22IU/ml for bovine and porcine insulin, respectively, at a total
concentration of insulin (dissolved and suspended) of 40 IU /mI. At zinc ion con-
centrations above 1 Jlg/IU insulin is practically insoluble at neutral pH, irrespec-
tive of species (SchlichtkruIl1958, Schlichtkrull et al. 1975 a, b).
A reinvestigation ofthe solubility ofporcine insulin ofvery high purity (MC-
insulin), is shown in Fig. 6. In comparison with the earlier results it is evident that
the removal ofthe impurities (proinsulin and derivatives, desamido insulins, etc.),
as would be expected, lowered the solubility of the insulin.

IU/ml in solution

Fig.7. Solubility ofporcine MC

insulin and porcine monodesamido-
\ (A21)-insulin in neutral aqueous
\
25 \ solution (pH 7) as a function of the
X
\ total zinc content (Brange,
\
\ unpublished). Insulin or
20 \
,, monodesamido·insulin in a total
concentration of 40 IU /mI was
15 ,,
X
dissolved followed by addition of the
,, predetermined amount of zinc acetate,
"'le, and allowed to stand for 3 days at
10 ...
4 ec with frequent agitation.
Monodesamido insulin (---), Insulin
5 (--); with addition of
methylparaben ( x - - x ), without
o preservative (o-o)
0.25 0.5 0.75 \19 Zn 2 + (total)/IU
20 Insulin Preparations

Calcium ions have been observed to precipitate insulin in neutral solution, al-
though not as effectively as Zn2+ (Emdin et al. 1980).
The solubility of insulin is also influenced by a variety of organic compounds.
In neutral solution positively charged molecules may combine with negatively
charged insulin molecules resulting in the precipitation of complexes. Thus, addi-
tion of basic proteins or peptides (cf. C. IV) or small amounts of cationic deter-
gents (Birdi 1973) will decrease the solubility. The presence of glucose has been
observed to augment the solubility of bovine insulin at pH 2 and 7 (Jeffrey 1974).
Even the low concentrations of preservative present in the pharmaceutical prep-
arations influence the solubility of insulin in neutral solution as shown in Fig. 7,
which also illustrates that deamidation of insulin enhances the solubility.
The three-dimensional structure of insulin in solution is essentially identical to
the structure of insulin in the rhombohedral zinc insulin crystals (cf. C.II.4) and
assembly of monomers into dimers and hexamers (cf. C.I1.2) does not alter the
conformation ofthe monomeric unit in any important way (Bi et al. 1984, Hefford
et al. 1986). The transformation from the 2 to 4 zinc structure (cf. C.II.4) has been
found to occur in solution in a similar manner to that observed in the crystal
(Williamson and Williams 1979, Ramesh and Bradbury 1986).
Dissolution of zinc insulin crystals is best carried out in acid solvent. Due to
their zinc content the crystals dissolve very slowly at neutral and mildly alkaline
reaction. Strongly basic solutions cannot be recommended because the disulfide
bonds in insulin are very susceptible to degradation in alkaline medium.

2. Association of Insntin
Insulin exhibits a very complex association pattern in the crystalline state, as well
as in solution. In solution it exists as an equilibrium mixture of monomers,
dimers, tetramers, hexamers and, possibly, some higher associated states, depen-
dent on concentration, pH, metal ions and salts. The monomeric insulin (bovine
MW 5734) occurs in aqueous solution only at extremes of pH or in very diluted
solutions. At pH 2 the monomer predominates only at concentrations below
0.5 g/l (Jeffrey and Coates 1966a). The dimer is the sole association species in the
concentration range from 1 to 10 g/l when the ion strength is 0.05, but increased
association up to an apparent molecular weight of about 25,000 can be observed
when the ion strength is raised to 0.2 (Jeffrey and Coates 1966b).
At neutral reaction appreciable amounts of monomers only occur at insulin
concentrations below 0.1 g/l (Frank et al. 1972). In diluted solutions (1-3 g/l)
zinc-free insulin exists mainly as monomers at pH above 9 (Fredericq 1956).
The three-dimensional structure and the self-association of monomeric insulin
into dimeric, tetrameric and hexameric molecules have been comprehensively re-
viewed by Blundell et al. (1972). When zinc or other divalent metal ions are pres-
ent, the hexameric form prevails in neutral and moderately alkaline solutions. In
solutions of bovine insulin at pH 7.0 with 2 zinc ions per 6 monomers of insulin,
the hexamer comprises 75% or more ofthe total insulin at a concentration above
0.1 g/l (Milthorpe et al. 1977). Higher concentration of Zn 2 + leads to association
oftwo hexameric units (Fredericq 1956) and precipitation ofthe complex (Grant
et al. 1972).
Association of Insulin 21

Calcium ions bind to zinc insulin hexamers at specific sites cross-linking three
hexamers; this binding is most probably responsible for the decrease in solubility
of zinc insulin at neutral pH in the presence of Ca2+ ions (Pitts et al. 1980).
It is mainly the nonpolar residues of the insulin monomer that are involved
in the association into dimers and hexamers. The surface of the hexamer is there-
fore almost entirely polar (Blundell et al. 1972). About 40% of the surface of the
monomers is buried in the hexamer (Renneboog-Squilbin et al. 1981). The recep-
tor binding region of insulin has been found to include, in addition to more polar
surface residues, many of the hydrophobic residues important to dimerization
(Pullen et al. 1976). The molecular structure of the hexameric unit cell found in
the crystal (Blundell et al. 1972) is assumed to be essentially the same as that ob-
served in neutral solutions of zinc insulin (Frank et al. 1972, Goldman and Car-
penter 1974, Strickland and Mercola 1976, Bi et al. 1984). Other studies, however,
indicate that the crystal packing inf1uences the 3-dimensional conformation (Ren-
neboog-Squilbin et al. 1981, Bradbury et al. 1981). The ave rage helix content has
been found to de crease when the crystals are dissolved and conformational tran-
sition takes place when dissociation into monomeric insulin occurs in very dilute
solutions (Pocker and Biswas 1980). From studies of the chemical reactivity of
insulin at different concentrations it has been concluded that the monomeric unit
has essentially the same conformation in its free and associated states (Kaplan et

No zinc binding Zinc binding

c:
.2
~
C
a>
u
c:
o
u
u
c:
N

u
c:
o
u
c:-
'5~
"'
c:::I
-o
g,.<:
- <.:::::::::::::::::::::.: W
c:=

~
Oi~
"'~
a>
U )jm:lmlm:I!I\:llll:L
.E ~
· ----------~::-=
:::=
::::=
::::+
T :=
::::=
::::=
::::~
:: · _ ~ -------------+~I
5.4 pH

Fig. 8. Schematic diagram ofthe association and crystallization behaviour ofinsulin. The shaded
area represents the insulin precipitation zone
22 Insulin Preparations

al. 1984, Hefford et al. 1986), and Kap1an et al. pointed out that surface
adsorption may have affected the data reported by Pocker and Biswas.
The association constant for the monomer-dimer equilibrium at pH 7.0 is
1.4-7.5 X 10 5 M- 1 (Pekar and Frank 1972, Jeffrey et al. 1976, Pocker and Biswas
1981, Strazza et al. 1985). Strazza et al. also showed that the dimerization
properties are similar for bovine and porcine insulin and found that neither pH
(pH 7 and pH 2) nor ion strength (0.1-0.01) influences the equilibrium constant
significantly.
The va1ue for the dimer-hexamer association constant at pH 7.0 for zinc free
porcine insulin has been found to be 4 x 10 8 M- 2 (Pekar and Frank 1972). At
pH 7.4 Holladay et al. (1977) found the value for zinc free bovine insulin to be
9 x 109 M- 2 •
Pekar and Frank (1972) calculated the mo1e ratio of monomer to dimer at
physio10gical concentrations (fasting state) to be about 8 x 10 5 to 1.
A schematic survey ofthe association behaviour ofinsu1in under various con-
ditions appears from Fig. 8.

3. Metal Ion Binding

The importance of zinc for the crystallization ofinsulin in the rhombohedral form
was first recognized by Scott (1934) and later further explored by Schlichtkrull
(1958), who found that the minimum zinc content necessary for crystallization
was 2 atoms per hexamer.
Insulin is capable of binding large amounts of zinc either in the crystalline
state or in solution in the slightly acid to the slightly alkaline range. At low pH
insu1in does not bind zinc (Cunningham et al. 1955, Schlichtkrull1956a).
The interaction with zinc requires the formation of hexamers and is strongly
pH-dependent. The content of bound zinc rises steeply when pH increases from
4.5 to 7.4 (Hallas-M011er et al. 1952, Schlichtkrull 1958) and is a function of the
free zinc concentration (Cunningham et al. 1955). At pH 8.0 insulin is capable of
binding 2.7 atoms of zinc per insulin monomer (Goldman and Carpenter 1974).
Two classes of binding sites with very different affinities for Zn 2 + have been
reported. In the pH-range 5 to 7 strong binding of two zinc atoms per hexamer
of insulin occurs with an association constant of 10 5-10 6 M- 1 at pH 7 to 8 (Sum-
merell et al. 1965, Grant et al. 1972, Goldman and Carpenter 1974). The two zinc
ions are hexacoordinate. Three ligands of each zinc ion are imidazolyl moieties
of His (B 10) from three monomers in different dimers (Adams et al. 1969, Brad-
bury et al. 1981). The remaining ligands are water molecules (Dunn et al. 1980)
or small anions. Both metal binding sites are located on the trigonal symmetry
axis, the metals serving to hold three dimers together in the hexamer (Brill and
Venable 1968).
Above pH 7 an additional, weaker binding of zinc takes place with an appar-
ent association constant of 10 3 -10 4 M- 1 at pH 8 (Goldman and Carpenter 1974).
The involved binding sites have been found to include the a-amino groups in the
insulin molecule (Marcker 1960, Frank et al. 1972). Chasteen et al. (1973), how-
ever, reported evidence that these sites are most likely the carboxyl groups of the
glutamyl residues.
Crystals and Crystallization 23

The porcine zinc-insulin complexes are soluble at neutral reaction when the
zinc/insulin (hexameric) ratio is below 4, but at higher concentrations of zinc the
solubility decreases (Finger et al. 1967) and at a ratio of 6 Zn 2 + /hexamer com-
plete precipitation of the insulin is observed (Grant et al. 1972).
A dynamic equiliQrium exists between bound and free zinc ions, as aH the zinc
ions present in the crystalline structure are exchanged in the course of one hour
(SchlichtkruHI958).
During recent years it has been substantiated that the insulin hexamer in
addition to the zinc binding sites also has three internal Ca 2 + binding sites
involving the Gln (B 13) carboxylates (Sudmeier et al. 1981, Storm and Dunn
1985, Alameda et al. 1985). Such binding of calcium ions will neutralize the
negative charges present on the six B13 glutamates in the central cavity of the
hexamer, thus decreasing the solubility ofthehexamer and facilitating the crystal
formation in vivo in the p-ceH granula of the pancreas.

4. Crystals and Crystallization

At least 6 different crystalline modifications of insulin have been discovered. The
zinc-free orthorhombic crystals grown at acid pH by EHenbogen (1949) and
Sundby (1962) containing four dimers in the unit ceH (Low 1952), the monoclinic
crystals with 2 hexamers in the unit ceH isolated by Schiichtkrull (1958) in neutral
medium in the presence of zinc and phenol and further described by Harding et al.
(1966) and Low and Chen (1969) and the zinc-free cubic crystals, obtained
originalIy by Abel (1926), rediscovered by Schiichtkrull (1958) and structuralIy
defined by Dodson et al. (1978) have never been used in therapeutic preparations
(Fig.9a-c).
The tetragonal crystals of protamine zinc insulin (Fig. 9 d) used in NPH prep-
arations (Krayenbiihl and Rosenberg 1946) contain, in addition to small amounts
of protamine, at least 2 atoms of zinc and 20 molecules of phenol (m-cresol) per
hexameric insulin. The protamine appears to occupy interstices in the structure
between unusualIy packed hexamers (Hodgkin 1974).
The formation and shape of the rhombohedral zinc insulin crystals with 2 and
4 structural zinc atoms per hexameric unit of insulin have been extensively studied
by Schlichtkrull (1957 a, 1958). For review see SchlichtkruH et al. (1975 a).
Schlichtkrull demonstrated that rhombohedral insulin crystals cannot be ob-
tained without a content of zinc or certain other metal ions, thus finalIy elucidat-
ing the reasons for the previous difficulties in obtaining reproducible yields ofin-
sulin by crystallization. The minimum content of zinc ions necessary for crystal-
lization is 2 atoms per hexamer, irrespective of pH of crystalIization (pH 5.5-7.0)
and species of insulin, indicating that the two atoms are an integral part of the
crystal structure (SchlichtkruHI956a). If, however, the amount offree zinc ions
present in the crystallization medium prior to crystallization exceeds 6 atoms per
hexamer of insulin the crystallization will be very slow or not take place at alI.
When already formed the rhombohedral crystals are able to take up addi-
tional zinc ions, thus showing evidence for the existence of additional binding
sites for zinc ions. The number of zinc atoms/hexamer increases to about 12 cor-
responding to about 2% zinc when pH is increased from 5 to 7.
Crystals and Crystallization 25

The shape of the 2 zinc crystals with an internal rhombohedral structure

varies with the species of insulin. Porcine and human insulins form geometrically
perfect rhombohedra (Fig. 9 e), but bovine and ovine insulins have a tendency to-
wards formation of distorted twin crystals often with a star-like appearance.
(Fig. 9 f).
Another important factor determining the shape is the presence in the crystal-
lization medium of halides or other insulin solubilizing agents. At sodium chlor-
ide concentrations above 6%, true rhombohedra are obtained irrespective of the
species of insulin (Schlichtkrull 1956 b). Concurrently a steep rise in the structural
amount of zinc ions from 2 to 4 atoms per hexamer is observed, the concentration
of halide also being a relevant factor determining the shape (Schlichtkrull 1958,
Schlichtkrull et al. 1957 a).
The three-dimensional arrangement of the atoms in rhombohedral 2 zinc in-
sulin crystals has been elucidated by the extensive work by Adams et al. (1969),
Blundell et al. (1971) and Dodson et al. (1979). The two zinc atoms are placed
on the three-fold symmetry axis 9 A above and 9 A below the two-fold axes. The
dimensions of the unit cells of 2 and 4 zinc insulin crystals are very similar with
nearly identicallattice constants (Dodson et al. 1966), but surprisingly large struc-
tural differences in the protein conformation and organization of the zinc atoms
were reported (Bentley et al. 1976, Chotia et al. 1983). In the 4 zinc crystals only
one zinc ion appears on the three-fold axis and a considerable rearrangement of
the first eight residues of the B-chain in 3 of the 6 monomers in the unit cell takes
place. Despite the extensive structural differences a reversible interconversion of
the two forms is possible simply by adding or removing chloride or other
anions from the liquid surrounding the crystals. The transformation takes place
within one to four hours - dependent on the temperature - with cracks appearing
simultaneously in the crystals (Bentley et al. 1978). The 2 to 4 zinc conversion
seems to be most efficient with anions high in the Hofmeister series (Williamson
and Williams 1979, de Graaff et al. 1981).
The crystal structure of human insulin has been found to be nearly iso-
morphous to that ofporcine insulin in the 2 zinc vers ion (Chawdhury et al. 1983)
as well as in the 4 zinc vers ion (Smith et al. 1984).
The kinetics of insulin crystallization were investigated by Schlichtkrull
(1957 c, d, e, 1958), who found that initial nucleation may take place on surround-
ing surfaces, e.g. the interface between liquid and air. Most of the crystal seeds
are formed on the surface of existing crystals, the rate of self-nucleation being
proportional to the product of the total surface of crystals and their growth rate.
The growth of an insulin crystal is anisotropic, as insulin is only deposited on
three of the six crystal faces, the rate being a simple power function of the con-
centration of insulin in solution.

Fig. 9. a Orthorhombic insulin crystals. x 380. b Monoc1inic insulin crystals, x 1 800 (Scanning
e1ectron microscopy). c Cubic insulin crystals, x 4 500 (Scanning e1ectron microscopy). d Tetra-
gonal protamine zinc insulin (NPH) crystals, x 2250 (Scanning electron microscopy). e Rhom-
bohedral zinc insulin crystals, x 1050 (Scanning electron microscopy). f Twinned, distorted
rhombohedral beef insulin crystals ("stars"), x 380
26 Insulin Preparations

In the preparation of pharmaceutical crystal suspensions monodisperse

crystals are obtained by introducing a predetermined number ofnuc1ei at the start
ofthe crystallization (Schlichtkrull1957b). This seeding ensures that the crystals
will be of the desired size and that the self-nuc1eation becomes almost insignifi-
cant.
In the crystallization of insulin intended for pharmaceutical suspensions a
mixture is prepared containing approx. 1.5% of insulin, 7% of sodium chloride,
0.1 M of sodium acetate and a quantity of zinc ions adequate to give a total of
0.8-0.9% of zinc by weight of the insulin corresponding to 4 Zn atoms per
hexamer ofinsulin. The pH is adjusted to 5.5 and the seeding crystals added. Ini-
tialIy almost alI of the insulin is precipitated as amorphous partic1es. During the
next hour some of the insulin gradualIy goes into solution, initiating the growth
of the crystals, the fraction of which first increases rapidly and later more slowly.
At the same time the concentration of dissolved insulin decreases, indicating that
crystals grow from dissolved insulin (Fig. 10). The crystals formed are ofthe 4 Zn-
insulin structure due to the high chloride concentration in the medium (7%).
In aqueous suspension the crystals contain 40-50% ofwater and have a por-
ous structure allowing the diffusion oflow molecular substances from the suspen-
sion medium into the crystallattice.
Chemical compounds which interact with either the zinc ions or the insulin
will change the crystallization pattern. The presence of phenol or its derivatives
will produce non-rhombohedral insulin crystals, and zinc binding agents such as
phosphate will hamper the crystallization.
The rate of crystallization is decreased by increasing the zinc content or the
pH of the medium. At pH 7.4 and a total zinc ion concentration of 2 Ilg/IU the
amorphous precipitate contains about 2.3% zinc and there will be no crystal for-
mation for several years (Semilente).

O/o

100 A c

50

10 20 40 80 160 360 min

Logarithmic scale
Fig. 10. Kinetics of Lente insulin crystallization. The curves represent the fraction of insulin in
(A) amorphous, (8) dissolved and (C) crystalline state during the crystallization process. Insulin
is crystallized from a mixture (pH 5.5) containing 400 IUImi of insulin (bovine, recrystallized),
7% w/v of sodium chloride, 0.1 M sodium acetate and zinc chloride to give a total of 0.9% of
zinc by weight ofthe insulin. (After Schlichtkrull1957 c)
Mechanisms of Prolongation 27

If, however, crystals formed at pH 5.5 are transferred to pH 704, the precipi-
tate will remain crystalline (Ultralente). In the course ofneutralization additional
zinc ions are captured to the same extent as in the amorphous insulin particles.
In the manufacture of the crystalline fraction of the pharmaceutical prepara-
tion Rapitard a transformation into the 2 zinc structure ofthe bovine 4 Zn crystal
is performed at pH 5.5 (cf. C.IVA).
In the manufacture of Ultralente additional zinc ions are introduced parallel
with the pH-adjustment and a dilution of the sodium chloride concentration to
0.7%. Lange et al. (1979), studying porcine Ultralente crystals by electron and X-
ray diffraction, found a close resemblance to the 2 Zn insulin structure rather
than to the 4 Zn insulin structure.
Recently a new modification of the monoclinic crystal type of insulin has been
discovered (Brange, unpublished). Such crystals are easily formed when Ca 2 + (or
Mg 2 +, Ba 2 +) is added up to a 3-5 mM concentration to neutral solutions of
Zn insulin containing 0.2% phenol (i.e. neutral regular insulin USP, cf. C.III.2).
X-ray studies on these Ca-insulin crystals have shown that they are closely
related to the monoclinic type without Ca 2 + but distinct differences have been
observed (Dodson, personal communication). Experiments indicate that 5 Ca ions
per hexamer of insulin are bound to these crystals (Brange, unpublished).

5. Chemical Reactions
The chemical reactivity of the various functional groups in the molecule ob-
viously makes insulin susceptible to transformation. The variety of possible
chemical modifications have been summarized by Blundell et al. (1972) and
Schlichtkrull et al. (1975 b). More recent studies have indicated that some of the
amino groups of insulin are in a very reactive state at neutral pH va1ues (Sheffer
and Kaplan 1979, Friesen 1979, Chan et al. 1981, Kaplan et al. 1984). Besides
Sheffer and Kaplan demonstrated that the amino groups will readily react with
CO 2 at physiological pH. A reaction between penicillin and the amino groups of
porcine insulin in neutral solution has been described by Corran and Waley
(1975).
Amaya-F. et al. (1976) studied the reaction between insulin and glucose in the
solid state and found that during 15 days at 37 °C and 70% relative humidity an
average of IA hexose residues bind to the amino groups per molecule of insulin.
Addition of carbohydrates or glycerol (cf. C.XI.3) to neutral solutions of insulin
has been found to impair insulin chemical stability (Brange and Havelund
1983 b, Brange et al. 1982). Aldehyde impurities in the glycerol seem to be involved
in the chemical reactions seen when insulin is mixed with glycerol (Havelund,
unpublished).
The influence of heat, formulation, etc. on the different pharmaceutical
preparations is de alt with in section C.IX.

6. Mechanisms of Prolongation
The basic principle in prolonging the action of insulin is to delay the absorption
after subcutaneous injection by altering the solubility of the insulin at physiolog-
ical pH. The different approaches relate to the following:
28 Insulin Preparations

a) Cationic organic compounds, added to acid solutions ofinsulin. These will bind
the insulin at the neutral reaction of tissue fluids making heavily soluble com-
plexes. Examples of complexing agents are surfen and globin. The precipitates
formed in vivo are more readily absorbed than preformed complexes.
b) Neutral suspensions of insulin combined with basic proteins. Examples are
protamine zinc insulin (amorphous or crystalline variety) prepared with surplus
protamine and Isophane (NPH) having a stoichiometric ratio between insulin and
protamine.
c) Neutral suspensions of insulin complexed with small amounts of iinc ions. For
further details cf. C.IV.3. Examples are: the Lente type preparations with 21lg
zinc per IV in V -40 preparations and crystals made of bovine insulin containing
as little as 0.3 Ilg zinc per IV (Rapitard).
d) Alteration of the physical state and size of the suspended insulin-zinc particles
entails variations in the duration of action. Small amorphous partic1es (Semi-
lente) have a slightly slower effect than regular insulin, whereas crystalline par-
tic1es (Vltralente) have a substantially more retarded effect, which again will vary
with the size of the crystals (Fig. 11), emphasizing the importance of a uniform
size distribution from batch to batch.
e) Species of insulin of the crystalline zinc insulin suspensions. The duration of
action of the porcine crystals is somewhat shorter than that of the bovine variety
(Fig. 11). An example is Lente (bovine crystal phase) versus Monotard (porcine
crystal phase).
t) Chemical derivation ofinsulin. Hallas-M011er (1945) coupled the amino groups
of insulin with phenylisocyanate (Iso-insulin). Schlichtkrull (1958) observed that
special heat treatment of zinc insulin crystals at pH 5.5 (but not at pH 7) pro-

Residual fraction

100
75

50
40
30
Bovine 2511
20
Fig. 11. The mean absorption
course (method after Binder
10 1969) in six diabetics after
Porcine 2511 subcutaneous injection of
12 IU 125 1-labelled Ultralente
5 of different species and crystal
size into the femora1 region.
Porcine 1111
(Lauritzen and Brange,
, ~ unpublished)
o 24 48 72 96 Hours
Rapid-Acting Preparations 29

duced a substantial enhancement of the retardation. No explanation was found

at that time, but it has since become apparent that a chemi cal reaction between
hexamers in the crystals takes place under those specific conditions, resulting in
the formation of covalent dimers of insulin. The covalent link between monomers
located in different hexamers will cross-link the hexamers within the crystals and
thereby decrease the solubility of the crystals (Brange, unpubl.). Formation of
only a few percent of dimer will cause a substantial change in solubility and, con-
sequently, timing. A preparation of this type has been successfully used in treat-
ment of streptozotocin diabetic rats (Rasch 1979).

III. Rapid-Acting Preparations

1. Introduction
In the early years ofinsulin manufacture only acid solutions ofvery impure insu-
lin were available for diabetes therapy. A low pH was required to solubilize the
foreign proteins and to protect the insulin from degradation by contaminating
pancreatic enzymes. Removal of enzymes was achieved within the first decades
of insulin production, but forty years elapsed before the overall purity of insulin
was improved sufficiently to allow the manufacture of a neutral solution of insu-
lin (Schlichtkrull et al. 1961, 1965).
The introduction of a retarded preparation in 1936 (Hagedorn et al., Scott and
Fisher) ended the monopoly of the acid insulin solutions, but their great impor-
tance is reflected in the variety of synonyms (Table 3). Nowadays the term "reg-
ular insulin" designates a neutral insulin solution and is used in the following in
this sense.
In recent years the increased awareness of the importance of strict metabolic
control has resulted in multidose regimens featuring rapid acting neutral insulins
as an essential part. The important aspects of mixing rapid-acting with interme-
diate and long-acting preparations are treated in section C.VIII.
In the future, regular insulins are likely to play a new role in therapy in con-
nection with the increasing use of infusion pumps for insulin delivery. The need
for more stable insulin solutions suitable for this application is discussed in sec-
tion C.XI.

Table 3. Synonyms for acid solutions of insulin

Alt-( old) insulin Quick-acting insulin

Clear insulin Rapid-acting insulin
Crystalline insulin Regular insulin
Crystalline zinc insulin Short-acting insulin
Insulin injection Soluble insulin
Normal insulin Toronto insulin
Ordinary insulin Unmodified insulin
Plain insulin
30 Insulin Preparations

2. Acid aud Neutral Formulatious

The formulation of commercially available, rapid-acting insulin solutions varies
with respect to species, pH, zinc content, preservative, isotonic agent and buffer-
ing substance, if used. The compositions of different brands of neutral rapid-act-
ing preparations are given in Table 4.
The neutral solutions may be better tolerated at the injection site. Besides they
possess other advantages compared to the acid solutions. The chemical stability
of neutral insulin is substantially better. The rates of deamidation during storage
at 4 ec of two different acid formulations compared with a neutral solution are
shown in Fig. 20 (cf. C.IX.2). A greater loss of potency is seen in the acid than
in the neutral solutions during prolonged storage (Pingel and V01und 1972). An-
other obvious advantage is their miscibility with the long-acting neutral insulin
preparations without affecting the neutral pH of the mixture.
Andersen (1973) found that an acid solution gives rise to more insulin anti-
body formation than a neutral solution of the same batch of insulin.

Table 4. Composition of neutral regular insulin from various manufacturers (1983)

Manufacturer Brand name Species Preservative Isotonic Other

agent additives

BIM' Neusulin Beef Methylparaben Sodium Sodium

(Wellcome, chloride ace tate
Boots,
Evans) UK
Connaught Insulin- Beef m-cresol Glycerol None
Canada Toronto
CSL b Nuralin Pork Methylparaben Sodium Sodium
Australia chloride acetate
Hoechst Optisulin CR Beef Methylparaben Sodium Sodium
Germany (des-PheB1 ) chloride acetate
Lilly Regular Iletin Beef, pork, Phenol Glycerol None
US I, II mixed
Humulin S Human m-cresol Glycerol None
Nordisk Velosulin Pork m-cresol Glycerol Sodium
Denmark (Insulin Leo phosphate
Neutral)
Novo Actrapid MC Pork Methylparaben Sodium Sodium
Denmark HM Human chloride acetate
Novo Actrapid MC Pork, human Phenol Glycerol None
Denmark HM
Squibb Regular Insu- Pork Phenol Glycerol None
US lin Injection
USP
Weddel Hypurin Beef Phenol+ Glycerol Sodium
UK Neutral m-cresol phosphate

• British Insulin Manufacturers

b Commonwealth Serum Laboratories
Absorption 31

Table 5. Viscosity and density at 20 ac of neutral regular insulins and corresponding media.
Viscosity was measured with a Haake precision falling-sphere viscosimeter or a tube viscosime-
ter. Additives in the BP preparation: methylparaben 0.1 %, sodium chloride 0.7% and sodium
acetate 0.01 M; in the USP preparation: phenol 0.2% and glycerol 1.6%. (Skelba:k-Pedersen,
unpublished)

Insulin formulation Distilled

water
BP USP (table
values)
Insulin concentration IU/ml Insulin concentration IU/ml

O 40 80 O 500

Viscosity (cp) 1.01 1.02 1.02 1.04 1.09 1.002

Density (g/ml) 1.005 1.006 1.006 1.003 1.008 0.998

The pH probably does not influence the rate of absorption of insulin solutions
from the subcutaneous tissue (cf. C.1II.4).
Stable neutral solutions of porcine zinc insulin containing 0.4% zinc (corre-
sponding to approx. 2 Zn atoms per insulin hexamer) can be made in concentra-
tions up to 500 IVImI, which are used for treatment of patients with insulin resis-
tance (Nathan et al. 1981). Ifzinc-free insulin is used it is possible to obtain a con-
centration of 5 000 IV ImI in neutral medium.
The density and viscosity of different strengths and formulations of neutral
solutions of zinc insulin are given in Table 5.

