Terpolymerization kinetics of amino acid N-carboxy anhydrides

Article abstract of DOI:10.1002/pola.27109

Based on their versatility with respect to amino acid type and sequence, polypeptides have become attractive for a number of biological applications such as drug delivery, biomineralization, and drugs. N-carboxy anhydride (NCA) polymerization is a convenient way to rapidly prepare high-molecular weight polypeptides with good control over molecular weight and polydispersity. However, the kinetics of the incorporation of NCA monomers into copolypeptides during random copolymerization are poorly understood. Here, kinetic data is presented that allows insight into the NCA polymerization of a terpolymer composed of three commercially relevant amino acids, namely, glutamic acid, lysine, and tyrosine. Furthermore, kinetic data and copolymerization parameters from the copolymerization of binary mixtures of these three amino acid NCAs is used to make predictions of the terpolymer composition. This study provides access to the information necessary to prepare functional copolypeptides with better-defined sequence architecture that will be essential for the future development of polypeptide-based materials.

Full text of DOI:10.1002/pola.27109

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
Terpolymerization Kinetics of Amino Acid N-Carboxy Anhydrides
Mischa Zelzer,1* Andreas Heise1,2
1Department of Chemical Engineering and Chemistry, Technical University Eindhoven, Den Dolech 2, 5612 AZ, Eindhoven,
The Netherlands
2School of Chemical Sciences, Dublin City University, Glasnevin, Dublin 9, Ireland
Correspondence to: M. Zelzer (E-mail: mischa.zelzer@nottingham.ac.uk) or A. Heise (E-mail: andreas.heise@dcu.ie)
Received 11 November 2013; accepted 14 January 2014; published online 12 February 2014
DOI: 10.1002/pola.27109
ABSTRACT: Based on their versatility with respect to amino acid
type and sequence, polypeptides have become attractive for a
number of biological applications such as drug delivery, biomi-
neralization, and drugs. N-carboxy anhydride (NCA) polymer-
ization is a convenient way to rapidly prepare high-molecular
weight polypeptides with good control over molecular weight
and polydispersity. However, the kinetics of the incorporation
of NCA monomers into copolypeptides during random copoly-
merization are poorly understood. Here, kinetic data is pre-
sented that allows insight into the NCA polymerization of a
terpolymer composed of three commercially relevant amino
acids, namely, glutamic acid, lysine, and tyrosine. Furthermore,
kinetic data and copolymerization parameters from the copoly-
merization of binary mixtures of these three amino acid NCAs
is used to make predictions of the terpolymer composition.
This study provides access to the information necessary to pre-
pare functional copolypeptides with better-defined sequence
architecture that will be essential for the future development of
C
V
polypeptide-based materials.
2014 Wiley Periodicals, Inc. J.
Polym. Sci., Part A: Polym. Chem. 2014, 52, 1228–1236
KEYWORDS: copolymerization; glutamic acid; lysine; polymeriza-
tion kinetics; polypeptides; tyrosine
important aspect that has yet to be explored in detail is the
understanding of NCA copolymerization both in terms of reac-
tion kinetics and chain microstructure.
INTRODUCTION In biology, polypeptides are essential compo-
nents of cellular processes. Both the chemical variety of amino
acids and the synergistic functionality that arises from specific
amino acid sequences make polypeptides versatile molecules
that perform a multitude of functions such as enzymatic catal-
ysis, recognition events, and regulating processes.1 The poten-
tial of peptides in materials science has been recognized1–3
and over the last decades increasing efforts have been placed
into designing functional peptide-based materials such as
drug delivery vehicles,4–10 peptide surfaces,11–14 and self-
assembling peptide-based gels.15–21 As a result, preparation
methods of peptides become more advanced. The preparation
of polypeptides through N-carboxy anhydrides (NCA) poly-
merizations has been employed since the 1950s22 but has
recently attracted increased interest due to the successful
efforts of several groups to improve the control over the NCA
polymerization through reduced temperature,23–25 reduced
pressure,24,26,27 or the employment of specific initiators.14,28–
The random copolymerization of two amino acid NCAs has
been explored in several early publications (reviewed else-
where37) wherein the relative reactivity of the two NCAs
was determined through their copolymerization parameters.
Progress of the reactions was generally monitored by CO2
detection38,39 or infrared spectroscopy;40,41 however, both
methods are not able to distinguish between different amino
acid NCAs and hence detailed kinetic information of the con-
sumption of each amino acid in the copolymerization
remained elusive. Nuclear magnetic resonance (NMR) (1H-
NMR or 15N-NMR) has also been used to study amino acid
NCA polymerization kinetics.42–44 While it is able to distin-
guish between certain amino acids, it is limited by the fact
that it only allows the study of small numbers of monomers
due to significant overlap of signals in more complex copoly-
merization. To address this, we have recently reported a pro-
cedure through which we were able to follow the
36
Hence, polypeptides with narrow polydispersities, defined
architectures, and end-groups are now accessible via NCA
polymerization opening new avenues for material design. One
Additional Supporting Information may be found in the online version of this article.
*Present address: Mischa Zelzer, School of Pharmacy, University of Nottingham, Boots Science Building, University Park, Notting-
ham, NG7 2RD, United Kingdom
C
V
2014 Wiley Periodicals, Inc.
1228
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
consumption of two amino acid NCAs, Glu(Bzl)-NCA and
Lys(Z)-NCA, simultaneously during the polymerization.45 In the
present study, we have taken the next step in complexity in our
investigation and report on a detailed kinetic study of an NCA
terpolymerization. Firstly, we employ our previously developed
method to follow amino acid NCA consumption of a more com-
plex mixture containing three NCA monomers, allowing direct
access to amino acid NCA kinetics in a terpolymerization for
the first time. Secondly, we determine copolymerization param-
eters for the amino acid NCAs involved in the terpolymerization
with the aim to establish relative reactivities of our monomers
and evaluate the possibility to predict the composition of ter-
polymers from binary copolymerizations.
acid NCA can be found in the literature for Glu(Bzl)-NCA and
Lys(Z)-NCA45 and in the Supporting Information for
Tyr(Bzl)-NCA.
