21

Use of Functionalized
Poly(Ethylene Glycol)s for
Modification of Polypeptides
SAMUEL ZALIPSKY and CHYI LEE

21.1. INTRODUCTION

The unique properties of poly(ethylene glycol) (or PEG) and its general
compatibility with polypeptide materials facilitated development of a variety of
different applications of this polymer. 1- ll A marked proportion of these applications
involve the use of covalently linked polypeptide-PEG adducts (reviewed else-
where 5- 11 ). For example, a number of PEG-enzymes were shown to be useful as
catalysts, soluble and active in organic solvents. 7 Due to the affinity to the upper
phase of PEG/Dextran and PEG/salt two-phase systems, PEG-modified proteins were
proven useful both as diagnostic tools 8 and in preparative separations of biological
cells. 9
Unquestionably, the properties of polymer-polypeptide conjugates in vivo were
a significant reason for the substantial amount of research reported in the area during
the last two decades. It has been repeatedly demonstrated that covalent attachment of
multiple strands of PEG to proteins produces conjugates with dramatically reduced
immunogenicity and antigenicity. IO Such preparations also show great resistance to
proteolytic digestion, and remain present in the bloodstream a considerably longer
time than the parent polypeptides. These beneficial properties of modified polypep-
tides allowed development of a number of PEG-modified therapeutic proteins. IO
Slight to moderate modification of potent allergenic proteins with PEG can often be

SAMUEL ZALIPSKY and CHYI LEE· Enzon, Inc., South Plainfield, New Jersey 07080. Present
address for s. Z.: Department of Chemistry, Rutgers-The State University of New Jersey, Piscataway, New
Jersey 08855.

347
J. M. Harris (ed.), Poly(Ethylene Glycol) Chemistry
© Springer Science+Business Media New York 1992
348 Samuel Zalipsky and Chyi Lee

sufficient to convert them into tolerance inducers and/or substantially reduce their
allergenicity.l1
Although, in some instances, hydroxyl end groups of the polymer can be used
directly for covalent attachment of molecules of biological relevance,5,12-14 in most
cases suitable functionalization of the polymer prior to the conjugation is essential. 12
Here we will review the methods for PEG functionalization and its covalent
attachment to polypeptides. Attention will be given to comparison of different
methods for preparation of PEG conjugates, properties of the linkages between the
polymer and peptide components and their influence on biological/enzymatic activ-
ities of the conjugates. This review will not deal with formation of PEG-peptide
conjugates that are built by stepwise addition of amino acid residues. 5 However, some
of the methods for interlinking peptides and PEG, though originally devised for
chemical and/or enzymatic protocols for peptide synthesis, are of general appli-
cability and will be induded in our discussion.
First, we will briefly go over the relevant properties of PEG itself, so that the
rationale for its extensive use as a carrier for biological molecules will become
apparent.

21.2. PROPERTIES OF PEG

Poly(ethylene glycol) is a polyether-diol with the general structure HO-

(CH 2CH 20)n-H. For the purpose of modification of peptides and proteins the useful
molecular weight range is 2,000-20,000 daltons.
The polymer has a wide range of solubilities. It is soluble in most organic
solvents as well as in aqueous solutions. 5 For example, it is soluble in both water and
dichloromethane, to such an extent that very concentrated solutions (> 50%) can be
prepared. Most polypeptides, as a result of their conjugation with PEG, in addition to
retaining and in some cases enhancing their water solubility, also acquire solubility in
some organic solvents. 7
It was shown that ethylene oxide-based oligomers (MW < 400) can be oxidized
in vivo into toxic diacid and hydroxy acid metabolites L1fough a process initiated by
alcohol dehydrogenase l5 ; however, toxicity of PEGs of molecular weights above 1000
daltons is very low. 16 Extensive toxicity studies on PEG-4000 showed that this
polymer can be safely administered intravenously in 10% solution to rats, guinea
pigs, rabbits, and monkeys at a dose level of 16 g per kg body weight. 17 This
corresponded to at least 1000-fold the amount of the polymer present in PEG-purified
factor-VIII doses normally administered in humans. It was also reported that when
administered intravenously to humans, PEGs of molecular weight 1000 and 6000 are
readily excreted mainly via the kidney. IS Biological activities of PEG conjugates are
typically dominated by the non-PEG part of a molecule.l 9
The polymer by itself, even with a molecular weight as high as 5.9 X 106, is a
very poor immunogen. PEG-recognizing antibodies (specific to 6-7 oxyethylene
units) could be generated in rabbits by administration of PEG-modified allergenic
proteins together with Freund's complete adjuvant. 2o In the absence of Freund's
Functionalized PEGs and Polypeptides 349

complete adjuvant, PEG-modified proteins do not induce formation of immunoglob-

ulins against the polymer, and also reduce the response against antigenic determi-
nants of a protein molecule. lO ,21
The polyether backbone of PEG is not degradable by mammalian enzymes.
However, several types of bacteria can readily degrade the polymer up to Mn =
20,000. The topic of polyether biodegradation was recently reviewed by Kawai. 22
The presence of PEG in protein solutions, even at high concentrations, in free
or in conjugated form does not have any adverse effect on protein molecules. 2 The
observation that a single point attachment of PEG to a peptide does not restrict the
access of enzymes to a modified peptide residue was made by a number of
investigators,23-25 and was used for preparation of enzymatically altered polymer-
peptide adducts. Generally, the conformation of peptides or proteins does not change
as a result of covalent conjugation with PEG. 6,26
It was recognized in one of the first papers dealing with PEG-proteins 27 that the
key property of PEG as a protein modifier is its ability to bind molecules of water.
It was also suggested l that the ability of the polymer to influence a structure of several
molecular layers of water might be one of the causes of its appreciable exclusion
effect, which might explain such phenomena as: (i) the ability of PEG to act as a
protein precipitating agent,2 (ii) repulsion of proteins from PEG-modified surfaces,4
and (iii) reduced immunogenicity and antigenicity of PEG-protein conjugates. 10,21,27

21.3. METHODS FOR COVALENT ATTACHMENT OF PEG TO

POLYPEPTIDES

21.3.1. PEG Derivatives Reactive toward Amino Groups

Covalent attachment of PEG derivatives in the vast majority of cases has been
achieved utilizing amino groups of polypeptide molecules as sites of modification.
The first step in this process is substitution of the hydroxyl end-groups of the polymer
by electrophile-containing functional groups. This process is often referred to as
.. activation. " Derivatives of monofunctional polymers, capped on one end by methyl
ethers (mPEG), were usually the reagents of choice for protein modifications. I (}-'12
Such modifications are expected to be free from crosslinking and result in the
attachment of multiple strands of the polymer to the globular polypeptide core. A
number of popular ways to activate the polymer are shown in Scheme 1.

