Substituted Polyesters by Thiol-Ene Modification: Rapid Diversification for Therapeutic Protein Stabilization

Article abstract of DOI:10.1021/jacs.6b10776

Many proteins, especially those used as therapeutics, are unstable to storage and shipping temperatures, leading to increased costs in research and industry. Therefore, the design and synthesis of novel stabilizers is an important area of investigation. H

Full text of DOI:10.1021/jacs.6b10776

Article
pubs.acs.org/JACS
Substituted Polyesters by Thiol−Ene Modification: Rapid
Diversification for Therapeutic Protein Stabilization
Emma M. Pelegri-O’Day, Samantha J. Paluck, and Heather D. Maynard*
Department of Chemistry and Biochemistry and California NanoSystems Institute, University of California, Los Angeles, 607 Charles
E. Young Drive East, Los Angeles, California 90095, United States
S
* Supporting Information
ABSTRACT: Many proteins, especially those used as ther-
apeutics, are unstable to storage and shipping temperatures,
leading to increased costs in research and industry. Therefore,
the design and synthesis of novel stabilizers is an important area
of investigation. Herein we report new degradable polymers that
stabilize proteins to environmental stressors such as refrigeration
and elevated temperature. Specifically, polycaprolactones with
different pendant groups were synthesized and surveyed for
their ability to stabilize an important therapeutic protein to
storage and shipping conditions. Ring-opening polymerization
(ROP) of an allyl-substituted caprolactone monomer was
carried out using the organocatalyst 1,5,7-triazabicyclo[4.4.0]-
dec-5-ene (TBD) to yield a well-defined, alkene-substituted
degradable polymer, which was used as a common backbone to
control for the degree of polymerization. Relevant side chains such as trehalose, lactose, glucose, carboxybetaine, and
oligo(ethylene glycol) were installed via postpolymerization thiol−ene reactions. These degradable polymers were then
employed as excipients for the stabilization of the therapeutic protein granulocyte colony-stimulating factor (G-CSF) against
storage at 4 °C and shipping temperatures of 60 °C. The best stabilization was observed using the trehalose- and zwitterion-
substituted polyesters. Both the trehalose- and carboxybetaine-substituted pCL were further investigated with regard to molecular
weight dependence, and it was found that the molecular weight was minimally important for stabilization to refrigeration, but
critical for G-CSF stabilization at elevated temperatures. Both high performing zwitterionic and trehalose polyesters were also
degraded, and the polymers and degradation products were shown to be noncytotoxic. This work provides potential
biocompatible polymers for stabilization of the important therapeutic G-CSF, as well as a general platform for the future
discovery of new polymeric protein stabilizers.
other amino acids have also been shown to stabilize proteins
through binding interactions, buffering, or hydration mecha-
nisms.7−9 Moreover, proteins such as human serum albumin
(HSA) have been used as bulking agents or to prevent protein
adsorption.5 Furthermore, surfactants such as polysorbate
(Tween) or modified polysaccharides such as hydroxyethyl
starch (HES) have been employed to prevent protein unfolding
and aggregation.5,10,11 Excipients have also been used in
nonbiological therapeutics. For instance, the recently approved
hyperkalemia drug patiromer includes sorbitol in its for-
mulation to improve stability.12 However, therapeutics still
suffer from activity loss despite the presence of these excipients,
prompting further development of improved materials.
INTRODUCTION
■
Due to their substrate specificity and biological function,
proteins have unique and essential roles in various industries.
For example, proteins are used as reagents for improving
chemical transformations, as cosmetic additives, as supplements
for improving nutrition of animal feed, and as biological
therapeutics. However, the stabilization of certain proteins
during storage and transport, especially those used as
therapeutics, can be critical to maintain structure and activity.
Conditions such as UV exposure,1 heat,2 lyophilization,3 and
excessive agitation4 can lead to protein unfolding, aggregation,
or loss of biological activity. Measures to prevent this loss of
activity, such as the maintenance of a refrigeration chain for
delicate protein therapeutics, increase costs and may still result
in inactivated protein.
Synthetic polymers comprise another promising class of
excipients used to stabilize proteins against environmental
stressors. Polymers such as anionic polyacrylate, poly(glutamic
acid), carboxylated polyamidosaccharides, and block copoly-
As a result, a number of compounds are used as excipients or
additives to maintain protein activity.5 For instance, osmolytes
and carbohydrates such as trehalose, sorbitol, and sucrose have
been shown to maintain protein activity through preferential
hydration or protein interactions.6 Arginine, histidine, and
Received: October 14, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Illustrative scheme of pCL backbone and modification with thiols using thiol−ene chemistry to produce a small library of degradable
polymers.
mers of poly(ethylene glycol) (PEG) and poly(histadine) have
been shown to stabilize a variety of proteins to stressors such as
heat, aggregation, and lyophilization.13−18 Other charged
polymers such as poly(ethylenimine) or heparin mimicking
polymers can stabilize a variety of enzymes or growth factors
using electrostatic interactions.19−22 Zwitterions have also been
shown to have significant stabilizing ability due to their
hydration and protein repulsion properties.23 Additionally,
thermoresponsive copolymers have been used for refolding
denatured proteins.24 We have previously developed styrene-
and methacrylate-based polymers with trehalose side chains and
shown that these polymers protect lysozyme, horseradish
peroxidase (HRP), and glucose oxidase (GOX) against elevated
temperatures both as excipients and as protein−polymer
conjugates.25,26 And others have investigated the use of
trehalose in polyacrylamide polymers to inhibit amyloid protein
aggregation and in polycationic nanoparticles for delivery of
siRNA.27,28
Though synthetic polymers show promise in stabilization of
proteins, most are nondegradable and thus will not be cleared
from biological systems or will persist in the environment. For
instance, PEG is the most widely used biocompatible polymer,
but has been shown to induce the formation of antibodies in
32−46% of patients during a clinical trial because of its
persistence in vivo.29,30 Additionally, vacuolation in rats has
been reported upon injection with high molecular weight (40
kDa) PEG.31 Small-molecule excipients that have been widely
used for therapeutic formulation present other disadvantages.