NCA Polymerization
Polymerization of the amino acid NCAs was carried out with a
total concentration of 2 M. The relatively high concentration
was chosen to allow the acquisition of data at relatively low
conversion. The reaction was initiated with n-hexylamine in
dry DMF under N2 atmosphere in a Schlenk tube according to
the literature procedures.45 The initiator concentration (3.8
lM) was constant for all the reactions. The polymerization
was stirred at room temperature until completion. Samples
were taken for high-performance liquid chromatography
(HPLC) analysis at periodic intervals under N2 flow and
directly subjected to the quenching procedure as described in
the HPLC section below. After complete conversion, the final
polymer was precipitated in diethylether, washed with ether,
and dried in a vacuum oven at 40 ꢀC over night. To obtain
polymers for the polymer composition analysis at low conver-
sion, separate reactions were used that were stopped after 20
min by adding 700 mL of the reaction solution to 500 mL of
the quenching solution. After addition of 2 mL ethyl ether, the
precipitate was filtered and washed with 10 mL diethylether
before drying the product at 40 ꢀC overnight.
Currently, only a small number of publications investigating the
random copolymerizations of more than two amino acid NCA is
available. Sommerdijk, Heise, and coworkers prepared Glu(Bzl)/
Asp(Bzl)/Ala46 and Glu(Bzl)/Lys(Z)/Ala47 terpolymers to control
and study biomineralization. Yamamoto et al. prepared a poly-
peptide from 18 different amino acid NCAs.48 In these studies,
only the overall composition of the final polymers was analyzed,
no information on the copolymer kinetics or the chain structure
was obtained. Only two reports have studied NCA terpolymeri-
zations in more detail. Hull and Kricheldorf investigated the
copolymerization of Gly-NCA/Leu-NCA/Val-NCA and Gly-NCA/
Val-NCA/c-methyl-Glu-NCA with 15N-NMR spectroscopy and
thus determined the copolymerization parameters and the aver-
age homogenous sequence length of the terpolymers.44 Wams-
ley et al. investigated the terpolymerization of Leu-NCA/
Asp(Bzl)-NCA/Val-NCA by determining the copolymerization
NMR
The polymers were dissolved in deuterated trifluoroacetic acid
(TFA). NMR spectra were acquired on a Varian Mercury 400
Autosampler using automatic acquisition protocols. The spec-
tra were analyzed using MestReNova (v 7.1.1). Peaks from the
amino acid parts of the polymer and not from the protection
groups were used to determine the copolymer composition as
TFA can be expected to cause removal of some protection
groups. Peak fitting was performed after baseline correction
with a Bernstein polynomial fit. To fit the peaks, component
peaks were added manually followed by automatic fitting pro-
tocols of the analysis software. This process was repeated until
the envelope of the component peaks matched the spectrum
and a minimal residual error was obtained.
1
parameters of the binary mixtures (with H-NMR spectroscopy)
and comparing the experimentally obtained terpolymer compo-
sition with the predicted composition.43 No kinetic data of the
progress of a terpolymerization is available to date.
In this work, we first study the kinetics of the copolymeriza-
tion of Glu(Bzl)-NCA/Tyr(Bzl)-NCA (E/Y) and Lys(Z)-NCA/
Tyr(Bzl)-NCA (K/Y), two monomer pairs whose copolymeriza-
tion has not yet been investigated. Polymers consisting of
Glu(Bzl)/Lys(Z) have been prepared before39,40,49 and we
have recently reported the copolymerization kinetics for these
materials.45 We then move on to elucidate the kinetics of the
terpolymerization of three NCAs, Glu(Bzl)-NCA, Lys(Z)-NCA,
and Tyr(Bzl)-NCA to gain insight into the effects of monomer
composition on the reaction kinetics and consequently gener-
ate an improved understanding of the polymer composition.
Mass Spectrometry (MS)
The mass of Tyr(Bzl)-NCA was confirmed with a Voyager
DE-STR matrix-assisted laser desorption ionization time of
flight (MALDI-TOF) mass spectrometer (Applied Biosystems).
Tyr(Bzl)-NCA was dissolved in THF at 1 mg ml21. Trans-2-
[3-(4-tert-Butylphenyl)-2-methyl-2-propenylidene] malononi-
trile (8 mg in 20 lL THF) was used as matrix and sodium
hexadecanesulfonate (1 mg in 200 lL THF) was used as salt.
About 0.3 lL of a mixture consisting of salt, matrix, and sam-
ple (1:4:4) were spotted on the sample plate and analyzed
using a 337 nm laser with a frequency of 20 Hz and a volt-
age of 25 kV. The composition of the quenched Tyr(Bzl)-NCA
was analyzed with liquid chromatography–MS (LC–MS) as
described previously for other amino acid NCAs.45
EXPERIMENTAL
Materials
Amino acid derivatives were purchased from Bachem. Dry eth-
ylacetate was obtained from Acros and n-heptane, 1,1,1,3,3,3-
hexafluoroisopropanol, and acetonitrile from Biosolve. Hexyl-
amine, a-pinene, triphosgene, diethylether, and dry dimethyl
formamide (DMF) were obtained from Sigma Aldrich.