21.3.1.1. Most Commonly Used PEG-Based Reagents

In the original work of Davis and co-workers 27 trichloro-s-triazine (cyanuric

chloride) was reacted with the primary alcohol groups on PEG so that only one of the
chlorides is displaced from the triazine ring (1) and the remaining chlorides can be
used for subsequent reaction with the amino groups of a protein. This approach was
adopted by other investigators8 ,9,28-3o and is still one of the most popular. The
synthesis of 1 was recently optimized to assure the reproducible and complete
350 Samuel Zalipsky and Chyi Lee

o
O-PEG
N..J.. N PEG-o)lN:J

CIJlN~O-PEG 4

CI
N~N
~/, PEG-OH - - - - . PEG-o)lo-ONO,
ClN~O-PEG

Scheme 1. Commonly used methods for activation of polyethylene glycol. For the purpose of this and the
following schemes, the abbreviation PEG refers to the structures: RO-(CH2CH20)n-CH2CH2- or
-(CH2CH20)n-CH2CH2-' where R is a simple alkyl residue, usually CH 3 .

conversion of mPEG-OH to 1 as was proven by titration of chloride, !3C-NMR, GPC,

and elemental analysis.3! Although this approach provides a very effective one-step
activation of the polymer, it suffers from various disadvantages such as toxicity of
cyanuric chloride derivatives and excessive reactivity of 1 toward nucleophilic
functional groups other than amines (e.g., cysteinyl and tyrosyl residues). In fact,
cyanuric halides are considered to be the least suitable reagents for selective protein
modification. 32 Consequently, modification of some proteins with 1 is often accom-
panied by a substantial loss of biological activity.29.33-35
A variation of the cyanuric chloride method in which two of the three chlorides
originally present on the trichloro-s-triazine molecule are displaced by mPEG chains
(2), and the remaining chloride used for the reaction with amino groups of a protein,
was developed by Inada and co-workers. 7 Unfortunately, the communications de-
scribing the synthesis of 2 failed to provide proof of the structure of this activated
polymer. 36-38 A recent report from the laboratory of Sehon39 convincingly demon-
strated lack of reactivity of the third CI group on the triazine ring toward ethanolamine
and ovalbumin, which places serious doubt as to whether 2,4-bis-(mPEG-O-)6-
chioro- s-triazine (2) is capable of reacting with proteins. In light of this conflicting
evidence, the issue of reactivity of 2 clearly needs to be clarified. It is interesting to
mention that reactivity toward sulfhydryl groups reported for 1 by Wieder et al. 33 was
also found for reagent 2.40
The shortcomings of cyanuric chloride activation were overcome by using PEG-
succinimidyl succinate (SS-PEG, 3a). This form of activated PEG was first used for
Functionalized PEGs and Polypeptides 351

crosslinking of proteins4! and synthesis of peptides substituted by PEG chains at their

terminals42 but was later adapted for preparation of mPEG-protein conjugates. 43 The
reagent (38) is usually prepared by succinylation of the terminal hydroxyl groups of
PEG!4 followed by dicyclohexylcarbodiimide-mediated condensation with N-hydroxy-
succinimide. 42 ,43 It reacts with proteins within a short period of time under mild
conditions (30 min, pH < 7.8, 25°C), producing extensively modified proteins with
well-preserved biological activities.26,4~5 The ester linkage between the polymer
and the succinic ester residue has limited stability in aqueous media,1l,25,46 which
causes slow hydrolytic cleavage of PEG chains from SS-PEG-modified proteins
under physiological conditions. Substitution of the succinate residue in 38 by
glutarate (3b)47 produced an activated form of the polymer very similar to SS-PEG,
but with slightly improved resistance to hydrolysis of the ester linkage. Replacement
of the aliphatic ester in 38 by an amide bond (3c) improves the stability of the reagent
and its conjugates even further. 46 ,48 However, preparation of 3c, starting from
hydroxyl terminated mPEG, involves 4 to 5 synthetic steps. 48
Formation of urethane (carbamate) linkages between the amino groups of a
protein and PEG overcomes the problem of hydrolytic release of the polymer chains.
In fact, it was demonstrated on radioactively labeled PEG derivatives that urethane
linkages are completely stable under a variety of physiological conditions. 49 The
attachment of mPEG chains to polypeptides through carbamate linkages was first
accomplished using imidazolyl formate derivatives of the polymer (4).50 Several
versions of a one-step synthesis of 4 using carbonyldiimidazole were reported. 35 ,5!
Complete transformation of the hydroxyl groups of PEG was confirmed by both
elemental analysis and NMR data. The polymer activated in this manner has a rather
mild reactivity, and therefore long reaction times were required for protein modifica-
tion (48-72 h, pH 8.5). However, good preservation of activity in these modified
enzymes was usually observed. 35 ,50,52,53
The products of protein modification using phenylcarbonates of PEG (5 and 6)
also have PEG chains grafted onto the polypeptide backbone through urethane
linkages. 54 Single-step activation protocols with commercially available chlorofor-
mates of 4-nitrophenol and 2,4,5-trichlorophenol were used for the preparation of 5
and 6, respectively. These activated PEGs seem to react faster than 4. However, both
4-nitrophenol and 2,4,5-trichlorophenol produced during the PEG- attachment pro-
cess are toxic and hydrophobic molecules with affinities toward proteins.
Succinimidyl carbonates of PEG (SC-PEG, 7), developed in our laboratory,
constitute a further improvement in urethane-forming PEG reagents. 55 ,56 SC-PEG
and its bifunctional analogs (BSC-PEG) of different molecular weights were pre-
pared in a one-pot procedure as shown by Scheme 2. Polymeric chloroformate,
generated in situ by treatment of PEG with phosgene, was reacted with N-hydroxy-
succinimide (HOSu). Purified preparations of the SC-activated polymers were deter-
mined to contain the theoretical amounts of active carbonate groups. To estimate the
reactivity of SC-PEG and to compare it to SS-PEG, measurements of hydrolysis and
aminolysis rates of both activated polymers derived from mPEG-5000 were per-
formed. The results of these experiments showed that SC-PEG is slightly less
352 Samuel Zalipsky and Chyi Lee

R = -CH 3 0rH

1. Phosgene

2. HOSu, TEA

Scheme 2. Preparation of succinimidyl carbonate derivatives of PEG. Abbreviations: HOSu, N-hydroxy-

succinimide; TEA, triethylamine.

reactive yet a more selective reagent. Protein modification reactions with 7 can be
performed within short periods of time (= 30 min) over a broad pH range, with the
highest reactivity at pH = 9.3. The HOSu released during polypeptide modification
is a nontoxic material that is often used in bioconjugate and peptide chemistries as
a leaving group residue. Our experience was that, unlike 4-nitrophenol and 2,4,5-
trichlorophenol mentioned above, HOSu does not show affinity toward proteins and
can be readily removed from the reaction solution by either dialysis or diafiltration.
Under appropriate conditions BSC-PEG can be used as a homobifunctional cross-
linker of proteins. It is also useful as a reactive macromonomer in polycondensation
with diamines. 57
The activated polymer 8 was prepared by carbodiimide-mediated esterification
of carboxymethylated PEG with HOSu. 48 ,58,59 Carboxymethylation of PEG is one
of the most straightforward ways to introduce carboxylic acid end groups onto the
polymer. This is best accomplished by nucleophilic displacement of the bromide in
ethyl bromoacetate with a PEG alkoxide, followed by saponification of the es-
ter. 23 ,58-60 An alternative way to essentially the same derivative is by oxidation of the
terminal hydroxyls of PEG,48,61 but this process is often accompanied by polyether
backbone degradation. A number of enzymes were modified using 8 with very good
preservation of specific activity.48.62,63 It was also used for modification of human
hemoglobin. 46 •59 The reactivity of 8 was reported to be one order of magnitude higher
than that of 3, as was estimated by hydrolysis (t1/2 = 1 min at pH 7.5, 27°C) and
aminolysis rates of the activated mPEGS.64 The increased reactivity was explained
by the presence of electron-withdrawing oxygen in one carbon-proximity to the
carbonyl of the active ester in 8. This result, combined with our own data (on SS- and
SC-PEG)55 and other literature sources, 54 allows us to estimate the order of reactivity
of active acyl-bearing PEGs (assuming equality of all other variables, such as the
molecular weight of the reagents) shown on Scheme 1:
8 > > 38 = 3c > 3b ;a. 7 > 5 > 6 >4
Functionalized PEGs and Polypeptides 353

One has to bear in mind that the higher the reactivity of a reagent, the less likely it is to
be selective, and consequently the higher is the probability of side reactions.

21.3.1.2. Less Commonly Used PEG Reagents for Modification of Amino

Groups

In some instances another activated form of the polymer, PEG-carboxymethyl

azide (9), was generated in situ from PEG-carboxymethyl hydrazide and then reacted
with proteins according to the reaction shown in Scheme 3.21,41,65 Normally, the slow
reacting acyl azide functionality gets an additional boost of reactivity due to the
stronger electrophilicity of the carboxymethyl residue.