For instance, sorbitol is widely used and effectively maintains
protein activity, yet has been shown to result in GI tract
complications such as bleeding, ulcers, and necrosis.32 Other
high-performing excipients include the nonionic surfactants
Tween 20 and Tween 80, which effectively prevent protein
aggregation but have been shown to undergo auto-oxidation,
resulting in the formation of damaging peroxides.33 Therefore,
the development of novel, degradable, and functional polymers
has been a subject of recent interest, especially for biological
applications.34,35 Degradable polymers might alleviate immuno-
genic responses, while also enabling the use of higher molecular
weight polymers, which typically cannot be employed due to
difficulty in clearance. In addition, enzymes are widely
employed in applications such as in detergents or animal
feed, where the protein stabilizers should be biodegradable to
avoid unwanted environmental buildup. Therefore, there is
significant need for well-controlled, homogeneous, and
degradable synthetic materials for biological and environmental
concerns.
Herein we report the synthesis of degradable stabilizing
polyesters using ring-opening polymerization (ROP). The
polymers were prepared by first synthesizing alkene-function-
alized polycaprolactones, followed by the installation of desired
side chains using high-yielding thiol−ene reactions. A variety of
materials were easily synthesized by varying mercaptan identity
and the resulting materials protected granulocyte colony-
stimulating factor (G-CSF) against loss of biological activity
when added as excipients. We expect that these polymers can
function as protein stabilizers in a variety of fields due to their
combination of biodegradability and stabilization abilities.
RESULTS
■
The nature of the degradable polymer backbone was an
important consideration in the design of a modular system for
protein stabilization. We have previously observed that
trehalose polymers with hydrophobic backbones have demon-
strated good protein stabilization,26 and hypothesized that the
nonionic surfactant character of these materials was an
important contributor to their desirable properties.5 Therefore,
the FDA-approved polymer poly(caprolactone) (pCL) was
selected because of its hydrophobic and biodegradable nature.
Previous examples have introduced functional side chains onto
pCL using a variety of postpolymerization click chemistries to
avoid chemical incompatibilities with ROP conditions and also
to minimize steric interference during polymerization. For
instance, aminooxy-functionalized PEG chains have been added
to ketone-modified pCL through oxime click chemistry,
B
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
resulting in graft copolymers.36 Alkyne- and alkene-function-
alized valero- and caprolactone monomers have been
synthesized and polymerized to yield polyesters with reactive
handles for later installation of PEG and peptide side
chains.37,38 We chose to synthesize our polyester backbone
with reactive alkene side chains and use it as a common
precursor to introduce stabilizing functionalities via postpoly-
merization thiol−ene reactions. Thiol−ene is a particularly
attractive type of “click” modification because it combines
efficiency, a metal-free nature, and a tolerance of both water
and oxygen.39,40 Using this type of chemistry allows for the
ready introduction of different functional and potentially
stabilizing moieties onto the polymer side chains by using a
variety of mercaptans (Figure 1). Additionally, with this post-
polymerization approach, the backbone length would be the
same between the different classes to rule out differences in
stabilizing ability due to changes in degree of polymerization.
Synthesis of a Library of Functionalized Polyesters.
The desired alkene-functionalized caprolactone monomer was
synthesized by adding allyl bromide to CL in the presence of
lithium diisopropylamide following a literature procedure37 and
polymerized using ROP. A degree of polymerization (DP) of
40 was targeted because it would result in functionalized
polymers with molecular weights between 20.9 and 12.8 kDa.
The organic catalyst 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD)
was used due to its fast polymerization kinetics at room
temperature and narrow dispersity (Đ) for the ROP of
functional lactones.41−44 The initiator 3-methyl-1-butanol was
(Scheme 1). The photoinitiator 2,2-dimethoxyphenyl-
acetophenone (DMPA) was used because of its demonstrated
high efficiency in photoinitiated thiol−ene reactions.46 A series
of easily accessible thiols (A−D) were selected containing
sugars or oligo(PEG) that as small molecules are known
stabilizing excipients.5 Thiolated trehalose was synthesized in
five steps and 53% overall yield from trehalose using trityl and
acetate protecting groups (Scheme S1). Briefly, monohydroxyl
trehalose heptaacetate was synthesized as previously de-
scribed.26 A tosylate ester was installed and displaced using
potassium thioacetate. Selective cleavage of the thioester using
hydrazine acetate then led to thiolated trehalose A. Thiolated
lactose C35,47 and thiolated mPEG D48 were synthesized as
previously described.
Use of acetate-protected saccharide mercaptans was found to
be important for good miscibility between the pCL backbone
and the thiol, giving clean conversion to the acetylated
glycopolymers. In all cases, three equivalents of thiol per
alkene were used to ensure complete reaction of the alkene
side-chains. After polymer modification, removal of the acetate
esters using potassium carbonate in methanol/chloroform26 or
hydrazine hydrate49−52 led to the desired glycopolymers
without hydrolysis of the polyester backbone. Complete
modification was confirmed by disappearance of the alkene
1
peaks in the H NMR (representative data in Figure 2a, full
data provided in the Supporting Information (SI)) as well as
clean shifts in GPC molecular weight (representative data in
Figure 2b, full data provided in the SI).
1
employed because its distinctive H NMR peaks allowed for
The carboxybetaine zwitterionic pCL polymer was synthe-
sized taking inspiration from a literature procedure for a
nondegradable polymer (Scheme 2).53 2-(Dimethylamino)-
ethanethiol hydrochloride was added to the pCL-allyl40
backbone polymer using photoinitiated thiol−ene conditions
and subsequently treated with sodium bicarbonate to neutralize
the hydrochloride salt. Exposure to t-butyl bromoacetate
quaternized the amine and hydrolysis of the t-butyl ester with
trifluoroacetic acid led to the formation of the zwitterion. No
good characterization.45 Using a monomer concentration of 2
M, high conversion and good control over molecular weight
were achieved, with Đ = 1.08, a degree of polymerization (DP)
of 36, and a number-average molecular weight of 5600 Da by
1H NMR and 5400 by gel permeation chromatography (pCL-
allyl40, Table 1). Initially, the polymer was purified by dialysis in
Table 1. Molecular Weights and GPC Data for the Library of
Polyesters (DP40)
1
acidic backbone scission was observed by GPC or H NMR
analysis.