3. Manufacture
A neutral rapid-acting preparation may be made by dissolving the zinc insulin
crystals in diluted hydrochloric acid at pH 3. The acid solution with an insulin
concentration of about 4% is sterilized by filtration and mixed with solutions of
the preservative, isotonic agent and, if included, the buffering agent. The solution
is neutralized slowly with diluted alkali during which local excess of base is
avoided by stirring continuously (cf. C.lI.l.). The neutral solution is adjusted to
the correct strength by addition of sterile water and filled aseptically into sterile
vials.

4. Absorption
Binder et al. (1967) found neutral insulin to be absorbed faster from subcutis than
an acid solution, however, Galloway et al. (1973) and Poulsen and Deckert (1976)
found no difference in serum insulin concentration and hypoglycaemic activity,
respectively, following subcutaneous injection of acid and neutral insulin. The
discrepancy is possibly explained by the fact that the acid solution used by Binder
contained five times as much zinc as the neutral solution. An acid insul in prep-
aration will pass the iso-electric precipitation zone of insulin after injection, and
a high zinc content will impede the redissolution of precipitated insulin and
thereby delay the absorption. The high zinc content of acid solutions was found
32 Insulin Preparations

O/o
100

80

60
Fig. 12. Absorption of Actrapid MC. Solid line
disappearance of radioactivity after
40 subcutaneous injection of 12 units 12sl-labelled
Actrapid MC 40 IU/ml into the femoral region
of 33 diabetics (mean ± SEM) measured as
20 ,,-- ..... described by Binder (1969). Broken line rate of
I "
II ' ... .... _ absorption in percent obtained by derivation of
I
the disappearance curve. (After Schlichtkrull
1977 b)
o 2 4 6 8 Hours

to stabilize the insulin (Sahyun et al. 1939), most likely because of an inactiv ating
influence on contaminating enzymes. Lens (1947) was not able to confirm this sta-
bilizing effect of zinc, possibly because ilie insulin examined was devoid of en-
zymes.
The absorption course of Actrapid after subcutaneous injection into the femur
is shown in Fig. 12. The rate curve obtained by differentiation shows that it takes
almost two hours for the insulin to be absorbed at maximum rate, viz. approxi-
mately 25 % of the dose per hour. (Binder et al. 1967, Schlichtkrull 1977 b).
Comparing two neutral regular insulins of different formulation (Actrapid
with sodium chloride, methylparaben and Leo Regular with glycerol, m-cresol)
Berger et al. (1982) found that the absorption kinetics were virtually identical.

IV. Protracted Preparations

1. Introduction
Shortly after the introduction of the first acid insulin solutions the search for pro-
tracted insulin preparations started. The very first attempts to prolong the action
of insulin included combination with gum arabic (1923), lecithin (1923), oii sus-
pensions (1925), other proteins (1925) and cholesterol (1926); however, they were
unsuccessful due to poor stability ofthe preparations, pain upon injection or too
variable an absorption rate. Subsequently attempts were made to obtain pro-
longed effect by the combination of insulin with vasoconstrictory hormones, such
as adrenaline and vasopressin, but these attempts were also abortive due to great
variations in the clinical effect. Reference is made to D6rzbach and Miiller (1971)
for a review of the earliest attempts to prolong the action of insulin.
The first successful protracted insulin preparation was protamine insulin, in-
troduced in 1936 by Hagedorn and co-workers. The principle was to depress the
solubility of insulin at neutral pH using a basic compound. Protamines were,
Table 6. Formulation of protracted insulin preparations (1983) "O
...o........
Classification Preparations with examples of trade names pH Physical state Retarding Species of Insulin ()
after timing principle ...'"
(1)
o-
of action Acid Neutral d a c p b p(b h Other "O
...,
(1)
"O
Short-acting Insulin Zinc Suspension (amorphous), x x Zinc X x x ....
Semilente, Semitard 'c."
'o"
Intermediate- Insulin Zinc Suspension (mixed) Monotard, x x x Zinc x x x x des-Phe ::l
CA
acting Monotard HM, Lente, Lentard, Hypurin
Lente, Neulente, Optisulin Long
Isophane Insulin x x Protamine x x x x
NPH, Protaphane, Isophane, Insulatard,
Retard, Hypurin Isophane, Neuphane,
Humulin 1
Surfen-insulins, Komb-insulin, Depot-insulin x x Surfen x x
Long-insulin x x x Surfen x
Globin Zinc Insulin, HG-insulin, Globin x x Globin x x
Protamine Zinc Insulins, (amorphous) x x Protamine + Zinc x x
Long-acting Insulin Zinc Suspension (crystalline), x x Zinc x
Ultralente, Ultratard
Protamine Zinc Insulins, Hipurin PZ x x x Protamine + Zinc x x x
Biphasic Biphasic Insulin x x x Zinc x
Rapitard, Biphasic
Biphasic Protamine Insulins x x x Protamine x
Mixtard, Initard

Abbreviations: d = dissolved, a = amorphous, c = crystalline, p = porcine, b = bovine, h = human

w
w
34 Insulin Preparations

among other basic peptides (histones, globins, etc.), found to show the most pro-
mising effect. Since the first neutral protamine insulin suspension was not stable,
it was necessary to dispense this new preparation in two separate vials, one con-
taining a phosphate buffer and the other an acid solution of protamine and insu-
lin (showing the same timing of action as soluble rapid acting insulins when in-
jected separately). The patient would then prepare sufficient suspension for a few
days by injecting buffer into the vial with the acid solution of protamine and
insulin.
Later more stable protracted preparations were developed, stiU including the
combination of insulin with foreign proteins. It was not until the introduction of
the Lente preparations (Hallas-M0ller et al. 1951) that protracted insulin prepa-
rations without added foreign proteins or synthetic compounds were obtained.
The composition of commercially available protracted insulin preparations is
given in Table 6.

2. Protamine Insulins
a) Protamine
Protamine is the generic name of a group of strongly basic proteins present in the
sperm cell nuclei in saltlike combination with nucleic acids. Commercially avail-
able protamines are made from fish sperm and usually obtained as the sulphate
salt.
Since protamines emanating from different families, genera and species offish
vary as to peptide composition, it is desirable to specify the family, genus and spe-
cies of the fish from which the protamine is isolated. Protamines used together
with insulin are normaHy obtained from salmon (salmine) or trout (iridine).
Salmine and iridine are inhomogeneous and have been separated into two and
three main fractions, respectively (Ando and Watanabe 1969). Each ofthese frac-
tions is probably also heterogeneous as shown for iridine (Ling et al. 1971), but
the different peptides having about 30 amino acid residues are very similar in
structure. The average molecular weight of protamine has been found to be ap-
prox. 4,300 for iridine and 4,250 for one of the fractions of salmine (Ando and
Watanabe 1969).
Basic residues constitute about two thirds of aH the amino acid residues in
protamines resulting in an isoelectric point of above 12. Salmine and iridine be-
long to the monoprotamines, which contain arginine as the only basic residue. In
addition these protamines contain relatively few other residues predominantly
Ser, Pro, Val and Gly.
Being devoid of aromatic amino acid residues, protamines have no UV-ab-
sorption in the region 260-360 nm, thus any absorption in this range denotes the
presence of potential impurities, e.g. DNA or histones, which contain aromatic
residues and are of higher molecular weight.
Protamines were earlier considered non-immunogenic (Kern and Langner
1939, Jaques 1949) and the observed immunological reactions to protamine have
been attributed to contaminants (Caplan and Berkman 1976). AHergic reactions
due to the protamine content ofNPH preparations have, however, been reported
(Shore et al. 1975, Sanchez et al. 1982) and Samuel concluded that protamines
Protamine Insulins 35

can be immunogenic in man and their use for medical purposes may lead to for-
mation of antibodies (Samuel 1977, Samuel et al. 1978). Recently Kurtz et al.
(1983) found evidence that the protamine-insulin complex is itself immunogenic,
as they showed a high prevalence of concomitant circulating antibodies against
insulin and protamine in patients treated with protamine-containing insulin prep-
arations.
Anaphylactic reactions to protamine sulphate have also been reported
(Nordstrom et al. 1978, Moorthy et al. 1980, Knape et al. 1981, Vontz et al. 1982,
Weiler et al. 1985), and the possibility of an allergic reaction to protamine must be
considered in patients who are allergic to fish (Knape et al. 1981). Stewart et al.
(1984) found a 50-fold increased risk of a severe adverse reaction to protamine if
protamine is administered to diabetics receiving protamine-containing insulin
preparations. The toxicity of protamine has recent1y been reviewed by Horrow
(1985).

b) Protamine Zinc Insulin (PZI)

The first stable neutral suspension was developed by Scott and Fisher (1936), who
discovered that a surplus of protamine and zinc salt in small quantities (2 /lg zinc/
IU) could stabilize the neutral protamine insulin. This protamine zinc insulin
(PZI) has a much more prolonged effect, which may last for up to 72 h (Colwell,
1947). The diss01ution ofinsulin before absorption is presumably due to degrada-
tion of the' protamine by the fibrinolytic tissue enzymes (Hagedorn et al. 1936,
Hagedorn 1946, Bang 1946, Brunfeldt and Poulsen 1953). The protaminase activ-
ity of serum is inhibited by zinc (Brunfeldt and Poulsen 1953), which may explain
the different timing of PZI and NPH.
PZI made according to the United States or European Pharmacopoeias con-
tains amorphous as well as crystalline protamine zinc insulin. Freshly prepared
this PZI contains mainly amorphous precipitate which will gradually be trans-
formed to crystalline particles upon storage, leading to a more protracted effect.
Physically stable PZI preparations can be made either by avoiding the zinc bind-
ing phosphate buffer (resulting in intermediate-acting amorphous PZI)
(Schlichtkrull1958) or by crystallizing the protamine insu1in (cf. NPH) before the
surplus of protamine and zinc are added (resulting in long acting crystalline
PZI).

c) NPH
A stable PZI modification, NPH (Neutral Protamine Hagedorn) also called
"Isophane Insulin", was deve10ped by Krayenbiihl and Rosenberg (1946) at
N ordisk Insulinlaboratorium. They found that insulin and protamine brought to-
gether in isophane proportions (the condition in which neither insulin nor prot-
amine is found in excess) at neutral pH, in the presence of a small amount of zinc
and phenol and/or phenol derivatives (cresols) will form an amorphous precipi-
tate, which is gradually transformed into tetragonal oblong crystals limited at the
ends by pyramidal faces (Fig. 9 d). At pH 7.3 insulin and salmine co-precipitate
in a 5 : 1 molar ratio corresponding to about 0.13 mg protamine per mg insulin
(Simkin et al. 1970).
36 Insulin Preparations

The folIowing conditions are necessary for rapid and complete crystalIization
of protamine insulin: the protamine, insulin and auxiliary substances must be rea-
sonably pure and the proportion between protamine and insulin nearly isophane.
The isophane ratio varies with different protamine and insulin qualities and spe-
cies as well as with pH, temperature, content of zinc and auxiliary substances.
Zinc in a concentration of approx. 0.2 ~g/IU is necessary for the preparation of
the tetragonal crystals as is the presence of phenol or phenol derivatives. m-cresol
is more suitable for the crystalIization than the corresponding 0- or p-derivatives
(Krayenbiihl and Rosenberg 1946) or phenol (Fullerton and Low 1970).
Insulin-salmine crystals contain insulin and protamine combined in a complex
in a molar ratio of8.5 to 1 (Fullerton and Low 1970). It is assumed that protamine
is to be found in the interstices between hexamers but not as an ordered compo-
nent of the crystallattice (Simkin et al. 1970, Hodgkin 1974).
Zinc is present in the same amount as in rhombohedral crystals (cf. C.II.4)
(2 atoms of zinc per hexamer) together with about 0.5 x 10- 3 moI phenol (or
phenol derivatives) per gram of protamine insulin corresponding to 22 mols of
phenol per hexamer (Krayenbiihl and Rosenberg 1946). Suspensions of such
crystals are stable in the absence of protamine-degrading proteolytic enzymes,
such as pancreatic enzymes, which may be present in impure insulin.
Graham and Pomeroy (1984) found marked differences in duration of action
between different brands ofNPH largely due to variations in crystal size and shape
rather than to differences in insulin species.

3. Insutin Zinc Suspensions (Lente Insutins)

The Lente insulins were developed in the NOVO laboratories by HalIas-M01Ier
and co-workers (1951, HalIas-M01Ier 1956), who elucidated the influence of zinc
ions on the timing of insulin preparations. They showed that insulin preparations
with protracted effect could be obtained by addition ofsmall amounts of zinc ions
provided that the preparations had a neutral pH and that no interfering (zinc
binding) ions, like phosphate or citrate, were present.
Furthermore they showed that the degree of protraction at the same zinc con-
centration depended on the physical state of the suspended insulin particles,
amorphous insulin particles having a shorter time of action than crystalline insu-
lin particles.
This led to the introduction of three new protracted insulin preparations alI
containing approx. 80 ~g zinc/mI (40 IU/ml): Semilente containing only amor-
phous insulin particles, Ultralente containing only crystalline insulin particles and
Lente containing a 3 : 7 mixture of amorphous and crystalline insulin particles.
Bovine insulin was originalIy chosen for the crystalline part and porcine insulin
for the amorphous part. Later Monotard, having a composition like Lente with
the exception that Monotard contains porcine or human insulin only, and Lente
preparations, containing only insulin of bovine origin, have been introduced.
Preparations containing insulin derivatives (des-PheB1 insulin, cf. C.XIl) in
the amorphous part have been investigated clinicalIy (Bottermann et al. 1980).
The Lente insulins were originalIy developed in the search for preparations
able to cover diabetics' insulin requirement with one daily injection. Adjustments
may be made by mixing the three types of Lente preparations in different ratios.
Insulin Zinc Suspensions (Lente Insulins) 37

In modern clinic al treatment individual mixtures of intermediate-acting and

rapid-acting preparations are often given more than twice daily. A regime where
the basal insulin requirement is covered by Ultratard, while Actrapid is used to
cover meals, has been described by Phillips et al. (1979) and Ward et al. (1981).
In the Lente preparations a certain proportion of the total amount of zinc is
present as zinc ions bound within the suspended insulin particles, whereas the rest
is present as free zinc ions in solution.
Schlichtkrull (1958) established an empirical relationship between the amount
of bound zinc ions (in number of zinc atoms per insulin hexamer), the concentra-
tion of free zinc ions and the concentration ofhydrogen ions (constant tempera-
ture, pH around 7.3):

At the constant pH value of the preparations this means that at constant free zinc
concentration, the amount of bound zinc in mg/lOO IU of insulin (proportional
to the number of zinc atoms per insulin hexamer) will be constant independent
of total insulin concentration.
A concentration of free zinc of approx. 0.05 mg/ml at pH 7.4 was chosen for
the original Lente series. This leads to an amount of bound zinc of approx.
0.09 mg/lOO IU of insulin.
A calculation of bound and total zinc based on a concentration of free zinc
of 0.05 mg/ml is shown in Table 7 for three different insulin strengths; as seen the
total zinc content relative to insulin (in mg/100 IU) is adapted to the various
strengths in order to maintain constancy ofthe chemical composition ofthe solid
phase and of the suspension medium, aiming at an unchanged protracted effect.
The composition of Lente preparations with respect to other auxiliary sub-
stances (ef. Table 8) is described in section C.V, but it should be emphasized that,
for example, phenol is un accepta bie as preservative, since its presence leads to a
change of the physical state of the insulin particles and thus probably a change
of the dissolution and timing properties.
In the manufacture of Ultralente or the crystalline part of Lente and Mono-
tard, the insulin is dissolved in acid, sterilized by filtration and thereafter crystal-
lized as described in C.II.4. After the crystallization, pH is adjusted to 7.4 and the
correct concentrations of insulin, zinc and other auxiliary substances are estab-

Table 7. Ca1culated values of bound, free and total zinc content in Lente
preparations containing 40, 80 and 100 IV/mI with the presupposition that the
content of free zinc is constant

Insu1in [Zn2+] rZn 2 +] bound [Zn 2 +] total

strength free
IU/ml mg/m1 mg/lOOIU mg/m1 mg/m1 mg/lOOIU

40 0.05 0.09 0.036 0.086 0.22

80 0.05 0.09 0.072 0.122 0.15
100 0.05 0.09 0.090 0.140 0.14
38 Insulin Preparations

Fig.l3. Amorphous flocculation. Photomicrograph ofSemilente showing loose aggregates of in-

divid ual amorphous particles. 176 x magnification, using differential interference contrast tech-
nique

lished. The crystals formed during the crystallization are of the 4 Zn-insulin
structure (cf. C.IIA) but X-ray studies have shown that complete conversion into
the 2 Zn-insuJin structure takes place during dilution and pH adjustment of the
crystal suspension (Dodson, personal communication).
In order to obtain a constant and narrow size distribution of the insuJin
crystals a special seeding technique must be used (Schlichtkrulll957 b). This is ad-
vantageous because the timing of the preparations is to some degree dependent
on the size of the crystals (cf. C.II.6, Fig. 11).
In the manufacture of Semilente or the amorphous part of the Lente prepara-
tions, the insuJin is dissolved at acid pH and after sterile filtration the correct pH
and zinc concentration are established. This leads to a precipitation of insuJin as
amorphous particles.
The size of the individual amorphous particles is about I ~m. Microscopi-
cally, the amorphous particles are seen to form loose aggregates (Fig. 13). This
flocculation is of importance for the physical stabiJity of the preparations. For-
mation of lumps or flakes due to inappropriate storage has been shown to de-
crease with increased flocculation (Langkjrer, unpubl.). Similar results have been
reported by Haines and Martin (1961) for other suspensions.
The original Lente preparations were prepared using zinc chloride as the
source of zinc ions but, as the amount of zinc ions present is the important factor,
other zinc salts (e.g. acetate) may also be used.
For a de tai led survey of the physical chemistry of the Lente insulins reference
is made to Schlichtkrull et al. (1975 a).

4. Biphasic Preparations
When using intermediate-acting insuJin preparations the initial insulin effect is
often too slight. A stronger initial effect can be obtained by mixing rapid- or
Other Types of Protracted Preparations 39

short-acting preparations with intermediate-acting preparations (cf. C.VIII). The

need for a stronger initial effect led to the search for sta bIe mixtures of rapid-act-
ing (dissolved) and intermediate- to long-acting insulin resulting in the develop-
ment of Rapitard (Biphasic Insulin Injection) (Schlichtkrull 1959, Schlichtkrull
et al. 1965).
Rapitard contains 75% crystalline bovine insulin suspended in a solution of
25% ofmainly porcine insulin. The rationale for this preparation is the utilization
ofthe difference in solubility between porcine and bovine insulin at specific values
of pH and zinc concentration (cf. C.II.l). Due to the low solubility of the bovine
insulin, only a small amount of the dissolved insulin is bovine.
The insulin crystals are prepared in the same way as the Ultralente and the
crystalline part of the Lente preparation (cf. C.II.4 and C.IV.3). Crystallization
in the presence of 7% sodium chloride leads to crystals containing 4 zinc atoms
per insulin hexamer. After crystallization the preparation is diluted at constant
pH (5.5) to the final concentration of insulin and auxiliary substances (i.e. 0.7%
w/v of sodium chloride).
This leads to a crystal transformation from crystals containing 4 zinc atoms
per insulin hexamer to crystals containing 2 zinc atoms per hexamer. The crystal
suspension is later adjusted to pH 7 and mixed in the correct proportion with the
separately prepared neutral solution of insulin. Rapitard is physically stable, the
proportion of insulin in solution remaining constant during shelf life.
Two biphasic preparations have been introduced based on Isophane Insulin:
Initard, al: 1 combination of regular and NPH insulin, and Mixtard, prepared
by adding 3 parts ofregular insulin to 7 parts ofNPH insulin. These mixtures are
not physically stable, as some of the dissolved insulin is transferred to the solid
phase. In Mixtard only about half the amount of the added regular insulin is re-
covered in solution (Galloway et al. 1982), whereas in Initard two thirds of the
added regular insulin can be detected in the supernatant (Lykkeberg, unpubl.).
A biphasic insulin preparation utilizing the higher solubility of the insulin de-
rivative, des-PheB1 -insulin (cf. C.xIl), as compared to that ofthe parental insulin
has been introduced (Optisulin Depot) (Zoltobrocki et al. 1980).

5. Other Types of Protracted Preparations

Surfen insulin preparations of intermediate and long duration of action have been
widely used in Germany. The prolonged action of these preparations is based on
formation of a slightly soluble complex between insulin and the synthetically pro-
duced substance, 1,3-bis (4-amino-2-methyl-6-quinolyl) urea (surfen) at neutral
reaction (Lautenschliiger et al. 1937, Umber et al. 1938). The intermediate acting
Depot-Insulin with 4.2 llg surfen/IU is an acid solution from which the surfen in-
sulin complex precipitates as amorphous particles after injection. Komb-Insulin
is a mixture of one part acid Regular Insulin and two parts Depot-Insulin, which
is a neutral suspension of crystalline and amorphous surfen insulin in the ratio
3: 1.
Several cases of allergic reactions to surfen have been reported (Kulpe 1958,
Forck et al. 1975, Goerz et al. 1981) and the incidence of these reactions has in-
creased in recent years (Goerz et al. 1981).
40 lnsulin Preparations

Bovine and later also human globin, the protein part of haemoglobin with
four polypeptide chains varying slightly in structure from species to species, have
also been used as complexing agents. Globin insulin, developed by Reiner et al.
(1939) and dispensed in acid solution, is no longer widely used.

V. Auxiliary Substances

1. Introduction
The auxiliary substances used for the formulation of insulin preparations can be
divided into five groups: 1. retarding substances (cf. C.IV), 2. preservatives, 3. iso-
tonic agents, 4. buffering substances and 5. acid and basic substances for adjust-
ment ofpH.
A survey of auxiliary substances, other than retarding substances, used in
commercially available protracted preparations is shown in Table 8.

2. Preservatives
As the individual dosage varies considerably from patient to patient multidose
vials have to be used. Consequently, the preparations must be formulated with

Table 8. Auxiliary substances in protracted insulin preparations

Classification Preparation Preservative lsotonic agent Other additives

after timing includ ing buffering
of action substances

Short-acting lnsulin Zinc Methylparaben Sodium chloride Sodium ace tate

Suspension
amorphous
lntermediate- lnsulin Zinc Methylparabcn Sodium chloride Sodium acetate
acting Suspension
mixed
lsophane Insulin Phenol and Glycerol Sodium phosphate
m-cresol
Surfen-insulin Methylparaben Glycerol'
Sodium chloride b -
Globin Zinc Phenol, cresol Of Glycerol
lnsulin methylparaben
ZPI-Novo Methylparaben G1ycerol Sodium acetate
Long-acting lnsulin Zinc Methylparaben Sodium chloride Sodium acetate
Suspension
(crystalline)
Protamine Zinc Phenol Glycerol Sodium phosphate
lnsulins
Biphasic Biphasic Insulin Methylparaben Sodium chloride Sodium acetate
Biphasic Prot- Phenol and Glycerol Sodium phosphate
amine Insulins m-cresol

• acid forms= Depot, Komb.

b neutral form = Long
Preservatives 41

r-~---- __ ,---- . ------------------~----------~

Fig. 14. Photomicrograph of Monotard modified by substitution of methylparaben with phenol,

showing the appearance of abnormal crystals after storage for 6 months at 25 ne. (460 X magni-
fication)

a suitable antimicrobial preservative. Selection of a proper antimicrobial agent is

a compromise of efficacy, toxicity, compatibility with the insulin and interaction
with the vi al and the rubber closure (AlIwood 1978). For instance, preservation
of insulin zinc suspensions (Lente series) with phenol is not possible without in-
fluencing the physical stability ofthe preparation (Fig. 14), whereas methyI4-hy-
droxy benzoate (methylparaben) does not exhibit this effect.
On the other hand, the presence ofphenol (or a phenol derivative) is a prereq-
uisite for obtaining protamine insulin crystals (NPH) (cf. C.IV.2.c).
The antimicrobial efficacy of phenols is generally considered to be better than
that ofthe parabens (alkyI4-hydroxybenzoates). Thus, AlIwood (1982) compared
the efficacy ofvarious preservatives in different insulin injections and found neu-
trai preparations with phenol or cresol relatively well preserved when compared
to preparations containing methylparaben. Schade and Eaton (1982) inoculated
10 mi of neutral insulin solutions formulated with phenol with 5 x 10 5 bacteria
commonly found on skin and found that the bacteria were killed within 24 hours.
They also demonstrated a strong bactericidal effect of zinc ions at a concentration
similar to that of the Lente series. Comparing the antimicrobial properties of dif-
ferent parabens O'Neill and Mead (1982) demonstrated that methylparaben in
neutral solution is a potent preservative and superior to other parabens.
The concentration of the preservatives is affected by sorption (viz. adsorption
and absorption) and in the case of parabens also hydrolysis. The distribution be-
tween water and a rubber closure was measured by Wallhăusser (1974), who
found stronger sorption of phenol and m-cresol, especially, than of methyl-
paraben. Rowles et al (1971) showed that phenol and m-cresol are not only ab-
sorbed by but also permeate the closure.
The hydrolysis of methylparaben was studied by Blaug and Grant (1974) who
found that phosphate increases the rate of hydrolysis. The amount of methyl-
42 Insulin Preparations

Table 9. Loss ofmethylparaben during storage

of Lente when the preparation is in direct
contact with the rubber c10sure (storage upside
down). (Brange, unpublished)

Storage at 4 ac Storage
at 25 ac
l year 3 years l year

Hydrolysis 2% 4% 9%
Sorption 7% 10% 12%

paraben disappearing due to absorption and hydrolysis during storage of Lente

insulin is shown in Table 9.
The presence ofthe preservative has caused misinterpretations ofthe chroma-
tograms obtained by analytical gel filtration of insulin preparations (cf. B.I.2).

3. Isotonic Agents
As with the preservative, the selection of a suitable isotonic agent is not a free op-
tion. Thus the use of sodium chloride is mandatory in the Lente series, because
it is necessary during formation of the crystalline phase (cf. C.II.4 and C.IV.3).
The choice ofpreservative and isotonic agent will affect the properties ofthe prep-
aration, as revealed by the chemical stability (cf. C.IX.2) and the miscibility of
neutral solutions with intermediate-acting preparations (cf. C. VIII. 1).
The interaction when adding aprotinin to neutral insulin solutions has also
been shown to vary with the formulation. Thus when glycerol is used as isotonic
agent (low ion strength) a precipitation occurs which is not seen when using
sodium chloride (Havelund, unpublished).

4. Buffering Substances
If the vial and rubber closure are of sufficient1y good quality, proper pH is main-
tained without addition of buffer. Thus Jackson et al. (1972) found buffering of
neutral insulin solutions unnecessary. In the Lente series acetate is used as buffer
during crystallization at pH 5.5 and only negligible buffer capacity remains upon
adjustment to neutral reaction. However, the preservative, methylparaben (pK a
8.4), possesses a certain buffer capacity (Tammilehto and Buchi 1968).
The presence of free zinc ions is prohibitive for the use of phosphate buffer
in the Lente series due to the low solubility of zinc phosphate (Schlichtkru1l1958,
Schlichtkrull et al. 1975 a). This also makes rapid-acting preparations containing
phosphate buffer unsuitable for mixing with this type of preparation (cf.
C.VIII.l).
Analytical Control Specific for Special Preparations 43

VI. Characterization and Testing

1. Introduction
The different insulin preparations are subjected to a number of analytical tests in
order to ensure their identity, and specified characteristics and uniformity. The
various tests can be divided into groups after their application, one group cover-
ing the tests common to alI the preparations and the other groups specific for cer-
tain special preparations. A historical review of the analytical control of insulin
has been published by Stewart (1974).

2. Analytical Tests Applied to alI Preparations

a) Control of Insulin Concentration
In the early days of insulin therapy biological standardization of insulin (cf.
B.II.1) and insulin preparations was introduced as a necessary measure due to the
large and varying content of impurities. Today this control is stiH based on bio-
logical standardization, even of the highly purified insulins, however, work is in
progress to replace bioassay methods by analytical chemical methods (cf.
B.II.2).
b) Control of Homogeneity of Insulin Suspensions
Control of the homogeneity of the filled vials is performed by nitrogen determi-
nations on samples taken at regular intervals during the filling procedure. The re-
sults have to comply to specified limits.
c) Determination ofpH, Zinc, Isotonic Agent, Preservative, and Species
The pH is of great importance for the physicochemical properties of insulin (cf.
C.II), and it must be within narrow limits.
Like pH, the zinc concentration inf1uences the physicochemical properties of
insulin. The zinc concentration can be determined by atomic absorption spectros-
copy as described in the British Pharmacopoeia.
The concentration of isotonic agent can be determined non-specificalIy by de-
termining the freezing point depression of the solutionjsuspension or by specifi-
calIy determining the concentration of the isotonic agent actualIy applied.
As the preservatives used in insulin preparations are strongly UV -absorbing,
their concentration can be determined by UV -spectroscopy, if necessary after pre-
cipitating the insulin by addition of a surplus of zinc and adjustment of pH to a
value where zinc is capable of precipitating the insulin.
Bovine insulin can be distinguished from porcine and human insulin by its
ability to form so-called "stars" during crystalIization under proper conditions
(cf. C.II.4). A new, convenient and more sensitive method capable of separating
bovine, porcine and human insulin is HPLC (cf. B.I.S).