NCA Synthesis
Gel Permeation Chromatography (GPC)
Protected amino acids were converted into NCAs as previ-
ously described.45 Details for the preparation of each amino
GPC of the polymers was performed in 1,1,1,3,3,3-hexafluor-
oisopropanol containing 3 g l21 trifluoroacetate. The GPC
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236
1229

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
was equipped with a Waters 1515 Isocratic HPLC pump, a
Waters 2414 refractive index detector (40 ꢀC), a Waters
2707 autosampler, a PSS PFG guard column followed by two
PFG-linear-XL columns (8 3 300 mm, 7 m) in series (at 40
ꢀC). To calculate the relative molecular weights, polymethyl
methacrylate standards with MP of 580 Da up to 7100 kDa
(Polymer Laboratories) were used.
Fourier Transform Infrared Spectroscopy (FTIR)
Infrared spectra were measured on a Perkin Elmer FTIR Spec-
trometer equipped with a universal sample accessory. A scan
resolution of 1 cm21 was used and 10 scans were averaged
for each measurement. The liquid samples from the polymer-
ization reactions were deposited onto the crystal without any
further sample modification. The spectra were analyzed with
Varian Resolutions 4.0 to obtain peak areas for the signals
around 1785 cm21 (C@O of NCA) and 1720 cm21 (C@O of
protection group). The C@O peak of the protection groups
was used as internal standard; the reported data is the ratio
SCHEME 1 Reaction scheme for the synthesis of NCAs and
polypeptides from Glu(Bzl), Lys(Z), and Tyr(Bzl).
of the peak areas at 1785 cm21/1720 cm21
.
in a binary mixture is essential. Hence, before studying the
copolymerization of Glu(Bzl), Lys(Z), and Tyr(Bzl), we first
compare the kinetics of the copolymerization of the binary
systems Glu(Bzl)/Lys(Z), Glu(Bzl)/Tyr(Bzl), and Lys(Z)/
Tyr(Bzl). As we have recently reported the copolymerization
kinetics for the first monomer pair, Glu(Bzl)/Lys(Z), to estab-
lish the procedure for the kinetic analysis of amino acid NCA
polymerizations45 only data for Glu(Bzl)/Tyr(Bzl) and
Lys(Z)/Tyr(Bzl) will be presented here, even though compar-
isons with the Glu(Bzl)/Lys(Z) pair will be drawn, as well.
Kinetic Analysis (HPLC)
For quantification of the remaining monomer in the polymer-
ization mixture, a recently described procedure was used.45
In short, 50 lL of the reaction mixture were quenched in
950 lL of an acidic acetonitrile/water solution. This converts
the NCAs back into the original amino acids which were sub-
sequently separated by RP-HPLC via gradual elution with
acetonitrile/water using an Agilent Zorbax SB-C18 column in
an Agilent 1100 Series chromatograph. The peaks were iden-
tified and quantified using calibration curves of the respec-
tive individual amino acid NCAs. More details and the
calibration curves for Glu(Bzl) and Lys(Z) can be found in
the literature.45 For Tyr(Bzl), validation of the peak position
and the corresponding calibration plot can be found in the
Supporting Information. The kinetic experiments were per-
formed at least twice for each condition. Both repeats dis-
played similar values and the same overall trend of the
kinetic curve; the data presented in the figures are datasets
of one experiment for each condition.
The procedure used for the preparation of the NCA mono-
mers and the NCA polymerization is shown in Scheme 1. For
the polymerization, different monomer ratios were used,
keeping the overall NCA concentrations constant at 2 M. Dur-
ing the reaction, samples were drawn to follow the con-
sumption of each individual amino acid NCA using our
recently established procedure.45 In brief, this consisted of
the quenching of samples taken at specific timepoints in an
acidic solution to convert non-reacted amino acid NCAs back
into the original amino acids. These samples were subse-
quently analyzed by HPLC which allowed simultaneous
quantitative analysis of the amino acid NCA concentration in
the feed solution over time. Since this procedure’s validity
was only demonstrated for Gly(Bzl)-NCA and Lys(Z)-NCA,
here we first verified that it can be equally applied to
Tyr(Bzl). Supporting Information Figure S1 shows that acidic
quenching converts Tyr(Bzl)-NCA back into the original pro-
tected amino acid. The small shift in the HPLC retention
time between the original amino acid and the quenched NCA
is due to a small concentration dependence of the peak posi-
tion that has been noted previously.45 Supporting Informa-
tion Figure S2 shows the HPLC calibration graph used to
determine the monomer concentration. For Glu(Bzl) and
Lys(Z), the calibrations reported in our previous study45
were used. The ability of the HPLC method to distinguish
Tyr(Bzl) from either Glu(Bzl) or Lys(Z) is shown in Support-
ing Information Figure S3 where the peaks are well sepa-
rated from each other. The small peak at approximately 17
Determination of Copolymerization Parameters
To determine the instantaneous copolymerization parameters
for the Glu(Bzl)/Tyr(Bzl) and the Lys(Z)/Tyr(Bzl) system,
copolymerizations at different NCA ratios (1:1; 11:9; 3:2;
13:7; 7:3; 3:1, and 4:1) were carried out for 20 min. About
700 lL of the reaction were then quenched by addition to 1
mL 15% HCl in acetonitrile/water (1:1). About 2 mL diethy-
lether was added and the precipitate was filtered and
washed with 10 mL diethylether before drying in the vac-
uum oven at 40 ꢀC overnight. The resulting polymer was
then analyzed by NMR as described above.