~/NH-NH2 HCI/NaN02 ~/N3

PEG-O If PEG-O If
o in situ o
PEG - carboxymethyl hydrazide

Scheme 3. Modification of polypeptides using in situ generated acyl azide.

Garman 66 reported synthesis of a polymeric analog of dimethylmaleic an-

hydride (10) suitable for PEG attachment to amino groups of polypeptides through a
linkage which is slowly erodable under physiological conditions. The validity of this
approach was demonstrated on plasminogen activator proteins. The PEG-plasminogen
activator conjugates obtained by this method yielded active, completely deacylated
proteins after 44 h incubation at pH 7.4, 37°C.
o o
y<o )(0
BrJL.-..(o mPEG-O 0
10

Polyethylene glycols can be quantitatively converted into dithiocarbonate (xan-

thate) derivatives,67,68 11, which were shown to be useful for grafting PEG chains
onto proteins via thionourethane linkages (Scheme 4). The composition of PEGoxy-
thiocarbonylated proteins could be conveniently determined spectrophotometric ally
from the increase in absorption at 242 nm as compared with that of native protein. The
conjugates showed good chemical stability in a variety of aqueous buffers. Because of
the mild reactivity of 11, the modification reactions with bovine serum albumin and
354 Samuel Zalipsky and Chyi Lee

PEG-OH
2.CS,
3. CICH,cONH,
11

o
S~
[PEG-O.t NH-] kProtein I N-R
S-{
o
Dts=N-R

Scheme 4. Preparation and use of PEGoxy-S-carboxamidomethyldithiocarbonate. R represents a peptide

or an amino acid residue.

ragweed antigen E were carried out at pH range 9-10 for = 20 h. (All the common
free amino acids reacted quite readily with 11 within 30 min.) The xanthate 11 derived
from bifunctional PEG-2000 proved useful as a polymeric reagent for the introduc-
tion of dithiasuccinoyl (Dts) protecting group into amino acids and dipeptides
following the sequence of reactions shown on Scheme 4. 68
Esters of organic sulfonic acids and PEG, p-toluenesulfonates (tosylates) in
particular, have been useful as starting materials for the preparation of a variety of
functionalized polymers. 68 ,69 Tresylates (2,2,2-trifluoroethanesulfonates) of PEG
(U, PEG-OS02CH 2CF3) were shown to be sufficiently reactive toward amino groups
(= l00-fold more reactive than tosylates) to be considered useful as protein modify-
ing reagents. 70 Modification of bovine serum albumin (BSA) with mPEG-tresylate in
16-fold excess to the amino groups, performed at pH 7.5, was complete in approx-
imately 1 h and resulted in attachment of mPEG to 18 amino groups per protein
molecule. 71 The comparison of these results to our own data on BSA modifica-
tions 55 ,56 using 3 or 7 leads us to believe that PEG-tresylate is the less reactive
reagent. As a result of protein modification with PEG- tresylate, the polymer chains
become grafted onto the polypeptide through very stable secondary amine bonds.
There are two important consequences to this: (i) the total charge of the protein does
not change in the process of modification; (ii) the modified proteins could be
conveniently characterized by quantitation of lysine by amino acid analysis. The
chemical composition of such conjugates can also be determined by the fluo-
rescamine assay, which specifically measures primary amines and is not interfered
with by the presence of secondary amino groups.72 The assay based on the use of
trinitrobenzene sulfonate (TNBS), which was extensively used for determination of
the extent of modification in PEG-proteins,73 theoretically should not be suitable in
the case of PEG-tresylate-modified proteins. TNBS reacts with both primary and
secondary amines as well as with other nucleophiles. 32 In light of this, it is surprising
Functionalized PEGs and Polypeptides 355

that the extent of modification of alkaline phosphatase modified with 12 was deter-
mined by TNBS assay. 35
Although amino acid analysis has been used routinely for characterization of
PEG-peptides,5 it has only recently been recognized as a powerful analytical tool for
characterization of PEG-proteins.28,74,75 A single amino acid analysis run of a
protein conjugate can provide several valuable pieces of information: (i) protein
concentration, (ii) detection of possible side reactions that may occur during protein
modification, and (iii) determination of the conjugate composition, provided that the
linker is designed for this purpose. For example, the polymeric active ester 13

o
[ mPEG-O-
I/"..
......, f(,NletLYS-OSU mPEG-O A
R
N Ay0su
o 2 H 0
13 14 R - side-chain residue of an amino acid

composed of two mPEG-5000 chains and containing reference amino acid norleucine
(Nle) was specifically designed for convenient determination of the amount of PEG
in its conjugates by amino acid analysis. 74 The amount of Nle determined in hydro-
lysates of proteins modified with 13 provides an accurate measure of the extent of
modification. A similar rationale was used with regard to activated PEGs of general
structure 14.75 Synthesis of PEGoxycarbonyl-amino acids, needed for preparation
of 14, can be readily accomplished by using active carbonates 5-775 or by reacting
isocyanato derivatives of amino acids with hydroxyls of PEG-OH.68,76
Imidoesters of mPEG (15) useful for protein modification were recently de-
scribed in the patent literature,77 These activated polymers were prepared by acid-
catalyzed methanolysis of mPEG-cyanolkylethers and reacted with amino groups of
proteins at pH 7-9 (Scheme 5). Making use of the fact that imidyllinkages survive the
conditions of acidic protein hydrolysis, the extent of mPEG-amidination was deter-
mined by diminished lysine content in the conjugates. The advantage of this approach
to protein modification is that the amidinated proteins possess the same net charge
as the native ones.

NH·HCI

mpEG-O(CH2)n~OCH3
MeOH
HCI

15

Protein (NH 2 )m

Scheme S. Preparation and use of mPEG-imidoesters.

356 Samuel ZaJipsky and Chyi Lee

d:?S~XPEG
x = 0, NH or Nle-O
16 0

21.3.2. PEG Reagents Reactive toward Arginine Residues

Polymeric derivatives 16 for selective attachment of PEG to arginine residues
were introduced to serve as carriers during semi synthesis of peptides in aqueous
solutions. 78 The reagents were readily obtainable by reaction of camphorquinone-lO-
sulfonylchloride with PEG-OH, PEG-NH 2 , or preferably with the norleucine ester
of the polymer. The modified peptides were formed in borate buffer at pH 9.0, at
37°C. They were stable to a variety of acidic and nucleophilic reagents, including
hydroxylamine, the recommended agent for cleavage of cyclohexadione-arginine
adducts. Active peptides could be released from the polymeric carrier by o-phenyl-
enediamine. When norleucine was incorporated as a spacer between the camphor-
quinone moiety and the polymer, the extent of arginine modification was conveniently
determined by amino acid analysis of hydrolysates of the PEG-peptides.
A variety of proteins were modified with a mPEG-5000 analog of phenylglyoxal
(17),79 a well-known reagent for modification of guanidino groups. The synthesis of
17 was performed according to the reactions depicted on Scheme 6. Readilyobtain-
able mPEG-tosylate 68 ,69 was subjected to nucleophilic displacement by 4-hydroxy-
acetophenone, which was oxidized by selenium dioxide to yield mPEG-phenylglyoxai.
Attachment of 17 to proteins through arginyl residues proceeded at pH range 5.5-
9.3, at room temperature, and was measured by the decrease of arginine content in
hydrolysates of modified polypeptides. The conjugates of 17 were claimed to possess
very good stability and biological activity.

mPEG-0-S02 O C H 3 4-hydro<yace,ophenone mpEG_0~CH3

\d''o

o
mpEG-OO~H
17

Scheme 6. Preparation of mPEG-phenylglyoxal.