3. Analytical Control Specific for Special Preparations

a) Protamine Insulins
Contamination of isophane insulin by proteases (potential contaminants from
pancreas) can lead to digestion ofprotamine and a consequent release ofinsulin
44 Insulin Preparations

into solution. This phenomenon willlead to dramatic changes ofthe timing char-
acteristics of the preparations and an analytical test for protaminase activity is
therefore essential. The British and the European Pharmacopoeias describe a
method of analysis for determining the proteolytic activity in Isophane Insulin by
which the loss ofweight ofthe isolated, washed and dried insulin protamine com-
plex after incubation at 37 cC for 30 days is measured. Other more specific deter-
minations of the sort of proteolytic activity have been described (Caygill and Ayl-
ing 1979), but the non-specific pharmacopoeia method has the advantage that it
is a measure ofthe amount ofinsulin-protamine complex degraded. The determi-
nation of a rather smaU (relative) weight difference after isolating, washing and
drying of the insulin protamine complex is inaccurate. The analysis may be im-
proved considerably by determining the increase of nitrogen content in the super-
natant.
It is ofimportance for the timing ofaction ofthe preparations that no insulin
occurs in solution. The absence of insu/in in solution can be determined by bioas-
say for aU protamine insulins, or for Isophane Insulin by control of the isophane
ratio as described in the foUowing.
Deviations from the isophane ratio between insulin and protamine (cf. C.IV.2)
can be controlled by isolating samples of the supernatant liquid and adding neu-
traI solutions of protamine and insulin, respective1y. If the supernatant liquid
contains insulin the addition of protamine solution will cause the formation of
turbidity; if the supernatant contains protamine the addition of insulin will cause
the formation ofturbidity.
b) Lente Insulins
As mentioned previously the total zinc concentration is of significance for aU in-
sulin preparations. For the Lente insulins it is, however, not only the total zinc
concentration that is of interest but also the distribution of zinc between free zinc
ions in solution and zinc bound to the suspended insulin partic1es (cf. C.IV.3).
Therefore the zinc ion concentration in the supernatant liquid isolated after cen-
trifugation is also determined.
The absence of insulin in solution may be controlled by bioassay but, as insulin
is the sole protein present in the Lente insulins, the control can also be performed
by non-specific determination of protein or nitrogen in the supernatant liquid.
For Lente and Ultralente preparations the content of amorphous insulin par-
tic1es is determined (methods described in various pharmacopoeias). As the amor-
phous partic1es are soluble in a buffered acetone/water solution, the content of
amorphous insulin present is determined as the proportion of nitrogen (equiva-
lent to insulin) which can be extracted by this buffer. An alternative method of
analysis has been described by Orr and Spence (1977). By this method the dis-
tribution of the differently sized amorphous and crystalline partic1es is deter-
mined using an electronic partic1e counter, whereby the proportion of amorphous
partic1es can be estimated.
The crystal size distribution in Ultralente and the crystalline part of Lente,
which is of importance for the timing of action of the preparations (cf. C.II.6),
can be determined microscopically (Schlichtkrull1957 a) or by using an electronic
partic1e counter.
Prolongation Tests 45

c) Biphasic Insulin Preparations

For the biphasic preparations the proportion of insulin present in dissolved state
is of importance for the timing. This proportion is determined e.g. by nitrogen
determination.
The crystal size distribution of Rapitard is determined as described above for
Lente insulins.

4. Prolongation Tests
The timing of action of insulin preparations has been studied in rabbits, guinea
pigs, mice, depancreatized dogs, healthy students and diabetics (both with stable
and unstable blood sugar) using either strictly controlled conditions or conditions
simulating daily life. Biological tests in animals are used for the control of con-
stancy of the timing of action of an insulin preparation, e.g. from one batch to
another, or during its she1f1ife until the date of expiry. Such tests for prolongation
of insulin effect using either guinea pigs or rabbits are described in the pharma-
copeias (British Pharmacopoeia 1980, Eur. Pharmacopoeia 1975, US Pharmaco-
poeia 1980). In these tests the prolongation ofthe hypoglycaemic effect produced
by the prolonged acting insulin preparation is compared with the effect of the
rapid-acting standard preparation. Figure 15 shows an example of such a test for
prolongation of insulin effect where Insulin Monotard MC, fresh and stored, is
compared with Insulin Actrapid MC. The test is performed according to the
method described in BP (1980) using 36 rabbits in a cross-over design. The data
recorded are the mean blood glucose values ofthe groups ofrabbits obtained Y2,
1, 2, 4, 6, and 7 hours after the injection expressed in percent of the initial blood
glucose. Both Monotard preparations show prolonged effect compared to Insulin
Actrapid MC, and analyses ofvariance followed by paired t-tests show no statis-
tically significant (5 percent sign. leve1) differences between Monotard MC
freshly prepared and that stored for 3 years at 4 ac.

Cross·over assay in 36 rabbits

Dose: 1.6 units per rabbit
100
................................
~ 90
:§
'O
"ECI) 80

...
1.)

CI)
c. 70
CI) Actrapid MC ............•.
!Il
o Fig. 15. Cross-over assay on 36
1.) 60
:1 Monotard MC rabbits where the timing of action
C, ofInsulin Monotard MC after
fresh
'"CI
o 50 preparation and after 3 years' stor-
o
age at 4 ac is compared with that
Monotard MC ____ _
ai 3 years 4°C
40 of Insulin Actrapid MC. (Serensen
and Pingel, unpublished)
o 2 3 4 5 6 7
Hours after injection
46 Insulin Preparations

% of insulin eluted per 10 min

5. Fig. 16. In vitro elution profite of

Monotard 80 IV jml. Cell volume
10.0 mI. Eluent flow 0.19 ml/min.
3 Temperature 37 ac. Time to
maxima is dependent on the condi-
tions of elution. In this experiment,
approx. 1 and 4.5 h (Skelbrek-
L....!_ _ _ _ _ _ _ _ _ _ _ _ _ _-="",. Pedersen, unpublished)
Time

An in vUro absorption model has been developed (Skelbrek-Pedersen, unpub-

lished) to imitate the in vivo absorption of insulin preparations. The model con-
sists of a "flow-through" cell with stirrer thermostated at 37 °e, and a filter fine
enough to retain suspended insulin particles of different kinds. The cell is eluted
with buffer, the composition of which imitates the composition of interstitial
fluid. The eluate is collected in fractions in which the insulin concentration is de-
termined. The result of such an experiment can be seen in Fig. 16, showing the elu-
tion profile of Monotard. The in vitro model allows a high degree of sensitivity
and clearly differentiates the biphasic dissolution profile corresponding to the
amorphous and crystalline insulin particles, respectively. An in vitro test based on
similar principles, but without the same possibility to demonstrate biphasic
action, has been published by Graham and Pomeroy (1984).

VII. Insulin Strengths

1. Introduction
The strength of insulin preparations is expressed in international units per mI or
IU/ml. Since the discovery of insulin there have been a number of definitions of
the unit of insulin. In 1923 it was decided to prepare a standard of Insulin with
a defined specific activity. This led to the introduction in 1925 ofthe Ist Interna-
tional Standard of Insulin having a specific activity of 8 units per mg. This has
been followed by the 2nd, 3rd and 4th International Standards of Insulin intro-
duced in 1935, 1952 and 1958, the latter a mixture of bovine and porcine insulin
and with a potency of24 IU per mg. At present new international standards based
on twice chromatographed species pure insulin (cf. B.II.1) are under way
(WHO 1982).
The standardization of insulin samples relative to the international standards
is described in section B.II.l. For a thorough survey of the development of the
unit of insulin reference is made to the review by Lacey (1967).

2. Strengths of Insulin Preparations

The first inslilin preparations contained only 1 unit per mI, but as the isolation
and purification procedures became better, preparations of higher strength be-
carne available: 1922: up to 10 units/ml, 1923: 20 units/ml, 1924: 40 units/ml and
Strengths of Insulin Preparations 47

in 1925: 80 units/ml (Scheller and GalIoway 1975). For a long time preparations
containing 40 and 80 units/ml were the two standard strengths produced. Because
of reports of medication errors caused by the use of both U-40 and U-80 prepa-
rations (Watkins et al. 1967) the American Diabetes Association decided that
only one concentration of insulin should be available in the United States. The
concentration of 100 IU /ml was chosen by the American Diabetes Association,
FDA and the insulin manufacturers in the USA and Canada, based on the folIow-
ing arguments: 1) the availability of syringes capable of accurately delivering
small doses of concentrated insulin, 2) the trend toward adoption of the metric
system in the United States and 3) the desirability of reducing the injection vol-
umes. A trial with U-100 Lente and regular in children showed its safety and ef-
ficacy (Rosenbloom et al. 1971). Since then U-100 insulin preparations have been
introduced in a number of countries, e.g. USA 1973, Canada 1975, Australia and
New Zealand 1980, United Kingdom 1983, Denmark 1986 (in some cases
paralleled by the withdrawal of alI other strengths).
Arguments have been raised against the use ofU-100 as the only strength. The
main problem is the difficulty of accurately administering very small doses ofhigh
strength insulin, as in the case of children for whom dilution of the U -100 prep-
arations is often necessary (Jackson et al. 1977). Cogswell and Simpson (1980)
found that 24 out of 40 newly diagnosed patients required daily doses of 4 units
or less some time during the first few years folIowing diagnosis. The frequent ne-
cessity of diluting U-100 preparations makes for even greater confusion than with
therapy based on the U-40 and U-80 preparations. The trend towards improving
insulin therapy by dividing the daily insulin do se into two or more injections wilI
also require smaller doses (Rosenbloom 1974).
If syringes with significant dead space are used for mixing different insulin
preparations, gre ater problems arise with U-IOO than with U-40 or U-80 prepa-
rations because larger insulin doses (up to 10 units) are contained in the dead-
space (cf. C.x.3 and Table 18). These problems may to some degree be overcome
by the use of properly designed insulin syringes with minimal dead space (Rosen-
bloom 1974).
Besides U-40, U-80 and U-lOO, other strengths like U-500 are available on
special request and preparations containing up to 5,000 IU/ml have been de-
scribed (GalIoway et al. 1981 b).

Table 10. Insulin concentration expressed as

mg of dry insulin (containing approx. 5%
water) per mi and mmol/liter in preparations
of different strengths

Insulin strength Insulin concentration

IV/mi
mg/ml mmol/I

40 1.5 0.24
80 3.0 0.47
100 3.8 0.59
48 Insulin Preparations

The influence of strength on the rate of absorption of the insulin has been ex-
amined by Binder (1969) and Galloway et al. (1975, 1981 b) who concordantly
found that injection of high strength preparations leads to relatively lower ab-
sorption rates. Ifthe variation in strength is small, like the change between U-40,
U-80 and U-IOO, the changes in absorption rates will be insignificant (Galloway
et al. 1975, Lauritzen et al. 1984). Lauritzen et al. suggested that the decreased
insulin absorption rate observed with increasing concentration and increasing
volume of the injected insulin solution (Binder 1969) is balanced when identical
doses ofU-40 and of the more concentrated, but less voluminous, U -100 are given.
Corresponding values for insulin concentrations expressed in different units
are shown in Table 10.

VIII. Mixing Insulin

1. Insutin Mixtures
Many type 1 diabetics require a stronger initial effect in addition to the delayed
action obtained with the intermediate-acting insulins. When mixing rapid-acting
with intermediate- or long-acting preparations mutual compatibility must be
taken into account (Table 11). Today, when neutral regular insulin is generally
available, addition of the original acid regular insulin to the insulin zinc suspen-
sions should be avoided as it results in pH shift, which may affect the time action
of the insulin.
Phosphate-containing preparations (e.g. Velosulin and Hypurin) must not be
mixed with preparations of the Lente-type, as the phosphate binds and precipi-
tates the zinc ions whereby the insulin particles dissolve and the retarded action
is lost (Langkjrer, unpublished).

Table 11. Miscibility of insulin preparations

Rapid-acting insulin Intermediate- and

long-acting insulin

Physically unstable mixtures, Acid regular Lente series a

mixing is not recommended Rapitard, NPH, PZI
Neutral regular PZI
Neutral regular Lente series a
(phosphate buffer) Rapitard

Physically unstable mixtures, Neutral regular Lente series a

mixing possible (no buffer) NPH
Neutral regular NPH
(phosphate buffer)

Sta bie mixtures that can be stored Actrapid (BP) Rapitard

for at least 1 month Acid regular Globin, soluble surfen
insulin

a lnduding Lente, Lentard, Monotard, Ultralente and Ultratard

Insulin Mixtures 49

Regular insulin when mixed with Protamine Zinc Insulin, with agreat excess
of free protamine in the latter, promptly precipitates the soluble insulin. The de-
gree ofprecipitation depends on the physical state ofthe PZI; the amorphous type
binds a gre ater part of admixed regular insulin than the crystalline type (Krayen-
biihl and Poulsen 1959).
Physically stable mixtures can be obtained by mixing Semilente, Lente and/or
Ultralente in any ratio. Mixtures of Actrapid (BP-formulation) and Rapitard can
be stored refrigerated for one month without affecting the individual timing char-
acteristics.
The most commonly used insulin mixtures, regular/NPH and regular/Lente
are not physically stable, as some of the soluble insulin binds to the protamine
insulin crystals or precipitates with zinc, respectively. This is demonstrated in
Tables 12 and 13. It appears that in both combinations some ofthe regular insulin
precipitates immediately after mixing and the proportion of insulin remaining in
solution decreases with time, with decreasing ratio between regular and modified
insulin and, regarding regular/NPH combinations, with increasing strength ofthe
preparations. Corresponding mixing experiments carried out with Actrapid of
BP-formulation (methylparaben and sodium chloride) showed a higher degree of

Table 12. Percentage of added regular insulin remaining in solution after

mixing Actrapid MC (USP formulation) with Isophane MC (NPH).
The figures represent mean±SD (n=3). Insulin in solution was
determined after centrifuging the mixtures. (Lykkeberg, unpublished)

Actrapid : NPH IU/ml Time after mixing

ratio
5min I day 30days

I:1 40 78±1% 73±2% 66±2%

3:7 40 49±3% 47±3% 44±1%
100 38±5% 34±2% 29±3%

Table 13. Percentage of added regular insulin

remaining in solution after mixing Actrapid
MC (USP formulation) U-IOO with a Lente-
type preparation, Monotard MC U-IOO. The
figures represent mean ± SD (n = 3). The mix-
tures were filtered and the insulin concentra-
tion of the filtrate determined using Bio-
Rad protein assay (Bio-Rad Laboratories).
(Langkjler, unpublished)

Actrapid: Time after mixing

Monotard ratio
2min 10min

1: 1 93±4% 85±4%
1: 2 71±2% 61±2%
1: 3 47±3% 36±2%
50 Insulin Preparations

precipitation in the regularjLente combinations. Comparable results were re-

ported by Nolte et al. (1983), who also found variation between different brands
of insulin with respect to the amount of regular insulin remaining in solution after
mixing with modified insulins. For this reason change ofinsulin brand should not
be made without adequate adjustment of the mixing ratio.
The fact that some of the insulin precipitates does not mean that the quick-
acting effect is lost, because a fast absorption rate will be preserved when the in-
sulin is injected in freshly precipitated form. Thus, Berger et al. (1982) found no
difference in serum IRI values in a comparison of separate and mixed injections
of Actrapid and Monotard when injected immediately after mixing. A time-lag
of 5 min between mixing and injection caused a somewhat delayed rise of serum
insulin not observed when regular and NPH were mixed. Galloway et al. (1982)
investigated the physical stability of NPH or Lente combinations with regular
and found that the amount of insulin in solution decreased more rapidly in reg-
ularjNPH mixtures than in the corresponding regularjLente combinations. The
in vitro data correlated with clinical findings in a study, when normal fasting sub-
jects were given the same mixtures. Based on these findings it was suggested that
patients who use mixtures of regular with NPH or regular with Lente should be
advised to administer their doses immediately after filling the syringe. In contrast
to these findings Heine et al. (1984) reported that the onset of action is delayed
when regular insulin is combined in the syringe with Lente insulin even when
administered immediately after mixing. Mixing regular insulin with human
Ultratard, Francis et al. (1985) found a blunting ofthe onset of action ofthe short-
acting insulin and suggested compensating for this by increasing the proportion of
short-acting insulin in the mixture (see also Table 13). Recent1y Colagiuri and
Villalobos (1986) showed that the magnitude of loss of quick~acting effect in
diabetics when mixing regular and Lente insulins depends on the time between
mixing and injecting the insulin (see also Table 13), and Forlani et al. (1986), using
glucose-clamp technique, found that the difference in insulin profiles in diabetics
is not accompanied by a significant difference in glucose requirement and
concluded that the slight delay in quick-acting effect when mixing Actrapid with
Monotard seems clinically irrelevant.
When mixing two insulins in the syringe it is generally recommended to with-
draw the rapid-acting insulin first. However, provided appreciable mutual con-
tamination of the preparations during the withdrawal procedure is avoided, the
mixing order is not significant. But it is important to follow the same mixing
order, as a change often means that the actual mixing ratio will be changed too,
dependent on the size of the dead-space of the syringe (cf. C.X.3 for further de-
tails).

2. Dilutions
When necessary, for instance, in the treatment of small children, insulin prepara-
tions may be diluted. The diluting medium should be of similar pH and compo-
sition as used in the preparation in question regarding preservative, isotonic
agent, buffer and free zinc ions. Dilutions of insulin preparations with proper me-
dia can be stored at 4 cC for up to a month under adequate asepsis.
Addition to Intravenous Infusion Fluids 51

3. Addition to Intravenous Infusion Fluids

Addition of insulin to intravenous infusion fluids is used in the treatment of dia-
betic ketoacidosis, for insulin delivery during surgery in the diabetic patient, and
for pregnant diabetics during labour. The infusion fluids to which insulin is most
commonly added are glucose, sodium chloride, and sodium bicarbonate solu-
tions.
In order to ensure that the insulin treatment is safe and effective a number of
factors must be considered. As infusion fluids must be particle-free, only insulin
preparations in which the insulin is in solution can be added. No precipitation or
cloudiness should be present after the insulin has been mixed with the infusion
fluid. Mixing must be thorough because an uneven distribution of insulin in the
infusion fluid may expose the patient to a potentially dangerous, possibly lethal,
dose ofinsulin (Bergman and Vellar 1982).
For safety reasons most hospital routines require that infusion fluids are used
within 12 hours - or at the most 24 hours - after the container has been punc-
tured. In this short span of time insulin is sta bIe in most i.v. fluids. Other drugs
added together with insulin may cause degradation of the insulin. Some cortico-
steroid and heparin preparations contain sulphite in amounts sufficient to destroy
significant quantities of insulin. The stability of insulin in infusion fluids of eso-
teric or unknown composition cannot be predicted.
Adsorption of insulin to infusion equipment, probably as dimeric molecules
(Browne et al. 1973), is the major source of insulin loss from infusion fluids.
Although this problem has been treated in about 30 publications since 1951 no
straightforward solution has been found and many contradictory results have
been reported. Recent publications on this subject are those of Hoffmann and
Riihl (1982), Mitrano and Newton (1982) and Hirsch et al. (1981). Insulin losses
from 20% up to 80% have been found when 10-100 units ofinsulin is mixed with
infusion fluids in 500 or 1,000 mI infusion containers.
Adsorption takes place within minutes and varies with the material, more
insulin being adsorbed to hydrophobic than to hydrophilic materials (Sefton and
Antonacci 1984). In the majority of the studies it was found that the percentage
of insulin adsorbed decreases with increasing insulin concentration and that less
insulin is lost when equipment with a smaller internal surface are a is used. It has
been reported that the addition of 0.1-1 % of albumin to an infusion fluid will
prevent the loss of insulin by adsorption (Hoffmann and Riih11982), or, at least,
lower it to less than 20% (Schildt et al. 1978), and that adsorption of insulin
can also be prevented by adding 0.5% of polygeline (Kraegen et al. 1975). The
addition of the patient's own blood to insulin infusion fluid is effective in
minimizing adsorption of insulin (Kerchner et al. 1980). Adding extra insulin to
the infusion fluid in order to counteract the adsorption loss has been suggested
(Schild et al. 1978); with this technique the error in insulin dosage will probably
not exceed 10-20%. Recent1y Furberg et al. (1986), using a polyvinyl chloride
infusion system, reported that optimal delivery of insulin in a mixture containing
at least 100 IV of insulin in 500 mI of 0.9% sodium chloride can be achieved by
filling the administration set with the insulin mixture, storing the filled system for
30 min, and flushing 100 mI of the mixture through the system before starting the
infusion.
52 Insulin Preparations

The loss of insulin by adsorption to the infusion equipment is of minor prac-

tical importance provided a standard procedure is strictly followed (Berger et al.
1981, Fort 1981), as the optimal insulin dosage is not known at the time the treat-
ment is initiated and is adjusted periodically according to bedside glucose moni-
toring.

IX. Stability

1. Physical Stability
Stored as recommended the rapid-acting insulin preparations remain as clear, col-
ourless solutions during their shelf life. The neutral solutions stand exposure to
temperatures as high as 60-70 ac for up to an hour without physical changes. On
the other hand, when the insulin is exposed to elevated temperature (above 30 DC)
and concomitant movement, insulin fibril formation may take place resulting in
precipitation of biologically inactive insulin (cf. C.xI). In acid regular insulin the
fibrillation appears as an increase in viscosity, which Storvick and Henry (1968)
observed after storage at 37 °c for 2 years. If insulin solutions are frozen the in-
sulin will precipitate but will normally dissolve again when brought to ambient
tempera ture after thawing.
The prolonged-acting preparations containing suspended insulin are reconsti-
tuted as homogenous suspensions upon shaking when stored correctly (2-8 0c).
Insulin suspensions exposed to freezing usually change in appearance and become
lumpy or granular due to aggregation of the insulin particles. The increase in par-
ticle size results in an increased sedimentation rate (Fig. 17). This has also been
shown by Graham and Pomeroy (1978), who studied the effect of freezing on par-
ticle size distribution, sedimentation rate and bioactivity of different prepara-
tions. The crystalline insulin zinc suspensions showed the smallest changes, al-
though some crystal damage was observed. No differences between frozen and
unfrozen preparations were found in bioassay and prolongation tests. Havlova

AI Semilente MC U·100 BI Protaphane @ (N PHI MC U·100

5f 0 !l:----.-~~
E
50 !~
\ \.. .
i
CI)
O \~-----"
I 'II
. .. --- .. -- ...
! I , ! ! I , !
O
I
.... -- -- -~ ~ - ... ----- .
, , , ' 11 ' I ! ! ! , , !

o 0.5 1 5 8 o 0.5 1 5 8
Hours Hours
Fig. 17 A,B. Sedimentation rate ofpreviously frozen (---) and unfrozen (--) insulin suspen-
sions. At time O the preparations are dispersed by turning upside down and left to settle. At dif-
ferent time intervals the distance between the upper level of insulin particles and the bottom is
measured. Before exposure to freezing the insulin preparations were allowed to settle for
2 days
Physical Stability 53

Fig. 18. Heat treatment of insulin suspensions. Vials of Protaphane (NPH) M C 40 1U jml (lefi)
and Lente MC 40 IUjml (right) were exposed to the following tempcraturcjtimc combinalions
in a water bath (from left to right): Untrealed referencc, 70 UCj24 h, 90 °Cj2 min, 85 "Cj30 min

et al. (1969), however, tested the effect offreezing on crystalline insulin zinc sus-
pension and found that the frozen preparation was absorbed more rapidly than
the unfrozen preparation. Therefore, it is not advisable to use insulin prepara-
tions which have been frozen, because it may be difficult to withdraw reproduc-
ible doses from lumpy and granular suspensions, and the timing may have
changed as well.
Insulin suspensions stored at temperatures above 25 °C for extended periods
of time may become difficult to homogenize. On exposure to temperatures above
50 °C the insulin tends to coagulate and form large lumps. Figure 18 shows the
appearance of fresh and heat-treated Lente and NPH. The biphasic preparations
containing soluble and crystalline insulin are also heat sensitive, as some of the
crystalline material may dissolve upon heating and precipitate after cooling, pos-
sibly resulting in lump formation and deposits on the wall of the vial.
The physical appearance of insulin suspensions may also change if a drainage
of liquid from sedimented insulin is allowed to take place. This may occur when
vials in whi€h the insulin particles have settled during a long period of storage are
quickly turned upside down or revolved, leaving a part of the insulin isolated
from the suspension. If these vials remain untouched for several weeks at room
or higher temperatures the remaining liquid will gradually be drained from the
deposited insulin resulting in lumps or flakes that are difficult to disintegrate
upon shaking.
Slight changes in the colour of different preparations have been observed after
long storage at 25 °C (Storvick and Henry 1968). In preparations containing
phenol or cresol a yellow discoloration will occur after heat exposure possibly due
to oxidation of the phenols.
A slight decrease of pH during storage of insulin preparations containing
methylparaben is the result of hydrolysis of a small proportion of this substance
into p-hydroxybenzoic acid (cf.C.V.2 and Table 9). Change of pH of Monotard
during storage at different temperatures is illustrated in Fig. 19.
54 Insulin Preparations

Change in pH

o Monotard U40

-0.2
4°C

-0.4

-0.6
Fig. 19. Change in pH of Monotard
-0.8 MC 40 IU/ml during storage at different
temperatures. The curves are based on results
L-_...L.._.......I_ _.J-_-'-_---iL.--+. from a study including 8 batches
o 2 3 4 5 Years

2. ChemiCal Stability
Like other proteins, insu1in is not a stab1e entity but is 1iab1e to degradation by
chemi cal reactions with mo1ecu1es or ions in its vicinity, or to intra- and intermo-
1ecu1ar transformations within the insu1in mo1ecu1es.
Whereas the stabi1ity of the pharmaceutica1 insu1in preparations with respect
to potency has been extensive1y studied, very 1itt1e has appeared in the litera ture
about the underlying chemi cal reactions 1eading to the 10ss in bio10gica1 po-
tency.

% deamidation

Acid A

60

40

Acid B
20
Neutral

2 3 4
•
5 Years

Fig. 20. Formation of deamidation products during storage at 4 DC of insul in solutions of differ-
ent pH and composition. Acid A 0.1 % methylparaben, 5% glucose, pH 3. Acid B 0.2% phenol,
1.6% glycerol, pH 3 Neutral 0.1 % methylparaben, 0.7% sodium chloride, 0.1 M sodium acetate,
pH 7. The desamido insulin content formed in Neutral and Acid B formulations was determined
by basic disc electrophoresis followed by densitometric scanning of the amidoschwarz stained
gels. Each point is the mean of analyses of 2-5 batches (Acid B± SD). (Brange and Langkjrer,
unpublished). The deamidation products formed in Acid A were determined by paper electro-
phoresis followed by staining, elution and spectrophotometric measurement of each component
as described by Sundby (1962). Each point represents a single determination. (Pingel, unpub-
lished)
 Chemical Stability 55

Hydrolytic transformation of insulin in acid solution, with the liberation of

ammonium ions and formation of deamidation products, has been reported by
Sundby (1962), who demonstrated that monodesamido-(A21)-insulin is the pre-
vailing derivative formed.
Jackson et al. (1972) compared acid and neutral solutions and showed that
similar, but less pronounced deamidation can be observed in neutral regular in-
sulin. Slow deamidation of insulin in neutral solution as well as in two different
neutral suspensions was reported by Schlichtkrull et al. (1975 a). A comparison
of the deamidation during storage at 4 ce of two different formulations of acid
solutions and a neutral solution is shown in Fig. 20.
At higher storage temperatures significant deamidation can also be observed
in neutral solution (Fig.21), in which the hydrolytic transformation has been
shown to take place in the B-chain of insulin mainly at AsnB3 (Brange et al. 1983).
The rate of deamidation in neutral insulin preparations varies with the formula-
tion as shown in Fig. 22.
Formation of small amounts of covalent dimerization and polymerization
products of insulin during storage of neutral preparations was reported by
Schlichtkrull et al. (1975 a). In neutral solutions this transformation varies with
the formulation and the brand ofinsulin, as reported by Brange et al. (1982), the
BP-formulation with sodium chloride and methylparaben being more stable than
the USP-formulation with glycerol and phenol (or cresol). The reverse applies to

% deamidation % deamidation

2S0C
30 20

20
Monolard
1S0C 10

10
Rapllard

4°C

O
O
~ ~
o 1 2 Years O 6 12 Months
Fig. 21 Fig. 22
Fig. 21. Deamidation of insulin during storage of Actrapid MC (BP-formulation) at different
temperatures. Each point represents the mean of analyses of 4-6 different batches. The desamido
insulin content was determined by basic disc electrophoresis followed by densitometric scanning
of the amidoschwarz stained gels. (Brange et al. 1983)
Fig.22. Formation of deamidation products during storage at 25 °C of different MC insulin
preparations. Each point represents the mean of analyses of 2-11 different batches. The des-
amido insulin content was determined by basic disc electrophoresis followed by densitometric
scanning of the amidoschwarz stained gels. (Brange et al. 1985)
56 Insulin Preparations

Table 14. Influence of formulation on the chemical transformation of insulin in neutral solution
during storage at 25 °e for 6 months (Brange et al. 1982, 1983). For analytical methods cf.legends
to Fig. 21 and Table 15

Formulation % Deamidation % Dimerization

Mean SD Mean SD

Neutral regular USP 6.7 2.6 0.68 0.11

(phenol or m-cresol, glycerol) (n=5) (n=4)
Neutral regular BP 12.4 0.9 0.34 0.13
(methylparaben, sodium chloride) (n=4) (n= 11)

the rate of deamidation (Table 14). The chemical stability offive different neutral
formulations with regard to the formation of higher molecular weight products
is shown in Table 15, from which is seen that variations in the formation ofhigh
molecular weight transformation products can also be observed within different
formulations of protracted preparations.
The dimerization is mainly due to a reaction between an N-terminal amino
group in one insulin molecule with a carboxamide group of a glutamine or an as-
paragine residue in the A-chain of another insulin molecule (Brange, unpub-
lished). It is interesting that this dimer formation is of the same order of magni-
tude in the neutral solution (Actrapid) as in the insulin zinc suspensions (Mono-
tard and Ultralente) (cf. Table 15). Apparently the intermolecular transformation
proceeds within the hexamer, which is the common unit in the solution and in the
crystals. This is supported by the fact that the rate of dimerization in neutral so-
lution is independent of the concentration of insulin (Brange et al. 1983). Cova-
lent insulin dimers, probably formed during production of crude insulin, and con-
tained in the b-component (cf. A.IV, B.I.2), have been isolated from bovine b-
component and extensively characterized by Helbig (1976).
The rate of transformation in protamine-containing preparations is higher,
the difference being most pronounced at lower temperatures (Table 15). This is

Table 15. Formation (in percent) of high molecular weight

transformation products of insulin in different preparations
during 2 years of storage (Brange et al. 1984). The results are the
mean of ana1yses by gel filtration on Biogel P 30 in 1 M acetic
acid on 3-12 batches of each type of preparation

Type of preparation Species of Temperature of storage

insulin
4°e 15°e 25°e

Actrapid (BP) Pork 0.11 0.46 1.6

Monotard Pork 0.16 0.41 2.3
Isophane (NPH) Pork 1.10 3.7 8.8
Ultralente Beef 0.15 0.50 2.6
Zinc Protamine Pork 1.6 5.5 15.5
Chemical Stability 57

E276
E276

;~ / \ NPH

~~
/ \ Monolard

l
04~
L
O.4~

~ ~
0.3 0.3
0.2 0.2
0.1 0.1
O O
~ ~
Elution volume Elution volume

Fig. 23. Gel filtration ofporcine Isophane (NPH) MC and Monotard on Biogel P 30 in 1 M acetic
acid showing the appearance of an extra transformation product in NPH after storage of the
preparations at 25 °C for 6 and 12 months, respectively (Brange et al. 1984)

partly explained by an additional formation of a chemicallink between protamine

and insulin. This covalent compound is easily demonstrated in isophane insulin
stored for some time as a peak eluting before the covalent insulin dimer in ana-
lytical gel filtration (Fig. 23).
Formation of di-and polymerization products increases dramatically with in-
creasing tempera ture, as shown for Monotard in Fig. 24.
Covalent dimerization products of insulin have been shown to circulate in
diabetic patients treated with insulin (Robbins et al. 1985). These products have
been demonstrated to originate from the insulin preparation by Maislos et al.
(1986), who found that they accumulate in serum and constitute 28 % of the mean
fasting plasma immunoreactive insulin, probably due to slower metabolic
c1earance relative to insulin.