RESULTS AND DISCUSSION
Copolymerization Kinetics of Glu(Bzl)/Tyr(Bzl) and
Lys(Z)/Tyr(Bzl)
To study the kinetics of a terpolymer, a detailed understand-
ing of the copolymerization of the three involved monomers
1230
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
When fitting the kinetic data obtained from these reactions
to rate law equations, linear correlations were found when a
first order rate law equation was used (Supporting Informa-
tion Fig. S5). This enabled the determination of rate law con-
stants (k) for the whole reaction as well as for the
consumption of each individual monomer. Table 1 gives val-
ues for k for the two systems reported here as well as for
the Glu(Bzl)-NCA/Lys(Z)-NCA system which were calculated
from data we reported previously.45 From this data it is evi-
dent that the reaction rate for one amino acid NCA not only
depends on the second monomer but also on the relative
concentrations of the two monomers. While Glu(Bzl)-NCA
polymerizes faster than Tyr(Bzl)-NCA if E:Y 5 3:1, the reac-
tion rates are similar for both monomers when E:Y 5 1:1. In
contrast, for both ratios of Lys(Z)-NCA and Tyr(Bzl)-NCA,
both monomers polymerize with similar rates. When paired
with Glu(Bzl)-NCA, the rate of Lys(Z)-NCA incorporation into
the polymer is significantly lower than that of Glu(Bzl)-NCA
at all monomer ratios. Generally, it can be concluded that
Glu(Bzl)-NCA reacts faster than the other two NCA mono-
mers and is thus likely to be preferentially incorporated into
the polymer. In contrast, Lys(Z)-NCA and Tyr(Bzl)-NCA have
similar reaction rates and their copolymerization can be
FIGURE 1 Individual and total consumption of Glu(Bzl)-NCA
and Tyr(Bzl)-NCA during polymerization at two different mono-
mer ratios and overall conversion of NCAs during the reaction.
(a) Glu(Bzl):Tyr(Bzl) 5 3:1; (b) Glu(Bzl):Tyr(Bzl) 5 1:1. [Color fig-
ure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
min in the Glu(Bzl)/Tyr(Bzl) chromatogram in Supporting
Information Figure S3 was previously allocated to low
amounts of Glu that were produced through deprotection of
Glu(Bzl) in the acidic quenching solution.45
Figures 1 and 2 show the progress of the Glu(Bzl)/Tyr(Bzl)
and Lys(Z)/Tyr(Bzl) copolymerizations for two monomer
ratios (X:Tyr(Bzl) 5 3:1 and 1:1). Notably, copolymerizations
with a Tyr(Bzl)-NCA concentration of 1.5 M result in the for-
mation of a gel within minutes after initiation, making the
monitoring of copolymerizations with high Tyr(Bzl)-NCA con-
tent impossible. Hence, data was only collected for reactions
in which the Tyr(Bzl)-NCA concentration was 1 M or less.
The measured starting monomer concentrations match well
with the theoretical values for all four reactions. In addition,
the relative decrease of the total monomer concentration
(determined by the HPLC method) follows a very similar
trend to that observed by IR (see Supporting Information
Fig. S4), confirming the validity of the HPLC measurement as
an approach to follow these copolymerizations. The polymer-
izations reach full conversion after approximately 7–10 h, a
timeframe comparable to that of the Glu(Bzl)/Lys(Z) system
we investigated previously.45
FIGURE 2 Individual and total consumption of Lys(Z)-NCA and
Tyr(Bzl)-NCA during polymerization at two different monomer
ratios and overall conversion of NCAs during the reaction. (a)
Lys(Z):Tyr(Bzl) 5 3:1; (b) Lys(Z):Tyr(Bzl) 5 1:1. [Color figure can
be viewed in the online issue, which is available at wileyonline-
library.com.]
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236
1231

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 First Order Reaction Rate Constants for the Consump-
tion of NCAs in Binary Copolymerizationsa
TABLE 2 Copolymerization Parameters for Glu(Bzl)-NCA (rEY)
and Tyr(Bzl)-NCA (rYE) Determined From 1H-NMR Data and Cal-
culated Using Two Different Models
c,d
NCA 1c,d
1023 s21
3
NCA 2
3
Total NCAd
1023 s21
3
Reactionb
1023 s21
rEY
rYE
rEY 3 rYE
E:Y 5 3:1
E:Y 5 1:1
K:Y 5 3:1
K:Y 5 1:1
E:K 5 3:1
E:K 5 1:1
E:K 5 1:3
7.1 6 0.1
9.0 6 0.6
6.4 6 0.3
4.5 6 0.2
7.2 6 0.2
5.4 6 0.2
5.4 6 0.2
9.5 6 0.7
6.5 6 0.2
5.2 6 0.2
3.8 6 0.1
3.8 6 0.2
7.0 6 0.2
6.7 6 0.1
9.1 6 0.5
6.4 6 0.3
4.8 6 0.2
5.9 6 0.1
4.5 6 0.2
7.9 6 0.2
Fineman–Ross
1.33 6 0.07
1.39 6 0.29
0.04 6 0.10
0.02 6 0.11
0.05
0.02
€
€
Kelen–Tudos
tem. For the Lys(Z)/Tyr(Bzl) system in particular, the marker
signals of Tyr(Bzl) in the NMR spectrum suffer from the
proximity of other Lys(z) related peaks, thus making the
quantification more prone to errors which becomes more
pronounced as the Tyr(Bzl) concentration decreases. Hence,
twice as many data points were collected for the Lys(Z)/
Tyr(Bzl) system to improve the statistical fidelity of the
determination of the copolymerization parameter.
11.7 6 0.2
a
Initiator concentration was 3.8 mM.
E 5 Glu(Bzl)-NCA; K 5 Lys(Z)-NCA; Y 5 Tyr(Bzl)-NCA.
First and second amino acid NCA shown in the labels in the left hand
b
c
column.
d
Errors originate from the statistical fitting of
through one kinetic curve
a linear regression
For Glu(Bzl)-NCA/Tyr(Bzl)-NCA, the parameter rYE displays a
large uncertainty due to the relatively large fitting error.