21.3.3. PEG Reagents for Modification of Cysteine Residues

Maleimide-containing reagents are effective modifiers of free cysteinyl residues
of polypeptides. 32 Preparation of PEG derivatives (18) bearing maleimido functional
Functionalized PEGs and Polypeptides 357

groups was reported using readily accessible l4 ,58 PEG-NH 2 and active esters of
6-maleimidohexanoic8o and 3-maleimidopropionic8l acids as starting materials
(Scheme 7). Direct conversion of amino-PEG into a maleimido-PEG using maleic
anhydride was also described. 2l A mutant protein of interleukin-2, in which cysteine
at position - 3 replaced the naturally occurring glycosylated threonine residue, was
coupled with 18 (n = 5) to produce a well defined, fully bioactive PEG-
polypeptide. 80 The reagent 18 (n = 2) was used for selective entrapment of free thiol-
containing peptides on the polymer, in order to simplify purification of unsymmetri-
cal cystine-peptides from a reaction mixture. 8l

Scheme 7. Preparation and use of PEG-maleimides.

Glass and co-workers reported preparation of 4-phenoxy-3,5-dinitrobenzoyIPEG

(19).24 This PEG derivative reacted rapidly and selectively at neutral pH with

N0 2 °
o-0-og.O.PEG
N02 19 N02 °
Peptide·SH peptide.s-o-g.o.PEG
Thiolysis N02

sulfhydryl groups of peptides to yield polymer-peptide adducts in which the

components were linked by a thiol-sensitive dinitrophenylene linker. Thus it was
possible to attach a peptide to the polymer and, after performing some chemical and!
or enzymatic alterations on the conjugate, release the derivatized peptide from the
PEG carrier.

21.3.4. Coupling of PEG Derivatives to Carboxylic Groups of Polypeptides

Amino-PEG was reacted with l-ethyl-3-(3-dimethylaminopropyl)carbodiimide-
activated carboxyl groups of trypsin and other proteins. 2l The selectivity of such
reaction is rather poor, due to the fact that amino-PEG has similar reactivity to the
358 Samuel Zalipsky and Chyi Lee

lysiyl residues of proteins. In another variation of this reaction p-aminobenzylether of

PEG was coupled to carboxyl groups of D-glucose 6-phosphate dehydrogenase by
treatment with l-ethyl-3-(3-dimethylaminopropyl)carbodiimide. 82 The crude conju-
gate retained 22% of the original enzymatic activity. The aromatic amino groups on
the polymer have a lower pKa than the amino groups of lysine residues and therefore
would be expected to exhibit higher reactivity toward activated carboxyls under the
acidic conditions of such reactions (pH 4.8-6.0). In both types of modification
involving aliphatic and aromatic amino-PEGs, the composition of the conjugates was
not determined.
In a recent report83 bilirubin oxidase was modified by a two-step protocol. First,
1,4-diaminobutane was coupled to the carboxyl groups using water-soluble carbo-
diimide. Then, both the newly introduced amino groups and the original amines
of the protein were modified with reagent 2. The activity of PEG-bilirubin oxidase
was approximately 30% that ofthe native enzyme. Unfortunately, the composition of
the conjugate was not reported. When using carbodiimide-mediated coupling of an
appropriate nucleophile to a protein, one has to be aware of possible side reactions,
such as modification oftyrosyl and cysteinyl residues and the formation of N-acylurea
derivatives. 32

21.3.5. Coupling of PEG to Oligosaccharide Residue of Glycoproteins

We found only one reported case of this type of protein modification. In this
example,84 the specific reactivity of a carbohydrate residue of horseradish peroxidase
was used for anchoring PEG chains. Mild oxidation of the carbohydrate portion of the
enzyme molecule with sodium periodate resulted in the formation of six aldehyde
groups. Bifunctional amino-PEG of molecular weight 20,000 was reacted with the
oxidized peroxidase to form Schiff base links which were reduced by sodium
borohydride. The conjugate produced retained 91 % of the original specific activity
and contained on average three PEG chains per glycoprotein molecule. In addition to
water, it was soluble and active in tetrachloromethane, toluene, chloroform, and
dimethylformamide.

21.4. RELATIONSHIP BETWEEN COUPLING CHEMISTRY AND

BIOWGICAL ACTIVITY OF PEG-POLYPEPTIDE
CONJUGATES

Only a few attempts have been made to address the issue of the interrelationship
between the chemistry of conjugation or activation of PEG and the biological activity
of a particular PEG-polypeptide conjugate. 35 ,45,53,67,85 Therefore, in order to over-
view this topic we had to compare data published by a number of different research
groups. We compiled the characteristics of selected PEG-protein conjugates in
Table 1, showing the different functionalized polymers that were used to modify each
Functionalized PEGs and Polypeptides 359

protein. While the data in Table I are presented as reported in the primary sources, the
methods used by the various authors to determine the chemical composition of the
modified proteins were often different. For example, several variations of TNBS
assay, often used to determine the extent of modification of proteins, had been
employed by the researchers. While most authors used the procedure of Habeeb, 73 which
measures the total number of amino groups on a protein, in some cases50 ,52,54,62,87
other versions of the TNBS assay, yielding a measure of the readily accessible amino
groups only, were employed. The reader also has to be aware that in many cases
conditions for modification reactions and design of biological/enzymatic assays also
differed from one laboratory to another. Despite the above-mentioned drawbacks,
several conclusions can be drawn from the data summarized in Table I, which also
illustrates some of the interesting properties and applications of PEG-polypeptide
conjugates.
Proteins modified with cyanuric chloride activated PEGs (1 and 2) almost always
possessed lower enzymatic activity than the same enzymes modified using alternative
chemistries. This is most likely due to excessive reactivity and thus lack of selectivity
of these reagents, which results in modification of nucleophilic groups other than
amines. The pattern of asparaginase inactivation as a result of exposure to 1 or 2
provides the clearest illustration of this phenomenon and is consistent with the known
reactivity of cyanuryl halides toward tyrosyl residues. 32 Inactivation of asparaginase
by tyrosyl-modifying reagents is well documented. 102
King et al. reported that PEG conjugates of Antigen E obtained via use of 1 had
approximately one order of magnitude lower antigenic activity compared to those
conjugates derived from 11,28,67 even though about the same numbers of amino
groups were modified with both reagents. One might speculate that attachment of
mPEG chains to nucleophilic residues other than lysyls, in the case of reagent 1, could
be the reason for the substantiitI difference in the antigenic activities. The lower
antigenicity could actually be advantageous for potential therapeutic use in allergic
patients, since greater amounts of the modified allergen may be used safely.
Yoshinaga and Harris 35 examined the activity of PEG-alkaline phosphatase
conjugates obtained by four different coupling methods. The best preservation of
activity was observed when SS-PEG (3a) was used as a modifying reagent, and only
1 caused substantial loss of enzymatic activity. The same research group examined
alkaline phosphatase activity of conjugates obtained using cyanuric chloride-
activated mPEGs and a number of its bifunctional analogs. 86 Interestingly, in contrast
to the pattern observed with cyanuric chloride-activated mPEGs, protein modifica-
tions with the bifunctional PEGs resulted in better preservation of enzymatic activity
which was independent of the extent of the modification or the molecular weight of the
polymeric reagent used. It is pertinent to note that alkaline phosphatase in its active
form is known to be a dimeric enzyme ,103 present in eqUilibrium with the only slightly
active monomeric form. Therefore, the reported improved preservation of enzymatic
activity could be attributed to partial fixation of the active dimeric form of the enzyme
by the crosslinking of two monomers of alkaline phosphatase by the activated
 ~

Table 1. Chemical and Biological Characteristics of Selected PEG-Protein Conjugates

No. of PEGs % Native
Active per protein enzyme
PEG (% modified activity
Protein (MW)a amino groups) (substrate) Excerpts/applications Ref.