% di- and polymer formed

Fig. 24. Formation of covalent di- and polymeriza-

tion products of insulin during storage of Mono-
tard for 1 month at different temperatures. Each
O point is calculated on the basis of a study including
four different batches. (Brange, unpublished)
4 15 25 37 45
Temperature (CO)
 58 Insulin Preparations

3. Biological Stability
The biological potency of insulin is reduced during storage according to a first-
order reaction

(1)

where k follows the Arrhenius equation

k=e(a-p/T) (2)

Po is the original potency, P the potency after t months of storage at temperature

T CK), and rx and pare characteristic constants for the various insulin prepara-
tions. These constants for insulin preparations manufactured from recrystrallized
porcine and bovine insulins have been determined from accumulated stability

Biological stability
Biological stability Lente/Monotard MC/Monotard HM
Actrapid/Actrapid MC/Actrapid HM >-

f ::1 !t~·~··
o

tt·*
c:

11°f
Il 1
al
Ci

IIi
c. 100 ...... ........
"t:I
····4°C·::········· .. ···.::.. .. ·· .. ········: al 4°C
Oi 90
Ui
Oi 90[ 110
Ui 'O
'O 110 "Eal
"E ...o 100
~ 100 al
...al C.
>-
C. o 90
>- c:
o 90 al
c: Ci 80
al o.
Ci 80
o.

o 2 3 4 o 2 3 4
Fig. 25 Years Fig. 26 Years

Fig. 25. Biological potency expressed in percent of stated potency after storage for different pe-
riods at 4 De or 25 De ofInsulin Actrapid prepared from recrystallized porcine insulin (solid lines
Actrapid), monocomponent porcine insulin (dotted lines Actrapid MC) or monocomponent hu-
man insulin (broken lines Actrapid HM). The solid lines are estimated from accumulated stability
data obtained by the mouse convulsion assay (Pingel and V01und 1972), whereas points con-
nected with dotted and brokenlines are the mean biological potencies with 95 percent confidence
limits obtained by the mouse convulsion assay on 3-12 (mean 8) batches ofthe monocomponent
insulin preparations. (Pingel, S0rensen and S0rensen, unpublished)
Fig.26. Biological potency expressed in percent of stated potency after storage for different pe-
riods at 4 De or 25 De of Insulin LentejMonotard prepared from either recrystallized bovine and
porcine insulin (solid lines Lente), monocomponent porcine insulin (dotted lines Monotard MC)
or monocomponent human insulin (broken lines Actrapid HM). The solid lines are estimated
from accumulated stability data obtained by the mouse convulsion assay (Pingel and V01und
1972), whereas points connected with dotted and broken lines are the mean biological potencies
with 95 percent confidence limits obtained by the mouse convulsion assay on 3-11 (mean 7)
batches of the monocomponent insulin preparations. (Pinge1, S0rensen and S0rensen, unpub-
lished)
Biologica! Stability 59

Table 16. Time of storage of insulin preparations at various

temperatures until biological potency is reduced by 2 percent/
5 percent, respectively

Insulin Storage temperature

preparation
4°C 15°C 25°C 40°C

years years months weeks

Actrapid 36/92 5/13 12/31 5/14
Semilente 45/115 4/11 7/18 2/5
Lente 36/91 3/9 5/14 1/4
Rapitard 22/55 3/8 7/17 3/7
Ultralente 19/48 2/5 4/10 1/3

data by Pingel and V0lund (1972). For each type ofinsulin preparation equation
(1) and (2) were fitted to data of biological potency (mouse convulsion assay)
which were determined after storage ofthe insulin at different temperatures (4 ac,
15 ac, 25 ac, 37 ac, and 45 aC) for different periods ranging from a few months
to several years.
The biological stability of insulin preparations manufactured from monocom-
ponent (MC) bovine, porcine or human insulins is not essentially different from
that of insulin preparations manufactured from recrystallized porcine and bovine
insulins, as illustrated in Figs. 25 and 26. Hence the equations for the biological
stability of recrystallized insulin preparations can be used for MC insulin prepa-
rations as well. Table 16 shows how long insulin preparations can be stored at
various temperatures before they Iose 2 or 5 percent of their biological potency.
The figures are calculated from equations (1) and (2) using the estimated values
for IX and f3 (Pingel and V0lund 1972). These figures show that insulin prepara-
tions have a remarkably high biological stability especiaHy when they are stored
at a low temperature. For instance, aH the insulin preparations williose less than
or about 2 percent in biological potency after storage for 19 years at 4 ac, 2 years
at 15 ac, 4 months at 25 ac or one week at 40 ac.
The biological potencies of insulin degradation and transformation products
formed upon storage of different insulin preparations at elevated temperatures
are shown in Table 17. The products formed by hydro1ysis of insulin possess full
or nearly full biological potency, independent ofthe species ofinsu1in and site of
deamidation. Similar resu1ts were reported by Carpenter (1966) and Chance
(1972) for bovine and porcine monodesamido-(A21)-insulin, respectively. In con-
trast, Easter et al. (1978) reported only a relative potency of about 60% for bovine
monodesamido-(A21)-insulin. The full potency of the derivative formed in neu-
traI solution exp1ains why Actrapid in spite of significant chemical degradation
possesses eminent bio10gica1 stability even at higher storage temperatures
(Table 16). The di-and polymerization products as well as the protamine-insulin
product exhibit very low potency. However, the impact on the biological stability
of the preparations on1y becomes significant after storage at very high tempera-
tures (Fig. 24).
60 Insulin Preparations

Table 17. Biological potency oftransformation products formed in different insulin preparations.
(Brange, S0rensen and S0rensen, unpublished)

Isolated Species Bioassay Potency· Potency

from of origin method IU/mg relative
(9S% conf. Iim.) to insulin

Monodesamido- Acid Pork MBGa+MCA b 24.4 92%

(A-21)-insulin solution (23.S-2S.4)
Beef MBG+MCA 22.S 8S%
(21.4--23.4)
Monodesamido- Actrapid Pork MCA 2S.9 97%
(B-3)-insulin (22.6-30.0)
Covalent insulin Ultralente Pork MBG 4.4 d IS%
dimer (porcine) (3.S-S.4)
Covalent insulin Semilente Pork MBG 3.8 d 13%
dimer (2.9--4.7)
Covalent protamine Isophane Beef MBG 1.2 d 4%
insulin complex (NPH) (0.9-1.S)
Insulin polymeri- Semilente Pork MBG 0.4" <2%
zation product (0.2-0.6)
(MW 30.000)

a Mouse Blood Glucose Assay (modified BP method C)

b Mouse Convulsion Assay (BP method)
• Based on dry insulin-component containing 14.S% nitrogen
d Slight protracted effect
e Strong protracted effect, wherefore a PZI preparation was used as standard

When a vial of insulin is exposed to sunlight a hundredfold increased loss of

biological activity has been observed, as compared to storage in the dark at the
same ambient temperature for the same length of time (Sundby, unpublished).

4. Immunogenicity Studies
Monodesamido-(A2l)-insulin, dimerization- and polymerization products have
been tested in rabbit immunization experiments and found not to be significantly
more immunogenic than the parent insulin (Schlichtkrull et al. 1975 a). The low
immunogenicity of monodesamido-(A21)-insulin has later been confirmed by
Kasama et al. (1981).
Deamidation products isolated from Actrapid or Semilente and the covalent
protamine insulin compound isolated from isophane (NPH)-insulin after acceler-
ated storage have also been found to be of low immunogenicity in rabbits
(Figs.27 and 28).

x. Storage and Use of Insulin

1. Storage
To ensure optimal preservation ofthe quality (characteristics) during shelf-life in-
sulin preparations should be stored in the dark between 2 and 8 ce, freezing
Storage 61

Average bind ing

of 1251-insulin

60

40

•
Fig.27. Groups of 10 rabbits immunized twice weekly
with neutral solutions ofporcine MC insulin and porcine
20 • monodesamido insulin isolated from Actrapid MC stored
•• at 37 ac for 3 months, 20 IU with Freund's incomplete

-
adjuvant. Each dot represents the average binding of
••
"•
125I-insulin (cf.legend to Fig. 28) during the first 70 days
o of immunization of one rabbit. (Brange et al. 1983)
MC·insulin Mono·
desamido·
insulin

% bound 1251-insulin

80

60

40

20

prolamine-insulin produci
o

o 20 40 60 80 100 120 140 Days

Fig.28. Groups of 10 rabbits immunized twice weekly with a mixture (1: 1) of acid solutions of
porcine a- and b-component (cf. B.I.2) and high molecular transformation products, inc1uding
50% of covalent protamine-insulin product isolated by gel filtration from Isophane MC (NPH)
after storage ofthe preparation for 3 months at 37 ac, 40 Jlg with Freund's incomplete adjuvant.
The ordinate represents the antibody level as determined by the percent bound 125I-beefinsulin
in serum diluted 1:3 (Schlichtkrull et al. 1972), the abscissa represents the time ofsampling since
the start ofimmunization. Each point represents the mean ±SEM, (Brange et al. 1984)
62 Insulin Preparations

should be avoided. When necessary insulin can be stored for several weeks at tem-
peratures not exceeding 25 ac without any significant change in biological activ-
ity. A vial in use can be stored at an ambient temperature not exceeding 25 ac
to make the injection more comfortable.
Storage at higher temperatures should be avoided as it may alter the physical
properties ofthe insulin preparations (cf. C.IX.1). Thus, precipitation may occur
in regular insulin and the insulin suspensions may become granular in appearance
or c1umped in large partic1es resulting in loss of activity, changed timing charac-
teristics or, at least, in difficulties in withdrawing the proper doses.
It is important to store insulin in the dark as exposure to direct sunlight and
even diffuse daylight results in decreased biological activity (cf. C.IX.3).
Insulin preparations should not be allowed to freeze, as freezing induces par-
tic1e aggregation, which makes it difficult to withdraw the correct dosage (cf.
C.IX.l).
Sedimented insulin suspensions that have been stored for more than a few
weeks should be homogenized prior to restorage, as their resuspendability may
otherwise be affected (cf. C.IX.l).

2. Withdrawal of Insulin from the Vial

Before the withdrawal of insulin the vi al should be inspected to ensure that the
preparation has not changed in appearance. No discoloration must be observed,
nor in solutions any precipitation found. The insulin suspensions are usually ho-
mogenized by gentle shaking, avoiding froth or bubbles which may make it dif-
ficult to measure the correct dose as insulin partic1es accumulate in the froth. If,
however, the suspension is not completely dispersed it is necessary to shake the
vial vigorously until complete homogenization. Then the vial should be allowed
to stand until the froth has disappeared. Just before withdrawal the insulin should
be homogenized again by turning the vial upside down several times. If the sus-
pension has become granular or still contains lumps of insulin it should not be
used.
The sterilized and separately packed disposable syringes and needles are most
convenient to use. If reusable glass syringes are to be used, the proper sterilization
procedure must be followed. The syringe should preferably be boiled in water for
5-15 min before each injection, alternatively, the syringe may be stored in 70%
ethyl a1cohol or 91 % isopropyl a1cohol (Purintun and McGrane 1972). In order
to avoid dilution or contamination of the insulin before withdrawal, water or al-
cohol must be completely removed from the syringe by pushing the plunger up
and down several times. For disinfection of the rubber c10sure ethyl a1cohol (in-
dustrial methylated spirit) or isopropyl a1cohol may be used. Surgical or medi-
cated spirit which contains skin irritating additives must not be used; other dis-
infectants, e.g. solutions of benzalkonium chloride, cetrimonium bromide, chlor-
hexidine, etc. are not advisable, as even minimal contamination with these, in-
duced by the needle perforating the rubber surface, can cause changes ofthe prep-
aration, e.g. by precipitating soluble insulin.
Before withdrawal of the desired do se from the vial a corresponding volume
of air is injected into the air space of the vial, taking care to avoid bubble forma-
tion.
Preloading of Syringes 63

After use glass syringes should be rinsed to remove traces of insulin which,
when the syringe is boiled or immersed in alcohol, may coagulate and block the
needle during the next injection. Clogging ofthe needle has been observed in rare
cases, especially with U-IOO suspensions. Usually this c10gging can be prevented
by correcting the sterilization procedure or by using disposable syringes. In order
to prevent plugging during injection, the injection should be completed within
5 seconds after the skin has been penetrated by the needle (Galloway et al.
1976).

3. Mixing Techniques
When mixing two insulins in the syringe mutual contamination of the two vials
should be avoided (ef. C.VIII.1). The size of the dead space in the conventional
syringes varies from 0.05 to 0.1 mI and, as the first drawn insulin occupies the
dead space volume, a change in the mixing order, model of syringe or insulin
strength will also change the ratio between the two components. This is shown
in Table 18. Amott et al. (1982) compared mixing of Actrapid and Monotard in
syringes with significant dead space and minimal dead space. They conc1uded that
the latter type should be recommended for patients who inject mixed insulins, be-
cause the minimal dead space syringes are more accurate and the order of mixing
is less relevant.
Mixtures of regular and retarded insulins normally have to be injected im-
mediately after they have been drawn into the syringe in order to preserve their
individual timing characteristics. Regarding compatibility between the different
insulin preparations cf. C.VIII.l.

4. Preloading of Syringes
Insulin partic1es settle quite rapidly on standing. Therefore it is recommended to
inject suspensions immediately after they have been drawn into the syringe. Some
patients, however, are not able to withdraw the insulin by themselves and may
therefore need to have their syringe preloaded. The preferred type of syringe for
this purpose is the disposable syringe with minimal dead space, as this has been

Table 18. Effect of size of dead space, insulin strength and mixing
order on actual doses delivered after mixing in the syringe.
Intended dose: 10 U regular insulin with 20 U protracted insulin.
The figures represent the actual doses in international units
calculated under the assumption that the preparations are
completely mixed in the syringe

Insulin drawn 0.06 mI dead space 0.10 mI dead space

up first
U-40 U-IQO U-40 U-I00

Regular (R) 11.5 R 13.3 R 12.4 R 15.0 R

18.5 P 16.7 P 17.6 P 15.0 P
Protracted (P) 9.3 R 8.3 R 8.8 R 7.5 R
20.7 P 21.7 P 21.2 P 22.5 P
64 Insulin Preparations

A) horlzonlal slorage Bl vertical slorage

no homogenizalion no homogenizaUon
100 100

50 50

o 10 20
o
10 20 40
CII
CI 40
<O
rn
o Dl vertical
".5 e) horizonlal slorage slorage

"3 rolling for 155 rolling for 15s

.~ 100 100

50 50

o o
10 20 40 10 20 40
Fig. 29 A-D. Preloading three different doses (10, 20 and 40 IU) of 12sl-labelled Insulin Zinc Sus-
pension (mixed) 40 IU Imi (Lente MC and Monotard MC) in disposable syringes (dead space
0.06 mi). After storing the syringes in horizontal (A, C) or vertical (B, D) position for 24 hours
the insulin was expelled either without any previous homogenization (A, B) or immediately after
rolling the syringe for 15 seconds between the palms of the hands (C, D). The percentage of the
intended dosages actually delivered was determined by relating the radioactivity ofthe expelled
volumes to the activity of corresponding accurately measured homogeneous suspensions.
(Langkjrer and Brange 1982)

shown to ensure correct dosage after having been preloaded for up to 3 days with-
out taking special precautions (Langkj<er and Brange 1982). If disposable
syringes with significant dead space have to be used, a dose loss may occur due
to sedimented insulin being retained in the dead space. As shown in Fig. 29, the
relative loss varies with do se size and storage position, but the loss can be pre-
vented by storing the syringe in a vertical position with the needle upwards and
rolling the syringe between the palms of the hands for about 15 seconds to dis-
perse the insulin particles prior to injection. The syringe should never be stored
with the needle downwards allowing the particles to settle on top of the needle
as this may cause plugging of the needle during injection.
From a sterility point of view preloading of syringes for a short period is no
problem, provided adequate aseptic precautions are followed. Oberly et al.
(1979), who tested the sterility of syringes preloaded with NPH, did not find any
significant bacterial growth even after storage for up to 12 weeks.
Use of Present Insulin Solutions in Infusion Pumps 65

XI. Insulin for Delivery Systems

1. Introduction
The subcutaneous injection of insulin does not provide optimal metabolic con-
trol, as this therapy is not able to mimic the delicate minute by minute modulation
of insulin secretion which in normal persons occurs in relation to meals, exercise,
etc. The prospects of decreasing the risk of late complications have, however,
stimulated attempts to improve the therapy. Although good metabolic control
with near normal glycemia has been demonstrated by using multiple (4-8) daily
injections, the inconvenience of such a large number of injections precludes a
widespread use of this regime.
In the past several years steadily increasing efforts have been directed towards
developing sophisticated insulin delivery devices able to supply insulin con-
tinuously throughout 24 hours with alternating basal delivery and increments in
relation to meals.
Development of such devices has taken two paths. Closed-loop systems regu-
late insulin delivery on the basis of continuously measured glucose concentration,
but the lack of a stable portable glucose sensor has limited their use to hospital-
ized diabetics (Albisser et al. 1974, Pfeiffer et al. 1974). Open-loop systems (with-
out glucose sensor) infuse insulin according to a preprogrammed schedule (Slama
et al. 1974, Pickup et al. 1980).
Today only externally worn infusion systems (insulin pumps) are in wide-
spread use, but prototypes of implantable insulin pumps have already been used
in diabetics (Buchwald et al. 1981, Irsigler et al. 1981, Schade et al. 1982a).
Insulin pumps are regarded as experimental instruments to be used in con-
trolled research rather than a current therapeutic option (Marliss 1982, Lauritzen
and Pramming 1984), and their use requires careful patient selection and available
care by skilled professionals (American Diabetes Association 1985).
Recently a new portable insulin delivery system, the NovoPen, solving some
of the problems of multiple daily injections, has become available. Having the size
and appearance of a fountain pen it delivers a metered do se from a disposable
cartridge by pressing a button without the need for handling vials, syringes etc.
and it allows patients and, especially, diabetic children to inject themselves more
conveniently, rapidly and with better dosage accuracy. Using the NovoPen,
Jefferson et al. (1985) assessed a multiple injection regimen in diabetic adolescents
and found that after a three month study period 8 out of 10 patients wished to
continue with the regimen, finding that the advantages it offered outweighed the
inconvenience of multiple injections. In another study 27 of 31 patients preferred
NovoPen to a conventional twice daily regimen even though it necessitated four
injections (Walters et al. 1985).

2. Use of Present Insutin Solutions in Infusion Pumps

Although neutral solutions ofinsulin can be stored for several years at ro om tem-
perature without significant physical change (cf. C.lX) the use of insulin pumps
has been complicated by the precipitation ofinsulin (Lougheed et al. 1980, Buch-
wald et al. 1980), especially when the pump design is based upon a peristaltic sys-
66 Insulin Preparations

tem (Irsigler and Kritz 1979, Jackman et al. 1980, Prestele et al. 1980, James et
al. 1981, Schade et al. 1981).
Such precipitates, which may result in obstruction of the flow-system and
especially of the catheter, can be due to the formation of amorphous particles of
insulin (isoelectric precipitation), insulin crystals or insulin fibrils. The incon-
sistent terminology used when referring to these precipitates may have been a
deterrent to solving many of the problems encountered when using insulin in
pumps (Ege 1986).
The amorphous or crystalline precipitation is usually caused by zinc or vari-
ous other divalent metal ions leaching from materials or by a 10wering of the pH
due to carbon dioxide diffusion or to leaching of acid substances from reservoir
or tubing materials. (Melberg et al. 1987). This kind of precipitation can be
prevented by selection of suitable materials for construction of the pump.

Fig.30. Insulin fibrils x 100,000 formed at a 37 ac and b 80 ue. (Transmission eleetron mieros-
eopy)
Physical Stabilization of Insulin Solutions 67

The tendency of insulin to form insoluble fibrils under a variety of conditions

by a non-covalent polymerization is more difficult to avoid. The promotion ofin-
sulin fibril formation by heat was first described by du Vigneaud et al. (1928) and
Krogh and Hemmingsen (1928), and later extensively studied by Waugh (1946,
1948, 1954, 1957, Waugh et al. 1953). The interactions between molecules in
insulin fibrils are main1y hydrophobic (Waugh et al. 1953) and seem to be of the
paralle\ {J-sheet type (Burke and Rougvie 1972). The formation of fibrils requires a
monomerization ofthe insulin and a change in monomer conformation (Brange et
al. 1987). Different forms of insulin fibrils are illustrated in Fig. 30. Fibril
formation is also acce\erated if insulin is exposed to hydrophobic surfaces
(interfaces), which explains why different materia1s used in pumps inf1uence the
physical stability very differently (Brange et al. 1982, Lougheed et al. 1983,
Feingold et al. 1984).
Schade et al. (1982 b) examined insulin fi brils by electron microscopy and con-
c\uded that the morpho10gic appearance of the insu1in precipitated in vivo in the
pump reservoir was not entirely reproduced by in vitro methods of insulin precip-
itation.
The requirements to the purity of insulin for continuous subcutaneous infu-
sion can be expected to be very high if an immunological response, e.g., antibody
formation or allergy, is to be avoided. The stimu1ation of the immune system is
probably more intensive with an indwelling catheter supplying the insulin to a de-
pot (pool) at the same infusion site for several days or weeks. Thus, Lauritzen et
al. (1982) reported a surprising1y high antibody formation during subcutaneous
infusion, and Pietri and Raskin (1981) found that allergic skin reactions, which
occurred at the site of infusion with purified porcine insulin (Lilly) ceased on
transfer to monocomponent insulin (N ovo). Similar experience was reported by
Levandoski et al. (1982), who obtained a dramatic reduction in nodule formation
after changing from Iletin II to Actrapid. The physical stability of neutral insulin
solutions with respect to fibrillation has been found to decrease with increasing
purity of the insulin. This partly explains the difference in physica1 stability ob-
served between different brands of highly purified neutral regular insulin, which
were shown to differ considerably with respect to content of insulin dimers and
polymers (Brange et al. 1982).
The contact with various polymer materials used in infusion pump systems
impairs the chemical stability of insu1in (Grau 1985, Me\berg et al. 1987). The
deterioration of the insulin is particularly profound when the material is gamma
irradiated PVC (Me1berg et al. 1987). Substances leached from the po1ymer
material have been shown to be mainly responsible for the chemical trans-
formation of insulin (Havelund, unpublished).

3. Physical Stabilization of Insulin Solutions

Much effort is being invested in the deve\opment of stable insulin formulations,
particularly for implantable pumps. Evaluation of the physical stability at ele-
vated temperatures of insulin solutions in ampoules and vials has been performed
by rotation on a wheel (Thurow 1980) or by horizontal shaking of normal insu1in
vials placed in vertical positions (Brange and Have\und 1983 a).
68 Insulin Preparations

The requirements for introducing fibril-inhibitory additives include that the

substance must be physiologically accepta bIe and compatible not only with the
insulin and the auxiliary substances, but also with the pump materials. Further-
more it must be without negative inf1uence on the insulin quality, viz. chemical
stability, timing, immunogenicity, etc. Improvements in the physical stability of
neutral insulin solutions have been reported by addition of autologous serum (Al-
bisser et al. 1980), polypropylene glycol (Irsigler et al. 1981, Grau et al. 1982,
Thurow and Geisen 1984), glycerol (Buchwald et al. 1981, Blackshear et al.
1983), carbohydrate (Brange and Havelund 1983b), different non-ionic and ionic
surfactants containing a long aliphatic group (Lougheed et al. 1983) and
calcium ions (Brange and Havelund 1983 a). Bringer et al. (1981) prevented aggre-
gation in acid solutions by addition of dicarboxylic amino acids, which, however,
were without effect in neutral solution. Unfortunately, carbohydrates and gly-
cerol have been found to impair the chemical and biological stability ofthe insulin
(Brange et al. 1982, Brange and Havelund 1983b) and the use of synthetic
detergents for long term infusion is questionable. The damaging effect of glycerol
has been shown to be correIated with the purity of the glycerol, aldehyde
impurities being responsible for the degradation of insulin (HaveIund, un-
published). Sulphated insulin (ef. C.XIl) having a biological potency of 20% of
that of insulin has been found more resistant to fibrillation than other stabilized
insulins (Albisser et al. 1982), possibly due to a high degree of heterogeneity
(purity factor, cf. C.xI.2).
Calcium ions, which probably play an important role for the storage ofinsulin
in the pancreatic B-cell as well (Howell et al. 1978), are particularly effective in
stabilizing higher concentrations of insulin (Fig. 31), and improve the physical
stability without affecting the quality of the insulin. The positive effect of calcium
ions is probably explained by binding of these ions to specific sites within the
hexamer (Sudmeier et al. 1981), thereby protecting and stabilizing the hexameric
structure.
Relative physical stability

25

20

15

o
U 40
10

5
Fig. 31. The effect of calcium ions on the rela-
tive physical stability of neutral porcine insulin
Ca2+(mM) solutions (0.2% phenol, 1.6% glycerol) of
L...-_..J......_--L._---JL...-_..J......_... different strengths. (After Brange et al. 1982)
o 2 3 4
Insulin Derivatives 69

Table 19. Influence of zinc content in Actrapid

(USP-formulation) on the relative physical
stability. (Brange et al. 1986)

Zinc/hexamer U 40 U 100 U 500

o 1 1.1 3.2
2 2.5 3 15
3 5 35
4 75 100 500

The marked fibrillation-inhibitory effect of increasing the zinc ion concentra-

tion from 2 Zn/hexamer to 4 Zn/hexamer (Brange and Havelund 1983 c, Brange et
al. 1986) is probably also due to a stabilizing effect on the hexamer (Table 19).
The more precise requirements to physical and chemical stability of insulins
for use in pumps can only be defined when experience has been gained with the
implantable devices presently under development.