Nevertheless, it is clear that rEY is significantly larger than
expected to yield a more even distribution of monomer units
in the final polymer.
rYE. This indicates that Glu(Bzl) has a much higher propen-
sity to be incorporated into the copolymer than Tyr(Bzl),
making Glu(Bzl) the most reactive of the three amino acid
NCAs. In contrast to the Glu(Bzl)/Lys(Z) system, however,
where the product of rEK 3 rKE was only slightly less than
unity, the product of rEY 3 rYE is near zero, indicating that
the copolypeptide obtained from this reaction does not fol-
low a random sequence distribution. This is also shown in
Figure 3(a), where the plot of the mole fraction in the poly-
mer versus the initial mole fraction in the feed shows a com-
position drift away from that of an ideal copolymer toward a
higher Glu(Bzl) content.
Copolymerization Parameters
To predict the composition of a terpolymer, theoretical mod-
els such as the Alfrey–Goldfinger equations have been devel-
oped.50 These models require knowledge of the
copolymerization parameters of the three monomers from
binary reactions. We have recently reported the copolymer-
ization parameters for the Glu(Bzl)/Lys(Z) system,45 where
rEK 5 1.87 6 0.31 and rKE 5 0.52 6 0.11 according to the
€ €
Kelen–Tudos (KT) method. Here, we used the same proce-
dure to determine the copolymerization parameters for the
Glu(Bzl)/Tyr(Bzl) and Lys(Z)/Tyr(Bzl) systems.
In the Lys(Z)/Tyr(Bzl) system, Tyr(Bzl) is the more reactive
monomer (Table 3), although the difference is far less pro-
nounced than for the Glu(Bzl)/Tyr(Bzl) system. This indi-
cates that Lys(Z) is the least reactive of the three NCAs. In
contrast, to the Glu(Bzl)/Tyr(Bzl) system, however, the prod-
uct of rKY 3 rYK is much closer to unity, suggesting that the
sequence in the polypeptide is more random. This is con-
firmed in Figure 3(b), where no significant composition drift
was observed. Instead, the data matches closely with the line
indicating the situation for an ideal random copolymeriza-
tion. The copolymerization parameters match well with the
conclusions made on the basis of the HPLC-based kinetic
data and the rate constants. Both methods indicate the same
relative reactivities for the three monomers and suggest a
more random monomer sequence distribution for the
Lys(Z)/Tyr(Bzl) copolymer.
Copolymerizations were carried out at seven different mono-
mer ratios. Due to the previously mentioned issues with
gelation at Tyr(Bzl)-NCA concentrations higher than 1 M, the
concentration range was varied at 0.2 M intervals between 1
M and 0.4 M Tyr(Bzl)-NCA, while the concentration of the
second NCA was increased to give a constant total NCA con-
centration of 2 M. The copolymerizations were stopped at
low conversion (20 min reaction time) and NMR spectra of
the isolated copolymers were used to determine the relative
amount of each amino acid in the polypeptide. Typical 1H-
NMR spectra for these systems are shown in Supporting
Information Figures S6 and S7. Glu(Bzl) was identified using
the CH peak at 4.6 ppm, Lys(Z) by the CH peak at 4.9 ppm,
and Tyr(Bzl) was identified using the CH2 peak between 2.5
and 3 ppm. Using these experimental polymer compositions,
the copolymerization parameters in Tables 2 and 3 were cal-
culated via the Fineman–Ross (FR) and KT methods. The
numeric data obtained is provided in Supporting Information
Tables S1 and S2; and the regression lines produced from
these calculations are shown in Supporting Information Fig-
ures S8 and S9. Due to the low Tyr(Bzl)-NCA concentration,
the linear regressions obtained for these experiments display
a lower R2 than the previously reported Glu(Bzl)/Lys(Z) sys-
TABLE 3 Copolymerization Parameters for Lys(Z)-NCA (rKY
)
and Tyr(Bzl)-NCA (rYK) Determined From 1H-NMR Data and Cal-
culated Using Two Different Models
rKY
rYK
rKY 3 rYK
Fineman–Ross
0.90 6 0.09
0.86 6 0.29
1.43 6 0.24
1.33 6 0.23
1.29
1.14
€
€
Kelen–Tudos
1232
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
FIGURE 3 Copolymer composition (obtained from 1H-NMR) as
function of the initial mole fraction of the NCAs in the feed.
The line represents the case of an ideal copolymerization.
[Color figure can be viewed in the online issue, which is avail-
able at wileyonlinelibrary.com.]
It is well known that various factors including electronic and
steric properties of the NCA and the initiator as well as sol-
vent polarity have pronounced effects on the reaction
kinetics of NCA polymerizations.37,51 Under the conditions of
the present study, the incorporation of Glu(Bzl)-NCA is
always favored, regardless of the nature of the amino acid
on the end of the polymer. This indicates that the reaction
kinetics are mostly governed by the properties of the mono-
mer as opposed to those of the active chain end of the poly-
mer. As a primary amine initiator was used, it can be
assumed that the normal amine mechanism (NAM) is the
main propagation mechanism for the polymerization.37 In
this mechanism, the free amine of the polymer chain per-
forms a nucleophilic attack on the carbonyl group of the
NCA. The mechanistic steps involved in the propagation of
the NAM are very complex, involving a series of transition
states and proton rearrangements.52 While we can conclude
that the nature of the different reactivities observed here
must lie in the susceptibility of the NCA carbonyl to a
FIGURE 4 Individual and total consumption of Glu(Bzl)-NCA,
Lys(Z)-NCA, and Tyr(Bzl)-NCA during the terpolymerization at
selected monomer ratios and overall conversion of NCAs during
the reaction. Data for all seven monomer ratios can be found in the
Supporting Information (Figure S11). [Color figure can be viewed
in the online issue, which is available at wileyonlinelibrary.com.]