Mn (1H
NMR)
Mn
(GPC)
1
All polymers were characterized by GPC and H NMR to
polymer
Đ
determine molecular weight and dispersity (Table 1). During
the deprotection, there was unlikely to be chain scission, as the
library of substituted pCL all gave narrow molecular weight
distributions between 1.19 and 1.07 and the peak shapes were
generally well-defined and symmetrical (Figures S26, S35, S36).
Using poly(methyl methacrylate) standards, the MN by GPC
for the DMF-soluble polymers varied from 12.7 to 23.6 kDa.
The zwitterionic material was not DMF-soluble and was instead
analyzed using PEG standards, making direct molecular weight
comparison difficult. However, because a postpolymerization
approach was used to synthesize these materials, the same
backbone was used to construct all polymers in the study.
Therefore, while the molecular weight varied due to side chain
identity, the DPs of all polymers (i.e., backbone length)
compared were identical.
starting polymer pCL-allyl40
5600
5400
1.08
protected
pCL-trehalose-
OAc40
34 700
28 400
1.06
pCL-glucose-OAc40
pCL-lactose-OAc40
20 300
29 900
22 300
28 900
1.06
1.07
deprotected
pCL-trehalose-OH40
pCL-glucose40
pCL-lactose40
18 500
11 300
18 000
20 400
12 500
12 700
18 300
16 900
23 600
1.17
1.09
1.17
1.07
1.19
PCL-PEG40
a
a
pCL-zwitterion40
5100
a
GPC run in buffer/MeCN with PEG standards.
Assessment of Stabilizing Ability. Next the ability of the
polymer to protect protein activity against environmental
stressors was assessed. The therapeutic protein granulocyte
colony-stimulating factor (G-CSF) was selected to compare
excipient efficacy due to its clinical importance. G-CSF is FDA-
approved as filgrastim (Neupogen), and lenograstim and is
used therapeutically to increase the neutrophil granulocyte
count during chemotherapy.54 G-CSF is highly unstable at
dichloromethane/methanol. However, polymers purified using
this method had unidentified impurities, which resulted in
significant loss of protein activity in later experiments (data not
shown). Purification by silica gel column chromatography
successfully removed the impurities, and subsequent polymers
were therefore purified using this method.
The allylated polymer was then used in radical thiol−ene
reactions to install the desired pendant stabilizing groups
C
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 1. Synthetic Scheme of Thiol−Ene Modification of pCL-Allyl Polymers with Acetyl-Trehalose, Acetyl-Glucose, Acetyl-
Lactose, and PEG Thiols, Followed by Deprotection of the Acetylated Sugars
1
Figure 2. Characterization of trehalose modification of pCL using thiol−ene chemistries. (a) The H NMR traces before and after modification
showing a disappearance of the alkene resonance peaks at 5.0 and 5.7 ppm and the appearance of resonance peaks corresponding to trehalose
anomeric protons. (b) GPC characterization of pCL-allyl40 before and after modification showing a shift toward a higher molecular weight species.
After deprotection of the trehalose, a shift toward a lower molecular weight species was observed showing complete modification and deprotection of
the polymer.
physiological pH and is therefore stored at pH 4.0; still at this
pH the protein readily degrades upon storage or subjection to
heat.55 The side chain identity was varied to determine the
relative stabilizing ability of the functional groups. To
investigate storage at refrigeration temperatures, pCL polymers
were added to G-CSF at 100 wt equiv to protein and the
protein was stored for 90 min at 4 °C at 1 μg/mL and pH 4.0.
Protein activity was determined by measuring cell proliferation
in murine myeloid leukemia NFS-60 cells, which is enhanced in
the presence of G-CSF, and compared to the proliferation of
freshly diluted protein.56
Scheme 2. Synthetic Scheme for the Synthesis of
Zwitterionic Polymer pCL-Zwitterionn
After stressing at 4 °C, G-CSF with no additive only
exhibited 133
6% cell proliferation, a drastic reduction
compared to fresh G-CSF (Figure 3a). Addition of the pCL-
glucose and pCL-lactose polymers was not statistically different
than no additive, indicating that, for this protein, the polymers
were not effective stabilizers. Interestingly, when the pCL-PEG
polymer was added to the G-CSF solution, significantly lower
cell proliferation was observed. PEG has been shown to
associate with hydrophobic moieties on the protein surface due
to its amphiphilic nature57 and has been previously observed to
lower protein thermal stability.58 A similar mechanism may be a
D
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
sugar around flexible polar residues on the protein surface,60,61
we included a larger trehalose CL polymer (preparation vide
infra) in the heat study. In this case, a 40 kDa pCL-trehalose80
polymer was statistically the same as the zwitterionic polymer,
showing that the larger trehalose pCL stabilizes as well as the
zwitterionic polymer and suggesting a molecular weight
dependence of the trehalose polymer stabilization ability.
Testing of Different Molecular Weights and Compar-
ison to Common Excipients. To further test this potential
molecular weight dependence, various CL trehalose polymer
sizes were synthesized. Using previously optimized ROP
conditions, well-defined pCL-allyl polymers were synthesized
with DP between 10 and 80 and Đ < 1.25 (Table 2). These
Table 2. Molecular Weights and GPC Data for the Library of
pCL-Trehalose and pCL-Zwitterion Polymers with Variable
DP
Mn (1H
NMR)
Mn
(GPC)
polymer
Đ
a
a
starting
polymers
pCL-allyl10
1600
ND
ND
pCL-allyl20
pCL-allyl40
pCL-allyl80
3600
2400
1.21
1.08
1.08
b
5600
5400
12 400
12 200
trehalose-OAc
pCLtrehalose-OAc10
10 600
20 200
9600
1.07
1.06
pCL-trehalose-
OAc20
15 400
pCL-trehalose-
34 700
67 000
28 400
53 100
1.06
1.06
b
OAc40
pCL-trehalose-
OAc80
trehalose-OH
zwitterions
pCL-trehalose-OH10
pCL-trehalose-OH20
6200
5600
1.09
1.15
1.17
14 400
18 500
8100
pCL-trehalose-
12 700
b
OH40
pCL-trehalose-OH80
41 000
17 000
1.39
Figure 3. Effect of side chain identity on stabilization of G-CSF at pH
4.0 to (a) storage conditions at 4 °C for 90 min and (b) thermal stress
at 60 °C for 30 min. # = no statistical difference from the fresh sample
(p > 0.01), Student’s t test. Data shown as the average of six
experimental repeats and six well repeats with standard deviation.
c
c
c
c
c
pCL-zwitterion10
pCL-zwitterion20
3200
1700
1.17
1.12
1.19
1.19
c
6400
3000
b
c
pCL-zwitterion40
12 400
25 400
5100
c
pCL-zwitterion80
8900
a
b
c
Too small for GPC analysis. Same entry as in Table 1. GPC run in
buffer/MeCN with PEG standards.
factor for the destabilizing effect of this pCL-PEG polymer.