Adenosine 1 9(40) 28 No reaction between PEG-ADA and the antibodies raised against native ADA. 34
deaminase PEG-ADA may be suitable for treating human ADA deficiency because of
(ADA) long circulating life and the lack of detectable antibody formation in mice.
3a PEG-ADA was used successfully for the treatment of two children suffering 44
from adenosine deaminase deficiency. Neither toxic effects nor hypersen-
sitivity reactions were observed.
4 19(858) 76 Long circulating half-life and significant retention of enzymatic activity in mice. 52
However, no reduction of the immunogenecity was reported.
7 14(65 e ) 51 Used as a model for evaluation of the activated PEG. 56
Alkaline 1 19(88) 33 The reagent 1 caused significant loss of enzymatic acttvlty. Best enzymatic 35
phosphatase 14(62) 67 activities were obtained in conjugates derived from 3a.
3a 17(79) 93 rIl
DO
13(61) 98 a
12 17(77) 86 !!.
=
16(73) 82 N
4 17(78) 77
12(56) 91
~
rIJ

5(23)-18(82)
~
1 66-44 Modification with higher molecular weight mPEG gave more deactivation than 86 DO
l(mI900) 1(5)-19(86) 97-55 did modification with the lower molecular weight mPEG. Modification with Q.
=
1(4000) 9(41)-17(77) 72 bifunctional PEG gave highly active protein conjugates and there was little ("J

1(8000) 9(40)-16(72) 70 dependence on molecular weight or degree of modification. -!

1(20000) 12(54)-18.5(84) 80
i
Antigen E 1 81(44) I.()h PEG-antigen E with reduced allergenic activities, yet retaining the immunogenic 28 ~
1(m2000) 7/(39) 2.4h properties of antigen E. Substantially lower antigenicity was observed for ~
conjugates obtained with reagent I. 67 8=-
11 5.6'(31) 25 h !.
7.8;(43) 17h ...r
Q.
11 (m2000) 8.0'(44) ISh '"C
tol
c;':)
Arginase 6 55(60)8 90 The PEG-enzyme had an increased structural stability, a decreased digestion by 87
proteolytic enzymes, and an expanded clearance time in rats. It is potentially
'"
II>
Q.
=
useful for the therapy of arginine-dependent tumors or of familial hyper-
argininemia. ~
'<
49(53) 65 Antisera from mice injected with native arginase reacted against arginase but not 88
against PEG-arginase. In addition, antisera from animals injected with PEG-
~
arginase reacted with neither native arginase nor PEG-arginase.
s:.
7 56(61) 71 Used as a model for evaluation of the activated PEG. 56 '"
Asparaginase 1 73(79) 7(Asn) The modification of asparaginase with PEG(m5000) showed the reduction of the 29
(E. coli) 15(ANA) antigenicity and had a resistivity against trypsin. Interestingly, the aspar-
l(ml900) 70(76) 14(Asn) aginase modified with PEG of 750 and 1900 daltons did not show a substantial
25(ANA) change of the immunogenic properties.
l(m750) 77(84) 12(Asn)
20(ANA)
2 52(57) 8(Asn) The PEG-asparaginase conjugate has the same enzymatic properties (Km value 89
and optimal pH) as the native enzyme. The half-lives of the modified and
native enzymes in mice were 56 and 2.9 h, respectively.
2 30-48 30-11 PEG-asparaginase conjugate retained II % of the enzymatic activity but showed 36
(33)-(52) (Asn) no binding activity toward anti-asparaginase serum.
3a 59(64) 52(Asn) PEG-asparaginase conjugates were active, stable, without immunosuppressive 43
effect, and with extended half-lives in mice.
7 62(68') 54(Asn) Used as a model for evaluation of the activated PEG. 56
8'(4500) 8(Asn) The modified asparaginase contained magnetite, which was used for facile 90
removal of PEG-asparaginase from aqueous solution.
(continued)
~
c:I'I
...
 ~
~
N

Table 1. (Continued)

No. of PEGs % Native

Active per protein enzyme
PEG (% modified activity
Protein (MW)a amino groups) (substrate) Excerpts/applications Ref.
Chymotrypsin 1 16(95) 5(ATEE) PEG-chymotrypsin retained catalytic activity and had increased solubility in 85
O(GFNA) organic solvents. It is useful for improving the coupling yield of peptide
13(75) 25(ATEE) fragments and avoiding the risk of racemization.
IO(GFNA)
5 9(55) 35(ATEE)
25(GFNA)
8 13(75) 50(ATEE)
30(GFNA)
5(m2000) 8.5-17 75-69 The PEG-chymotrypsin was found to be soluble in benzene and DMF and 91
(50-1()()g) (BTNA) catalyzed transesterification in cyclohexane. The enzymatic activity in organic
6(m2000) 2-7 77-64 solvents was decreasing with increasing the extent of modification and ranged
(1l-4OK) (BTNA) 34-171 % of the native activity.
2 12.5(83) 57(ATEE) The yields of ester and amide formations both were 90% in 1.1, I-trichloroethane 92
and the reaction rate was linearly enhanced with increasing amount of the 38
PEG-chymotrypsin conjugate. ~
1 8(45) 6O(BTNA) Substrate specificity of PEG-chymotrypsin in organic solvents was altered since 30
..,
!3
arginine and lysine esters were found to be as effective as substrates as =
~
derivatives of aromatic amino acids. N
7 9(54') I3l(BTEE) Used as a model for evaluation of the activated PEG. 56 !.
.;'
15 I (BTNA) ~
14(82') I 22(BTEE) '<
..,
161(BTNA) 5.
n
Gulonolactone 38 18(47') 74 The modified enzyme was found to be more stable at 37°C than native enzyme. 45 =-
':l.
oxidase The PEG-enzymes retained immunogenicity and reacted with preformed I:""
1 15(38') 67 antibodies. The circulating half-life of the modified enzyme was not extended. ~
Lipase 2 (49) 43 b The PEG-lipase catalyzed ester synthesis in organic solvents. The highest 93 ~
(P seudOl'l'lOTUls activity was observed in I, I , I-trichloroethane. =
tragi) 8c(45OO) 59-15 b The modified enzyme contained magnetite, which was used for convenient 90 te..
removal of PEG-lipase from reaction mixtures. The preparation was used for
N'
ester synthesis. to
Q,
(Pseudomonas 2 (55) 8()b PEG-lipase catalyzed ester synthesis, transesterification and aminolysis reac- 94
jluorescens) tions in organic solvents.
~
~
CIl
2 (52) 67 d PEG-lipase was used in organic solvents with indoxyl acetate as a substrate; it 95 10
was possible to determine the Michaelis-Menten constants for water. S.
2 (60) 70 b Two types of lipase were modified with two different reagents (2 & 8). In contrast 96 l
to the case of Pseudomonasjluorescens, the enzyme from Candida cylindracea '<
(Candida 8(m45OO) (47) 561> when modified with 8 catalyzed ester synthesis form short-chain alcohols and
1
'C
cylindracea) 0.- or l3-substituted carboxylic acids in benzene. E
to
CIl
8(m45OO) (95) 68 b The PEG-lipase catalyzed ester exchange reaction between dipalmitoyl phospha- 63
tidylcholine and eicosapentaeuoic acid.
Superoxide 4 18 or 19 >95 SOD coupled to PEG increased its plasma half-life from 3.5 minutes to 9 or more 50
dismutase (90 or 95) hours depending on the PEG derivative studied.
(SOD) 8 3(15)-18(90) 90-72 SOD-PEG conjugate showed longer half-life in rats than native SOD. 48
(bovine 3c 3(13)-18(90) 90-70
erythrocyte)
1 12(60)-14(70) 100 The PEG-modified enzyme increased cellular enzyme activities and provided 97
3a 100 prolonged protection from partially reduced oxygen species.
1 19(95) 51 No evidence of an immune response to repetitive injections of PEG-SOD was 98
observed.
6 1O(5()g) 80 The PEG-phenylcarbonate derivatives were stable in neutral aqueous solution 54
and were reactive enough to modify proteins extensively in reasonable time
5 periods.
8 16(828) 75 No structural modification occurred at the metal active site region and, in fact, the 62
metal binding was higher in PEG-SOD than native SOD. The biological life of
PEG-SOD decreased in the order i.v.> i.p.> i.m.> s.c.
3a 11 (57) 47 PEG-modified SOD resulted in high heterogeneity and substantial changes in 99
isoelectric point and hydrophobicity.
~
~
(continued) ~
 t