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236
1233

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 4 First Order Reaction Rate Constants for the Consumption of NCAs in Terpolymerizations with Various Monomer Starting
Concentrationsa
Reactionb
Glu(Bzl)-NCAc 3 1023 s21
Lys(Z)-NCAc 3 1023 s21
Tyr(Bzl)-NCAc 3 1023 s21
Total NCAc 3 1023 s21
E:K:Y 5 1:2:3
E:K:Y 5 1:3:2
E:K:Y 5 2:1:3
E:K:Y 5 2:3:1
E:K:Y 5 3:1:2
E:K:Y 5 3:2:1
E:K:Y 5 1:1:1
14.0 6 0.6
8.4 6 0.4
8.6 6 0.5
6.8 6 0.2
6.9 6 0.3
6.1 6 0.2
10.7 6 0.6
6.2 6 0.3
5.4 6 0.2
4.6 6 0.6
3.9 6 0.3
3.5 6 0.2
2.9 6 0.2
6.1 6 0.3
8.5 6 0.4
6.8 6 2.2
7.3 6 0.5
4.9 6 0.2
5.1 6 0.4
4.5 6 0.2
9.7 6 0.4
7.7 6 0.4
6.2 6 2.3
8.9 6 0.7
4.9 6 0.3
5.4 6 0.3
4.5 6 0.2
8.3 6 3.1
a
c
Initiator concentration was 3.8 mM.
E 5 Glu(Bzl)-NCA; K 5 Lys(Z)-NCA; Y 5 Tyr(Bzl)-NCA.
Errors originate from the statistical fitting of
through one kinetic curve.
a linear regression
b
nucleophilic attack, to date there is no systematic study
available that would provide more insight into the effect of
the side group on the monomer reactivity.
The reaction rate constants of the terpolymerizations were
determined by fitting the data to a first order rate law (Sup-
porting Information Fig. S12). While linear trends were
observed, it should be noted that the fit for the binary
copolymerizations appeared to be better, suggesting that the
addition of a third amino acid decreases the precision of the
experimental procedure. The values for the rate constants
are given in Table 4. For all reactions, the rate constants for
Glu(Bzl)-NCA are highest, followed by those of Tyr(Bzl)-NCA.
This reflects the same trend in monomer reactivity suggested
by the binary copolymerizations. As with the binary reac-
tions, a concentration dependence of the rate constant can
be observed; with a few exceptions, the rate constants
decrease as the initial concentration of the respective mono-
mer is reduced. We propose that this observation is in parts
due to the relative probability of the monomer to collide
with an active polymer chain end.
Terpolymerization Kinetics
The terpolymerization of Glu(Bzl), Lys(Z), and Tyr(Bzl) was
studied in the same manner as the binary systems. The three
NCAs were polymerized at different ratios, given subse-
quently in the order Glu(Bzl):Lys(Z):Tyr(Bzl). The total NCA
concentration was kept constant at 2 M. Samples were taken
periodically during the polymerizations and analyzed by
HPLC. After completion of the reaction, the isolated polymers
were analyzed by GPC (Supporting Information Fig. S10).
The GPC data shows that the polymers were monomodal,
suggesting that they are indeed copolymers and not polymer
blends. Small shoulders in the traces are attributed to the
interference introduced by drawing samples.
Examples of the kinetic curves obtained from the HPLC
measurements are shown in Figure 4; the data for all seven
monomer ratios can be found in Supporting Information Fig-
ure S11. As with the binary systems, the measured starting
concentrations are similar to the expected theoretical values
with small deviations being attributed to experimental limi-
tations due to sample preparation and measurement. Full
conversion was reached after approximately 7–8 h.
Terpolymer Composition
To investigate the composition of the terpolymers, 1H-NMR
of polypeptides obtained after 20 min reaction were carried
1
out as described above. A typical H-NMR spectrum of a ter-
polymer is shown in Supporting Information Figure S13.
Curve fits to the backbone CH of Glu(Bzl) and Lys(Z) and the
TABLE 5 Comparison of the Experimental and Predicted Terpolymer Compositions
Experimental Polymer Compositiona
Predicted Polymer Compositionb
Feed Composition (E:K:Y)
E
K
Y
E
K
Y
1:1:1
1:2:3
1:3:2
2:1:3
2:3:1
3:1:2
3:2:1
0.60
0.46
0.44
0.62
0.55
0.68
0.70
0.16
0.18
0.31
0.07
0.34
0.09
0.20
0.25
0.35
0.26
0.31
0.11
0.22
0.10
0.57
0.48
0.43
0.60
0.54
0.69
0.68
0.19
0.18
0.31
0.08
0.32
0.08
0.20
0.24
0.35
0.26
0.32
0.14
0.22
0.13
a
b
From NMR spectra after 20 min reaction time (ꢁ10%–30%
Calculated using the Alfrey–Goldfinger equations and the copolymer-
ization parameters from the binary polymerization.
conversion).
1234
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
immediately evident that all polypeptides have different
compositions from that of the initial feed concentrations. In
addition, the initial polymer composition is shifted directly
toward higher Glu(Bzl) content in all reactions. This is con-
sistent with the observations in the kinetic experiments
where Glu(Bzl)-NCA was identified as the most reactive
monomer. The preferential incorporation of Glu(Bzl) at the
start of the reaction suggests that the starting end of the
polypeptides contain longer Glu(Bzl) sequences than the rest
of the polymer. Thus, the observed shift in the initial poly-
peptide composition during the terpolymerization can be
well rationalized using the kinetic data and the copolymer-
ization parameters of the NCAs.