Both the zwitterionic and trehalose side chains significantly
outperformed the other polymers, stimulating 171 7% and
DPs were selected so that, after modification with thiolated
trehalose, the molecular weight of the pCL-trehalose polymers
would be between 5 and 40 kDa, assuming quantitative
conversion. For the smallest pCL-allyl polymer, matrix assisted
laser desorption ionization (MALDI) was used to confirm the
molecular weight (Figure S13). Modification was again carried
out using photoinitiated thiol−ene chemistry, yielding a series
of trehalose-modified pCL polymers. This series demonstrated
increased dispersity (Đ) with increasing molecular weight. At
high molecular weights (Table 2, pCL-trehalose80) Đ was
increased to 1.39 and the GPC molecular weight was
168
3% cell proliferation, respectively. Both stabilizing
polymers were not statistically different than the fresh sample,
indicating that both are equally effective at preventing G-CSF
activity loss under these conditions. G-CSF was also stressed at
60 °C for 30 min; this is representative of the maximum
temperature inside truck and shipping containers during
transport.59 As expected, G-CSF lost more than 95% of the
native activity after heating; addition of the pCL-PEG polymer
was not statistically different than no additive (Figure 3b). The
pCL-glucose and the pCL-lactose polymers were moderately
1
stabilizing. Addition of the trehalose (133
8%) and
correspondingly lower than that predicted by H NMR. The
zwitterionic (179 3%) side chain polymers resulted in the
highest cell proliferation, and the pCL-zwitterion80 was not
statistically different than the fresh sample. We observed that
the zwitterionic polymers retained greater activity than the
trehalose side chain polymers to heat. Since one hypothesis of
why trehalose provides stabilization is due to clustering of the
peak shape was also asymmetrical and extended toward the low
molecular weight side (Figure S37a). To confirm that this peak
broadening was not due to hydrolysis of the backbone esters,
the molecular weight of pCL-trehalose80 was also measured on
an aqueous size exclusion chromatography (SEC) system
(Figure S37b). In aqueous solution, no asymmetry was
E
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 4. (A) Effect of pCL-trehalose molecular weight on G-CSF stabilization at pH 4.0 to storage at 4 °C for 90 min. (B) G-CSF stabilization to
thermal stress at 60 °C for 30 min. Data shown as the average of six experimental repeats and six well repeats with standard deviation. (C) Effect of
pCL-zwitterion molecular weight on G-CSF stabilization at pH 4.0 to storage at 4 °C for 90 min. (D) G-CSF stabilization to thermal stress at 60 °C
for 30 min. Data shown as the average of four experimental repeats and six well repeats with standard deviation. All polymers exhibited statistically
significant stabilization (p < 0.05) relative to no stabilizing additive. A dependence on molecular weight was also observed in that greater molecular
weight polymers showed greater stabilization (# = p < 0.01 relative to DP10 polymers, ‡ = p < 0.01 relative to DP20 polymers, * = p < 0.01 relative
to DP40 polymers, Student’s t test).
observed and the calculated dispersity was lower (1.26). The
dragging observed at high molecular weights was therefore
hypothesized to be a result of interactions with the stationary
phase of the GPC column. Similarly, a series of pCL-zwitterion
polymers were synthesized using photoinitiated thiol−ene
chemistry on the pCL-allyl backbones. Analysis by GPC
showed that they were well-defined and demonstrated clear
shifts in molecular weight with increasing pCL-allyl DP (Table
2).
Both sets of polymeric backbones were then subjected to the
same stability tests using 100 wt equiv of polymer. First the
trehalose polymers were tested and a very slight dependence of
protein activity on molecular weight was observed upon storage
at 4 °C (Figure 4a). Larger polymers offered improved
stabilization compared to smaller polymers, but there was no
significant difference between the stabilizing effects of DP40
and DP80 polymers, or between the DP10 and DP20 polymers.
The series of pCL-trehalose polymers were also used as
stabilizers against 60 °C heating (Figure 4b). In this case, a
drastic molecular weight dependence was observed, with the
pCL-trehalose80 polymer exhibiting the highest cell prolifer-
ation. It should be noted that despite the increase in molecular
weight, the concentration of stabilizing units in solution
remained constant at 69 weight equivalents of trehalose or
190 μM, indicating that the observed changes in stabilizing
ability were solely due to the molecular weights of the
polymers.
Similar experiments were carried out using the zwitterionic
backbone. Upon exposure to the milder 4 °C stressor, only a
moderate dependence on molecular weight was observed
(Figure 4c). While the pCL-zwitterion10 polymer sample
exhibited reduced cell proliferation, there was no statistical
difference between the DP20, DP40, and DP80 polymers. They
were statistically the same as the fresh sample, indicating the
presence of a molecular weight threshold for complete
stabilization ability. However, when the protein was heated to
60 °C for 30 min, separation between the polymer additives
was observed (Figure 4d). At this temperature, the performance
of the DP20, DP40, and DP80 polymers was significantly
different from each other, and only the two largest polymers
retained comparable activity to the pristine sample.
Additional experiments were carried out to better understand
the observed dependence on molecular weight. To determine if
shorter polymers could demonstrate improved stabilization at
higher weight equivalents, we stressed G-CSF at 60 °C for 30
min and added pCL-trehalose40 and pCL-zwitterion40, varying
the amount of polymer in solution between 1 and 500 wt equiv
(Figure S38). We were curious to determine if the DP40
F
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
polymers would match the stabilizing performance of higher
molecular weight DP80 polymers when more weight
equivalents were used. Instead, there was a distinct plateau,
and for both polymers only 100 wt equiv were required to see
the best stabilization, without further improvement at the
higher concentrations tested. This is strong evidence that the
number of repeat units on the polymer chain has a distinct
effect on the polymer’s stabilizing ability. Trehalose has been
previously shown to demonstrate a clustering effect in
computational studies, self-organizing near polar residues on
proteins.61 The molecular weight trends observed support a
multivalency effect in these materials, where increased
equivalents offer inferior protection compared to a preor-
ganized or pregrouped set of stabilizing units. This sort of
molecular weight effect has been previously reported in other
systems.62
of 80 are as good as the currently utilized additives for
therapeutic G-CSF and better than other common protein
excipients at the concentrations tested.