XII. Insutin Derivatives

Numerous attempts have been made to modify the primary structure of insulin
in an endeavour to: 1. gain more detailed knowledge of the spatial organization
of the molecule and the role of the functional groups in stabilizing the secondary,
tertiary and quaternary structure; 2. localize the biologically active part of the
molecule in order to understand the molecular mechanisms ofbinding and action;
3. obtain analogues with a prolonged timing of action; 4. reduce the antigenicity
and/or immunogenicity in therapeutic use. The first two aspects have been re-
viewed by Blundell et al. (1972) and Pullen et al. (1976) and will not be dealt with
here.
An insulin analogue possessing a protracted effect was prepared for the first
time by Hallas-M011er (1945), who coupled insulin with phenylisocyanate and
found that the resulting phenylcarbamoyl insulin (Iso-insulin), although soluble
at physiological pH, showed a slow onset and prolongation of effect. Iso-insulin
and a mixture of Iso-insulin and regular insulin (Di-insulin) were in therapeutic
use for many years, but have been withdrawn. Later Bradbury et al. (1973)
prepared a series of reversible derivatives of insulin in an attempt to obtain a
retarded action by changing the isoelectric point and lowering the solubility at
tissue pH, but sufficiently long duration of action was not obtained.
Various modified insulins have been tried for the treatment of the rare
cases of immunological insulin resistance, the objective being to reduce the
affinity to insulin antibodies in the patient's serum. The successful use of des-
alanine(B-30) porcine insulin was reported by Boshell et al. (1964), Kumar and
Miller (1970), and Burman et al. (1973). The immunological significance of the
B-chain carboxy-terminal amina acid (Ala B-30) was reviewed by Kumar (1979),
who found that two thirds of diabetics resistant to insulin had antibodies differ-
entiating between porcine and desalanine porcine insulin, with a lower affinity for
70 Insulin Preparations

the latter. The immunological properties of insulin analogues substituted at

position B-30 by different amino acid residues have been investigated by Neu-
bauer et al. (1984), who found that the individual analogues stimulated antibody
development (immunogenicity) to the same extent as did the parent insulin,
whereas the binding capacity to pre-formed antibodies (antigenicity) was reduced
to 80-90% of that of the parent insulin.
However, desalanine (B-30) porcine insulin appears less efficacious when
compared with sulphated insulin (Davidson and DeBra 1978). Sulphated insulin
is a preparation containing porcine or bovine insulin treated with sulphuric acid
(Moloney et al. 1964), resulting in a heterogenous mixture of at least nine different
derivatives containing an average 4.5 sulphate ester groups per molecule.
The sulphation of beef insulin results in more than 50 percent loss of potency
and 400 percent increase in cost (Little et al. 1977). Zeuzem et al. (1985) found
that 22 times and 4 times the concentration of sulphated insulin was required to
achieve equivalent biologica1 effect in rat and human adipocytes, respectively.
Pongor et al. (1983) using mild reaction conditions have prepared sulphated
insulin with 2 sulphate groups per molecule and a bioactivity of about 80%.
Successful treatment of insulin resistance has been reported by Maloney (1964),
Mencze1 et al. (1966), and Little and Amott (1966).
Des-phenylalanine insulin, which is an analogue obtained by removal of
the N-terminal amino acid of the B-chain (Brandenburg 1969, Kerp et al. 1974)
has been introduced in a variety of different formulations (Sachse et al. 1982). The
immunogenicity of this derivative has been found to be comparable with native
insulin ofthe same species, whereas the binding capacity to pre-existing antibodies
seemed to be less than that ofunmodified insulin (Santeusanio et al. 1981). Thus,
Valdenaire and Klein (1979) treated 11 patients with insulin resistance and
achieved a considerable decrease of insulin requirement in more than half of the
patients. A c1inicallong-term trial (Him et al. 1982) and a comparative study
with conventional preparations (Sachse and Federlin 1981), have not revealed
significant clinical advantages of Des-Phe-preparations with respect to metabolic
control.
Recently, the physiologic effects of human proinsulin, the biosynthetic pre-
cursor ofhuman insulin (Steiner and Oyer 1967) and now available through re-
combinant DNA technology (cf. D.I), have been studied in normals and diabetic
patients (Revers et al. 1984, Bergenstal et al. 1984, Glauber et al. 1986).

XIII. Alternative Administration of Insulin

The chemical character of insulin was not known during the first years after the
discovery of the hormone and it was natural to try different ways of administra-
tion; therefore a number of parenteral as well as several non-parenteral routes,
inc1uding cutaneous, nas al, lingual, tracheal, oral, intestinal, rectal and vaginal,
were tested.
It soon became c1ear that insulin was administered most effectively by par-
enteral injection, but due to the discomfort and inconvenience of injection ther-
apy agreat many efforts were made to improve the non-parenteral administration
Alternative Administration of Insulin 71

forms by protecting the insulin molecule from enzymatic degradation and facili-
tating the absorption by addition of various substances. At least 60 reports were
published during the first 15 years ofinsulin therapy (Jensen 1938), but alI these
attempts were without or with limited or varying success.
In recent years new formulation techniques have been used in attempt to de-
velop dosage forms of insulin applicable for absorption by oral, nasal, rectal or
other mucosal routes. Recent results are reviewed in the folIowing.
Nasal absorption of insulin has been demonstrated in animals and humans.
Hirai et al. found that the absorption through the nasal mucosa in dogs (1978)
or in rats (1981) was promoted when a surfactant was present and showed that
addition of 1% polyoxyethylene-9-lauryl ether or bile salts to the insulin solution
resulted in a bioavailability of about 30% compared with intravenous adminis-
tration. Using one of the same bile salts (sodium glycocholate) Pontiroli et al.
(1982) found only about 10% bioavailability when insulin was given intranasalIy
to normal subjects and diabetic patients and Hirata et. al. (1983) observed damage
to the microvilli of the nasa1 mucous membrane upon repeated administration of
sodium glycocholate. Compared with subcutaneous administration a more rapid
increase in serum insulin concentration resulting in a prompt reduction of blood
glucose concentration was seen with bile salt promoted nasal absorption in
normal and diabetic subjects (Moses et al. 1983), but large differences in
absorption were observed correlating positively with increasing hydrophobicity
ofthe steroid nucleus ofthe bile salts (Gordon et al. 1985). Salzman et al. (1985)
assessed the efficacy of 1% polyoxyethylene-9-lauryl ether in long-term studies in
diabetics and found that the absorption profile more closely simulated that of
endogeneously secreted insulin in response to a meal than that of subcutaneously
injected insulin, but less than 10% bioavailability was obtained compared with
intravenous administration.
Absorption across respiratory mucosae on delivery of regular insulin by aero-
sol inhalation was reported in humans by Wigley et al. (1971), but only low and
variable efficiency was obtained. Administered as powdered aerosol and de-
livered directly into the trachea of rabbits the bioavailability has been shown to
be approx. 40% (Y oshida et al. 1979). Ishida et al. (1981) investigated the absorp-
tion through oral mucosa (buccal cavity) of dogs and found that the percentage
ofinsulin absorbed from a new dosage form was about 0.5% compared with in-
tramuscular injection of insulin.
A variety of different new formulations have been tested in order to increase
the effectiveness of oral administration of insulin by use of surfactants, insulin de-
rivatives, adsorption of insulin to small particles or entrapment of insulin in or-
ganic vehicles. These attempts include administration of insulin: 1. together with
polyethylene-20-oley1 ether (Brij 98) to normals and diabetic patients (GalIoway
and Root 1972); 2. adsorbed physically on p-naphthyl-azo-polystyrene particles
(Shichiri et al. 1971) or cross-linked with glutaraldehyde (Oppenheim et al. 1982)
to rabbits and mice or rats, respectively; 3. entrapped in phosphatidyl·choline li-
posomes (Dapergolas and Gregoriadis 1976, TragI et al. 1979, and Arrieta-Mo-
lero et al. 1982) or in water-in-oil-in-water emulsions (Shichiri et al. 1976) to rats
or rabbits. Only low and inconsistent efficiency (~ 1% bioavailability) has been
reported in these articles, and no dose response relationship was demonstrated.
72 Insulin Preparations

Furthermore, the stability and effectiveness of liposome-entrapped insulin was

unpredictable (Arrieta-Molero et al. 1982, Shenfield and Hill1982). U sing coating
of insulin with polymers cross-linked with an azoaromatic group, Saffran et al.
(1986) observed a certain blood sugar lowering effect in normal rats after oral
administration, but the maximal responses occurred at variable times, from 1 to
9 hours after administration of the protected insulin.
Direct intestinal administration of insulin has been tried in a number of cases.
Crane et al. (1968) showed that injected into the upper jejunum of a pancreatec-
tomized man, the fraction of insulin absorbed did not exceed about 0.5% of the
dose given in spite of the total absence of pancreatic proteolytic enzymes. Given
together with absorption-promoting substances to diabetic rats the efficiency of
intestinal absorption was 3-10% compared with parenteral administration (Shi-
chiri et al. 1975, Touitou et al. 1980).
Rectal absorption of insulin formulated as suppositories prepared by mixing
insulin with surfactants has been described by a number of groups. Shichiri et al.
(1978) found the absorption in depancreatized dogs more rapid than after intra-
muscular injection, but the comparative efficiency was only about 10%. Ichikawa
et al. (1980) tested the absorption promoting effect in rabbits of a variety of non-
iemie, anionic, cationic and amphoteric surfactants including bile acids and found
an optimal effect of suppositories with 1% polyoxyethylene-9-lauryl ether. How-
ever, the do se of rectally administered insulin needed to produce hypoglycaemia
of similar magnitude was about two or three times the intravenous dose.
Using the same optimal surfactant, Yamasaki et al. investigated rectal appli-
cation of insulin suppositories in diabetic dogs (1981 a) and in normal and dia-
betie human subjects (1981 b). They found evidence that 30% of the insulin ab-
sorbed from the rectum was distributed to the portal vein, but the total bioavail-
ability was not more than about 10%. By rect al administration of microenema
containing 5-methoxysalicylate Nishihata et al. (1983) found maximum plasma
insulin levels in dogs after 45 minutes and a bioavailability of approximately 25%
compared with i.m. administration. An increase ofbioavailability from 19 to 38%
compared with intramuscular administration was observed when the suppository
with insulin plus adjuvant was followed by another suppository with adjuvant
only (Nishihata et al. 1985). Using sodium salicylate as primary adjuvant
Liversidge et al. (1985), after rectal administration in dogs, observed a greater
decrease of blood glucose levels with increasing content of acetic acid in the
suppository. Hildebrandt et al. (1984) found in diabetic patients approximately
5% uptake of insulin from suppositories containing Brij 58 compared with sub-
cutaneous injection. Rectal administration of insulin has been reviewed by
Ritschel and Ritschel (1984) who also discussed the pharmacologieal aspects of
rectal insulin absorption.
Thus, even by use of modern application forms the nonparenteral routes of
administration of insulin have not so far resulted in reliable alternatives to the
parenteral routes.
In recent years new approaches have been suggested for parenteral insulin de-
livery. These include subcutaneous implantation in rats of small vinyl-ethylene
copolymer pellets (Creque et al. 1980, Brown et al. 1986) or biodegradable
insulin-albumin microbeads (Goosen et al. 1983), which can release insulin at
Alternative Administration of Insulin 73

constant rate for several weeks, and intravenous injection of maltose-insulin

derivatives bound to lectin in complexes which allow displacement of varying
amounts of hormone when exposed to glucose solutions of different concentra-
tions (Brownlee and Cerami 1979, 1983, Sato et al. 1984). The clinic al usefulness
of these approaches remains, however, to be seen.
D. Human Insulin

1. Manufacture

1. Sources
Human insulin has been prepared from four sources:
a) From human pancreas. The amount of human insulin that can be prepared
from cadaveric pancreas glands is totally inadequate for therapy.
b) From amino acids by peptide synthesis (Sieber et al. 1974, 1977). The 200 reac-
tion steps of a total synthesis make the product extremely costly.
c) From porcine insulin by a semisynthetic conversion into human insulin. Por-
cine insulin can be made in sufficient quantities to meet the demands for the
next decades and a quantitative conversion process for large-scale production
has been developed (cf. D.I.2).
d) From fermentation of E. coli bacteria or Saccharomyces cerevisiae yeast,
suitably encoded by DNA recombinant methods. This source is, in theory,
unlimited (cf. D.I.3).
The methods relevant for the manufacture ofhuman insulin are c) developed
by NOVO and d) developed by E. Lilly (E. coli) and later by NOVO (Saccharo-
myces cerevisiae). Sources and characteristics ofhuman insulins of different origin
have been reviewed by Ege (1985).

2. Conversion of Porcine Insutin

Porcine insulin is converted to an ester of human insulin in a transpeptidation
reaction, in which the alanine residue No. 30 ofthe B-chain is replaced by a threo-
nine ester. The reaction is catalysed by porcine trypsin and takes place in a mix-
ture of water and organic solvents in the presence of a large excess of the threo-
ni ne ester (Markussen 1980, Markussen 1982, Markussen and Schaumburg
1983). The reaction mechanism is depicted in Fig. 32. First, porcine insulin and
trypsin associate under formation ofthe Michaelis complex. Then the alanine res-
idue (AlaB30 ) is split off under formation of des(AlaB3 °)-insulinyl-trypsin ester,
(DAI - trypsin), the ester being formed from the carboxyl group oflysine (Lys B29)
and the hydroxyl group of serine (Ser 183 ) in the active site of trypsin.
Hydrolysis of this intermediate is suppressed in the reaction medium, and in-
stead the threonine ester brings about an aminolysis rendering the human insulin
ester, first as the Michaelis complex which eventually dissociates. The yield ofhu-
76 H uman Insulin

Porcine Insulin Des (AlaB 30) -Insulin (DAI)

r-----------...,I-Ala +H20~ - - - - - - - - - - - ~

I~ k3 1 1
Porcine Insulin, DAI-Trypsin ---lI-+ 1 DAI,. 1
Trypsin : k2 (acyl-enzyme) +--1 TrYPsJn 1

i::
L.. _ _ _ _ _ _ _ _ _ _ _ ...J1 1<-. 1 1
' - - - - - - - - - - - ' _ H230 l... - - - - - - - - - - - ..J

+Thr-O~l hr OR
-

.-----------.,
1 1
1 HI-OR, 1
: Trypsin :
1 ___________
L.. ....I1

Human Insulin Ester (HI-OR)

Fig. 32. Reaction mechanism of the trypsin catalyzed transpeptidation of porcine insulin into hu-
man insulin ester. Michaelis complexes are in boxes with broken lines. The uptake ofwater (k 3 )
under formation of desalanine insulin (DAI) is suppressed in the reaction medium. (Markussen
et al. 1982)

man insulin ester in the transpeptidation reaction is about 97%. Proinsulin and
related compounds present in crude insulin are likewise transpeptidized to human
insulin ester, thus making a contribution to the yield of about 3%.
The further processing to human monocomponent insulin is shown in Fig. 33
(Markussen 1982, Markussen et al. 1982). Trypsin is removed by column chroma-
tography at a low pH where it is inactive. Residual amounts of unconverted por-
cine insulin are removed by anion exchange chromatography. As porcine insulin
contains an additional negative charge relative to the human insulin ester, the
product is the first to emerge from the column. The ester group is cleaved render-
ing human insulin. A final ion exchange chromatography removes any residual
human insulin ester and ensures that the final product fulfills the specifications
for the monocomponent insulins. The data and purity tests of relevance to human
insulin originating from porcine insulin are compiled in the second column of
Table 20.

3. Biosynthesis in E. coli
The strategy developed for biosynthesis of human insulin is depicted in Fig. 34
(Goeddel et al. 1979, Chance et al. 1981 a--c). DNA sequences encoding for the
A and B chains of insulin were synthesized chemically. A codon for methionine
was introduced at the N-terminals of the A and B chain genes. The tryptophan
synthetase operon, trp E, was placed in front of the methionine codon, and the
two constructed genes were inserted into plasmids and cloned separately in
E. coli. Fermentations of the bacteria containing the plasmids result in the pro-
Biosynthesis in E. coli 77

Crude porcine insulin

Thr-OR ~ Transpeptidation
~Ala
Trypsin ~ reaction

Human insulin ester (H-I-OR)

Gelfiltration at
~Trypsin
lowpH

Chromatography

Ion exchange ~ Unconverted

Chromatography porcine insulin

Cleavage of H-I-OR

Human insulin

Chromatography Chromatography I~ H-I-OR

MC-porcine insulin Human Monocomponent insulin

(HM-insulin)

Fig. 33. Flow sheet ofthe production of Human Monocomponent insulin from crude porcine in-
sulin. (After Markussen et al. 1982)

duction of chimeric proteins consisting of a tryptophan synthetase fragment fol-

lowed by methionine and either the A or the B chain. The large chimeric pro-
teins precipitate in the cytoplasm of E. coli, the precipitation thus preventing pro-
teolysis of the insulin chains by the proteases of the cytoplasm. The chimeric pro-
tein amounts to 20% of the total protein in the cells and a yield of 10 5 molecules
per cell can be achieved (Goeddel et al. 1979).
After the fermentations the cells are harvested and the chimeric proteins are
isolated from the cells. The insulin chains are released from the chimeric partner
by treatment with cyanogen bromide, which specifically cleaves at methionine
residues, Fig. 34. This cleavage is possible since human insulin has nomethionine
residues. The methionine residues are converted into homoserine residues stiH at-
tached to the tryptophan synthetase fragment.
The insulin chains are then converted to the S-sulphonates, which are easy to
handle and purify in aqueous solution. The next step is the critic al chain combi-
78 Human Insulin

1 Separate fermentation of chains 1

~ ~
~ - Met -1 A-chain 1 1 trp-E 1 - Met - 1 B-chain 1

~ ~
1 CNBr cleavage in formic acid 1

~ ~
~ - hSer + A-chain 1 trp-E 1 - hSer + B-chain
~ ~
Oxidative sulfitolysis, air oxidation. Purification

A-(SS03kchain B-(SS03lz-chain

2.6 eqUiValent~ / 1 equivalent

Chain combination reaction.

Concentration of chains: 5-10 mg/ml
0.1 M glycine buffer, pH 10.5
Reaction time: 24 hours at 0-4°C
Thiol reagent: Dithioerythritol,
1.1 equivalent of SH relative to SS03

Human Insulin
(Yield -60% relative to B-chain)
+ isomers + polymers.

-+ Isomers, polymers

~ ~
A-chain polymers Human Insulin (97.34% pure)

Fig.34. Flow sheet of the production of human insulin from the fermentations of the insulin
chains in separate clones of E. coli. Purity according to Chance et al. (1981 b). M et methionine;
h-Ser homoserine; CNBr cyanogen bromide; trp E tryptophansynthetase E

nation reaetion. From one A-ehain and one B-ehain it is, in theory, possible to
form 12 isomerie eompounds, only one ofwhieh is insulin.
By applying an exehange reaetion between S-sulphonates and thiols (Chanee
et al. 1981 e) eonditions were found that yielded satisfaetory yields of insulin, ef.
Fig. 34.
The reaetions are formally:
/SH /s
A-SS0 3 +R(SH)z --t A-SH+R --t R,,---~ + HS03
""'--SS03
A-SH+B-SS0 3 --t A-S-S-B+HS0 3
HS0 3 +OH- --t H 2 0+S0 3 .
Biosynthesis in Saccharomyces cerevisiae 79

Table 20. Data and purity tests of the two human insulins, originating from porcine pancreatic
glands and fermentation of E. coli bacteria, respectively

Origin Porcine insulin E. coli bacteria

Manufacturer Novo E. Lilly

Potency of dry substance 28.5±0.8 IU/mg 27.5± 1.7 IU/mg
Method of bioassay Mouse convulsion Rabbit hypoglycemia
Purity by HPLC' >99% >97%
Porcine insulin n.d., d.1. < 0.1 %
Porcine PLI b <lppm
Porcine trypsin < 1: 1022 (w/w)
"E. coli proteins", RIA C <4ppm e
Endotoxin, LAL d test <0.6ppm

• High pressure liquid chromatography

b Proinsulin like immunoreactivity
C Radio immunoassay

d Limulus amebocyte lysate

C Quantitation of a heterogeneous contamination dependent upon antiserum and standard.
Abbreviations: n.d. non-detectable; d.1. detection limit

After purification, Le. removal of isomers and polymers, human insulin with
a purity of about 97% is obtained (Chance et al. 1981 b). Data and purity tests
indicative for human insulin produced in E. coli are compiled in the third column
of Table 20. The E. coli protein contaminants have been estimated by aRIA and
four batches contained less than 1 ppm, while detectable amounts (3.3 ppm) were
found in one batch (Baker et al. 1981). It must, however, be realized that attempt-
ing to quantify the heterogenous mixtures ofE. coli contaminants as a single value
by RIA poses certain problems, as the results may vary dependent on the antisera,
tracer and standard used, as well as on the composition of the sample.
More recently a process using DNA sequences encoding for human proinsulin
has been described for production of human insulin in E. coli (Frank et al. 1981).
The fermentation steps in this method are in principle similar to the steps
described for the A- and B-chain biosynthesis, but the post-fermentation
chemistry is simpler as the chain combination process now is directed by the
proinsulin C-peptide.

4. Biosynthesis in Saccharomyces cerevisiae

An alternative method ofmanufacture ofhuman insulin by biosynthesis in yeast

has recently been described (Markussen et al. 1986a, b, Thim et al. 1986).
Recombinant plasmids coding for single-chain insulin precursors with only a
fewamino acid residues linking the A- and B-chain are inserted into a S. cerevisiae
strain, which is grown in continuous fermentation. During fermentation the
single-chain insulin precursor is exported to the broth with free B-chain N-
terminal and with the correct disulfide bridges formed. After removal of yeast
cells, it is isolated from the broth by adsorption on an ion exchange column. The
precursor is converted by tryptic transpeptidation to a double-chain human
insulin ester, followed by hydrolysis of the ester, a process similar to the con-
80 Human Insulin

version of porcine to human insulin (cf. D.I.2). Purification including 2 crys-

tallizations, gel filtration, 2 anion chromatographies and a final purification by
HPLC resulted in a purity of 99.5-99.9% as assessed by analytical HPLC. Yeast
protein contaminants were detected to be less than 1 ppm (detection limit 1 ppm)
by an ELISA method (Markussen et al. 1986a, b).

II. PreparatiollS
The avai1abi1ity of human insu1in in 1arge quantities has resulted in the introduc-
tion of severa1 different pharmaceutical preparations covering the need for short-,
intermediate- and long-acting types. In principle the formulations ofthese human
insulin preparations with respect to content of auxilliary substances (cf. C.V) and
mechanisms of pro10ngation (cf. C.II.6, C.IV) are the same as those of their
porcine or bovine counterparts.
It has been observed that the regular human insu1ins are absorbed somewhat
quicker than porcine insulin, irrespective of the source and brand of the human
insulin (Federlin et al. 1981, Ciippers et al. 1982, Owens 1986). This quicker
absorption might be due to the fact that human insu1in is more hydrophilic than
porcine insulin, but a tendency to less association ofhuman insu1in into hexamers
than that of porcine insulin, as revealed in a gel filtration experiment (Brange,
unpublished), can a1so be a possib1e explanation.
The same tendency of a quicker initial effect of human intermediate- and
long-acting preparations has also been observed when compared with the
corresponding porcine or bovine formu1ations. Thus, a quicker initial effect and
shorter duration of action of NPH made from biosynthetic human insulin than
porcine NPH have been reported (Enzmann at al. 1982, Clark et al. 1982, Owens
et al. 1984). In contrast to these results, Hi1debrandt et al. (1984) found no
difference between absorption rate of NPH insu1ins from different species when
given in equal doses but observed a relatively lower absorption rate for a large
than for a small dose. Monotard HM was not different from the porcine
Monotard MC in a short-term, doub1e-b1ind, cross-over study (Sestoft et al. 1982)
nor in longer, open switch-over studies (Egstrup and 01sen 1982, Vestermark
1982). In a fifteen-month doub1e-b1ind crossover study Home et al. (1984) found a
small, but clinically significant, pharmacokinetic difference between human and
porcine insulin in patients treated with Actrapid and Monotard.
Human U1tra1ente (Ultratard HM) was studied in a double blind crossover
study in 18 diabetics and found as effective as bovine Ultra1ente in controlling
basal plasma glucose concentrations (Holman et al. 1984). Hi1debrandt et al.
(1985) reported a substantially faster absorption of human than bovine Ultra-
lente, but conduded that human U1tralente is sui table as the basal insulin
preparation for multiple insulin injection regimens. The faster absorption of the
human relative to bovine Ultralente was confirmed by Owens et al. (1986).
C1inica1 and pharmacologica1 studies of human insulin have recent1y been
reviewed by Owens (1986).
Immungenicity studies (cf. B.III) have shown that human MC insulin induced
significant1y less insulin antibodies than porcine MC insulin (Fankhauser et al.
Preparations 81

1982, Schernthaner et al. 1983, Heding et al. 1984, Luyckx et al. 1986) and this
may have clinical implications, as even small amounts of antibodies are associated
with a reduced C-peptide secretion and a higher insulin requirement after about
one year of insulin treatment of insulin-dependent diabetic children (Ludvigsson
1984, Schernthaner et al. 1984).
Introducing pure human insulin for the treatment of diabetes has finally made
it possible to substitute diabetics' lack of endogenous insulin with an identical
molecule.
References

Aaby P (1979) Concentration ofporcine proinsulin-like material in plasma ofinsulin-treated dia-

betics in relation to purity ofinsulin preparations. Horm Metab Res 11:455--457
Abel JJ (1926) Crystalline insulin. Proc Natl Acad Sci (Wash) 12:132-136
Adams MJ, Blundell TL, Dodson EJ, Dodson GG, Vijayan M, Baker EN, Harding MM,
Hodgkin DC, Rimmer B, Sheat S (1969) Structure ofrhombohedral2 zinc insulin crystals.
Nature 224:491--495
Alameda GK, Evelhoch JL, Sudmeier JL, Birge RR (1985) Characterization ofthe internal cal-
cium(II) binding sites in dissolved insulin hexamer using europium(III) fluorescence. Bioche-
mistry 24:1757-1762
Albisser AM, Leibel BS, Ewart TG, Davidovac Z, Botz CK, Zingg W (1974) An artificial endoc-
rine pancreas. Diabetes 23:389-396
Albisser AM, Lougheed W, Perlman K, Bahoric A (1980) Nonaggregating insulin solutions for
long-term glucose control in experimental and human diabetes. Diabetes 29:241-243
Albisser AM, Lougheed WD, Chow JC, Tung AK (1982) A modified insulin for pumps. Diabe-
tes 31 (Suppl. 2, part 2):67 A (abstract)
Allwood MC (1978) Antimicrobial agents in single- and multi-dose injections. J Appl Bacteriol
44:Svii-Sxviii
Allwood MC (1982) The effectiveness ofpreservatives in insulin injections. Pharm J 229:340
Amaya FJ, Lee T-C, Chichester CO (1976) Biological inactivation ofproteins by the Maillard
reaction. Effect ofmild heat on the tertiary structure ofinsulin. J Agric Food Chem 24:465-
467
American Diabetes Association (1985) Policy statement, continuous subcutaneous insulin infu-
sion. Diabetes 34:946-947
Andersen 00 (1973) Insulin antibody formation. II The influence of species difference and me-
thod of administration. Acta Endocrinol 72:33--45
Ando T, Watanabe S (1969) A new method for fractionation ofprotamines and the amino acid
sequences of salmine and three components of iridine. Int J Protein Res 1:221-224
Arnott RD, Cameron MA, Stepanas TV, Cohen M (1982) Insulin syringes. Dangers of dead spa-
ce. Med J Aust 2:39--40
Arrieta-Molero JF, Aleck K, Sinha MK, Brownscheidle CM, Shapiro LJ, Sperling MA (1982)
Orally administered liposome-entrapped insulin in diabetic animals. A critical assessment.
Horm Res 16:249-256
Ashford WR, Campbell J, Davidson JK, Fisher AM, Haist RE, Lacey AH, Lin B, Martin JM,
Morley NH, Rastogi KS, Storvick WO (1969) A consideration ofmethods ofinsulin assay.
Diabetes 18:828-833
Baker RS, Schmidtke JR, Ross JW, Smith WC (1981) Preliminary studies on the immunogenicity
and amounts of Escherichia coli polypeptides in biosynthetic human insulin produced by re-
combinant DNA technology. Lancet 2:1139-1142
Bang HO (1946) Some investigations on the absorption mechanism of protamine insulin. Acta
Pharmacol Toxicol (Copenh) 2:79-88 .
Bangham DR, Mussett MV (1959) The fourth international standard for insulin. Bull WHO
20:1209-1220
Bangham DR, Jonge H de, Noordwijk J van (1978) The collaborative assay of the European
pharmacopoeia biological reference preparation for insulin. J Biol Stand 6:301-314
Banting FG, ·Best CH (1922) Pancreatic extracts. J Lab Clin Med 7:464--472 .
84 References