CONCLUSIONS
This study demonstrates for the first time that it is possible
to determine the kinetics of individual amino acid NCAs dur-
ing an NCA terpolymerization. We have employed this
method for the elucidation of the reaction kinetics of three
amino acid NCAs, Glu(Bzl)-NCA, Lys(Z)-NCA, and Tyr(Bzl)-
NCA; and presented copolymerization parameters and reac-
tion rate constants for these polymerizations. The kinetic
data provides a higher level of understanding of NCA copoly-
merizations in complex mixtures, paving the way for a better
structure/property understanding. We believe that the
insight and control gained from this approach will contribute
to the design of well-defined copolypeptides for applications
as drugs and responsive biomaterials.
FIGURE 5 Ternary diagram showing the composition of the
feed (solid symbols) and the polymer (open symbols) obtained
after polymerization for 20 min. Corresponding feed/polymer
compositions were given the same symbol-shape. Arrows
highlight the direction of the composition shift between the
feed and the polymer. [Color figure can be viewed in the online
issue, which is available at wileyonlinelibrary.com.]
CH2 of the tyrosine side chain allowed an estimation of rela-
tive composition of the terpolymers (Table 5).
The ability of the copolymerization parameters determined
from binary copolymerizations to predict the composition of
the terpolymer was explored using the Alfrey–Goldfinger
equation. Here, the change in NCA concentrations, d[E], d[K],
and d[Y], is related directly to the amount of amino acid
incorporated into the polymer. This can be expressed as a
function of the initial NCA concentrations ([E], [K], and [Y])
and their copolymerizations parameters:
ACKNOWLEDGMENTS
We are grateful to Anneke Delsing for help with the HPLC meas-
urements and to Joost van Dongen for the LC-MS measure-
ments. MZ acknowledges financial support through a Marie
Curie fellowship (FP7-PEOPLE-2011-IEF 299280). AH is a SFI
Stokes Senior Lecturer (07/SK/B1241).
d½Eꢂ:d½Kꢂ:d½Yꢂ
ꢀ
ꢁꢀ
ꢁ
½Eꢂ
½Kꢂ
½Yꢂ
½Kꢂ ½Yꢂ
1
REFERENCES AND NOTES
5½Eꢂ
ꢀ
1
1
½Eꢂ1
:
r
YE rKE
r
KE rYK
r
YE rKY
rEK rEY
1 R. de la Rica, H. Matsui, Chem. Soc. Rev. 2010, 39, 3499–
3509.
ꢁꢀ
ꢁ
ꢁ
½Eꢂ
EK rYE
½Eꢂ
½Kꢂ
EK rYK
½Kꢂ
½Yꢂ
½Eꢂ ½Yꢂ
½Kꢂ
1
1
½Kꢂ1
1
:
r
r
r
YK rEY
½Yꢂ
rKE rKY
2 M. Zelzer, M. R. Alexander, J. Phys. Chem. B 2010, 114, 569–
576.
ꢀ
ꢁꢀ
½Eꢂ ½Kꢂ
½Yꢂ
1
1
½Yꢂ1
1
3 D. W. P. M. Lowik, E. H. P. Leunissen, M. van den Heuvel, M.
B. Hansen, J. C. M. van Hest, Chem. Soc. Rev. 2010, 39, 3394–
3412.
r
EY rKE
r
KY rEK
r
EY rKY
rYE rYK
To calculate the predicted terpolymer compositions, the
copolymerization parameters obtained from the KT method
were used because they are more robust to the effect of sin-
gle outlying data points.43,45 The results obtained from this
approach are given in Table 5. The predicted values are in
very good agreement with the experimental data, thus vali-
dating the copolymerization parameters calculated for the
binary systems and supporting the experimental results for
the terpolymer composition.
4 J. Ding, J. Chen, D. Li, C. Xiao, J. Zhang, C. He, X. Zhuang,
X. Chen, J. Mater. Chem. B 2013, 1, 69–81.
5 Y. C. Huang, Y. S. Yang, T. Y. Lai, J. S. Jan, Polymer 2012,
53, 913–922.
6 T. B. Ren, W. J. Xia, H. Q. Dong, Y. Y. Li, Polymer 2011, 52,
3580–3586.
7 J. P. Lin, J. Q. Zhu, T. Chen, S. L. Lin, C. H. Cai, L. S. Zhang,
Y. Zhuang, X. S. Wang, Biomaterials 2009, 30, 108–117.
8 E. G. Bellomo, M. D. Wyrsta, L. Pakstis, D. J. Pochan, T. J.
Deming, Nat. Mater. 2004, 3, 244–248.
To visualize the correlation between the mole fraction in the
feed and the experimentally observed terpolymer composi-
tion, a ternary diagram is presented in Figure 5. It becomes
9 E. P. Holowka, V. Z. Sun, D. T. Kamei, T. J. Deming, Nat.
Mater. 2007, 6, 52–57.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236
1235

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
10 C. Zheng, M. Zheng, P. Gong, J. Deng, H. Yi, P. Zhang, Y.
Zhang, P. Liu, Y. Ma, L. Cai, Biomaterials 2013, 34, 3431–3438.
31 T. J. Deming, J. Polym. Sci. Polym. Chem. 2000, 38, 3011–
3018.
11 M. Zelzer, D. J. Scurr, M. R. Alexander, R. V. Ulijn, ACS
Appl. Mater. Interfaces 2012, 4, 53–58.
32 G. J. M. Habraken, C. E. Koning, A. Heise, J. Polym. Sci.
Polym. Chem. 2009, 47, 6883–6893.
12 X. L. Liao, R. T. Petty, M. Mrksich, Angew. Chem. Int. Ed.
2011, 50, 706–708.