Additionally, the half-life of G-CSF at 60 °C was tested with
the DP80 polymers as excipients (Figure S39). When the pCL-
trehalose80 polymer was added, G-CSF retained 50% of the
native activity until 48 min of heating, whereas when the pCL-
zwitterion80 polymer was used, the half-life was calculated to be
90 min, almost double. In contrast, with no additive G-CSF was
already inactive after 30 min (first time point tested). These
data show that that both pCL scaffolds, especially the
zwitterion-substituted polymer, provide a significant increase
in thermal stability.
Degradation, Biological Compatibility, TEM, and DSC
Analysis of the Polymers. To confirm that the poly-
caprolactone was still degradable, pCL-trehalose40 and pCL-
zwitterion40 were treated with 5% KOH to hydrolytically cleave
the backbone esters (Scheme S2). The molecular weight of the
polymeric materials was determined post-cleavage by aqueous
SEC (Figure S41). In both cases, a complete shift in molecular
weight toward a lower molecular weight species was observed
after 24 h, confirming hydrolytic degradation. No hydrolytic
degradation was observed under more moderate degradation
conditions (cell media at 37 °C) for up to 49 days, consistent
with the slow hydrolysis rates observed for polycaprolactone in
vivo.64 Additionally, experiments were carried out to confirm
that the pCL polymers remained stable under the acidic
conditions of the cell assay. pCL-trehose and -zwitterion
polymers were heated to 60 °C for 30 min to mimic the
thermal stress conditions, then buffer was removed and the
materials were analyzed by GPC (Figure S42). No shift was
observed in the chromatogram, confirming that the polymers
were intact throughout the experiment.
The stabilizing abilities of the pCL polymers to protect G-
CSF from 60 °C thermal stress were additionally compared to
commonly used small-molecule excipients: sucrose, trehalose,
betaine, sorbitol, and Tween 80 (Figure 5). These compounds
Cytotoxicity and biocompatibility of the trehalose and
zwitterion based polycaprolactone polymers were also assessed
in human umbilical vein endothelial cells (HUVECs) as a
primary, noncancerous cell line. HUVECs were cultured in the
presence of pCL-trehalose20, pCL-zwitterion20, and their
polymeric degradation products. Compared to the control, no
reduction in cell viability was observed upon addition of either
polymer (pCL-trehalose and pCL-zwitterion) or polymeric
degradation products, up to 1 mg/mL, confirming that the
substituted polyesters and their eventual degradation products
are noncytotoxic (Figure S43).
Figure 5. Stabilization of G-CSF against thermal stress at 60 °C for 30
min and comparison of pCL-trehalose80 and pCL-zwitterion80 with
relevant small molecule controls. Data shown as the average of three
experimental repeats and six well repeats. # = no statistical difference
from the fresh control (p > 0.05). ‡ = no statistical difference from
sorbitol (p > 0.05). * = no statistical difference from Tween (p >
0.05).
Analysis of the substituted pCL polymers using transmission
electron microscopy (TEM) indicated the presence of
aggregated structures in both samples of pCL-zwitterion80
and pCL-trehalose80 alone and in the presence of G-CSF
(Figure S40). This self-assembly may play an important role in
the mechanism of stabilization and shows that the polymers are
nonionic surfactants, an important class of excipients.5 Similar
aggregates have been observed for tyloxapol, a polymeric
material with an aryl backbone and poly(ethylene oxide) side
chains that is structurally similar to the pCL polymers.65 We
additionally investigated the osmolyte character of the
synthesized materials using differential scanning calorimetry
(DSC). Both polymers changed the enthalpy of melting and
crystallization of water (Table S1) suggesting the polymers are
able to depress ice formation.26
were chosen to represent the materials present in the high-
performing pCL scaffolds, with the addition of sorbitol and
polysorbate 80 (Tween 80), which are used industrially in the
formulation of Neupogen (therapeutic GCSF)63 and sucrose,
which is a widely used excipient.5 The pCL-trehalose80 and
pCL-zwitterion80 polymers were selected because they were the
highest-performing pCL polymers in the experiments described
above and were added at 100 wt equiv. Small molecules were
added to be equivalent to the concentration of stabilizing units
in the pCL-zwitterion80 polymer except for Tween 80, which
was added at 100 wt equiv because of its larger molecular
weight, similar to the CL polymers. After heating to 60 °C for
30 min, sucrose, betaine, and sorbitol had little stabilizing effect
and the cell proliferation was low. However, the sorbitol and
Tween 80 maintained high protein activity that was statistically
equivalent to pCL-trehalose80 and pCL-zwitterion80, respec-
tively. The results show that the degradable polymers with DP
DISCUSSION
■
We have demonstrated that a biodegradable backbone can be
transformed into a library of potential protein stabilizers using
G
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
thiol−ene chemistry. This approach allowed us to survey the
effectiveness of various side chains without complications due
to differences in the number of backbone repeat units. We
tested five relevant side chains: three sugars, oligoPEG, and a
zwitterion. Yet, one can readily envision the synthesis of larger
libraries of side chains using this approach. To investigate side
chain effectiveness, the therapeutic protein G-CSF was
stabilized against refrigeration and heat, two stresses that are
relevant to the storage and shipping of many proteins; for
example, nearly 80% of current protein therapeutics need to be
refrigerated or frozen.66 This temperature requirement causes
inconvenience and increased costs to patients and may make
some therapeutics impossible for use in parts of the world that
do not have an effective cold chain. In addition, in some
industries such as personal care where products are stored at
room temperature, the instability of some proteins of interest
may preclude their use.