Table 1. (Continued)
No. of PEGs % Native
Active per protein enzyme
PEG (% modified activity
Protein (MW)a amino groups) (substrate) Excerpts/applications Ref.
Superoxide dismutase
(SOD) (cont.)
(Human) 17 2.1 or 2.5 Arginine residues of SOD were selectively modified. 79 r.n
(Serratia) 2 5(24) 52 The PEG (m5000) modified SOD showed enhanced anti-inflammatory activities 100
2(ml900) 10(48) 41 and radioprotective effects in mice.
2(m750) 10(48) 41
i
2(m350) 10(48) 39
Tissue 4 10(44) 80 The reaction of rt-PA with activated-PEG 4 was much slower than the reaction 53 i
plasminogen 13(60) 30 with activated-PEG Ja. The conjugate of PEG-rt-PA has a potential to be used
activator 16.5(75) 20 as thrombolytic agent in human. ~
t"l
3a 12(55) 36 <3,
14.5(66) 14
22(100) 0 ~
 10 13(59) 50-70 Reversible conjugation oft-PA has been achieved with the PEG-contianing maleic 66
anhydride reagent. l
17 8(36) or 9(41) Arginine residues of t-PA were selectively modified. 79
Trypsin 1 4(24) 95(BAEE) Proteolytic activity of the conjugate was markedly reduced. PEG-trypsin conju- 101
9(59) 150(BAEE) gate (59% modified) dissolved soft blood clots at one-fourth the rate of trypsin. I
7 7(46 e ) 95(BAEE) Used as a model for evaluation of the activated PEG. 56 ~
224(ZAPA) ~
12(78 e ) 92(BAEE)
326(ZAPA) i
8 12.5(83) 110(BAEE) Used as a model for evaluation of the activated PEG. 48 l
Abbreviations: ANA, Aspartic acid j3-p-nitroanilide; Asn, Asparagine; ATEE, Acetyl tyrosine ethyl ester; BTEE, Na-benzoyl-L-tyrosine ethyl ester; BTNA, Na-benzoyl-L-tyrosine-p-
nitroanilide; GFNA, Glutaryl phenyl alanine p-nitroanilide; ZAPA, Na-benzyloxycarbonyl-L-arginine-p-nitroanilide; rt-PA, Recombinant tissure plasminogen activator; Lv., intravenous;
Lp., intraperitoneal; Lm., intramuscular; s.c., subcutaneous.
"Molecular weight given only for PEG derivatives different from mPEG-5000. Letter m appearing in parentheses prior to the number indicates mPEG derivative.
bHydrolysis of olive oil in emulsified aqueous system.
i
<Bifunctional PEG-4500 was activated in presence of magnetite.
dlndoxyl acetate hydrolysis in emulsified aqueous solution.
eFluorescamine assay (Ref. 72).
f Amino acid analysis.
RTNBS assay: version measuring only the readily accessible amino groups.
h% Antigenic activity of antigen E.
'The number of lysine residues modified were determined by a spectrophotometrical method and amino acid analysis.

~
366 Samuel Zalipsky and Chyi Lee

bifunctional PEGs. Thus, the degree of this intermonomeric crosslinking, and not the
number of modified amino groups, could have been the dominant factor determining
enzymatic activity of a given preparation.
Superoxide dismutase (SOD) has been modified with PEG in a number of
laboratories. With one exception,99 the attachment of PEG through amide (reagents
3, 8) or urethane (reagents 4, 6) linkages caused only minimal inactivation of the
enzyme. In the work of McGoff et al. ,99 which is the exception, the modified and the
native enzymes were from two different sources. Consequently, the comparison of the
two can hardly be valid. Interestingly, while Pyatak et al. 98 reported a 50% loss of
SOD activity after modification with 1, Beckman et al. 97 observed complete preser-
vation of enzymatic activity of PEG-SOD derived from the same reagent. Unfor-
tunately, no activity was reported for SOD derivatives obtained by attachment oft7 to
arginyl residues of the protein. 79
Preparation of functionally active, yet extensively modified, PEG conjugates
derived from proteins having large-size substrates proved more difficult than with
enzymes acting on low molecular weight substrates. For example, several PEG-tissue
plasminogen activator (tPA) conjugates were prepared using succinimidyl succinate
(3a) and imidazolyl formate (4) mPEGs as modifying reagents. 53 Preference was
given to mPEG-imidazolyl formate-derived conjugates, due to their somewhat
higher fibrinolytic activities. Regardless of the activated PEG employed, the activity
of the conjugates decreased with increased extent of modification. Similarly to the
case of tPA, proteolytic activity of PEG-modified trypsin was also drastically
reduced lOl in contrast to the well preserved and in some cases even enhanced activity
towards low molecular weight substrates (Table 1). Other proteolytic enzymes have
shown similar behavior. Using far-ultraviolet circular dichroism and intrinsic protein
fluorescence, Pasta et al. 26 showed that the serine protease, subtilisin, modified with
3a maintains its native secondary structure and thus the integrity of its catalytic site.
It is generally believed that steric hindrance is responsible for the diminished
proteolytic activity of PEG-modified proteases. We believe that the well-documented
ability of PEG to exclude proteins from its environmentl -4 is also partially responsible
for this phenomenon.
From the examples shown in Table I the choice of the "best" performing
activated PEG is not obvious. Overall, the acylating reagents (3-8) performed
comparably well. In some cases the ease of preparation and shelf-life of reagents are
very important considerations. 55 Based on these two criteria the urethane-forming
functionalized PEGs (4-7) are clearly superior.

21.S. FUTURE PERSPECTIVES

Activity in the area of PEG-modified polypeptides has increased over the last
two decades, as evidenced by growing numbers of research groups that have joined
this field, as well as by the total number of relevant publications and patents. We
expect this trend to continue during the nineties. It is clear that recombinant proteins,
Functionalized PEGs and Polypeptides 367

which have become more available and in many cases have potential for therapeutic
use, can benefit from the increased stability, resistance to proteolysis, and extended
plasma lifetime that conjugation with PEG is almost certain to provide. More
sophisticated PEG-based reagents, that modify selective sites or residues of polypep-
tide molecules, will certainly emerge, accompanied by the development of the new
analytical methods for the characterization of PEG-polypeptide conjugates. Superior
quality of commercially available PEGs and their functionalized derivatives are
already being developed by a number of companies. For example, the undesirable
presence of bifunctional PEG contaminants in mPEG preparations 104 will have to be
dramatically reduced to minimize the possibilities for crosslinking and heterogeneity
of PEG-protein preparations. Some recently developed methods for selective intro-
duction of one reactive functional group per polymer chain60 as well as for synthesis
of heterobifunctional PEG derivatives using readily available PEG-diols as starting
materials 76 might minimize such complications and facilitate a more controlled and
rationale design of PEG-polypeptide conjugates and their applications. 105