Bentley G, Dodson E, Dodson G, Hodgkin D, Mercola D (1976) Structure of insulin in 4-zinc

insulin. Nature 261:166-168
Bentley G, Dodson G, Lewitova A (1978) Rhombohedral insulin crystal transformation. J Moi
BioI126:871-875
Bergenstal RM, Cohen RM, Lever E, Polonsky K, Jaspan J, Blix PM, Revers R, Olefsky JM,
Kolterman O, Steiner K, Cherrington A, Frank B, Galloway J, Rubenstein AH (1984) The
metabolic effects ofbiosynthetic human proinsulin in individuals with type 1 diabetes. J Clin
Endocrinol Metab 58:973-979
Berger M, Ciippers HJ, Hegner H, Jorgens V, Berchtold P (1982) Absorption kinetics and bio-
logic effects of subcutaneously injected insulin preparations. Diabetes Care 5:77-91
Berger W, Keller U, Vorster D (1981) Verlauf und Therapie des Coma diabeticum. Internist
(Berlin) 22:219-228
Bergman N, Vella ID (1982) Potentiallife-threatening variations of drug concentrations in in-
travenous infusion systems. Potassium chloride, insulin, and heparin. Med J Aust 2:270-
272
Berson SA, Yalow RS (1959) Quantitative aspects of the reaction between insulin and insulin-
binding antibody. J Clin Invest 38:1996-2016
Berson SA, Yalow RS (1960) Insulin inhibitors and insulin resistance. NY State J Med 60:3658-
3665
Berson SA, Yalow RS (1964) The present status of insulin antagonists in plasma. Diabetes
13:247-259
Berson SA, Yalow RS, Bauman A, Rothschild MA, Newerly K (1956) Insulin-I 131 metabolism
in human subjects: Demonstration of insulin binding globulin in the circulation of insulin
treated subjects. J Clin Invest 35:170-190
Bi RC, Dauter Z, Dodson E, Dodson G, Giordano F, Reynolds C (1984) Insulin's structure as
a modified and monomeric molecule. Biopolymers 23:391-395
Biemond MEF, Sipman WA, Olivie J (1979) Quantitative determination of polypeptides by
gradient elution high pressure liquid chromatography. J Liq Chromatogr 2:1407-1435
Binder C (1969) Absorption of injected insulin. Acta Pharmacol Toxicol (Copenh) (Suppl. 2)
27:1-87
Binder C, Nielsen AA, J0rgensen K (1967) The absorption of an acid and a neutral insulin so-
lution after subcutaneous injection into different regions in diabetic patients. Scand J Clin
Lab Invest 19:156-163
Birdi KS (1973) The binding of cationic detergent to insulin, zinc-insulin and Iso-insulin. Bio-
chem J 135:253-255
Blackshear PJ, Rohde TD, Palmer JL, Wigness BD, Rupp WM, Buchwald H (1983) Glycerol
prevents insulin precipitation and interruption of flow in an implantable insulin infusion
pump. Diabetes Care 6:387-392
Blaug SM, Grant DE (1974) Kinetics of degradation of the parabens. J Soc Cosmet Chem
25:495-506
Bloom SR (1979) Vasoactive intestinal peptide. In: Jaffe BM, Behrman HR (eds) Methods of
hormone radioimmunoassay. Academic Press, New York, pp 553-565
Bloom SR, West AM, Polak JM, Barnes AJ, Adrian TE (1978) Hormonal contaminants of in-
sulin. In: Bloom SR (ed) Gut hormones. Churchill Livingstone, Edinburgh London New
York, pp 318-322
Bloom SR, Barnes AJ, Adrian TE, Polak JM (1979) Autoimmunity in diabetics induced by hor-
monal contaminants ofinsulin. Lancet 1:14--17
Blundell TL, Cutfield JF, Cutfield SM, Dodson EJ, Dodson GG, Hodgkin DC, Mercola DA,
Vijayan M (1971) Atomic positions in rhombohedral2-zinc insulin crystals. Nature 231:506-
511
Blundell T, Dodson G, Hodgkin D, Mercola D (1972) Insulin: The structure in the crystal and
its reflection in chemistry and biology. Adv Protein Chem 26:279-402
Boshell BR, Barrett JC, Wilensky AS, Patton TB (1964) Insulin resistance. Response to insulin
from various animal sources, inc1uding human. Diabetes 13:144--152
Bottermann P, Hiigler P, Schweigart U, Zilker T, Giebeler K, Enzmann F (1980) Ermittlungvon
Insulin-Wirkprofilen mit Hilfe des "Glukose-kontrollierten Insulin-Infusions-Systems"
(Biostator®). Akt Endokrin 1:337-345
References 85

Bradbury AF, Ko ASC, Massey DE, Salokangas A, Smyth DG, Sabey GA, Webb FW, Stewart
GA (1973) The quest for a latent action insulin. Postgrad Med J 49 (Suppl):945-948
Bradbury JH, Ramesh V, Dodson G (1981) 1H nuclear magnetic resonance study of the histidine
residues ofinsulin. J MoI BioI150:609-613
Brandenburg D (1969) Des-PheB1-insulin, ein kristallines Analogon des Rinderinsulins. Hoppe-
Seyler's Z Physiol Chem 350:741-750
Brange J, Havelund S (1983a) Properties ofinsulin in solution. In: Brunetti P, Alberti KGMM,
Albisser AM, Hepp KD, Benedetti MM (eds) Artificial systems for insulin delivery.
Raven Press, New York, pp 83-88
Brange J, Havelund S (1983 b) Insulin pumps and insulin quality: Requirements and problems.
Acta Med Scand (Suppl) 671:135-138
Brange J, Havelund S (1983c) Stabilized insulin preparations. US Patent 4476118
Brange J, Havelund S, Hansen P, Langkjrer L, S0rensen E, Hildebrandt P (1982) Formulation
of physically stable neutral insulin solution for continuous infusion by delivery systems. In:
Gueriguian JL, Bransome ED, Outschoom AS (eds) Hormone Drugs. United States Phar-
macopeial Convention, Rockville, Maryland, pp 96-105
Brange J, Langkjrer L, Havelund S, S0rensen E (1983) Chemical stability ofinsulin: Neutral in-
sulin solutions. Diabetologia 25:193 (abstract)
Brange J, Langkjrer L, Havelund S, S0rensen E (1984) Chemical stability of insulin: formation
of covalent insulin dimers and other higher molecular weight transformation products in in-
termediate- and long-acting insulin preparations. Diabetologia 27:259A-260A (abstract)
Brange J, Langkjrer L, Havelund S, S0rensen E (1985) Chemical stability of insulin: formation
of desamido insulins and other hydrolytic products in intermediate and long acting insulin
preparations. Diabetes Res Clin Pract (Suppl1) p 67 (abstract)
Brange J, Havelund S, Hommel E, S0rensen E, Kiihl C (1986) Neutral insulin solutions physi-
cally stabilized by addition of Zn 2 + . Diabetic Medicine 3:532-536
Brange J, Hansen JF, Havelund S, Melberg SG (1987) Studies ofthe insulin fibrillation process.
In: Brunetti P, Waldhiiusl W (eds) Advanced models for the therapy ofinsulin-dependent di-
abetes. Raven Press, New York, pp 85-90
Brill AS, Venable JH (1968) The binding of transition metal ions in insulin crystals. J MoI Biol
36:343-353
Bringer J, Heldt A, Grodsky GM (1981) Prevention ofinsulin aggregation by dicarboxylic amino
acids during prolonged infusion. Diabetes 30:83-85
British Pharmacopoeia (1980) voI 1+11. Her Majesty's Stationery Oftice, London
Brown L, Munoz C, Siemer L, Edelman E, Langer R (1986) Controlled release ofinsulin from
polymer matrices. Control of diabetes in rats. Diabetes 35:692-697
Browne M, Cecil R, Miller JC (1973) Some reactions ofinsulin at non-polar surfaces. Eur J Bio-
chem 33:233-240
Brownlee M, Cerami A (1979) A glucose-controlled insulin-delivery system: Semisynthetic insu-
lin bound to lectin. Science 206:1190-1191
Brownlee M, Cerami A (1983) Glycosylated insulin complexed to concanavalin A. Biochemical
basis for a closed-Ioop insulin delivery system. Diabetes 32:499-504
Brunfeldt K, Poulsen JE (1953) Protamine-splitting enzyme from serum and subcutaneous tis-
sue. Rep Steno Mem Hosp Nord Insulinlab 5:51-61
Buchwald H, Rohde TD, Dorman FD, Skakoon JG, Wigness BD, Prosl FR, Tucker EM,
Rublein TG, Blackshear PJ, Varco RL (1980) A totally implantable drug infusion device:
Laboratory and clinical experience using a model with single flow rate and new design for
modulated insulin infusion. Diabetes Care 3:351-358
Buchwald H, Varco RL, Rupp WM, Goldenberg FJ, Barbosa J, Rohde TD, Schwartz RA,
Rublein TG, Blackshear PJ (1981) Treatment of a type II diabetic by a totally implantable
insulin infusion device. Lancet 1:1233-1235
Burgermeister W, Enzmann F, Schone H-H (1975) The isolation of insulin from the pancreas.
In: Hasselblatt A, Bruchhausen F von (eds) Insulin: Handbook of Experimental Pharmacol-
ogy, voI XXXII/2. Springer, Berlin Heidelberg New York, pp 715-727
Burke MJ, Rougvie MA (1972) Cross-p protein structures. 1. Insulin fibrils. Biochemistry
11:2435-2439
Burman KD, Cunningham EJ, Klachko DM, Bums TW (1973) Successful treatment ofinsulin
resistance with dealaninated pork insulin (DPI). Mo Med 70:363-366
86 References

Capi an SN, Berkman EM (1976) Protamine sulfate and fish allergy. N Engl J Med 295:172
Carpenter FH (1958) Partition column chromatography of insulin in 2-butanol-aqueous acid
systems. Arch Biochem Biophys 78:539-545
Carpenter FH (1966) Relationship of structure to biological activity of insulin as revealed by de-
gradative studies. Am J Med 40:750-758
Cavallini D, Federici G, Barboni E, Marcucci M (1970) Formation ofpersulfide groups in alka-
line treated insulin. FEBS Lett 10:125-128
Caygill CPJ, Ayling CM (1979) Effect of contaminating proteolytic activity upon insulin and in-
sulin-protamine complex. J Pharm PharmacoI31:849-852
Chan Y-K, Oda G, Kaplan H (1981) Chemical properties of the functional groups of insulin.
Biochem J 193:419-425
Chance RE (1972) Amino acid sequences of proinsulins and intermediates. Diabetes 21
(SuppI2):461-467
Chance RE, Ellis RM, Bromer WW (1968) Porcine proinsulin: Characterization and amino acid
sequence. Science 161:165-167
Chance RE, Root MA, Galloway JA (1976) The immunogenicity ofinsulin preparations. Acta
Endocrinol (SuppI205) (Copenh) 83:185-196
Chance RE, Moon NE, Johnson MG (1979) Human pancreatic polypeptide (HPP) and bovine
pancreatic polypeptide (BPP). In: Jaffe BM, Behrman HR (eds) Methods ofhormone radio-
immunoassay. Academic Press, New York, pp 657-672
Chance RE, Kroeff EP, Hoffmann JA (1981 a) Chemical, physical, and biological properties of
recombinant human insulin. In: Gueriguian JL (ed) Insulins, growth hormone, and recom-
binant DNA technology. Raven Press, New York, pp 71-86
Chance RE, Kroeff EP, Hoffmann JA, Frank BH (1981 b) Chemical, physical and biological
properties ofbiosynthetic human insulin. Diabetes Care 4:147-154
Chance RE, Hoffmann JA, Kroeff EP, Johnson MG, Schirmer EW, Bromer WW (1981 c) The
production of human insulin using recombinant DNA technology and a new chain combi-
nation procedure. In: Rick DH, Gross E (eds) Peptides: Synthesis-Structure-Function. Pro-
ceedings of the seventh American peptide symposium. Pierce Chemical Company, Rockford,
Illinois, pp 721-728
Chasteen ND, DeKoch RJ, Rogers BL, Hanna MW (1973) Use ofvanadyl (IV) ion as a new spec-
troscopic probe of metal binding to proteins. Vanadyl insulin. J Am Chem Soc 95:1301-
1309
Chawdhury SA, Dodson EJ, Dodson GG, Reynolds CD, Tolley SP, Blundell TL, Cleasby A,
Pitts JE, Tickle IJ, Wood SP (1983) The crystal structures of three non-pancreatic human
insulins. Diabetologia 25:460-464
Chothia C, Lesk AM, Dodson GG, Hodgkin DC (1983) Transmission ofconformationalchange
in insulin. Nature 302:500-505
Clark AJL, Knight G, Wiles PG, Keen H, Ward JD, Cauldwell JM, Adeniyi-Jones RO, Leiper
JM, Jones RH, MacCuish AC, Watkins PJ, Glynne A, Scotton JB (1982) Biosynthetic hu-
man insulin in the treatment of diabetes. Lancet 2:354-357
Cogswell JJ, Simpson RM (1980) Introduction ofU-100 insulin. Br Med J 280:566
Colagiuri S, Villalobos S (1986) Assessing effect ofmixing insulins by glucose-clamp technique
in subjects with diabetes mellitus. Diabetes Care 9:579-586
Cole RD (1960) The chromatography of insulin in urea-containing buffer. J Biol Chem
235:2294-2299
Colwell AR (1947) Protamine insulin mixtures in the treatment of diabetes mellitus. NY State
J Med 47:1103-1110
Corran PH, Waley SG (1975) The reaction ofpenicillin with proteins. Biochem J 149:357-364
Crane CW, Path MC, Luntz GRWN (1986) Absorption ofinsulin from the human small inte-
stine. Diabetes 17:625-627
Creque HM, Langer R, Folkman J (1980) One month of sustained release ofinsulin from a po-
lymer implant. Diabetes 29:37-40
Cunningham LW, Fischer RL, Vestling CS (1955) A study of the binding of zinc and cobalt by
insulin. J Am Chem Soc 77:5703-5707
Ciippers HJ, Franzke D, Esken P, J6rgens V, Berger M (1982) Pharmakokinetik und biologische
Aktivităt von semisynthetischen und biosynthetischen Humaninsulin-Prăparaten. Aktuel
Endokrinol Stoffwechsel3:102
References 87

Damgaard U, Kruse V (1982) Inherent problems in radioimmunoassay exemplified by determi-

nation of proinsulin-like immunoreactivity in bovine insulin. In: Gueriguian JL, Bransome
ED, Outschoom AS (eds) Hormone Drugs. United States Pharmacopeial Convention,
Rockville, Maryland, pp 187-191
Damgaard U, Markussen J (1979) Analysis ofinsulins and related compounds by HPLC. Horm
Metab Res 11:580-581
Dapergolas G, Gregoriadis G (1976) Hypoglycaemic effect of liposome-entrapped insulin ad-
ministered intragastrically into rats. Lancet 11:824-827
Davidson JK, DeBra DW (1978) Immunologic insulin resistance. Diabetes 27:307-318
Davis BJ (1964) Disc electrophoresis - II. Ann NY Acad Sci 121:404-427
Diezel W, Kopperschliiger G, Hofmann E (1972) An improved procedure for protein staining
in polyacrylamide gels with a new type of Coomassie Brilliant Blue. Anal Biochem 48:617-
620
Dinner A, Lorenz L (1979) High performance liquid chromatographic determination of bovine
insulin. Anal Chem 51:1872-1873
Dodson E, Harding MM, Hodgkin DC, Rossmann MG (1966) The crystal structure ofinsulin.
III. Evidence for a 2-fold axis in rhombohedral zinc insulin. J Moi BioI16:227-241
Dodson EJ, Dodson GG, Lewitova A, Sabesan M (1978) Zinc-free cubic pig insulin: Crystalli-
zation and structure determination. J Moi BioI125:387-396
Dodson EJ, Dodson GG, Hodgkin DC, Reynolds CD (1979) Structural relationships in the two-
zinc insulin hexamer. Can J Biochem 57:469-479
Dorzbach E von, Millier R (1971) Die Insulintherapie: Die Insulinpriiparate. In: Pfeiffer EF
(Hrsg) Handbook of Diabetes Mellitus. Pathophysiology and clinical considerations. VoI. II.
J. F. Lehmann's Veriag, Miinchen, pp 1087-1111
Dunn MF, Pattison SE, Storm MC, Quiel E (1980) Comparison ofthe zinc binding domains in
the 7 S nerve growth factor and the zinc-insulin hexamer. Biochemistry 19:718-725
Easter BRD, Sutton DA, Drewes SE (1978) Crystalline (A21-desamido) bovine insulin. Hoppe-
Seyler's Z Physiol Chem 359:1229-1236
Ege H (1985) Sources and characteristics ofthe human insulins. In: Breimer DD, Speiser P (eds)
Topics in pharmaceutical sciences. Elsevier Science Publishers B.V., Amsterdam New York
Oxford, pp 57-70
Ege H (1986) Semantics ofinsulin. Diabetic Medicine 3:212-215
Egstrup K , Olsen J (1982) Treatment with human insulin in outpatient diabetics. Lancet 11:222-
223
Ellenbogen E (1949) The determination of the physical-chemical properties of insulin and their
application to the equilibrium between insulin of molecular weights 12,000 and 36,000. Dis-
sertation, Harvard University
Emdin SO, Dodson GG, Cutfield JM, Cutfield SM (1980) Role of zinc in insulin biosynthesis.
Some possible zinc-insulin interactions in the pancreatic B-cell. Diabetologia 19:174-182
Eneroth G, Aahlund K (1968) Biological assay ofinsulin by blood sugar determination in mice.
Acta Pharm Suec 5:591-594
Eneroth G, Aahlund K (1970a) A twin cross-over method for bioassay of insulin using blood
glucose levels in mice - a comparison with a rabbit method. Acta Pharm Suec 7:457-462
Eneroth G, Aahlund K (1970 b) Exogenous insulin and blood glucose levels in mice. Acta Pharm
Suec 7:491-500
Enzmann F, Balikcioglu S, Glynne A, Scotton J, Marsden J (1982) Zusammenfassende Betrach-
tung der klinischen Erprobung von Biosynthetischem Human-Insulin in Europa. In: Peter-
sen KG, Schliiter KJ, Kerp L (Hrsg) Neue Insuline. Freiburger Graphische Betriebe, Frei-
burg, pp 181-186
European Pharmacopoeia (1975) III Counci! of Europe (European Treaty Series No. 50) Mai-
sonneuve S.A., Sainte-Ruffine, p 78
Fahrenkrug J, Schaffalitzky de Muckadell OB (1977) R,.adioimmunoassay ofvaspactive intesti-
nal polypeptide (VIP) in plasma. J Lab Clin Med 89:1379-1388
Falholt K (1982) Determination of insulin specific IgE in serum of diabetic patients by solid-pha-
se radioimmunoassay. Diabetologia 22:254-257
Falholt K, Hoskam JAM, Karamanos BG, Siisstrunk H, Viswanathan M, Heding LG (1983)
Insulin-specific IgE in serum of 67 diabetic patients against human insulin (Novo), porcine
insulin, and bovine insulin. Four case reports. Diabetes Care 6:61-65
88 References

Fankhauser S, Herz G, Zuppinger K (1982) Experience avec l'insuline humaine chez des enfants
diabetiques. Med Hyg 40:1378-1384
Federlin K, Laube H, Veicovsky HG (1981) Biologic and immunologic in vivo and in vitro stud-
ies with biosynthetic human insulin. Diabetes Care 4:170-174
Feingold V, lenkins AB, Kraegen EW (1984) Effect of contact material on vibration-induced in-
sulin aggregation. Diabetologia 27:373-378
Finger H, Schaeg W, Niemann H (1967) Untersuchungen iiber den EinfluB von iiberschiissigem
Zink auf die Aggregation von nativem und photooxidiertem Insulin. Experientia 23:435
Fisher BV, Porter PB (1981) Stability of bovine insulin. 1 Pharm PharmacoI33:203-206
Fisher BV, Smith D (1986) HPLC as a replacement for the animal response assays for insulin.
1 Pharm Biomed Anal 4:377-387
Fitz-Patrick D, Patel YC (1981) Antibodies to insulin, pancreatic polypeptide, glucagon and
somatostatin in insulin-treated diabetics. 1 Clin Endocrinol Metab 52:948-952
Forck G, Wilhelm E, Baumeister G, Westerloh K (1975) Scheinbare Insulinvertrăglichkeit bei
einer Arthusreaktion durch Surfen. Med Welt 26:664-669
Forlani G, Santacroce G, Ciavarella A, Capelli M, Mattioli L, Vannini P (1986) Effects ofmixing
short- and intermediate-acting insulins on absorption course and biologic effect of short-act-
ing preparation. Diabetes Care 9:587-590
Fort P (1981) Low-dose insulin infusion for diabetic ketoacidosis. Editorial correspondence. 1
Pediatr 98:169
Francis Al, Hanning I, Alberti KGMM (1985) The effect ofmixing human soluble and human
crystalline zinc-suspension insulin: plasma insulin and blood glucose profiles after subcuta-
neous injection. Diabetic Medicine 2: 177-180
Frank BH, Pekar AH, Veros Al (1972) Insulin and proinsulin conformation in solution. Dia-
betes 21 (Suppl 2):486-491
Frank BH, Pettee lM, Zimmerman RE, Burck Pl (1981) The production ofhuman proinsulin
and its transformation to human insulin and C-peptide. In: Rich DH, Gross E (eds) Peptides:
Synthesis-Structure-Function. Pierce Chemical Company, Rockford, pp 729-738
Fraser DT (1923) White mice and the assay ofinsulin. 1 Lab Clin Med 8:425-428
Fredericq E (1956) The association of insulin molecular units in aqueous solutions. Arch Bio-
chem Biophys 65:218-228
Fredericq E, Neurath H (1950) The interaction ofinsulin with thiocyanate and other anions. The
minimum molecular weight ofinsulin. 1 Am Chem Soc 72:2684-2691
Friesen H-l (1980) The relative reactivity of insulin amino groups as an indicator of structural
accessibility and its use for synthetic approaches for structure-function studies. In: Branden-
burg D, Wollmer A (eds) Insulin: Chemistry, structure and function of insulin and related
hormones. Walter de Gruyter, Berlin New York, pp 125-133
Fullerton WW, Low BW (1970) Insulin crystallization in the presence ofbasic proteins and pep-
tides. Biochim Biophys Acta 214:141-147
Furberg H, lensen AK, Salbu B (1986) Effect ofpretreatment with 0.9% sodium chloride or in-
sulin solutions on the delivery of insulin from an infusion system. Am 1 Hosp Pharm
43:2209-2213
Galloway lA, Root MA (1972) New forms ofinsulin. Diabetes 21 (SuppI2):637-648
Galloway lA, Root MA, Rathmacher RP, Carmichael RH (1973) A comparison of acid regular
and neutral regular insulin. Responses of normal fasted subjects to varying doses of regular
insulin. Diabetes 22:471-479
Galloway lA, Root MA, Chance RE, lackson RL, Wentworth SM, Davidson lA (1975) New
forms ofinsulin. In: Kryston LJ, Shaw RA (eds) Endocrinology and Diabetes, 30th Hahne-
mann Symposium. Grune & Stratton, New York, pp 329-342
Galloway lA, Kennedy ME, Marshall KA, Michael KR, Nimoy M, Raines KC, Spitzer TQ
(1976) How to avoid dogging of insulin syringes. Diabetes Forecast 29, No.6 (Nov-
Dec):27
Galloway lA, Spradlin CT, Root MA, Fineberg SE (1981 a) The plasma glucose response of nor-
mal fasting subjects to neutral regular and NPH biosynthetic human and purified pork in-
sulins. Diabetes Care 4:183-188
Galloway lA, Spradlin CT, Nelson RL, Wentworth SM, Davidson lA, Swarner lL (1981 b) Fac-
tors influencing the absorption, serum insulin concentration, and blood glucose responses
after injections of regular insulin and various insulin mixtures. Diabetes Care 4:366-376
References 89

Galloway JA, Spradlin CT, Jackson RL, Otto DC, Bechtel LD (1982) Mixtures ofintermediate-
acting insulin (NPH and Lente) with reguiar insulin: an update. In: Skyler JS (ed) Insulin
Update, Excerpta Medica. Amsterdam Oxford Princeton, pp 111-119
Glauber HS, Revers RR, Henry R, Schmeiser L, Wallace P, Kolterman OG, Cohen RM, Ru-
benstein AH, Ga1loway JA, Frank BH, Olefsky JM (1986) In vivo deactivation ofproinsulin
action on glucose disposal and hepatic glucose production in norma! man. Diabetes 35:311-
317
Goeddel DV, Kleid DG, Bolivar F, Heyneker HL, Yansura DG, Crea R, Hirose T, Kraszewski
A, ltakura K, Riggs AD (1979) Expression in Escherichia coli of chemically synthesized
genes for human insulin. Proc Natl Acad Sci USA 76:106-110
Goerz G, Ruzicka T, Hofmann N, Drost H, Griineklee D (1981) Granulomatose allergische Re-
aktion vom verzogerten Typ aufSurfen. Hautarzt 32:187-190
Goldman J, Carpenter FH (1974) Zinc binding, circular dichroism, and equilibrium sedimenta-
tion studies on insulin (bovine) and several ofits derivatives. Biochemistry 13:4566-4574
Goosen MFA, Leung YF, O'Shea GM, Chou S, Sun AM (1983) Long-acting insulin. Slow re-
lease of insulin from a biodegradable matrix implanted in diabetic rats. Diabetes 32:478-
481
Gordon GS, Moses AC, Silver RD, Flier JS, Carey MC (1985) Nasal absorption ofinsulin: En-
hancement by hydrophobic bile salts. Proc Natl Acad Sci USA 82:7419-7423
Graaff RAG de, Lewit-Bentley A, Tolley SP (1981) Effects of destabilizing agents on the insulin
hexamer structure. In: Dodson G, Glusker J, Sayre D (eds) Structural studies on molecules
ofbiological interest. Clarendon Press, Oxford, pp 547-556
Graham DT, Pomeroy AR (1978) The effects offreezing on commercial insulin suspensions. Int
J Pharm 1:315-322
Graham DT, Pomeroy AR (1984) An in-vitro test for the duration of action ofinsulin suspen-
sions. J Pharm Pharmacol 36:427-430
Grant PT, Coombs TL, Frank BH (1972) Differences in the nature ofthe interaction ofinsulin
and proinsulin with zinc. Biochem J 126:433-440
Grau U (1985) Chemical stability of insulin in a delivery system environment. Diabetologia
28:458-463
Grau U, Seipke G, Obermeier R, Thurow H (1982) Stabile Insulinlosungen fUr automatische Do-
siergeriite. In: Petersen K-G, Schliiter KJ, Kerp L (Hrsg) Neue Insuline. Freiburg Graphi-
sche Betriebe, Freiburg, pp 411-418
Hagedorn HC (1946) The absorption ofprotamine insulin. Rep Steno Hosp (Cph) 1:25-28
Hagedorn HC, Jensen BN, Krarup NB, Wodstrup 1 (1936) Protamine insulinate. J Am Med As-
soc 106:177-180
Haines BA Jr., Martin AN (1961) Interfacial properties ofpowdered material; caking in liquid
dispersions 1. Caking and flocculation studies. J Pharm Sci 50:228-232
Hallas-MlIJller K (1945) Chemical and biological insulin studies 1 and II. Dissertation. Copenha-
gen University
Hallas-MlIJller K (1956) The Lente insulins. Diabetes 5:7-14
Hallas-MlIJller K, Petersen K, Schlichtkrull J (1951) Crystalline and amorphous insulin-zinc com-
pounds with prolonged action (in Danish). Ugeskr Laeger 113:1761-1767
Hallas-MlIJller K, Petersen K, Schlichtkrull J (1952) Crystalline and amorphous insulin-zinc com-
pounds with prolonged action. Science 116:394-399
Harding MM, Hodgkin DC, Kennedy AF, O'Connor A, Weitzmann PDJ (1966) The crystal
structure of insulin. II. An investigation of rhombohedral zinc insulin crystals and a report
of other crystalline forms. J Moi BioI16:212-226
Harfenist EJ, Craig LC (1952) Countercurrent distribution studies with insulin. J Am Chem Soc
74:3083-3087
Havlova M, Nobilis M, Patocka L (1969) Stability of Superdep insulin (in Czechoslovakian).
Cesk Farm 18:390-392
Heding LG (1971) Radioimmunological determination ofpancreatic and gut glucagon in plas-
ma. Diabetologia 7:10-19
Heding LG, Larsen UD, Markussen J, JlIJrgensen KH, Hallund O (1974) Radioimmunoassays
for human, pork and ox C-peptides and related substances. In: Levine R,Pfeiffer EF (eds)
Radioimmunoassay: Methodology and applications in physiology and in clinica! studies.
Georg Thieme Publishers, Stuttgart, pp 40-44
90 References