We found that pCL with trehalose- and zwitterion-
substituted side chains were the most effective stabilizers to
G-CSF to refrigeration and heating, with the zwitterion
polymers as the most effective over different time and
molecular weight ranges. Activity loss in G-CSF has been
reported to be a result of both methionine oxidation and
aggregation.55,67,68 Trehalose as an excipient has been shown to
have no effect on methionine oxidation of G-CSF, presumably
because it is preferentially excluded from the protein surface,
but has been shown to broadly inhibit aggregation of various
proteins.68,69 Zwitterionic materials are known to be nonfouling
and to repel proteins due to their high hydrophilicity and
strong hydration.70,71Additionally, the nonionic surfactant
Tween has been shown to reduce G-CSF aggregation through
micelle formation.72 Initial analyses by TEM and DSC suggest
that the trehalose- and zwitterion-substituted polymers form
structured aggregates alone and in the presence of G-CSF likely
due to the nonionic surfactant character of the polymers.
Additionally, the materials have the capability to reduce the
enthalpy of water crystallization and melting, equivalent to the
thermodynamic effects that have been previously observed for
their constituent side chain materials. Many current studies of
osmolyte−protein interactions hypothesize that their stabilizing
effect is in fact due to water−osmolyte interactions,73,74 and the
pCL materials are likely to be similar to small-molecule
osmolytes in this manner. The materials therefore combine two
different classes of known excipients, namely osmolytes and
nonionic surfactants. We also found the polymers to be as good
as excipients currently used in the formulation for Neulasta, a
therapeutic G-CSF. However, as has been previously noted,
sorbitol and polysorbate both present downsides to large-scale
and repeated applications in therapeutics. Namely, sorbitol has
been linked to GI tract problems and polysorbate has been
shown to undergo auto-oxidation.32,33 The substituted pCL
polymers offer equivalent stabilities and may be potential
alternatives to the clinically used additives for G-CSF.
weights. As we have demonstrated, the effect of molecular
weight on stabilization can be quite important, and the ability
to add molecular weight variation to a library of polymeric
stabilizers is significant. Additionally, the excellent control
provided by ROP conditions allows for delicate tuning of the
hydrolytic stability and degradability through selection of a
variety of cyclic monomers or even using copolymerization. We
anticipate this will greatly expand the possible applications for
these materials, and this work is underway.
Polymers have the additional advantage that they may be
processed as bulk materials and are widely used in biomedical
applications. For example, polycaprolactone is FDA-approved
as a copolymer with glycolide in the absorbable suture
Monocryl.75 We have previously shown that polystyrene with
trehalose side chains stabilizes proteins in the solid state.76
Therefore, it may also be possible to utilize these substituted
pCL polymers as solid-state protein stabilizers for a myriad of
applications where degradability is required. Thermal gravi-
metric analysis (TGA) has shown that pCL-trehalose20 is stable
to over 250 °C when heated (Figure S27), permitting use of
these materials at high temperature. Furthermore, polymers
such as PEG have been conjugated to proteins to increase their
in vivo stability via enhanced pharmacokinetic effects.34,77 It
should be possible to conjugate these polymers to a variety of
proteins to additionally stabilize them to environmental
stressors, and this work is underway.
CONCLUSIONS
■
A series of alkene-functionalized polyesters were synthesized by
organocatalyzed ring-opening polymerization. Postpolymeriza-
tion thiol−ene modification with a series of thiols led to well-
defined trehalose-, lactose-, glucose-, PEG-, and zwitterion-
based biodegradable polyesters. These biodegradable stabilizers
were investigated as to their ability to protect the therapeutic
protein G-CSF from storage and heat stressors. Side chains
containing trehalose and a zwitterionic carboxybetaine were
found to be the most effective at maintaining G-CSF activity.
Molecular weight studies of pCL-trehalose and pCL-zwitterion
were explored, and the polymers were shown to have moderate
molecular weight dependence to refrigeration, where larger
polymers (DP40 and DP80) demonstrated greater protein
stabilization to heat. Both high-performing polymer scaffolds
and their degradation products were also not cytotoxic up to at
least 1 mg/mL. These materials could be used for stabilization
of protein activity in therapeutic and industrial applications,
leading to improved performance and lowered cost.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b10776.
■
S
Experimental details and characterization of small
molecules and polymers, experimental protocol of
stability studies and results not in the main paper,
polymer stability, degradation studies, TEM data, DSC
data, and in vitro cytotoxicity (PDF)
Although we looked at G-CSF, it should be possible to utilize
this library approach to investigate a wide variety of proteins,
and the outcome may be different depending on the individual
protein degradation mechanism and the stress imposed. Using
the versatile thiol−ene strategy it should be possible to readily
alter the polymer side chains to identify stabilizers for a wide
variety of stressors. The use of controlled ring-opening
polymerization allows for the rapid synthesis of a variety of
molecular weights to compare to commercially available
additives, which may also be available in multiple molecular
AUTHOR INFORMATION
Corresponding Author
*E-mail: maynard@chem.ucla.edu.
■
ORCID
Heather D. Maynard: 0000-0003-3692-6289
H
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(25) Mancini, R. J.; Lee, J.; Maynard, H. D. J. Am. Chem. Soc. 2012,
134, 8474.
(26) Lee, J.; Lin, E.-W.; Lau, U. Y.; Hedrick, J. L.; Bat, E.; Maynard,
Notes
The authors declare the following competing financial
interest(s): A patent application for these materials is pending.
H. D. Biomacromolecules 2013, 14, 2561.
(27) Wada, M.; Miyazawa, Y.; Miura, Y. Polym. Chem. 2011, 2, 1822.
(28) Sizovs, A.; Xue, L.; Tolstyka, Z. P.; Ingle, N. P.; Wu, Y.; Cortez,
M.; Reineke, T. M. J. Am. Chem. Soc. 2013, 135, 15417.
(29) Hershfield, M. S.; Ganson, N. J.; Kelly, S. J.; Scarlett, E. L.;
Jaggers, D. A.; Sundy, J. S. Arthritis Res. Ther. 2014, 16, R63.
(30) Armstrong, J. K.; Hempel, G.; Koling, S.; Chan, L. S.; Fisher, T.;
Meiselman, H. J.; Garratty, G. Cancer 2007, 110, 103.
(31) Rudmann, D. G.; Alston, J. T.; Hanson, J. C.; Heidel, S. Toxicol.
Pathol. 2013, 41, 970.
(32) Abraham, S. C.; Bhagavan, B. S.; Lee, L. A.; Rashid, A.; Wu, T.