REFERENCES
I. I. N. Topchieva, Usp. Khim. 49, 494; Russ. Chem. Rev. 49, 260 (1980).
2. K. C. Ingham, Methods Enzymol. 104, 351 (1984).
3. H. Walter and G. Johansson, Anal. Biochem. 155, 215 (1986).
4. E. Merrill and E. W. Salzman, Am. Soc. Artif. Internal Organs 1. 6, 60 (1983).
5. M. Mutter and E. Bayer, in: The Peptides (E. Gross and 1. Meienhofer, eds.), Vol. 2, p. 285,
Academic Press, New York (1979).
6. V N. R. Pillai and M. Mutter, Acc. Chem. Res. 14, 122 (1981).
7. Y. Inada, K. Takahashi, T. Yoshimoto, A. Ajima, A. Matsushima, and Y. Saito, Trends in
Biotechnol. 4, 190 (1986).
8. B. Mattiason, Methods Enzymol. 92,498 (1983).
9. D. E. Brooks, K. A. Sharp, and S. 1. Stocks, Makromol. Chem., Macromol. Symp. 17, 387 (1988).
10. A. Abuchowski and F. F. Davis, in: Enzymes as Drugs (1. Holsenberg and 1. Roberts, eds.), p. 367,
Wiley, New York (1981).
II. S. Dreborg and E. B. Akerblom, Crit. Rev. Therap. Drug Carrier Syst. 6, 315 (1990).
12. 1. M. Harris, 1. Macromol. Sci., Rev. Macromol. Chem. Phys. C25, 325 (1985).
13. S. Zalipsky, C. Gilon, and A. Zilkha, 1. Macromol. Sci. Chem. A21, 839 (1984).
14. S. Zalipsky, C. Gilon, and A. Zilkha, Eur. Polym. 1. 19, 1177 (1983).
15. D. A. Herold, K. Keil, and D. E. Bruns, Biochem. Pharmacol. 38, 73 (1989).
16. H. F. Smyth, Jr., C. P. Carpenter, and C. S. Weil, 1. Am. Pharm. Assoc. 39, 349 (1950).
17. A. 1. Johnson, M. H. Karpatkin, and 1. Newman, Br. 1. Hematol. 21, 21 (1971).
18. c. B. Shaffer and F. H. Critchfield, 1. Am. Pharm. Assoc. 36, 152 (1947).
19. E. C. Dittmann, Naunyn-Schmeideberg's Arch. Pharmacol. 276, 199 (1973).
20. A. W. Richter and E. Akerblom, Int. Arch. Allergy Appl. Immunol. 70, 124 (1983).
21. F. F. Davis, T. Van Es, and N. C. Palczuk, U.S. Patent 4,179,337 (1979).
22. F. Kawai, CRC Crit. Rev. Biotechnol. 6, 273 (1987).
23. G. P. Royer and G. M. Ananthramaiah, 1. Am. Chem. Soc. 101, 3394 (1979).
24. 1. D. Glass, L. Silver, 1. Sondheimer, C. S. Pande, and 1. Coderre, Biopolymers 18, 383 (1979).
25. K. Ulbrich, 1. Strohalm, and 1. Kopecek, Makromol. Chem. 187, 1131 (1986).
26. P. Pasta, S. Riva, and G. Carrea, FEBS Lett. 236, 329 (1988).
27. A. Abuchowski, T. Van Es, N. C. Palczuk, and F. F. Davis, 1. Bioi. Chem. 252, 3578 (1977).
368 Samuel Zalipsky and Chyi Lee

28. T. P. King, L. Kochoumian, and L. M. Lichtenstein, Arch. Biochem. Biophys. 178, 442 (1977);
T. P. King, L. Kochoumian, and N. Chiorazzi, J. Exp. Med. 149,424 (1979).
29. Y. Ashihara, T. Kono, S. Yamazaki, and Y. Inada, Biochem. Biophys. Res. Commun. 93, 385 (1978).
30. H. F. Gaertner and A. 1. Puigserver, Proteins 3, 130 (1988).
31. S. G. Shafer and 1. M. Harris,1. Polym. Sci., Polym. Chem. Ed. 24, 375 (1986).
32. M. Z. Atassi, in: Immunochemistry of Proteins (M. Z. Atassi, ed.), Vol. 1, p. 1, Plenum Press, New
York (1977).
33. K. 1. Wieder, N. C. Palczuk, T. Van Es, and F. F. Davis,1. Bioi. Chem. 254, 12579 (1979).
34. S. Davis, A. Abuchowski, Y. K. Park, and F. F. Davis, Clin. Exp. Immunol. 46, 649 (1981).
35. K. Yoshinaga and 1. M. Harris, 1. Bioact. Compatible Polym. 4, 17 (1989).
36. A. Matsushima, H. Nishimura, Y. Ashihara, Y. Yokota, and Y. Inada, Chem. Lett., 773 (1980).
37. H. Nishimura, K. Takahashi, K. Sakurai, K. Fujinuma, Y. Imamura, M. Ooba, and Y. Inada, Life
Sci. 33, 1467 (1983).
38. K. Takahashi, A. Ajima, T. Yoshimoto, M. Okada, A. Matsushima, Y. Tamaura, and Y. Inada,
1. Org. Chem. 50, 3414 (1985).
39. C. C. Jackson, 1. L. Charlton, K. Kuzminski, G. M. Lang, and A. H. Sehon, Anal. Biochem.165,
114 (1987).
40. A. Yoshimoto, S. G. Chao, Y. Saito, I. Imamura, H. Wada, and Y. Inada, Enzyme 36,261 (1986).
41. M. Rubinstein, U.S. Patent 4,101,380 (1978).
42. M. Joppich and P. L. Luisi, Makromol. Chem. 180, 1381 (1979).
43. A. Abuchowski, G. Kazo, C. R. Verhoest, T. Van Es, D. Kafkewitz, M. L. Nucci, A. T. Viau, and F.
F. Davis, Cancer Biochem. Biophys. 7, 175 (1984).
44. M. S. Hershfield, R. H. Buckley, M. L. Greenberg, A. L. Melton, R. Schiff, C. Hatem, 1. Kirtzberg, M.
L. Markert, R. H. Kobayashi, A. L. Kobayashi, and A. Abuchowski, N. Eng/. 1. Med. 316, 589 (1987).
45. K. B. Hadley and P. H. Sato, Enzyme 42,225 (1989).
46. K. Iwasaki, Y. Iwashita, and T. Okami, U.S. Patent 4,670,417 (1987).
47. v. N. Katre, M. 1. Knauf, and W. 1. Laird, Proc. Natl. Acad. Sci. U.S.A. 84, 1487 (1987).
48. E. Boccu, R. Largajolli, and F. M. Veronese, Z. Naturforsch. 38c, 94 (1983).
49. D. Larwood and F. Szoka, J. Labelled Compd. Radiopharm. 21, 603 (1984).
50. C. O. Beauchamp, S. L. Gonias, D. P. Menapace, and S. V. Pizzo, Anal. Biochem. 131, 25 (1983).
51. L. Tondelli, M. Laus, A. S. Angeloni, and P. Ferruti, 1. Controlled Release 1, 251 (1985).
52. C. Beauchamp, P. E. Daddona, and D. P. Menapace, in: Adv. Exp. Med. Bioi. (C.H.M.M. DeDruyn,
H. A. Simmons, and M. Muller, eds.), 16SA, 47 (1984).
53. H. Berger and S. V. Pizzo, Blood 71, 1641 (1988).
54. F. M. Veronese, R. Largajolli, E. Boccu, C. A. Benassi, and O. Schiavon, Appl. Biochem.
Biotechnol. 11, 141 (1985).
55. S. Zalipsky, Patent pending; S. Zalipsky, R. Seltzer, and S. Meron-Rudolph, Biotechnol. Appl.
Biochem. 15, 100 (1992).
56. S. ZaIipsky, R. Seltzer, and K. Nho, Polym. Prepr., Am. Chem. Soc., Div. Polym. Chem. 31(2), 173
(1990); in: Polymeric Drugs and Drug Delivery Systems (R. L. Dunn and R. M. Ottenbrite, eds.),
p. 91, ACS Symposium Series 469, Washington, DC (1991).
57. A. Nathan, S. Zalipsky, and 1. Kohn, Polym. Prepr., Am. Chem. Soc., Div. Polym. Chem. 31(2), 213
(1990).
58. A. Buckmann, M. Morr, and G. Johansson, Makromol. Chem. 182, 1379 (1981).
59. M. Leonard, 1. Neel, and E. Dellacherie, Tetrahedron 40, 1581 (1984).
60. S. ZaIipsky and G. Barany,1. Bioact. Compat. Polym. 5, 227 (1990).
61. G. Johansson, J. Chromatogr. 368, 309 (1986).
62. F. M. Veronese, P. Caliceti, A. Pastorino, O. Schiavon, L. Sartore, L. Banci, and L. M. Scolaro,
J. Controlled Release 10, 145 (1989).
63. T. Yoshimoto, M. Nakata, S. Yamaguchi, T. Funada, Y. Saito, and Y. Inada, Biotechnol. Lett. 8, 771
(1986).
Functionalized PEGs and Polypeptides 369

64. D. E. Nitecki, L. Aldwin, and M. Moreland in: Peptide Chemistry 1987 (T. Shiba and S. Sakakibara,
eds.), p. 243, Protein Res. Foundation, Osaka (1988).
65. K. Shimizu, T. Nakahara, and T. Kinoshita, U.S. Patent 4,495,285 (1985); K. Shimizu, T. Nakahara,
T. Kinoshita, 1. Takatsuka, and M. Igarashi, U.S. Patent 4,640,835 (1987).
66. A. 1. Garman and S. B. Kalindjian, FEBS Lett. 223, 361 (1987); A. 1. Garman, U.S. Patent
4,935,465 (1990).
67. T. P. King and C. Weiner, Int. 1. Peptide Protein Res. 16, 147 (1980).
68. S. Zalipsky, F. Albericio, U. Slomczynska, and G. Barany, Int. 1. Peptide Protein Res. 30, 740 (1987).
69. 1. M. Harris, E. C. Struck, M. G. Case, M. S. Paley, M. Yalpani, 1. M. Van Alstine, and
D. E. Brooks, 1. Polym. Sci .• Polym. Chem. Ed. 22, 341 (1984).
70. K. Nilsson and K. Mosbach, Methods Enzymol. 104, 56 (1984).
71. C. Delgado, 1. N. Patel, G. E. Francis, and D. Fisher, Biotechnol. Appl. Biochem. 12,119 (1990).
72. P. Bohlen, S. Stein, W. Dairman, and S. Udenfriend, Arch. Biochem. Biophys. 155, 213 (1973);
S. 1. Stocks, A. 1. M. Jones, C. W. Ramey, and D. E. Brooks, Anal. Biochem. 154, 232 (1986).
73. A. S. F. A. Habeeb, Anal. Biochem. 14, 328 (1966).
74. N. Yamasaki, A. Matsuo, and H. Isobe, Agric. Bioi. Chem. 52, 2125 (1988).
75. L. Sartore, P. Caliceti, O. Schiavon, P. Ferruti, E. Ranucci, and F. M. Veronese, Proc. Int. Symp.
Control. Rei. Bioact. Mater. 17, 208 (1990).
76. S. Zalipsky and G. Barany, Polym. Prepr.. Am. Chem. Soc .. Div. Polym. Chem. 27(1), 1(1986).
77. H. Ueno and M. Fujino, Eur. Patent Appl. 0236987 (1987).
78. C. S. Pande, M. Pelzig, and 1. D. Glass, Proc. Natl. Acad. Sci. U.S.A. 77,895 (1980); 1. D. Glass,
R. Miller, and G. Wesolowski, in: Peptides Structure and Function (D. Rich, ed.), p. 209, Pierce
Chemical Co., Rockford, IL (1981); J. D. Glass, M. Pelzig, and C. S. Pande, in: Peptides 1978
(I. Z. Siemion and G. Kupryszewski, eds.), p. 235, Wroclav University Press, Poland (1979).
79. A. Sano, H. Maeda, Y. Kai, and K. Ono, Eur. Patent Appl. 0340741 (1989).
80. R. J. Goodson and N. V. Katre, Bio/Technology 8,343 (1990).
81. S. Romani, W. Gohring, L. Moroder, and E. Wunsch, in: Chemistry of Peptides and Proteins
(w. Voelter, E. Bayer, Y. A. Ovchinnikov, and E. Wunsch, eds.), Vol. 2, p. 29, Walter de Gruyter,
Berlin (1984).
82. A. Pollak and G. M. Whitesides, 1. Am. Chem. Soc. 98, 289 (1976).
83. M. Kimura, Y. Matsumura, Y. Miyauchi, and H. Maeda, Proc. Soc. Exp. Bioi. Med. 188, 364
(1988).
84. M. Urrutigoity and J. Souppe, Biocatalysis 2, 145 (1989).
85. M.-T. Babonneau, R. Jacquier, R. Lazaro, and P. Viallefont, Tetrahedron Lett. 30, 2787 (1989).
86. K. Yoshinaga, S. G. Shafer, and J. M. Harris, 1. Bioact. Compatible Polym. 2, 49 (1987).
87. C. Visco, C. A. Benassi, F. M. Veronese, and P. A. Maglioli, II Farmaco 42,549 (1987).
88. K. V. Savoca, A. Abuchowski, T. Van Es, F. F. Davis, and N. C. Palczuk, Biochem. Biophys. Acta
578, 47 (1979).
89. Y. Kamisaki, H. Wada, T. Yaguara, A. Matsushima, and Y. Inada, 1. Pharmacol. Exp. Ther. 216,410
(1981).
90. T. Yoshimoto, T. Mihama, K. Takahashi, Y. Saito, Y. Tamaura, and Y. Inada, Biochem. Biophys. Res.
Commun. 145, 908 (1987).
91. C. Pina, D. Clark, and H. Blanch, Biotechnol. Tech. 3, 333 (1989).
92. A. Matsushima, M. Okada, and Y. Inada, FEBS Lett. 178, 275 (1984).
93. T. Nishio, K. Takahashi, T. Tsuzuki, T. Yoshimoto, Y. Kodera, A. Matsushima, Y. Saito, and
Y. Inada, 1. Biotechnol. 8, 39 (1988).
94. Y. Inada, H. Nishimura, K. Takahashi, T. Yoshimoto, A. R. Saha, and Y. Inada, Biochem. Biophys.
Res. Commun. 122, 845 (1984).
95. A. Matsushima, M. Okada, K. Takahashi, T. Yoshimoto, and Y. Inada, Biochem.lnt.ll, 551 (1985).
96. Y. Kodera, K. Takahashi, H. Nishimura, A. Matsushima, Y. Saito, and Y. Inada, Biotechnol. Lett.
8, 881 (1986).
370 Samuel Zalipsky and Chyi Lee

97. J. S. Beckman, R. L. Minor, C. W. White, 1. E. Repine, G. M. Rosen, and B. A. Freeman, J. Bioi.

Chern. 263, 6884 (1988).
98. P. S. Pyatak, A. Abuchowski, and F. F. Davis, Res. Cornrnun. Chern. Pathol. Pharmacol. 29, 113
(1980).
99. P. McGoff, A. C. Baziotis, and R. Maskiewicz, Chem. Pharm. Bull. 36, 3079 (1988).
100. K. Miyata, Y. Nakagawa, M. Nakamura, T. Ito, K. Sugo, T. Fujita, and K. Tomoda, Agric. Bioi.
Chern. 52, 1575 (1988).
101. A. Abuchowski and F. F. Davis, Biochem. Biophys. Acta 578, 41 (1979).
102. U. Bagert and K.-H. Rohm, Biochern. Biophys. Acta 999, 36 (1989).
103. H. Neumann and A. Lustig, Eur. 1. Biochem. 109, 475 (1980).
104. M. Leonard and E. Dellacherie, Makromol. Chern. 189, 1809 (1988).
105. S. Zalipsky, F. Albericio, and G. Barany, in: Peptides: Structure and Function (V, J. Hruby and
K. H. Kopple, eds.), p. 257, Pierce Chemical Co., Rockford, IL (1985).