33 Y. Liu, C. Li, H. Y. Wang, X. Z. Zhang, R. X. Zhuo, Chem.
Eur. J. 2012, 18, 2297–2304.
13 H. Duran, B. Yameen, H. U. Khan, R. Forch, W. Knoll, React.
Funct. Polym. 2013, 73, 606–612.
34 M. Byrne, P. D. Thornton, S. A. Cryan, A. Heise, Polym.
Chem. 2012, 3, 2825–2831.
14 T. Borase, M. Iacono, S. I. Ali, P. D. Thornton, A. Heise,
Polym. Chem. 2012, 3, 1267–1275.
35 I. Dimitrov, H. Schlaad, Chem. Commun. 2003, 2944–2945.
36 H. Lu, J. J. Cheng, J. Am. Chem. Soc. 2007, 129, 14114–
15 A. R. Hirst, S. Roy, M. Arora, A. K. Das, N. Hodson, P.
Murray, S. Marshall, N. Javid, J. Sefcik, J. Boekhoven, J. H.
van Esch, S. Santabarbara, N. T. Hunt, R. V. Ulijn, Nat. Chem.
2010, 2, 1089–1094.
14115.
37 N. Hadjichristidis, H. Iatrou, M. Pitsikalis, G. Sakellariou,
Chem. Rev. 2009, 109, 5528–5578.
38 M. Oya, T. Takahashi, J. Polym. Sci. Polym. Chem. 1982, 20,
529–539.
16 S. Sur, C. J. Newcomb, M. J. Webber, S. I. Stupp, Biomate-
rials 2013, 34, 4749–4757.
39 Y. Shalitin, E. Katchalski, J. Am. Chem. Soc. 1960, 82, 1630–
1636.
17 J. Huang, C. L. Hastings, G. P. Duffy, H. M. Kelly, J.
Raeburn, D. J. Adams, A. Heise, Biomacromolecules 2013, 14,
200–206.
40 A. Nakajima, T. Hayashi, Bull. Inst. Chem. Res. 1972, 50,
303–317.
18 K. L. Niece, J. D. Hartgerink, J. Donners, S. I. Stupp, J. Am.
Chem. Soc. 2003, 125, 7146–7147.
41 V. Saudek, J. Stejskal, P. Schmidt, K. Zimmermann, V.
Skarda, P. Kratochvil, J. Drobnik, Biopolymers 1987, 26, 705–
725.
19 D. L. Liu, X. Chang, C. M. Dong, Chem. Commun. 2013, 49,
1229–1231.
42 H. R. Kricheldorf, D. Muller, W. E. Hull, Biopolymers 1985,
24, 2113–2129.
20 A. P. Nowak, V. Breedveld, L. Pakstis, B. Ozbas, D. J. Pine,
D. Pochan, T. J. Deming, Nature 2002, 417, 424–428.
43 A. Wamsley, B. Jasti, P. Phiasivongsa, X. L. Li, J. Polym.
Sci. Polym. Chem. 2004, 42, 317–325.
21 D. Pati, N. Kalva, S. Das, G. Kumaraswamy, S. Sen Gupta,
A. V. Ambade, J. Am. Chem. Soc. 2012, 134, 7796–7802.
44 W. E. Hull, H. R. Kricheldorf, Macromol. Chem. Phys. 1980,
181, 1949–1966.
22 R. B. Woodward, C. H. Schramm, J. Am. Chem. Soc. 1947,
69, 1551–1552.
45 M. Zelzer, A. Heise, Polym. Chem. 2013, 4, 3896–3904.
23 G. J. M. Habraken, M. Peeters, C. Dietz, C. E. Koning, A.
Heise, Polym. Chem. 2010, 1, 514–524.
46 Z. W. Deng, G. J. M. Habraken, M. Peeters, A. Heise, G. de
With, N. Sommerdijk, Soft Matter 2011, 7, 9685–9694.
24 G. J. M. Habraken, K. Wilsens, C. E. Koning, A. Heise,
Polym. Chem. 2011, 2, 1322–1330.
47 V. Dmitrovic, G. J. M. Habraken, M. Hendrix, W. Habraken,
A. Heise, G. de With, N. Sommerdijk, Polymers 2012, 4, 1195–
1210.
25 W. Vayaboury, O. Giani, H. Cottet, S. Bonaric, F. Schue,
Macromol. Chem. Phys. 2008, 209, 1628–1637.
48 H. Yamamoto, A. Nagai, T. Okada, A. Nishida, Mar. Chem.
1989, 26, 331–338.
26 T. Aliferis, H. Iatrou, N. Hadjichristidis, Biomacromolecules
2004, 5, 1653–1656.
49 Y. Huang, Z. Tang, X. Zhang, H. Yu, H. Sun, X. Pang, X.
Chen, Biomacromolecules 2013, 14, 2023–2032.
27 D. L. Pickel, N. Politakos, A. Avgeropoulos, J. M. Messman,
Macromolecules 2009, 42, 7781–7788.
50 T. Alfrey, G. Goldfinger, J. Chem. Phys. 1944, 12, 322–322.
28 C. S. Cho, J. B. Cheon, Y. I. Jeong, I. S. Kim, S. H. Kim, T.
Akaike, Macromol. Rapid Commun. 1997, 18, 361–369.
51 K. Ishiwari, T. Hayashi, A. Nakajima, Polym. J. 1978, 10, 87–
91.
29 Y. Y. Choi, J. H. Jang, M. H. Park, B. G. Choi, B. Chi, B.
Jeong, J. Mater. Chem. 2010, 20, 3416–3421.
52 J. Ling, Y. Huang, Macromol. Chem. Phys. 2010, 211, 1708–
30 T. J. Deming, Nature 1997, 390, 386–389.
1711.
1236
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 1228–1236