Heding LG, Larsson Y, Ludvigsson J (1980) The immunogenicity ofinsulin preparation. Anti-
body levels before and after transfer to highly purified porcine insulin. Diabetologia 19:511-
515
Heding LG, Marshall MO, Persson B, Dahlquist G, Thalme B, Lindgren F, Âkerblom HK, Ril-
va A, Knip M, Ludvigsson J, Stenhammar L, Str6mberg L, S0vik O, Brevre H, Wefring K,
Vidnes J, Kjrergard 11, Bro P, Kaad PH (1984) Immunogenicity ofmonocomponent human
and porcine insulin in newly diagnosed type 1 (insulin-dependent) diabetic children. Diabe-
tologia 27:96--98
Hefford MA, Oda G, Kaplan H (1986) Structure-function relationships in the free insulin mo-
nomer. Biochem J 237:663-668
Heine RJ, Bilo HJG, Fonk T, Veen EA van der, Meer J van der (1984) Adsorption kinetics and
action profiles of mixtures of short- and intermediate-acting insulins. Diabetologia 27:558-
562
Helbig H-J (1976) Insulindimere aus der b-Komponente von Insulinprăparationen. Rheinisch-
WestfăIische Technische Hochschule, Aachen (Dissertation)
Hemmingsen AM, Krogh A (1926) The assay ofinsulin by the convulsive-dose method on white
mice. League of Nations Health III. 7:40-46
Hildebrandt P, Birch K, Sestoft L, V0lund Aa (1984) Dose-dependent subcutaneous absorption
ofporcine, bovine and human NPH insulins. Acta Med Scand 215:69-73
Hildebrandt P, Berger A, V0lund Aa, Kiihl C (1985) The subcutaneous absorption of human
and bovine ultralente insulin formulations. Diabetic Medicine 2:355-359
Hildebrandt R, Ilius A, Lotz U, Schliack V (1984) Effect ofinsulin suppositories in type I diabetic
patients (preliminary communication). Exp Clin Endocrinol 83:168-172
Hirai S, Ikenaga T, Matsuzawa T (1978) Nasal absorption of insulin in dogs. Diabetes 27:296--
299
Hirai S, Yashiki T, Mima H (1981) Effect of surfactants on the nasal absorption of insulin in
rats. Int J Pharm 9: 165-172
Hirata Y, Kohama T, Ooi K (1983) Nasal administration ofinsulin in healthy subjects and dia-
betic patients. In: Sakamoto N, Alberti KGMM (eds) Current and future therapies with in-
sulin. Excerpta Medica, Amsterdam Oxford Princeton, pp 263-267
Hirn S, K6nigstein RP, Pietschmann H (1982) Langzeitanwendung von Des-Phe-Insulin. In: Pe-
tersen K-G, Schliiter KJ, Kerp L (Hrsg) Neue Insuline. Freiburger Graphische Betriebe,
Freiburg, pp 384-389
Hirsch JI, Wood JH, Thomas RB (1981) Insulin adsorption to polyolefin infusion bottles and
polyvinyl chloride administration sets. Am J Hosp Pharm 38:995-997
Hodgkin DC (1974) Varieties ofinsulin. J EndocrinoI63:3P-14P
Hoffmann R, Riihl J (1982) Insulinverluste an Infusionsmaterial. Medizinische Fakultăt der
Universităt Diisseldorf (Dissertation)
Holladay LA, Ascoli M, Puett D (1977) Conformational stability and self-association of zinc-
free bovine insulin at neutral pH. Biochim Biophys Acta 494:245-254
Holman RR, Steemson J, Darling P, Reeves WG, Turner RC (1984) Human ultralente insulin.
Br Med J 288:665-668
Home PD, Mann NP, Hutchison AS, Park R, Walford S, Murphy M, Reeves WG (1984) A fif-
teen-month double-blind cross-over study of the efficacy and antigenicity of human and
pork insulins. Diabetic Medicine 1:93-98
Horrow JC (1985) Protamine: a review ofits toxicity. Anesth Analg 64:348-361
Howell SL, Tyhurst M, Duvefelt H, Andersson A, Hellerstr6m C (1978) Role of zinc and ca1cium
in the formation and storage of insulin in the pancreatic B-cell. Cell Tissue Res 188:107-
118
Humbel RE, Bosshard HR, Zahn H (1972) Chemistry of insulin. In: Greep RO, Astwood EB
(eds) Endocrinology. Handbook ofPhysiology, section 7, voI 1,. American Physiological So-
ciety, Washington DC, pp 111-132
Ichikawa K, Ohata I, Mitomi M, Kawamura S, Maeno H, Kawata H (1980) Rectal absorption
ofinsulin suppositories in rabbits. J Pharm PharmacoI32:314-318
Irsigler K, Kritz H (1979) Long-term continuous intravenous insulin therapy with a portable in-
sulin dosage-regulating apparatus. Diabetes 28:196-203
References 91

Irsigler K, Kritz H, Hagmiiller G, Franetzki M, Prestele K, Thurow H, Geisen K (1981) Long-

term continuous intraperitoneal insulin infusion with an implanted remote-controlled insu-
lin infusion device. Diabetes 30:1072-1075
Ishida M, Machida Y, Nambu N, Nagai T (1981) New mucosal dosage form ofinsulin. Chem
Pharm Bull (Tokyo) 29:810-816
Jackman WS, Lougheed W, Marliss EB, Zinman B, Albisser AM (1980) For insulin infusion:
A miniature precision peristaltic pump and silicone rubber reservoir. Diabetes Care 3:322-
331
Jackson RL, Storvick WO, Hollinden CS, Stroeh LE, Stilz JG (1972) Neutral regular insulin.
Diabetes 21 :235-245
Jackson RL, Bilginturan N, Murthy DYN, England J (1977) New improved insulins. In: Laron
Z (ed) Pediatric adolescent endocrinology, voi 2. F. Karger, Basel, pp 105-109
James DE, Jenkins AB, Kraegen EW, Chisholm DJ (1981) Insulin precipitation in artificial in-
fusion devices. Diabetologia 21:554-557
Jaques LB (1949) A study of the toxicity ofthe protamine, salmine. Br J PharmacoI4:135-144
Jefferson IG, Marteau TM, Smith MA, Baum JD (1985) A multiple injection regimen using an
insulin injection pen and pre-filled cartridged soluble human insulin in adolescents with di-
abetes. Diabetic Medicine 2:493-497
Jeffrey PD (1974) Polymerization behaviour of bovine zinc-insulin at neutral pH. Molecular
weight ofthe subunit and the effect of glucose. Biochemistry 13:4441-4447
Jeffrey PD, Coates JH (1966a) An equilibrium ultracentrifuge study of the self-association of
bovine insulin. Biochemistry 5:489-498
Jeffrey PD, Coates JH (1966b) An equilibrium ultracentrifuge study of the effect of ionic
strength on the self-association of bovine insulin. Biochemistry 5:3820-3824
Jeffrey PD, Milthorpe BK, Nichol L W (1976) Polymerization pattern of insulin at pH 7.0. Bio-
chemistry 15:4660-4665
Jensen HF (1938) Administration ofinsulin. In: Insulin. Its chemistry and physiology. Common-
wealth Foundation, New York, pp 92-111
Jorgensen KH, Brange J, Hallund 0, Pingel M (1970) A method for the preparation ofessentially
pure insulin. In: Rodrigues RR, Ebling FJG, Henderson 1, Assan R (eds) Seventh Congress
of the International Diabetes Federation. Excerpta Medica Foundation, Amsterdam New
York London Milan Tokyo Buenos Aires (International Congress Series No. 209, Abstracts,
p 149)
Jorgensen KH, Hallund 0, Heding LG, Tronier B, Falholt K, Damgaard U, Thim L, Brange
J (1982) Estimation ofinsulin purity in light of developments in analytical methods. In: Gue-
riguian JL, Bransome ED, Outschoorn AS (eds) Hormone Drugs. United States Pharmaco-
peial Convention, Rockville, Maryland, pp 139-147
Jorpes JE (1949) Recrystallized insulin for diabetic patients with insulin allergy. Arch Intern Med
83:363-371
Kaplan H, Hefford MA, Chan AM-L, Oda G (1984) Chemical reactivity ofthe functional groups
ofinsulin. Concentration-dependence studies. Biochem J 217:135-143
Kasama T, Iwata Y, Okubo T, Sakaguchi Y, Sugiura M (1980) Determination of purity and
identification of animal sources of insulin in various insulin preparations. Jpn J Pharmacol
30:293-300
Kasama T, Iwata Y, Oshiro K, Uchida M, Sakaguchi Y, Namie K, Sugiura M (1981) Antige-
nicity of desamido-insulin and monocomponent insulin. Diabetologia 21 :65-69
Kerchner J, Colaluca DM, Juhl RP (1980) Effect ofwhole blood on insulin adsorption onto in-
travenous infusion systems. Am Jm Hosp Pharm 37:1323-1325
Kern RA, Langner PH (1939) Protamine and allergy. J Am Med Assoc 113:198-200
Kerp L, Steinhilber S, Kasemir H, Han J, Henrichs HR, Geiger R (1974) Changes in immuno-
specificity and biologic activity of bovine insulin due to subsequent removal of the amina
acids B!, B2 and B3' Diabetes 23:651-656
Klostermeyer H, Zahn H (1971) Struktur, Eigenschaften und Synthese des Insulins. In: D6rz-
bach E (Hrsg) Insulin. Handbuch der experimentellen Pharmakologie, Bd XXXII/1. Sprin-
ger, Berlin Heidelberg New York, pp 273-312
Knape JTA, Schuller JL, de Haan P, de Jong AP, Bovill JG (1981) An anaphylactic reaction to
protamine in a patient allergic to fish. Anesthesiology 55:324-325
92 References

Kraegen EW, Lazarus L, Meler H, Campbell L, Chia YO (1975) Carrier solutions for low-Ievel
intravenous insulin infusion. Br Med 1 3:464-466
Krause U, Beyer 1 (1975) Die Reinheit handelsiiblicher Insulinzubereitungen. Dtsch Med Wo-
chenschr 100:238-240
Krayenbiihl C, Poulsen lE (1959) Protamine-zinc-insulin in crystalline suspension. Dan Med
Bull 6:270-272
Krayenbiihl C, Rosenberg T (1946) Crystalline protamine insulin. Rep Steno Mem Hosp Nord
Insulinlab 1:60-73
Kroeff EP, Chance RE (1982) Applications of high performance liquid chromatography for the
analysis of insulins. In: Gueriguian lL, Bransome ED, Outschoorn AS (eds) Hormone
Drugs. United States Pharmacopeial Convention, Rockville, Maryland, pp 148-162
Krogh A, Hemmingsen AM (1928) The destructive action ofheat on insul in solutions. Biochem
122:1231-1238
Kruse V, Heding LG, lorgensen KH, Tronier B, Christensen M, Thim L, Frank BH, Root MA,
Cohen RM, Rubenstein AH (1984) Human proinsulin standards. Diabetologia 27:414-415
Kulpe W (1958) Hautnekrosen bei der Insulinbehandlung durch Surfen-Uberempfindlichkeit.
Muench Med Wochenschr 100:998-999
Kumar D (1979) Immunoreactivity ofinsulin antibodies in insulin-treated diabetics. Significance
of the beta-chain carboxyterminal amin o acid (B-30) of insulin. Diabetes 28:994--1000
Kumar D, Miller LV (1970) Pork insulin resistance treated with dealaninated insulin. Diabetes
(Suppl) 19:392 (abstract)
Kurtz AB, Gray RS, Markanday S, Nabarro lDN (1983) Circulating IgG antibody to protamine
in patients treated with protamine-insulins. Diabetologia 25:322-324
Lacey AH (1967) The unit ofinsulin. Diabetes 16:198-200
Lange RH, BIădorn 1, Magdowski G, Trampisch Hl (1979) Crystalline preparations ofrhom-
bohedral porcine insulin as studied by electron diffraction. 1 Ultrastruct Res 68:81-91
Langkjrer L, Brange 1 (1982) Preloading of insulin syringes. Diabetologia 23:182-183 (ab-
stract)
Lauritzen T, Pramming S (1984) Insulin pumps - still a research tool? Ann Clin Res 16:98-106
Lauritzen T, Deckert T, Frost-Larsen K, Svendsen PAa, Larsen H-W, Christiansen lS, Parving
H-H, Binder C, Nerup 1, Deckert M, Larsen A, Lorup B, Bojsen 1, Beck-lansen L (1982)
One year's experience of insulin pumps in diabetes. Nord Med 97: 130-133
Lauritzen T, Thorsteinsson B, Pramming S, Sorensen L, Binder C (1984) Subcutaneous absorp-
tion ofU-40 and U-I00 insulin. Horm Metabol Res 16:611-612
Lautenschliiger KL, Dărzbach E, Schaumann O (1937) Verfahren zur Herstellung von Priipara-
ten aus dem blutzuckersenkenden Hormon der Bauchspeicheldriise. D R Patent 727888
Lens 1 (1947) The inactivation ofinsulin solutions. 1 Biol Chem 169:313-322
Levandoski LA, White NH, Santiago lV (1982) Localized skin reactions to insulin: Insulin lipo-
dystrophies and skin reactions to pumped subcutaneous insulin therapy. Diabetes Care 5:6-
10
Li CH (1954) Protein hormones. In: Neurath H, Bailey K (eds) The proteins, voI II/A. Acad
Press, New York, p 636
Ling V, lergil B, Dixon GH (1971) The biosynthesis ofprotamine in trout testis. III. Character-
ization of protamine components and their synthesis during testis development. 1 Biol Chem
246:1168-1176
Little lA, Arnott lH (1966) Sulfated insulin in mild, moderate, severe, and insulin-resistant di-
abetes mellitus. Diabetes 15:457-465
Little lA, Lee R, Sebriakova M, Csima A (1977) Insulin antibodies and clinical complications
in diabetics treated for five years with Lente or sulfated insulin. Diabetes 26:980-988
Liversidge GG, Nishihata T, Engle KK, Higuchi T (1985) Effect ofrectal suppository formula-
tion on the release of insulin and on the glucose plasma levels in dogs. Int 1 Pharm 23:87-
95
Lloyd LF (1982) Analysis of insulin preparations by high-performance liquid chromatography.
Anal Proc (London) 19:131-133
Lloyd LF, Corran PH (1982) Analysis of insulin preparations by reversed-phase high-per-
formance liquid chromatography. 1 Chromatogr 240:445-454
Lougheed WD, Woulfe-Flanagan H, Clement lR, Albisser AM (1980) Insulin aggregation in ar-
tifical delivery systems. Diabetologia 19:1-9
References 93

Lougheed WD, Albisser AM, Martindale HM, Chow JC, Clement JR (1983) Physical stability
of insulin formulations. Diabetes 32:424-432
Low BW (1952) Orientation ofthe polypeptide chains in crystals of acid insulin sulphate. Nature
169:955-956
Low BW, Chen CCH (1969) Monoclinic insulin crystals. J MoI BioI43:227-229
Ludvigsson J (1984) Insulin antibodies in diabetic children treated with monocomponent porcine
insulin from the onset: relationship to B-cell function and partial remission. Diabetologia
26:138-141
Luyckx AS, Daubresse J-C, Jaminet C, Scheen A, Lefebvre PJ (1986) Immunogenicity of semi-
synthetic human insulin in man. Long-term comparison with porcine monocomponent insu-
lin. Acta Diabetol Lat 23:101-106
Maislos M, Mead PM, Gaynor DH, Robbins DC (1986) The source ofthe circulating aggregate
of insulin in type I diabetic patients is therapeutic insulin. J Clin Invest 77:717-723
Marcker K (1960) The binding of the "structural" zinc ions in crystalline insulin. Acta Chem
Scand 14:2071-2074
Marks HP (1925) The biological assay ofinsulin preparations in comparison with a stable stan-
dard. Br Med J 11:1102-1104
Markussen J (1980) US Patent 4343898
Markussen J (1982) Human Monocomponent aus Schweine-Rohinsulin. In: Petersen K-G,
Schliiter KJ, Kerp L (Hrsg) Neue Insuline. Freiburger Graphische Betriebe, Freiburg,
pp 38-44
Markussen J, Schaumburg K (1983) Reaction mechanism in trypsin catalyzed synthesis ofhu-
man insulin studied by 170_NMR spectroscopy. In: Peptides 1982. Blaha K, Malon P (eds)
Proceedings 17th European Peptide Symposium, Prague. Walter de Gruyter, Berlin New
York, pp 387-394
Markussen J, J0rgensen KH, Heding LG (1970) Preparation of bovine 1251-tyrosyl-C-peptide.
Horm Metab Res 2:53-55
Markussen J, Damgaard U, J0rgensen KH, Rasmussen E, Snel L, Thim L, Voigt HO (1982) Pro-
duction ofhuman monocomponent insulin. In: Gueriguian JL, Bransome ED, Outschoorn
AS (eds) Hormone Drugs. United States Pharmacopeial Convention, Rockville, Maryland,
pp 116-126
Markussen J, Damgaard U, Diers 1, FiiI N, Hansen MT, Larsen P, Norris F, Norris K, Schou
O, Snel L, Thim L, Voigt HO (1986 a) Biosynthesis ofhuman insulin in yeast via single chain
precursors. Diabetologia 29:568A-569A (abstract)
Markussen J, Damgaard U, Diers 1, FiiI N, Hansen MT, Larsen P, Norris F, Norris K, Schou
O, Snel L, Thim L, Voigt HO (1986 b) Biosynthesis ofhuman insulin in yeast via single-chain
precursors. In: Theodoropoulos (ed) Peptides. 1986: Walter de Gruyter, Berlin (in press)
Marliss EB (1982) Insulin: Sixty years ofuse. N Engl J Med 306:362-364
Melberg SG, Havelund S, Villumsen, J, Brange J (1987) Insulin compatibility with polymer ma-
terials used in external pump infusion systems. Diabetic Medicine. Submitted for publica-
tion
Menczel J, Levy M, Bentwich Z (1966) Insulin resistant diabetes treated with sulphated insulin.
Isr J Med Sci 2:764-768
Miles AA, Mussett MV, Perry WLM (1952) Third international standard for insulin. Bull WHO
7:445-459
Milthorpe BK, Nichol LW, Jeffrey PD (1977) The polymerization pattern ofzinc(U)-insulin at
pH 7.0. Biochim Biophys Acta 495:195-202
Mirsky IA, Kawamura K (1966) Heterogeneity of crystalline insulin. Endocrinology 78:1115-
1119
Mitrano FP, Newton DW (1982) Factors affecting insulin adherence to type I glass bottles. Am
J Hosp Pharm 39:1491-1495
Mizuno N, Ogawa T, Ishida M, Okada K, Shigeaki B (980) Pancreatic polypeptide binding an-
tibodies in insulin-treated diabetics. J Jpn Diabetic Soc 23:219-226
Moloney PJ, Aprile MA, Wilson S (1964) Sulfated insulin for treatment of insulin-resistant dia-
betics. J New Drugs 4:258-263
Moorthy SS, Pond W, Rowland RG (1980) Severe circulatory shock following protamine (an
anaphylactic reaction). Anesth Analg 59:77-78 .
94 References

Moses AC, Oordon OS, Carey MC, Flier JS (1983) Insulin administered intranasally as an in-
sulin-bile salt aerosol. Effectiveness and reproducibility in normal and diabetic subjects. Di-
abetes 32:1040-1047
Nathan DM, Axelrod L, Flier JS, Carr DB (1981) U-500 insulin in the treatment of antibody-
mediated insulin resistance. Ann Int Med 94:653-656
Neubauer HP, Obermeier R, Schnorr O (1984) Immunological properties and biological effec-
tiveness ofinsulin analogues substituted at position B30. Diabetologia 27:129-131
Nishihata T, Rytting JH, Kamada A, Higuchi T, Routh M, Caldwell L (1983) Enhancement of
rectal absorption ofinsulin using salicylates in dogs. J Pharm PharmacoI35:148-151
Nishihata T, Okamura Y, Kamada A, Higuchi T, Yagi T, Kawamori R, Shichiri M (1985) En-
hanced bioavailability of insulin after rectal administration with enamine as adjuvant in de-
pancreatized dogs. J Pharm Pharmacol 37:22-26
Nolte MS, Poon V, Orodsky OM, Forsham PH, Karam JH (1983) Reduced solubility of short-
acting soluble insulins when mixed with longer-acting insulins. Diabetes 32: 1177-1181
Nordstriim L, Fletcher R, Pavek K (1978) Shock of anaphylactoid type induced by protamine:
a continuous cardiorespiratory record. Acta Anaesthesiol Scand 22:195-201
Oberly K, Clark JL, Paulshock BZ (1979) Sterility of insulin stored in syringes. Diabetes Care
2:531
Ohta M, Tokunaga H, Kimura T, Satoh H, Kawamura J (1983) Analysis ofinsulins by high-
performance liquid chromatography. II. Separation of various species of insulins. Chem
Pharm Bull (Tokyo) 31:3566-3570
O'Neill JJ, Mead CA (1982) The parabens: Bacterial adaptation and preservative capacity. J Soc
Cosmet Chem 33:75-84
Oppenheim RC, Stewart NF, Oordon L, Patel HM (1982) The production and evaluation of
orally administered insulin nanopartic1es. Drug Dev Ind Pharm 8:531-546
Ornstein L (1964) Disc electrophoresis-I. Ann NY Acad Sci 121 :321-349
Orr NA, Spence J (1977) Applications of partic1e size analysis in the pharmaceutical industry.
Analyst 102:466-472
Owens DR (1986) Human insulin: Clinical pharmacological studies in normal man. MTP Press
Limited, Lancaster Boston The Hague Dordrecht. Thesis
Owens DR, Jones MK, Hayes TM, Heding LO, Alberti KOMM, Home PD, Burrin JM, New-
combe RO (1981) Human insulin: study of safety and efficacy in man. Br Med J 282:1264-
1266
Owens DR, Jones IR, Birtwell AJ, Burge CTR, Luzio S, Davies CJ, Heyburn P, Heding LO
(1984) Study of porcine and human isophane (NPH) insulins in normal subjects. Diabeto-
logia 26:261-265
Owens DR, Vora JP, Heding LO, Luzio S, Ryder REJ, Atiea J, Hayes TM (1986) Human, por-
cine and bovine ultralente insulin: Subcutaneous administration in normal man. Diabetic
Medicine 3:326-329
Patel YC, Reichlin S (1979) Somatostatin. In: Jaffe BM, Behrman HR (eds) Methods of hor-
mone radioimmunoassay. Academic Press, New York, pp 77-99
Pekar AH, Frank BH (1972) Conformation of proinsulin. A comparison of insulin and proin-
sulin self-association at neutral pH. Biochemistry 11 :4013-4016
Petersen K (1945) Method ofproducing crystalline insulin. US Patent 2626228
Pfeiffer EF, Thum C, Clemens AH (1974) The artificial {J-cell- a continuous control of blood
sugar by external regulation of insulin infusion (glucose controlled insulin infusion system).
Horm Metab Res 6:339-342
Phillips M, Simpson RW, Holman RR, Turner RC (1979) A simple and rational twice daily in-
sulin regime. Q J Med New series 48:493-506
Pickup JC, Keen H, Viberti OC, White MC, Kohner EM, Parsons JA, Alberti KOMM (1980)
Continuous subcutaneous insulin infusion in the treatment of diabetes mellitus. Diabetes
Care 2:290-300
Pietri A, Raskin P (1981) Cutaneous complications of chronic continuous subcutaneous insulin
infusion therapy. Diabetes Care 4:624-626
Pingel M, V0lund Aa (1972) Stability ofinsulin preparations. Diabetes 21:805-813
Pingel M, V0lund Aa, S0rensen E, S0rensen AR (1982) Assessment ofinsulin potency by chemi-
cal and biological methods. In: Oueriguian JL, Bransome ED, Outschoorn AS (eds)
Hormone Drugs. United States Pharmacopeial Convention. Rockville, Maryland, pp 200-207
References 95

Pingel M, V01und Aa, S0renSen E, Collins JE, Dieter CT (1985) Biological potency of porcine,
bovine and human insulins in the rabbit bioassay system. Diabetologia 28:862-869
Pitts JE, Wood SP, Horuk R, Bedarkar S, Blundell TL (1980) Pancreatic hormone storage gran-
ules: The role of metal ions and polypeptide oligomers. In: Brandenburg D, Wollmer A (eds)
Insulin. Chemistry, structure and function of insulin and related hormones. Walther de
Gruyter, Berlin New York, pp 673-682
Pocker Y, Biswas SB (1980) Conformational dynamics of insulin in solution. Circular dichroic
studies. Biochemistry 19:5043-5049
Pocker Y, Biswas SB (1981) Self-association ofinsulin and the role ofhydrophobic bonding: A
thermodynamic model of insulin dimerization. Biochemistry 20:4354-4361
Pongor S, Brownlee M, Cerami A (1983) Preparation ofhigh-potency, non-aggregating insulins
using a novel sulfation procedure. Diabetes 32:1087-1091
Pontiroli AE, Alberetto M, Secchi A, Dossi G, Bosi 1, Pozza G (1982) Insulin given intranasally
induces hypoglycaemia in normal and diabetic subjects. Brit Med J 284:303-306
Poulsen JE, Deckert T (1976) Insulin preparations and the clinical us ofinsulin. Acta Med Scand
(Suppl) 601:197-245
Prestele K, Franetzki M, Kresse H (1980) Development of program-controlled portable insulin
delivery devices. Diabetes Care 3:362-368
Pullen RA, Lindsay DG, Wood SP, Tickle IJ, Blundell TL, Wollmer A, Krail G, Brandenburg
D, Zahn H, Gliemann J, Gammeltoft S (1976) Receptor-binding region of insulin. Nature
259:369-373
Purintun LR, McGrane HF (1972) A survey of sterilization procedures recommended to diabetic
patients. Health Serv Rep 87:357-364
Ramesh V, Bradbury JH (1986) lH NMR studies ofinsulin. Reversible transformation of2-zinc
to 4-zinc insulin hexamer. Int J Pept Protein Res 28:146-153
Rasch R (1979) Control of blood glucose levels in the streptozotocin diabetic rat using a long-
acting heat-treated insulin. Diabetologia 16:185-190
Reiner L, Searle DS, Lang EH (1939) On the hypoglycemic activity of globin insulin. J Pharma-
col Exp Ther 67:330-340
Renneboog-Squilbin C, Delhaise P, Wodak S (1981) The inf1uence of crystal packing on the 3-
dimensional conformation of insulin. Arch Int Physiol Biochim 89:Bl92-B193
Revers RR, Henry R, Schmeiser L, Kolterman O, Cohen R, Rubenstein A, Frank B, Galloway
J, Olefsky JM (1984) Biosynthetic human insulin and proinsulin have additive but not syner-
gistic effects on total body glucose disposal. J Clin Endocrinol Metab 58:1094-1098
Ritschel WA, Ritschel GB (1984) Rectal administration ofinsulin. Methods Find Exp Clin Phar-
macoI6:513-529
Rivier J, McClintock R (1983) Reversed-phase high-performance liquid chromatography of in-
sulins from different species. J Chromatogr 268:112-119
Robbins DC, Shoelson SE, Tager HS, Mead PM, Gaynor DH (1985) Products of therapeutic
insulins in the blood ofinsulin-dependent (type 1) diabetic patients. Diabetes 34:510-519
Rolando RL, Torroba D (1972) Heterogeneity of the fourth international standard for insulin
by gel-chromatography on Sephadex. Experientia 28:1169
Romans RG, Scott DA, Fisher AM (1940) Preparation of crystalline insulin. Ind Eng Chem
32:908-910
Rosenbloom AL (1974) Advances in commercial insulin preparations. Am J Dis Child 128:631-
633
Rosenbloom AL, Londono JH, Jordan J, Rosenbloom EK (1971) U 100 insulin: Trial of Lente
and Regular in 50 children with diabetes. Physicians Drug Man 2:142-144
Rowles B, Sperandio GJ, Shaw SM (1971) Effects of elastomer closures on the sorption of certain
14C-Iabelled drug and preservative combinations. Bull Parenteral Drug Assoc 25:2-22
Sachse G, Federlin K (1981) Sind Des-Phe-Insulin-haltige Insulinmischungen handelsiiblichen
Prăparaten vorzuziehen? Med Klin 76:319-323
Sachse G, Măser E, Federlin K (1982) Kurz- und Langzeitherapie mit Des-Phe-insulinhaltigen
Insulinen. In: Petersen KG, Schliiter KJ, Kerp L (Hrsg) Neue Insuline. Freiburger Graphi-
sche Betriebe, Freiburg, pp 317-324
Saffran M, Kumar GS, Savariar C, Burnham JC, Williams F, Neckers DC (1986) A new ap-
proach to the oral administration of insulin and other peptide drugs. Science 233:1081-
1084
96 References