T. Am. J. Surg. Pathol. 2001, 25, 637.
(33) Kerwin, B. A. J. Pharm. Sci. 2008, 97, 2924.
(34) Pelegri-O’Day, E. M.; Lin, E. W.; Maynard, H. D. J. Am. Chem.
Soc. 2014, 136, 14323.
ACKNOWLEDGMENTS
■
This work was funded by the National Science Foundation
(NSF CHE-1507735). The AV500 NMR was supported by the
NSF under equipment grant number CHE-1049904. The MRL
Shared Experimental Facilities are supported by the MRSEC
Program of the NSF under Award No. DMR 1121053; a
member of the NSF-funded Materials Research Facilities
Network (www.mrfn.org). E.M.P.-O. thanks the NSF Graduate
Research Fellowship (DGE-1144087) and the Christopher S.
Foote Graduate Research Fellowship in Organic Chemistry.
S.J.P. thanks the NIH Chemistry−Biology Interface Training
Fellowship (T32-GM-008496) and the UCLA Graduate
Division for funding. The authors thank Dr. Uland Lau
(UCLA) for providing the G-CSF used in this study. The
authors additionally thank Dr. Rachel Behrens (MRL) for the
DSC analysis, Dr. Jacquelin Kammeyer (UCLA) for the TEM
analysis, and Jeong Hoon Ko (UCLA) for the MALDI analysis.
(35) Pelegri-O’Day, E. M.; Maynard, H. D. Acc. Chem. Res. 2016, 49,
1777.
(36) Taniguchi, I.; Mayes, A. M.; Chan, E. W. L.; Griffith, L. G.
Macromolecules 2005, 38, 216.
(37) Parrish, B.; Quansah, J. K.; Emrick, T. J. Polym. Sci., Part A:
Polym. Chem. 2002, 40, 1983.
(38) Parrish, B.; Breitenkamp, R. B.; Emrick, T. J. Am. Chem. Soc.
2005, 127, 7404.
REFERENCES
■
(1) Maity, H.; O’Dell, C.; Srivastava, A.; Goldstein, J. Curr. Pharm.
Biotechnol. 2009, 10, 761.
(39) Lowe, A. B. Polym. Chem. 2010, 1, 17.
(40) Hoyle, C. E.; Bowman, C. N. Angew. Chem., Int. Ed. 2010, 49,
1540.
(2) Bischof, J. C.; He, X. M. Ann. N. Y. Acad. Sci. 2005, 1066, 12.
(3) Pikal-Cleland, K. A.; Carpenter, J. F. J. Pharm. Sci. 2001, 90, 1255.
(4) Sluzky, V.; Tamada, J. A.; Klibanov, A. M.; Langer, R. Proc. Natl.
Acad. Sci. U. S. A. 1991, 88, 9377.
(5) Kamerzell, T. J.; Esfandiary, R.; Joshi, S. B.; Middaugh, C. R.;
Volkin, D. B. Adv. Drug Delivery Rev. 2011, 63, 1118.
(6) Arakawa, T.; Timasheff, S. N. Biophys. J. 1985, 47, 411.
(7) Falconer, R. J.; Chan, C.; Hughes, K.; Munro, T. P. J. Chem.
Technol. Biotechnol. 2011, 86, 942.
(8) Arakawa, T.; Tsumoto, K.; Kita, Y.; Chang, B.; Ejima, D. Amino
Acids 2007, 33, 587.
(9) Chen, B.; Bautista, R.; Yu, K.; Zapata, G. A.; Mulkerrin, M. G.;
Chamow, S. M. Pharm. Res. 2003, 20, 1952.
(10) Fang, W.-J.; Qi, W.; Kinzell, J.; Prestrelski, S.; Carpenter, J. F.
Pharm. Res. 2012, 29, 3278.
(11) Garzon-Rodriguez, W.; Koval, R. L.; Chongprasert, S.; Krishnan,
S.; Randolph, T. W.; Warne, N. W.; Carpenter, J. F. J. Pharm. Sci.
2004, 93, 684.
(12) Montaperto, A. G.; Gandhi, M. A.; Gashlin, L. Z.; Symoniak, M.
R. Curr. Med. Res. Opin. 2016, 32, 155.
(13) Martin, N.; Ma, D.; Herbet, A.; Boquet, D.; Winnik, F. M.;
Tribet, C. Biomacromolecules 2014, 15, 2952.
(14) Lee, E.-H.; Tsujimoto, T.; Uyama, H.; Sung, M.-H.; Kim, K.;
Kuramitsu, S. Polym. J. 2010, 42, 818.
(15) Izaki, S.; Kurinomaru, T.; Handa, K.; Kimoto, T.; Shiraki, K. J.
Pharm. Sci. 2015, 104, 2457.
(16) Gombotz, W. R.; Pankey, S. C.; Phan, D.; Drager, R.;
Donaldson, K.; Antonsen, K. P.; Hoffman, A. S.; Raff, H. V. Pharm.
Res. 1994, 11, 624.
(17) Stidham, S. E.; Chin, S. L.; Dane, E. L.; Grinstaff, M. W. J. Am.
Chem. Soc. 2014, 136, 9544.
(41) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Waymouth, R.
M.; Hedrick, J. L. J. Am. Chem. Soc. 2006, 128, 4556.
(42) Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.;
Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.;
Waymouth, R. M.; Hedrick, J. L. Macromolecules 2006, 39, 8574.
(43) Silvers, A. L.; Chang, C.-C.; Emrick, T. J. Polym. Sci., Part A:
Polym. Chem. 2012, 50, 3517.
(44) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.;
Lohmeijer, B. G. G.; Hedrick, J. L. Chem. Rev. 2007, 107, 5813.
(45) Stevens, D. M.; Watson, H. A.; LeBlanc, M.-A.; Wang, R. Y.;
Chou, J.; Bauer, W. S.; Harth, E. Polym. Chem. 2013, 4, 2470.
(46) Campos, L. M.; Killops, K. L.; Sakai, R.; Paulusse, J. M. J.;
Damiron, D.; Drockenmuller, E.; Messmore, B. W.; Hawker, C. J.
Macromolecules 2008, 41, 7063.