Sahyun M, Nixon A, Goodell M (1939) Inf1uence of certain metals on the stability of insulin.
J Pharmacol Exp Ther 65:143-149
Salzman R, Manson JE, Griffing GT, Kimmerle R, Ruderman N, McCall A, Stoltz EI. Mullin
C, Small D, Armstrong J, Melby JC (1985) Intranasal aerosolized insulin. Mixed-meal stud-
ies and long-term use in type 1 diabetes. N Engl J Med 312:1078-1084
Samuel T (1977) Antibodies reacting with salmon and human protamines in sera from in fertile
men and from vasectomized men and monkeys. Clin Exp ImmunoI30:181-187
Samuel T, Kolk AHJ, Riimke P (1978) Studies on the immunogenicity ofprotamines in humans
and experimental animals by means of a micro-complement fixation test. Clin Exp Immunol
33:252-260
Sanchez MB, Paolillo M, Chacon RS, Camejo M (1982) Protamine as a cause of generali sed al-
lergic reactions to NPH insulin. Lancet 1:1243
Sanger F (1959) Chemistry ofinsulin. Science 129:1340-1344
Sanger F, Thompson EOP, Kitai R (1955) The amide groups ofinsulin. Biochem J 59:509-518
Santeusanio F, Massi-Benedetti M, Clementi A, Calabrese G, Bueti A, Picchio E, Brunetti P
(1981) Clinical trial with porcine Des-PheBl insulin. A comparative study with unmodified
insulin on therapeutical efficacy, biological activity and immunogenicity. Diabete Metab
7:173-179
Sato S, Jeong SY, McRea JC, Kim SW (1984) Self-regulating insulin delivery systems. II. In vitro
studies. J Controlled Release 1:67-77
Schade DS, Eaton RP (1982) Bactericidal properties of commercial USP formulated insulin. Di-
abetes 31:36-39
Schade DS, Eaton RP, Spencer W (1981) Implantation of an artificial pancreas. Current perspec-
tives. J Am Med Assoc 245:709-710
Schade DS, Eaton RP, Edwards WS, Doberneck RC, Spencer WJ, Carlson GA, Bair RE, Love
JT, Urenda RS, Gaona JI (1982a) A remotely programmable insulin delivery system. Suc-
cessful short-term implantation in man. J Amer Med Assoc 247:1848-1853
Schade DS, Eaton RP, DeLongo J, Saland LC, Ladman AJ, Carlson GA (1982b) Electron mi-
croscopy of insulin precipitates. Diabetes Care 5:25-30
Scheller JC, Galloway JA (1975) The development of the insulin unit. Am J Pharm 147:29-32
Schernthaner G, Borkenstein M, Fink M, Mayr WR, Menzel J, Schober E (1983) Immunogenic-
ity ofhuman monocomponent or pork monocomponent insulin in HLA-DR-typed insulin-
dependent diabetic individuals. Diabetes Care 6:43-48
Schernthaner G, Schober E, Borkenstein M (1984) Insulin-antibody-free state in human-insulin
treated type-I diabetics is associated with increased endogenous insulin production. Dia-
betes 33, Supp11:12A (abstract)
Schildt B, Ahlgren T, Berghem L, Wendet Y (1978) Adsorption ofinsulin by infusion materials.
Acta Anaesthesiol Scand 22:556-562
Schlichtkrull J (1956a) Insulin crystals 1. The minimum mole-fraction of metal in infusion crystals
prepared with Zn++, Cd++, Co++, Ni++, Cu++, Mn++, or Fe++. Acta Chem Scand
10:1455-1458
Schlichtkrull J (1956b) Insulin crystals II. Shape ofrhombohedral zinc-insulin crystals in relation
to species and crystallization media. Acta Chem Scand 10:1459-1464
Schlichtkrull J (1957a) Insulin Crystals III. Determination of the rhombehedral zinc-insulin
unit-cell by combined microscopical and chemical examinations. Acta Chem Scand 11:291-
298
Schlichtkrull J (1957b) Insulin Crystals IV. The Preparation of nuclei, seeds and monodisperse
insulin crystal suspensions. Acta Chem Scand 11:299-302
Schlichtkrull J (1957c) Insulin Crystals V. The nucleation and growth of insulin crystals. Acta
Chem Scand 11 :439-460
Schlichtkrull J (1957d) Insulin Crystals VI. The anisotropic growth of insulin crystals. Acta
Chem Scand 11 :484-486
Schlichtkrull J (1957e) Insulin Crystals VII. The growth ofinsulin crystals. Acta Chem Scand
11 :1248-1256
Schlichtkrull J (1958) Insulin crystals. Chemical and biologic al studies on insulin crystals and in-
sulin zinc suspensions. Thesis, first edn. Ejnar Munksgaard Publisher, Copenhagen
References 97

Schlichtkrull J (1959) New insulin crystal suspensions with various timings ofaction and contain-
ing no added zinc. In: Oberdisse K, Jahnke K (eds) Diabetes Mellitus III. Kongress der In-
ternational Diabetes Federation Dusseldorf. 21-25. Juni 1958, Georg Thieme, Stuttgart,
pp 773-777
Schlichtkrull J (1974) Antigenicity of monocomponent insulins. Lancet 11:1260--1261
Schlichtkrull J (1977a) Purity and antigenicity of insulin preparations. Acta Paediatr Scand
[Suppl] 270:37-40

°
Schlichtkrull J (1977b) The absorption of insulin. Acta Paediatr Scand [Suppl] 270:97-102
Schlichtkrull J, Funder J, Munck (1961) Clinical evaluation of a new insulin preparation. In:
Demole M (ed) 4e Congres de la F€:deration internationale du Diabete, Geneve. Editions
M€:decine et Hygiene, Geneve, 1: pp 303-305
Schlichtkrull J, Munck 0, Jersild M (1965) Insulin Rapitard and Insulin Actrapid. Acta Med
Scand 177:103-113
Schlichtkrull J, Brange J, Ege H, Hallund 0, Heding LG, J0rgensen K, Markussen J, Stahnke
P, Sundby F, V0lund Aa (1970) Proinsulin and related proteins. Diabetologia 6:80--81
(abstract)
Schlichtkrull J, Brange J, Christiansen AaH, Hallund 0, Heding LG, J 0rgensen KH (1972) Clini-
cal aspects of insulin - antigenicity. Diabetes 21 (Suppl 2):649--656
Schlichtkrull J, Brange J, Christiansen AaH, Hallund 0, Heding LG, J0rgensen KH, Rasmussen
SM, S0rensen E, V0lund Aa (1974) Monocomponent insulin and its clinical implications.
Horm Metab Res (Suppl Ser) 5:134--143
Schlichtkrull J, Pingel M, Heding LG, Brange J, J0rgensen KH (1975a) Insulin preparations
with prolonged effect. In: Hasselblatt A, Bruchhausen F von (eds) Handbook of Experimen-
tal Pharmacology, New Series, voI XXXII/2, Springer-Verlag, Berlin Heidelberg New York,
pp 729-777
Schlichtkrull J, Ege H, J 0rgensen KH, Markussen J, Sundby F (1975 b) Die Chemie des Insulins.
In: Oberdisse K (Hrsg) Diabetes mellitus A (Handbuch der Inneren Medizin, Bd 7/2A).
Springer, Berlin Heidelberg New York, pp 77-127
Schmidt DD, Arens A (1968) Proinsulin vom Rind. Isolierung, Eigenschaften und seine Aktivie-
rung durch Trypsin. Hoppe-Seyler's Z Physiol Chem 349:1157-1168
Scott DA (1934) Crystalline insulin. Biochem J 28:1592-1602
Scott DA, Fisher AM (1936) Studies on insulin with protamine. J Pharmacol Exp Ther 58:78-
92
Sefton MV, Antonacci GM (1984) Adsorption isotherms of insulin onto various materials. Di-
abetes 33:674--680
Sestoft L, V0lund Aa, Gammeltoft S, Birch K, Hildebrandt P (1982) The biological properties
of human insulin. Subcutaneous absorption, receptor binding and the clinical effect in dia-
betics assessed by a new statistical method. Acta Med Scand 212:21-28
Sheffer MG, Kaplan H (1979) Unusual chemical properties ofthe amino groups ofinsulin: im-
plications for structure-function relationship. Can J Biochem 57:489-496
Shenfield GM, Hill JC (1982) Infrequent response by diabetic rats to insulin-liposomes. Clin Exp
Pharmacol PhysioI9:355-361
Shichiri M, Okada A, Kikkawa R, Kawamori R, Shigeta Y, Abe H (1971) B-Naphthyl-azo-poly-
styrene-insulin as a means of protecting insulin molecule from digestive enzymes. Biochem
Biophys Res Comm 44:51-56
Shichiri M, Kawamori R, Yoshida M, Etani N, Hoshi M, Izumi K, Shigeta Y, Abe H (1975)
Short-term treatment of alloxan-diabetic rats with intrajejunal administration of water-in-
oil-in-water insulin emulsions. Diabetes 24:971-976
Shichiri M, Kawamori R, Goriya Y, Oji N, Shigeta Y, Abe H (1976) A model for evaluation
ofthe peroral insulin therapy: short-term treatment ofalloxan diabetic rats with oral water-
in-oil-in-water insulin emulsions. Endocrinol Jpn 23:493-498
Shichiri M, Yamasaki Y, Kawamori R, Kikuchi M, Hakui N, Abe H (1978) Increased intestinal
absorption ofinsulin: an insulin suppository. J Pharm PharmacoI30:806-808
Shore RN, Shelley WB, Kyle GC (1975) Chronic urticaria from isophane insulin therapy. Sen-
sitivity associated with noninsulin components in commercial preparations. Arch Dermatol
111 :94--97 .
98 References

Sieber P, Kamber B, Hartmann A, J ahi A, Riniker B, Rittel W (1974) Totalsynthese von H uman-
insulin unter gezielter Bildung der Disulfidbindungen. Helv Chim Acta 57:2617-2621
Sieber P, Kamber B, Hartmann A, Jahl A, Riniker B, Rittel W (1977) Totalsynthese von Human-
insulin IV. Beschreibung der Endstufen. Helv Chim Acta 60:27-37
Simkin RD, Cole SA, Ozawa H, Magdoff-Fairchild B, Eggena P, Rudko A, Low BW (1970) Pre-
cipitation and crystallization of insulin in the presence of lysozyme and salmine. Biochem
Biophys Acta 200:385-394
Slama G, Hautecouverture M, Assan R, Tchobroutsky G (1974) One to five days of continuous
intravenous insulin infusion on seven diabetic patients. Diabetes 23:732-738
Smith DJ, Venable RM, Collins J (1985) Separation and quantitation ofinsulins and related sub-
stances in bulk insulin crystals and in injectables by reversed-phase high performance liquid
chromatography and the effect of temperature on the separation. J Chromatogr Sci 23:81-
88
Smith GD, Swenson DC, Dodson EJ, Dodson GG, Reynolds CD (1984) Structural stability in
the 4-zinc human insulin hexamer. Proc Natl Acad Sci USA 81:7093-7097
Smith KL (1969) Insulin. In: Dorfmann RI (ed) Methods in hormone research voi IIA. Bioassay.
Academic Press, New York, pp 365--414
Steiner DF (1967) Evidence for a precursor in the biosynthesis of insulin. Trans NY Acad Sci
(Ser II) 30:60-68
Steiner DF, Oyer PE (1967) The biosynthesis of insulin and a probable precursor of insulin by
a human islet cell adenoma. Proc Natl Acad Sci USA 57:473--480
Steiner DF, Hallund O, Rubenstein A, Cho S, Bayliss C (1968) Isolation and properties ofproin-
sulin, intermediate forms, and other minor components from crystalline bovine insulin. Di-
abetes 17:725-736
Stewart GA (1974) Historical review of the analytical control of insulin. Analyst 99:913-928
Stewart WJ, McSweeney SM, Mirle BS, Kellett MA, Faxon DP, Ryan TJ (1984) Increased risk
of severe protamine reactions in NPH insulin-dependent diabetics undergoing cardiac cath-
eterization. Circulation 70:788-792
Storm MC, Dunn MF (1985) The Glu(B13) carboxylates ofthe insulin hexamer form a cage for
Cd2+ and Ca2+ ions. Biochemistry 24:1749-1756
Storvick WO, Henry HJ (1968) Effect of storage temperature on stability of commercial insulin
preparations. Diabetes 17:499-502
Strazza S, Hunter R, Walker E, Darnall DW (1985) The thermodynamics of bovine and porcine
insulin and proinsulin association determined by concentration difference spectroscopy.
Arch Biochem Biophys 238:30--42
Strickland EH, Mercola D (1976) Near-ultraviolet tyrosyl circular dichroism of pig insulin
monomers, dimers, and hexamers. Dipole-dipole coupling ca1culations in the monopole ap-
proximation. Biochemistry 15:3875-3884
Sudmeier JL, Bell SJ, Storm MC, Dunn MF (1981) Cadmium-1l3 nuclear magnetic resonance
studies of bovine insulin: Two-zinc insulin hexamer specifically binds calcium. Science
212:560-562
Summerell JM, Osmand A, Smith GH (1965) An equilibrium-dialysis study ofthe binding of zinc
to insulin. Biochem J 95:31P
Sundby F (1962) Separation and characterization of acid-induced insulin transformation prod-
ucts by paper electrophoresis in 7 Murea. J Biol Chem 237:3406-3411
Sutcliffe N, Bristow AF (1984) The proinsulin content of commercial bovine insulin formula-
tions. J Pharm Pharmacol 36:163-166
Szepesi G, Gazdag M (1981) Improved high-performance liquid chromatographic method for
the analysis ofinsulins and related compounds. J Chromatogr 218:597-602
Tammilehto S, Buchi J (1968) Untersuchungen uber die p-Hydroxybenzoesiiureester (Nipagi-
ne®). Pharm Acta Helv 43:726-738
Tanford C, Epstein J (1954) The physical chemistry of insulin. 1. Hydrogen ion titration curve
of zinc-free insulin. J Am Chem Soc 76:2163-2169
Terabe S, Konaka R, Inouye K (1979) Separation of some polypeptide hormones by high-per-
formance liquid chromatography. J Chromatogr 172:163-177
Thim L, Hansen MT, Norris K, Hoegh 1, Boei E, Forstrom J, Ammerer G, Fiii NP (1986) Se-
cretion and processing of insulin precursors in yeast. Proc Natl Acad Sci USA 83:6766-
6770
References 99

Thompson EOP, O'Donnel! IJ (1960) The chromatography of insulin on DEAE-cel!u!ose in

buffers containing 8 Murea. Aust J Bio! Sci 13:393-400
Thurow H (1980) Studies on the denaturation of disso!ved insulin. In: Brandenburg D, Wol!mer
A (eds) Insulin chemistry; ·structure and function ofinsulin and re!ated hormones; ·Walther
de Gruyter, Berlin New York, pp 215-221
Thurow H, Geisen K (1984) Stabilisation of disso!ved proteins against denaturation at hydro-
phobic interfaces. Diabeto!ogia 27:212-218
Tjioe TO, Wacker A (1972) Reinheitspriifung von im Hande! befindlichen Insulinpriiparaten mit
Hilfe der diskontinuierlichen Po!yacry!amidge!-Elektrophorese. Klin Wochenschr 50:882-
884
Touitou E, Donbrow M, Rubenstein A (1980) Effective intestina! absorption of insulin in dia-
betic rats using a new formu!ation approach. J Pharm Pharmaco!32:108-110
Trag! KH, Pohl A, Kinast H (1979) Zur perora!en Verabreichung von Insulin mitte!s Liposomen
im Tierversuch. Wien Klin Wochenschr 91:448-451
Trevan JW, Boock E (1962) The standardization ofinsulin by the determination ofthe convul-
sive dose for mice. League of Nations Health III 7:45-56
Tronier B, Larsen UD (1982) Somatostatin-like immunoreactivity in man. Measurement in pe-
ripheral plasma. Diabete Metab 8:35-40
Umber F, Storring FK, Fol!mer W (1938) Erfolge mit einem neuartigen Depotinsulin ohne Prot-
aminzusatz (Surfen-Insulin). Klin Wochenschr 17:443-446
Unger RH, Aquilar-Parada E, Miil!er WA, Eisentraut AM (1970) Studies of pancreatic alpha
cel! function in normal and diabetic subjects. J Clin Invest 49:837-848
United States Pharmacopeia (1980) XX The United States Pharmacopeial Convention, Inc,
Rockville. Mack Printing Company, Easton, p 405
Valdenaire K, Klein W (1979) Behandlung von Diabetikern mit Insulinresistenz und -al!ergie mit
Des-Phe-Insulin. Dtsch Med Wochenschr 104:1637-1639
Velcovsky HG, Federlin KF (1984) Experiences in immunological aspects ofhuman insulin. In:
Diabetes mellitus: achievements and scepticism. (Royal Society of Medicine International
Congress and Symposium Series; no. 77.) Oxford University Press, Oxford, pp 73-81
Vestermark S (1982) Human-Monokomponent(HM)-Insuline in klinischer routinemiiBiger An-
wendung. Aktuel Endokrino! StoffwechseI3:114
Vigneaud V du, Geiling EMK, Eddy CA (1928) Studies on crystalline insulin VI. Further con-
tributions to the question whether or not crystalline insulin is an adsorption product. J Phar-
macol Exp Ther 33:497-509
Villalpando S, Drash A (1979) Circulating glucagon antibodies in children who have insulin-de-
pendent diabetes mellitus. Clinical significance and characterization. Diabetes 28:294--299
V0lund Aa, Pingel M, S0rensen E (1982) Differential potency of pork and beef insulins in the
USP rabbit bioassay system. In: Gueriguian JL, Bransome ED, Outschoorn AS (eds)
Hormone Drugs. United States Pharmacopeial Convention. Rockvil!e, Maryland, pp 208-215
Vontz FK, Puestow EC, Cahil! DJ (1982) Anaphylactic shock following protamine administra-
tion. Am Surg 48:549-551
Wallhiiusser KH (1974) Antimicrobial preservatives in biologics. Pharm Ind 36:716-722
Walters DP, Smith PA, Marteau TM, Brimble A, Borthwick LJ (1985) Experience with
NovoPen, an injection device using cartridged insulin, for diabetic patients. Diabetic
Medicine 2:490-497
Ward GM, Simpson RW, Ward EA, Tumer RC (1981) Comparison oftwo twice-daily insulin
regimens: Ultralente/Soluble and Soluble/Isophane. Diabetologia 21:383-386
Watkins JD, Roberts DE, Williams TF, Martin DA, Coyle V (1967) Observation ofmedication
errors made by diabetic patients in the home. Diabetes 16:882-885
Waugh DF (1946) A fibrous modification ofinsulin 1. The heat precipitate ofinsulin. J Am Chem
Soc 68:247-250
Waugh DF (1948) Regeneration ofinsulin from insulin fibrils by the action ofalkali. J Am Chem
Soc 70:1850-1857
Waugh DF (1954) Protein-protein interactions. Adv Protein Chem 9:325-437
Waugh DF (1957) A mechanism for the formation offibrils from protein molecules. J Cel! Comp
Physiol (SuppI1) 49:145-164
Waugh DF, Wilhelmson DF, Commerford SL, Sack!er ML (1953) Studies ofthe nuc!eation and
growth reactions of selected types of insulin fibrils. J Am Chem Soc 75:2592-2600
100 References

Weiler JM, Freiman P, Sharath MD, Metzger WJ, Smith JM, Richerson HB, Ballas ZK, Hal-
verson PC, Shulan DJ, Matsuo S, Wilson RL (1985) Serious adverse reactions to protamine
sulfate: Are alternatives needed? J Allergy Clin Immunol 75:297-303
Weitzel G, Fretzdorff A-M, Strecker F-J, Roester U (1953) Zinkgehalt und Glukagoneffekt kri-
stallisierter Insulinprăparate. Hoppe-Seyler's Z Physiol Chem 293:190-215
Welinder BS (1980) Gel permeation chromatography of insulin. J Liq Chromatogr 3:1399-
1416
Welinder BS (1984) Homogeneity of crystalline insulin estimated by GPC and reversed phase
HPLC. In: Hancock WS (ed) CRC Handbook of HPLC for the separation of amino acids,
peptides, and proteins, voi II. CRC Press, Inc, Boca Raton, pp 413-419
Welinder BS, S0rensen HH, Hansen B (1986) Reversed-phase high-performance liquid chroma-
tography ofinsulin. Resolution and recovery in relation to column geometry and buffer com-
ponents. J Chromatogr 361:357-367
WHO (1980) World Health Organisation. Expert Committee on Diabetes Mellitus. Second Re-
port. Technical Report Series 646, WHO, Geneva, pp 62-63
WHO (1982) World Health Organisation. Expert Committee on Biological Standardization.
Technical Report Series 673:29
Wigley FM, Londono JH, Wood SH, Shipp JC, Waldman RH (1971) Insulin across respiratory
mucosae by aerosol delivery. Diabetes 20:552-556
Williamson KL, Williams RJP (1979) Conformational analysis by nuclear magnetic resonance:
Insulin. Biochemistry 18:5966-5972
Wintersteiner 0, Vigneaud V du, Jensen H (1928) Studies on crystalline insulin. V. The distribu-
tion of nitrogen in crystalline insulin. J Pharmacol Exp Ther 32:397-411
Yamasaki Y, Shichiri M, Kawamori R, Morishima T, Hakui N, Vagi T, Abe H (1981 a) The ef-
fect of rectal administration of insulin on the short-term treatment of alloxan-diabetic dogs.
Can J Physiol PharmacoI59:1-6
Yamasaki Y, Shichiri M, Kawamori R, Kikuchi M, Vagi T, Arai S, Tohdo R, Hakui N, Oji N,
Abe H (1981 b) The effectiveness of rectal administration of insulin suppository on normal
and diabetic subjects. Diabetes Care 4:454-458
Yip CC, Logothetopoulus J (1969) A specific anti-proinsulin serum and the presence of proin-
sulin in calf serum. Proc Nat Acad Sci 62:415-419
Yoshida H, Okumura K, Hori R (1979) Absorption of insulin delivered to rabbit trachea using
aerosol dosage form. J Pharm Sci 68:670-671
Zeuzem S, Taylor R, Agius L, Schoeffling K, Albisser AM, Alberti KGMM (1985) Biological
effects ofsulphated insulin in adipocytes and hepatocytes. Moi Cell Biochem 68:161-168
Zoltobrocki M, Enzmann F, Vanderbeke 0, Federlin K, Laube H, Klein W, Valdenaire K, Sauer
H, Riedesel G, Sch5ffling K, Neubauer M, Willms B, Ahlhausen M (1980) Erste kontrollier-
te Multizenterstudie zur Erfassung des Wirkprofils ei ner neuartigen Insulinzubereitung mit
Des-Phe-Insulin (Hoe 03 R=Optisulin Depot). Akt Endokrin 1:41-51
Subject Index

a-component 5, 6, 7, 14, 61 continuous subcutaneous infusion 65-67

absorption - see timing of action covalent dimerization 7,12,29,55,59-60,67
acid formulation 30-31, 34, 54 covalent polymerization 55-57, 59-60
Actrapid 30, 32, 37, 45, 48, 49-50, 55-56, covalent protamine-insulin product 57, 60,
58-61, 63, 67, 80 61
additives - see auxiliary substances cresol - see phenol
administration, alternative 70-73 crystal, size distribution 44
adsorption 51-52 crystal structure 23-25
aggregation - see non-covalent polymerization crystallization 3--4, 21, 23-27
alkaline medium 20 crystals of insulin 2-3, 21, 23-27
allergic reaction 5, 14--15, 34--35, 39, 67
amorphous insulin 26-27, 35, 36, 38, 44, 66 deamidated insulin - see deamidation
analogue - see insulin derivatives deamidation 5,8-9,11-12,19,30,55-56,
analytical control 43--46 59-60,61
antibody formation 14--15,30,35,67,80-80 delivery systems 65-67
antigenicity 70, see also immunogenicity density 31
antimicrobial agent - see preservative derivation - see insulin derivatives
aprotinin 10, 42 des-alanine insulin 70
arginine insulin 5, 7, 12 des-phenylalanine insulin 33, 36, 70
association 2, 20--22 desamido insulin - see deamidation
association constants 22 detergent 20 - see also surfactant
auxiliary substances 40--42 dilution 50
dimer (non-convertible) - see covalent
b-component 5, 6, 7-8, 56, 61 dimerization
bile salts 71 dimer (non-covalent association) 20-22
bioassay 12-13 disc electrophoresis 5, 8-9, 54, 55
biological potency 12-13,46, 58-60
biological stability 58-60 enzymes - see proteolytic enzymes
biosynthetic production 76-80 ethylester insulin 5, 9, 12
biphasic insulin 33, 38-39, 40, 45, 53
bovine insulin 1,11-12,13,18-19,25,26, fibrils 52, 66-69
36,39 formulation
buffer 3, 30, 42 - see also phosphate and protracted preparations 33
citrate regular insulin 30
freezing 52-53, 62
c-component 7-8
ca1cium ions 20,21,27,68 gel filtration 5-6, 7-8, 56-57
charge of insulin 2 GLI (glucagon-like immunoreactivity)
chemical reactions 27, 54--57 10-11
chemical stability 54--57 glob in insulin 33, 40
chromatography 5, 6, 76, 77 glucagon 3, 10, 14
citrate 3, 36 glycerol 30, 40, 42, 55-56, 68
composition 30 - see also formulation
concentration - see strength heat, exposure to 52-53
contaminants 5, 10--11, 14 hexamer 20--22, 56, 68-69
102 Subject Index

hormonal contaminants - see contaminants monoarginyl-insulin - see arginine insulin

HPLC (high performance liquid monocomponent (MC) insulin 5,11,14-15,
chromatography) 8,11-12,79-80 67, 76-77
human insulin 11-12, 13, 14-15,25, 36, 58, monodesamido insulin - see deamidation
75-81 monomer insulin 20, 67
hydrolysis 42, 55-56, 75 Monotard 28,33,36-37,45--46,48,49-50,
53-57, 58, 63, 64, 80
immunogenicity 5, 14-15, 34-35, 60-61,
69-70, 80-81 nas al administration 71
infusion pumps 65-67 neutral insulin solution - see regular insulin
Initard 33, 39 non-covalent polymerization 67 - see also
insulin antibodies 14-15 - see also antibody fibrils
formation and immunogenicity non-parenteral routes of administration -
insulin, bovine - see bovine insulin see administration, alternative
insulin crystals - see crystals NovoPen 65
insulin derivatives 28-29,69-70 - see also NPH - see isophane insulin
chemical stability
insulin, dissolution of 20 oral administration 71-72
insulin fibrils - see fibrils ordinary insulin - see regular insulin
insulin hexamer - see hexamer
insulin, human - see human insulin pancreatic enzymes - see proteolytic enzymes
insulin injection - see regular insulin para ben - see methylparaben
insulin mixtures 48-50 pH 18-21,27,30,42--48, 50, 53-54
insulin, polymerization products - see covalent phenol 8,26,30,35-36,37,40--41,53,55-56
polymerization phosphate 26, 34, 35, 36, 42, 48
insulin, porcine - see porcine insulin physical stability 38, 48-50, 52-54, 65-69
insulin in solution 18-20, 29-32 physicochemical properties 1-3, 18-27
insulin zinc suspensions 33,36-38,40,41,52, PLI (proinsulin-like immunoreactivity)
64 10-11
intermediate-acting insulin 33, 38-39, 40, 48 polyacrylamide disc gel electrophoresis - see
see also protracted preparations disc electrophoresis
international standard - see standard of polymerization - see (a) covalent polymeriza-
insulin tion, (b) non-covalent polymerization
iridine 34 porcine insulin 1, 11, 12, 13, 14-15, 18-20,
isoelectric point 2, 18, 34 25, 31, 36, 39, 58, 69, 75-76
isoelectric precipitation 31, 66 potency - see biological potency
isophane insulin (NPH) 28, 33, 35-36, 39, PP (pancreatic polypeptide) 10-11, 14
48-50, 52-53, 55, 60, 61, 64, 80 precipitation 2, 21 - see also isoelectric
isophane ratio 35-36, 44 precipitation
isotonic agent 30, 40, 42, 43 preloading of syringes 63-64
prcservative 8, 19-20, 30, 40--42, 43 -
Lentard - see Lente see also phenol and methylparaben
Lente 28,33,36-38,41,42,44,48,50,52-53, primary structure 1
58-59,64 production of insulin 1-3 - see also
long-acting insulin 33, 40, 48 - see also manufacture
protracted preparations proinsulin 5, 7, 9, 10, 12, 19, 70, 79
prolongation mechanism 27-29
manufacture, prolongation test 45--46
Lente insulins 37-38 protamine 23, 28, 33, 34-36, 44, 49, 57
regular insulin 31 protamine insulins 23, 28, 34-36, 43-44, 49
MC-insulin - see monocomponent insulin protamine zinc insulin (PZI) 33, 35, 56
metal-ion binding 22-23, 23-27 proteolytic enzymes 2, 3, 32, 36, 44
methylparaben 8, 30, 40--42, 49, 53, 55-56 protracted preparations 32--40, 63
mixing 48-52, 63 pump delivery 65-67
Mixtard 33, 39 purification 5-6
molecular weight 1 purity 4,7-12,67, 78-80
mono-pic insulin 6 PZI - see protamine zinc insulin
Subject Index 103

radioimmunoassay (RIA) 9-11 storage 60-62

rapid-acting insulin 29-32, 48, 50, 52 - strength of insulin 37, 46-48, 49
see also regular insulin structure - see primary structure and tertiary
Rapitard 27, 28, 33, 39, 45, 48-49, 55, 59 structure
rectal administration 72 sulphated insulin 68, 70
regular insulin 29-32, 54, 56, 63, 67 - sunlight 60, 62
see also Actrapid surfactants 68,71-72 - see also detergent
retarded insulin - see protracted preparations, surfen insulin 33, 39
intermediate-acting and long-acting insulin
retarding principle 33 - see also tertiary structure 20-22, 25
prolongation mechanism three-dimensional structure - see tertiary
rhombohedra 23-26 structure
RI insulin 6 timing of action (protracted preparations)
RIA - see radioimmunoassay 33, 40, 45, 48
timing of action (regular insulin) 31-32,
salmine 34, 36 48-50
sedimentation 52
Semilente 26, 28, 33, 36-38, 49, 52, 59-60 Ultralente 27,28, 33, 36-38, 44, 48, 56,
short-acting insulin 33, 40 - see also 59-60, 80
Semilente Ultratard - see Ultralente
single component (SC) insulin 5 use of insulin 60-64
single peak (SP) insulin 6
size exc1usion chromatography - VIP (vasoactive intestinal polypeptide)
see gel filtration 10-11,14
solubility 2, 18-20,27,31, 39 viscosi ty 31
somatostatin 10-11
stability - see biological stability, chemical yield 4,75-76, 77-78
stability and physical stability
standard of insulin 12, 46 Zinc ions 2-3, 18-28, 30-32, 35-38, 42-44,
standardization 12-14 48-50, 66, 69