(47) Yu, M.; Yang, Y.; Han, R.; Zheng, Q.; Wang, L.; Hong, Y.; Li, Z.;
Sha, Y. Langmuir 2010, 26, 8534.
(48) Erdem, S. S.; Nesterova, I. V.; Soper, S. A.; Hammer, R. P. J.
Org. Chem. 2009, 74, 9280.
(49) Aoi, K.; Tsutsumiuchi, K.; Okada, M. Macromolecules 1994, 27,
875.
(50) Takasu, A.; Kojima, H. J. Polym. Sci., Part A: Polym. Chem. 2010,
48, 5953.
(51) Dong, C. M.; Faucher, K. M.; Chaikof, E. L. J. Polym. Sci., Part
A: Polym. Chem. 2004, 42, 5754.
(52) Takasu, A.; Houjyou, T.; Inai, Y.; Hirabayashi, T.
Biomacromolecules 2002, 3, 775.
(53) Cao, Z.; Yu, Q.; Xue, H.; Cheng, G.; Jiang, S. Angew. Chem., Int.
Ed. 2010, 49, 3771.
(54) Aapro, M. S.; Cameron, D. A.; Pettengell, R.; Bohlius, J.;
Crawford, J.; Ellis, M.; Kearney, N.; Lyman, G. H.; Tjan-Heijnen, V.
C.; Walewski, J.; Weber, D. C.; Zielinski, C.; Eortc, G. C. Eur. J. Cancer
2006, 42, 2433.
(55) Krishnan, S.; Chi, E. Y.; Webb, J. N.; Chang, B. S.; Shan, D. X.;
Goldenberg, M.; Manning, M. C.; Randolph, T. W.; Carpenter, J. F.
Biochemistry 2002, 41, 6422.
(18) Taluja, A.; Bae, Y. H. Mol. Pharmaceutics 2007, 4, 561.
(19) Mazzaferro, L.; Breccia, J. D.; Andersson, M. M.; Hitzmann, B.;
Hatti-Kaul, R. Int. J. Biol. Macromol. 2010, 47, 15.
(20) Andersson, M. A.; Hatti-Kaul, R. J. Biotechnol. 1999, 72, 21.
(21) Nguyen, T. H.; Kim, S.-H.; Decker, C. G.; Wong, D. Y.; Loo, J.
A.; Maynard, H. D. Nat. Chem. 2013, 5, 221.
(22) Matsusaki, M.; Serizawa, T.; Kishida, A.; Akashi, M.
Biomacromolecules 2005, 6, 400.
(56) Shirafuji, N.; Asano, S.; Matsuda, S.; Watari, K.; Takaku, F.;
Nagata, S. Exp. Hematol. 1989, 17, 116.
(57) Wu, J.; Zhao, C.; Lin, W.; Hu, R.; Wang, Q.; Chen, H.; Li, L.;
Chen, S.; Zheng, J. J. Mater. Chem. B 2014, 2, 2983.
(23) Keefe, A. J.; Jiang, S. Nat. Chem. 2011, 4, 59.
(24) Yoshimoto, N.; Hashimoto, T.; Felix, M. M.; Umakoshi, H.;
Kuboi, R. Biomacromolecules 2003, 4, 1530.
I
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(58) Jayaraman, S.; Gantz, D. L.; Gursky, O. Biochemistry 2004, 43,
5520.
(59) Young, D. ISTA Temperature Project- Data Summary;
International Safe Transit Association: East Lansing, MI, 2002.
(60) Katyal, N.; Deep, S. Phys. Chem. Chem. Phys. 2014, 16, 26746.
(61) Fedorov, M. V.; Goodman, J. M.; Nerukh, D.; Schumm, S. Phys.
Chem. Chem. Phys. 2011, 13, 2294.
(62) Lees, W. J.; Spaltenstein, A.; Kingerywood, J. E.; Whitesides, G.
M. J. Med. Chem. 1994, 37, 3419.
(63) Alebouyeh, M.; Tahzibi, A.; Yaghoobzadeh, S.; Zahedy, E. T.;
Kiumarsi, S.; Soltanabad, M. H.; Shahbazi, S.; Amini, H. Biologicals
2016, 44, 150.
(64) Woodruff, M. A.; Hutmacher, D. W. Prog. Polym. Sci. 2010, 35,
1217.
(65) Regev, O.; Zana, R. J. Colloid Interface Sci. 1999, 210, 8.
(66) Leader, B.; Baca, Q. J.; Golan, D. E. Nat. Rev. Drug Discovery
2008, 7, 21.
(67) Pan, B.; Abel, J.; Ricci, M. S.; Brems, D. N.; Wang, D. I. C.;
Trout, B. L. Biochemistry 2006, 45, 15430.
(68) Yin, J.; Chu, J. W.; Ricci, M. S.; Brems, D. N.; Wang, D. I. C.;
Trout, B. L. Pharm. Res. 2005, 22, 141.
(69) Ohtake, S.; Wang, Y. J. J. Pharm. Sci. 2011, 100, 2020.
(70) Chen, S. F.; Zheng, J.; Li, L. Y.; Jiang, S. Y. J. Am. Chem. Soc.
2005, 127, 14473.
(71) Shao, Q.; Jiang, S. Y. Adv. Mater. 2015, 27, 15.
(72) Niven, R. W.; Prestrelski, S. J.; Treuheit, M. J.; Ip, A. Y.;
Arakawa, T. Int. J. Pharm. 1996, 127, 191.
(73) Bruzdziak, P.; Panuszko, A.; Stangret, J. J. Phys. Chem. B 2013,
117, 11502.
(74) Street, T. O.; Bolen, D. W.; Rose, G. D. Proc. Natl. Acad. Sci. U.
S. A. 2006, 103, 13997.
(75) Bezwada, R. S.; Jamiolkowski, D. D.; Lee, I. Y.; Agarwal, V.;
Persivale, J.; Trenkabenthin, S.; Erneta, M.; Suryadevara, J.; Yang, A.;
Liu, S. Biomaterials 1995, 16, 1141.
(76) Bat, E.; Lee, J.; Lau, U. Y.; Maynard, H. D. Nat. Commun. 2015,
6, 6654.
(77) Caliceti, P.; Veronese, F. M. Adv. Drug Delivery Rev. 2003, 55,
1261.
J
DOI: 10.1021/jacs.6b10776
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX