Cholesterol modification of (bio)polymers using UV-vis traceable chemistry in aqueous solutions

Article abstract of DOI:10.1002/mabi.201300286

Cholesterol modification of synthetic and biological polymers is achieved using of thiocholesterol (TC) and thiol-disulfide exchange. TC is reacted with Ellman's reagent to produce a mixed disulfide (TC-ER) which is activated towards thiol-disulfide exchange. TC-ER is used to obtain an inclusion complex with methyl-β-cyclodextrin, which is then employed to achieve cholesterol functionalization of a model peptide, synthetic polymers, and physical hydrogels based on poly(vinyl alcohol). It is anticipated that the established techniques will significantly broaden the use of cholesterol in bio- and nanotechnology and specifically biomedicine.

Full text of DOI:10.1002/mabi.201300286

Full Paper
Cholesterol Modification of (Bio)Polymers
Using UV-Vis Traceable Chemistry in
Aqueous Solutions
Kasper F. Rasmussen, Anton A. A. Smith, Pau Ruiz-Sanchis, Katrine Edlund,
Alexander N. Zelikin*
Cholesterol modification of synthetic and biological polymers is achieved using of
thiocholesterol (TC) and thiol-disulfide exchange. TC is reacted with Ellman’s reagent to
produce a mixed disulfide (TC-ER) which is activated towards thiol-disulfide exchange. TC-ER
is used to obtain an inclusion complex with
methyl-b-cyclodextrin, which is then employed
to achieve cholesterol functionalization of a
model peptide, synthetic polymers, and physical
hydrogels based on poly(vinyl alcohol). It is
anticipated that the established techniques will
significantly broaden the use of cholesterol
in bio- and nanotechnology and specifically
biomedicine.
1. Introduction
the liver. Modification of water-soluble polymers with
hydrophobic groups such as cholesterol aims to induce
formation of micelles for drug delivery.[8] Also, cholesterol-
aided linkages between polymers and liposomes are
used to create sub-compartmentalized assemblies as cell
mimics.[9] These applications necessitate facile, reliable
techniques in bioconjugation using cholesterol. However,
while multiple approaches are indeed available, there
are several fundamental challenges associated with
cholesterol chemistry. These include i) solubility profile
of cholesterol – which is not compatible with physio-
logical conditions, ii) ‘‘invisible’’ nature or an ill-defined
signature of this molecule in most characterization
techniques (UV-Vis, NMR, FTIR, etc.), and iii) purification
difficulties, that is, removal of non-specifically bound
cholesterol from chemically conjugated product. The main
objective of this work was to investigate opportunities in
Cholesterol is an important building block in nature and a
recognizedtoolofbio/nanotechnology.[1] Functionalization
with cholesterol is pursued for diverse biomedical and
biotechnological needs such as PEGylation of liposomes
through cholesterol-aided anchoring of the polymer into
the lipid bilayer,[2–4] RNAi techniques wherein conjugation
of the nucleic acid with cholesterol enhances cellular
uptake,[5] increases stability to degradation by nucleases,[6]
and aids in targeting to the desired tissues,[7] specifically
K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund,
A. N. Zelikin
Department of Chemistry, Aarhus University, Langelandsgade
140, DK-8000 Aarhus C, Denmark
E-mail: zelikin@chem.au.dk
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
Macromol. Biosci. 2014, DOI: 10.1002/mabi.201300286
33

K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund, A. N. Zelikin
www.mbs-journal.de
bioconjugation with cholesterol and (bio)polymers in
aqueous conditions and using chemistries with facile
tracing such as UV-Vis spectra.
was synthesized as a linker for immobilization of anti-
bodies on the surface of polymer hydrogel capsules
for targeting applications.[10] For PVA, we used the
same RAFT agent as mentioned above and produced
amine-terminated PVA chains for facile fluorescence
labeling of chains,[11] site-specific thiolation,[22] and
bioconjugation with oligopeptides.[28] In the latter case,
bioconjugation was performed under physiological con-
ditions using both, polymer chains in solution and
physical hydrogels as surface-adhered biointerfaces
thus establishing novel opportunities in functionaliza-
tion of these matrices for biomedical applications. We
note that for cholesterol functionalization in particular,
RAFT-associated techniques have already documented
success in growing polymer chains from cholesterol-
derived RAFT agents[29,30] and using cholesterol (meth)
acrylate for direct polymer synthesis.[1,24] However, due
to the above mentioned limitations of synthesis, these
techniques have not been implemented for cholesterol-
functionalization of PVP and PVA. For RAFT derived
polymers, a readily available site for conjugation is
the thiol functionality produced via hydrolysis of the
RAFT agent[31,32] yet this approach too is futile for PVP
and PVA due to a well-documented rearrangement of
the Z-group-derived terminal thiol.[33,34]
With a broader interest in macromolecular design,
bioconjugation, and hydrogel biomaterials,[10–12] as
partners to cholesterol in developing conjugation techni-
ques we used water-soluble synthetic polymers – poly-
(vinylpyrrolidone) (PVP), and poly(vinyl alcohol) (PVA).
These polymers are among the most well-characterized
materials in biomedicine and have regulatory approval
for clinical applications in humans yet scant opportu-
nities in bioconjugation. For PVP, opportunities in the
synthesis of copolymers with ‘‘activated esters’’ for
subsequent modification of the polymer chains via
polymer analogous reaction[13] are limited due to
incompatibility of the monomers. Inertness of the
lactam ring to modification or ring opening further
explains failure of this polymer to enter the mainstream
of polymer therapeutics or polymer-protein conjugation
techniques. Success of ‘‘NVP-alike’’ monomers in produc-
tion of copolymers amenable for subsequent chemical
modification and/or bioconjugation[14–16] was also mod-
erate, largely due to laborious synthesis of monomers.
For PVA, polymer synthesis proceeds via hydrolysis
of the parent polymer, poly(vinyl acetate), which is
typically performed at conditions which are incompatible
with labile esters or activated disulfide functionali-
ties[13,17–19] thus minimizing post-conjugation opportu-
nities.[20] PVA offers numerous hydroxyl groups for
bioconjugation and indeed, conjugation of PVA with
cholesterol has been achieved,[8] however, this conjuga-
tion strategy relied on non-physiological conditions for
bioconjugation. Also, hydroxyls are significantly disad-
vantaged compared to the classic sites of bioconjugation
(amines and thiols) in terms of specificity of reactions.
Finally, our specific interests lie in investigating
physical hydrogels based on poly(vinyl alcohol) toward
their use in creation of intelligent biointerfaces.[21–25]
These matrices are labile and their chemical modification
is possible only in aqueous conditions.
The outline of this report is as follows. First, we
revisit the synthesis of amine-functionalized PVP and
PVA with an emphasis on removal of the phthalimide
group from the RAFT derived polymer chains. Subse-
quently, we investigate cholesterol-modification of
polymers via carbamate linkages. Next, we outline
several approaches to polymer conjugation with thio-
cholesterol via thiol-disulfide reshuffling with an em-
phasis on producing water-soluble form of cholesterol
and performing reactions in aqueous conditions. Finally,
we achieve cholesterol attachment onto matrices com-
prised of labile, physical hydrogels based on PVA. Taken
together, this work contributes significantly to the
development of methods to literally link materials
science and a nature-derive building block of nano-
technology, cholesterol.
As an approach to bioconjugation with PVP and
PVA, we use RAFT derived polymers and associated
opportunities in the design of terminal groups.[26] For PVP,
we have previously achieved polymer conjugation with
oligopeptides and oligomeric nucleic acids, specifically
using O-ethylphthalimidomethyl xanthate RAFT agent
and removal of the terminal phthalimide (Phth) groups
using hydrazine hydrate to afford amine-functionalized
polymer chains.[27] We took advantage over adsorptive
behavior of PVP toward silica surfaces and performed
bioconjugation reactions as on-bead, solid support
reactions and this afforded ease of purification and
facilitated product analysis via flow cytometry. Subse-
quently, amine- and thiol-containing telechelic PVP
2. Experimental Section
2.1. Materials and Instrumentation
Unless stated otherwise, all chemical were obtained from Sigma-
Aldrich and used without purification. 1H NMR spectra were
recorded on a Varian Mercury 400 MHz spectrometer. Spectra
were referenced to the residual solvent peak: CDCl3 (d ¼ 7.26),
D2O (d ¼ 4.79), and DMSO-d6 (d ¼ 2.50). IR spectra were obtained
on an iD5 ATR integrated (Thermo Scientific, USA) in a Nicolet iS5
(Thermo Scientific, USA). HRMS was performed on a Micromass
LC-TOF spectrometer with positive electrospray ionization.
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
34
www.MaterialsViews.com

Cholesterol Modification of (Bio)Polymers Using . . .
www.mbs-journal.de
Bz-Gly-Phe-OMe (C19H20N2O4Na, m/z 363.132) or Boc-Ser(OBn)-
Ser-Leu-OMe (C25H39N3O8Na, m/z 532.2635) were used as internal
standards. Analytical HPLC was performed on a Shimadzu LC-
2010A HT equipped with a Shimadzu SPD-M-20A PDA detector and
Peptide ES-C18 column with 2.7 mm particles, a length of 150 mm
and an internal diameter of 3 mm from Sigma-Aldrich. The
temperature was fixed at 40 8C; acetonitrile and water [with
0.045% trifluoroacetic acid (TFA)] were used as mobile phases. The
gradient was changed from 30 to 100% MeCN within 15 min and
then 100% MeCN was used for an additionally 30 min. Polymer
characterizations were performed on a Shimadzu gel permeation
chromatography (GPC) instrument consisting of: an auto sampler
(SIL-10AF), a solvent pump (LC-20AD), a column oven (CTO-20A), a
light scattering (LS) detector (DAWN HELEOS 8), a differential
refractive index detector (RID-10A), and last a PDA detector (SPD-
M-20A). The used column for DMF measurements was a Mz-Gel
SDplus Linear column with 5 mm particles, a length of 300 mm and
an internal diameter of 8 mm. The eluent was DMF (with 0.010 M
LiBr) and the samples were run at 30 8C with a flow rate of
1 mL minꢀ1. The column used for measurements in water was a
HEMA-Bio Linear column with 10 mm particles, a length of 300 mm
and an internal diameter of 8 mm. The eluent was purified MQ
water (with NaN3 300 ppm).
2.2.3. a-Aminomethyl-PVP (3)
2 (2.010 g, 13 kDa) and NH2NH2 ꢂ H2O (270 mg, 263 mL, 5.39 mmol)
were combined in MeOH (8 mL). The mixture was heated for 1 h
at 60 8C. The product was obtained by precipitation in Et2O and
filtration to give 1.742 g of 3, a colorless solid. 1H NMR (CDCl3):
d ¼ 0.50–2.50 (—CH2CH2CH2CON—, CH2CH—N, CH3CH2O—); 3.20
(—CH2CH2CH2CON—); 3.71 (CH2CH—N).
2.2.4. a-Cholesterylcarbonylaminomethyl-PVP (4)
A solution of cholesteryl chloroformate (22.2 mg, 24.7 mmol) in dry
dichloromethane (DCM, 1 mL) was added dropwise to a solution
of triethylamine (TEA, 2 mL, 1.45 mg, 14.3 mmol) and 3 (204.4 mg,
13.3 kDa) in dry DCM (1 mL) at 0 8C. The reaction was stirred for
12 h at room temperature. The solvent was removed and the
crude was re-dissolved in CHCl3; the polymer was precipitated
in Et2O and filtrated. The obtained compound 4 was a colorless
solid (71.0 mg, cholesteryl ester loading ¼ 91%). 1H NMR (CDCl3):
d ¼ 0.67 (s, 3H, —C—CH3); 0.85 (d, J ¼ 1.6 Hz, 3H, —CH(CH3)2); 0.86
(d, J ¼ 1.6 Hz, 3H, —CH(CH3)2); 0.92 (d, J ¼ 6.9 Hz, 3H, —CH—CH3);
0.99 (s, 3H, —CH—CH3); 1.00–2.50 (—CH2CH2CH2CON—, CH2CH—
N—, CH3CH2O—); 3.20 (—CH2CH2CH2CON—); 3.72 (CH2CH—N);
5.35 (m, 1H, —C55CH—). IR (ATR, solid): n ¼ 1 662 cmꢀ1 (vs, >N—
C55O).
2.2. Syntheses
2.2.5. a-N-{4-[2-(3-carboxy-4-nitrophenyl)dithio]-1-
iminobutyl]aminomethyl}-PVP (5)
2.2.1. O-Ethyl-S-Phthalimidylmethyl Xanthate (1)
3 (253.6 mg), Traut’s reagent (20.1 mg, 146 mmol) and Ellman’s
reagent (80.9 mg, 204 mmol) were suspended in buffer 0.1
NaHCO3 (10 mL, pH ¼ 8.3). The reaction was stirred for 14 h.
A NAP-25 column was used for purification and the collected
fraction was freeze-dried to give a pale red solid 5 (163.2 mg). The
Ellman’s reagent loading was estimated by UV-measurements
(179%). 1H NMR (CDCl3): d ¼ 1.20–2.50 (—CH2CH2CH2CON—,
CH2CH—N, CH3CH2O—, —S—CH2—CH2—CH2—C—); 3.13
(—CH2CH2CH2CON—); 3.66 (CH2CH—N); 7.74 (b s, 2H, ar); 8.06
(b s, 1H, CHar).
A solution of N-(bromomethyl)phthalimide (233.8 mg, 0.98 mmol)
in CHCl3 (6 mL) was added dropwise to a solution of O-ethylxanthic
acid potassium salt (296.1 mg, 1.85 mmol) in CHCl3 (5 mL). The
solution was stirred at room temperature for 14 h. The reactionwas
followed by thin-layer chromatography (TLC, Rf ¼ 0.6–0.7 CHCl3).
The reaction was diluted with CHCl3 (5 mL) and the solution was
washed with water (2 ꢁ 10 mL) and brine (15 mL). The organic layer
was dried over magnesium sulfate and then filtered, and the
solvent was removed under vacuum affording the target
compound, a pale yellow solid product (170.4 mg, 0.6 mmol,
72%,). 1H NMR (CDCl3): d ¼ 1.45 (t, J ¼ 7.1 Hz, 3H, CH3CH2—); 4.67
(q, J ¼ 7.1 and 7.1 Hz, 2H, CH3CH2—); 5.33 (s, 2H, S—CH2—N); 7.74
(q, J ¼ 3.1 and 5.4 Hz, 2H, CHar), 7.87 (q, J ¼ 3.1 and 5.4 Hz, 2H, CHar).
13C NMR (CDCl3): d ¼ 13.7 (q, CH3CH2—); 41.2 (t, N—CH2—S); 70.5
(t, CH3CH2—); 123.6 (2d, CHar); 131.81 (2s); 134.39 (2d, CHar); 166.6
(2s, C55O); 210.2 (s, C55S).
2.2.6. a-{N-[4-(thiocholesteryl)thio]-1-iminobutyl]
aminomethyl}-PVP (6)
5 (100 mg) was suspended in CHCl3 (2 mL) and degassed for 30min.
A solution of thiocholesterol (47.2 mg, 11.7 m mol) in CHCl3 (5 mL),
which has been previously degassed for 30 min, was added to the
PVP solution. The reaction wasstirred for 14 h. A red solid, 6 (72.4 mg)
wasobtainedbyprecipitationinEt2O. 1HNMR(CDCl3): d¼ 0.65(s,3H,
—C—CH3); 0.84 (s, 3H, —CH(CH3)2); 0.86 (s, 3H, —CH(CH3)2); 0.89
(d, J ¼ 6.6 Hz, 3H, —CH—CH3); 0.98 (s, 3H, —CH—CH3); 1.00–2.50
(—CH2CH2CH2CON—, CH2CH—N); 3.19 (—CH2CH2CH2CON—); 3.72
(CH2CH—N); 5.31 (b s, 1H, —C55CH—).
2.2.2. a-Phthalymidyl-v-(O-ethylcarbonodithiotyl)-PVP (2)
1
(132.5 mg, 471 mmol), azobisisobutyronitrile (AIBN) (7.3mg,
44.4 mmol), and NVP (10 mL, 1.043 g, 9.9 mmol) were combined in
toluene (20 mL). The reaction mixture was degassed by three freeze-
pump-thawcyclesandthenheatedfor16hat608C.Thereactionwas
quenched by rapidly cooling with liquid nitrogen. The product was
isolated by precipitation into diethyl ether, filtration and removal of
solvent on the vacuum line to give 2 in 4.5 g, as colorless solid. GPC
analysis yielded a Mn of 13 kDa and a Ð of 1.13. 1H NMR (CDCl3):
d¼ 0.50–2.50 (—CH2CH2CH2CON—, CH2CH—N, CH3CH2O—); 3.22
(—CH2CH2CH2CON—); 3.74 (CH2CH—N); 7.71 (m, 2H, CHar); 7.82
(m, 2H, CHar).
2.2.7. 5-[(Thiocholesteryl)thio]-2-nitrobenzoic acid (7)
A solution of Ellman’s reagent (99.8 mg, 0.252 mmol) in degassed
tetrahydrofuran (THF, 5 mL) and TEA (200 mL) were added to a
degassed solution of thiocholesterol in THF (103.4 mg in 5 mL,
0.257 mmol). The color of the solution immediately changed from
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
35
www.MaterialsViews.com

K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund, A. N. Zelikin
www.mbs-journal.de
colorless to strong orange. The solution was stirred for 14 h at room
temperature.THFwasremovedundervacuumgivinganorangeoil.
The product was isolated trough a column chromatography (DCM/
MeOH, 9:1) affording 7, a pale yellow solid (76.0 mg, 51%). 1H NMR
(CDCl3): d ¼ 0.65 (s, 3H, —CH3); 0.85 (s, 3H, —CH3); 0.87 (s, 3H,
—CH3); 0.89 (s, 3H, —CH3); 0.91 (s, 3H, —CH3); 1.71–2.04 (m, CH2);
2.26 (b s, 2H, CH2); 2.67 (s, 1H, —CH—S—);5.21 (s, 1H,55CH—);7.32–
7.65 (m, 3H, CHar). HRMS (ESIþ) calculated for C34H49NO4S2Na m/z
(M þ Naþ) 622.2995, found 622.3001.
d¼ 0.67 (s, 3H, —C—CH3); 0.84–0.86 (m, 6H, —CH—(CH3)2); 0.89–0.91
(m, 6H, —CH—CH3); 0.99–2.50 (—CH2CH2CH2CON—, CH2CH—N);
3.19 (—CH2CH2CH2CON—); 3.40–4.10 (CD); 4.99 (CD).
2.2.12. Aqueous Conjugation of Thiocholesterol with GSH
For kinetic measurements, in an Eppendorf tube, stock solutions
of 10 and GSH in 0.1 M carbonate buffer, pH ¼ 8.3 were combined
to afford 10 g Lꢀ1 concentration of the inclusion complex and
5 ꢁ 10ꢀ3 M GSH. Reaction mixture was incubated on a rotary shaker
taking UV-Vis measurements on a NanoDrop instrument at
indicated time points. To determine an optical density correspond-
ing to a full conversion of thiol-disulfide exchange, dithiothreitol
(DTT, 1 mL, 100 g Lꢀ1) was added instead of GSH.
2.2.8. a-{N-[4-(thiocholesteryl)thio]-1-iminobutyl}
aminomethyl-PVP (8)
Traut’s reagent (9.8 mg, 71 mmol) and 3 (229.1mg, 22.9mmol) were
dissolved in 0.1 M NaHCO3 (4mL, pH ¼ 8.28). A solution of 7 (42 mg,
70 mmol) in THF (2 mL) and buffer (0.1mL, NaHCO3, 0.1 M, pH ¼ 8.28)
wasadded.Thereactionwasstirredovernightatroomtemperature.
THF was removed under vacuum, the precipitate was centrifuged
(14 000 rpm, 2 min) and the supernatant filtered (0.2 mm filter).
The product was isolated by NAP-25 column and freeze-dried to
give 8, a light brown solid (57.4mg). 1H NMR (CDCl3): d ¼ 0.66
(d, J ¼ 6.1 Hz, 3H, —C—CH3); 0.84–0.86 (m, 6H, —CH—(CH3)2); 0.88–
0.94 (m, 3H, —CH—CH3); 0.99 (s, 3H, —CH—CH3); 1.00–2.70
(—CH2CH2CH2CON—, CH2CH—N); 3.20 (—CH2CH2CH2CON—);
3.72 (CH2CH—N); 5.31 (m, 1H, —C55CH—).
For estimation of GSH thiol-to-disulfide conversion, aqueous
solutions were prepared in 0.1M NaHCO3 buffer (pH ¼ 8.3) to a final
concentration of the inclusion complex of 10g Lꢀ1 and a varied
concentration of GSH. Mixtures were incubated on a rotary shaker
for14 huponwhichUV-measurementswereperformedtoestimate
the reaction conversions. Optical density corresponding to a full
reaction conversion was estimated using DTT as described above.
2.2.13. Hydrogel Functionalization
Surface adhered, micro-structured PVA hydrogels were prepared
via micro-transfer molding using a sample of reversible addition/
fragmentation chain transfer (RAFT)-derived, amine-containing
PVA (31 kDa) as described in detail previously.[28] In brief, a drop
of 12 wt% polymer solution was placed between a glass cover slip
and a poly(dimethylsiloxane) stamp with cylindrical cavities
(diameter 5 mm, center-to-center spacing 10 mm, depth 2 mm) and
clamped at finger tight pressure overnight. Disassembly of the
clamps afforded surfaced adhered materials which underwent a
treatment using 0.5 M sodium sulfate for 1 h and rehydration
in phosphate-buffered saline (PBS, pH ¼ 7.4) to afford a final
preparation, surface adhered hydrogels.
2.2.9. Inclusion Complex of Thiocholesterol in Methyl-b-
cyclodextrin (9)
Thiocholesterol (1.4 mg) was dissolved in chloroform/methanol
2:1 (1.4 mL). Methyl-b-cyclodextrin (200 mg) was dissolved in
deionized water (4 mL) to give a 5 wt% solution. Thiocholesterol
solution was added in small (0.2 mL) aliquots to the solution at
80 8C, and reacted for 2 min per addition. The solution was filtered
through a 0.2 mm filter and then the water was removed by freeze-
drying to give 9, a white solid (178.3 mg, thiocholesterol inclusion
¼ 3 mg per 1 g solid).
Terminal amine groups on polymer chains comprising hydrogel
samples were converted into thiol groups via immersion of the
specimen into a 0.010 M solution of iminothiolane in carbonate
buffer pH ¼ 8.3 supplemented with sodium sulfate to 0.25M
(hereandbelowdenotedascarbonate-sulfatebuffer)andincubation
for 30min. Following this, cover slips were washed with copious
amountsofcarbonate-sulfatebuffer.ReactionwithEllman’sreagent
was performed via adding 10mL of 10ꢀ3 M solution of Ellman’s
reagent to 200 mL of carbonate-sulfate buffer covering the hydrogel
sample and reaction times 15 or 60min. Reaction with inclusion
complex7wasperformedthroughadding10mLofa63g Lꢀ1 solution
of 7 (corresponding to ꢃ 10ꢀ3 M TC-ER) to 200 mL carbonate-sulfate
buffer covering the hydrogel samples and incubation time 15
or 60min. Progression of reactions was monitored via UV-Vis
spectroscopy, specifically quantifying optical density at 412 nm.
2.2.10. Inclusion Complex of 5-[(Thiocholesteryl)thio]-2-
nitrobenzoic acid in Methyl-b-cyclodextrin (10)
Methyl-b-cyclodextrin (1.0 mg, 0.76 mmol) was dissolved in
H2O (20 mL) affording a 5 wt% solution. A solution of 7 (35.0 mg,
58.3 mmol) in a mixture of CHCl3/MeOH (2:1, 3.5 mL) was added in
small aliquots (0.25 mL) at 80 8C, and reacted for 2 min per addition.
The solution was filtered (0.2 mm filter) and freeze-dried to give 10,
a white solid (978 mg, 30 mg of TC-TNB per 1 g solid).
2.2.11. Aqueous Thiocholesterol Modification of
Amine-Containing PVP (Thiol Disulfide Exchange
Between 10 and 3)
Solutions of amine-containing PVP (100 g Lꢀ1, 100mL), Traut’s reagent
(1 g Lꢀ1, 860ml) and 10 (50 g Lꢀ1, 3ml) 0.1M carbonate buffer, pH ¼ 8.3
were combined in a reaction vial and incubated on a rotary shaker for
4 h. Thesolutionwaspurifiedbydialyzingwith, first, carbonatebuffer
(38 h), then, MilliQ water (36 h). The crude was freeze-dried giving
a colorless solid, re-dissolved, filtered (0.4 mm filter), and freeze-
dried again affording a colorless solid (14.77 mg). 1H NMR (CDCl3):
3. Results and Discussion
3.1. Polymer Synthesis
The synthesis of PVP with controlled molecular weight and
appropriate end-group for further synthetic modifications
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
36
www.MaterialsViews.com

Cholesterol Modification of (Bio)Polymers Using . . .
www.mbs-journal.de
was accomplished using O-ethyl-S-(phthalimidylmethyl)
xanthate as a RAFT agent. The latter was synthesized
starting from N-(bromomethyl)phthalimide and potassi-
um ethyl xanthate, Scheme 1.[35] Polymerizations were
conducted combining RAFT agent, 1-vinyl-2-pyrrolidone,
and AIBN as initiator in toluene at 60 8C. Molecular
weights of the resulting polymers were independently
estimated on the basis of 1H NMR, comparing integrals
corresponding to phthalimide and pyrrolidone protons,
and via a gel permeation chromatography performed in
DMF containing lithium bromide, as described in the
Experimental Section.
of that registered for Phth-PVP (Figure 1B). Removal of the
terminal group was also verified via 1H NMR spectroscopy.
This method revealed otherwise identical spectra for
polymer before and after hydrazinolysis and confirmed
disappearance of the signals corresponding to Phth group
(Figure 1C) thus, corroborating above discussed GPC
analysis.
3.2. Conjugation with Cholesteryl Chloroformate
For proof-of-concept functionalization of polymers with
cholesterol via terminal amine groups, we chose a well-
established synthetic approach based on cholesterol
chloroformate, a derivative which is reactive toward classic
sitesof bioconjugation, amines. This reagent haspreviously
been used in the design of lipopolymers,[36,37] pH-respon-
sive polymeric micelles,[38] and cholesterol-containing
polyamine lipids.[39] Reaction with amine-containing PVP
was conducted in dichloromethane in the presence of
triethylamine (Scheme 2), resulting polymer was recovered
via precipitation into diethyl ether. NMR characterization
(Figure 2A) revealed a 90% loading of polymer chains with
cholesterol. The recovered solid did not contain cholesteryl
chloroformate, which was verified via FTIR spectroscopy
(Figure 2B). We note that reaction yield obtained herein (in
terms of number of ligands per polymer chain) is
comparable to that achieved in conjugation using terminal
thiol groups obtained via hydrolysis of RAFT agent
thioester.[40] This notion illustrates a practical utility of
the conjugationapproach developed inthis work. However,
this reaction also highlights the limitations and challenges
of conjugation reactions with cholesterol. Thus, reaction
analysis relies on rather ill-defined NMR signature of
cholesterol and a semi-quantitative FTIR spectroscopy.
Furthermore, progression of reaction does not produce or
For removal of the terminal Phth group, polymer
samples were incubated in methanolic hydrazine. Result-
ing polymer was recovered via precipitation into diethyl
ether, however, this purification technique alone was
insufficient and resulting polymer samples contained
significant amount of impurities. Polymer was dissolved
in water and additionally purified on a size exclusion
column (NAP-25) and recovered via freeze-drying. Polymer
analysis was performed using GPC and a combination of
multi-angle light scattering and full spectrum UV-Vis
detectors. We have previously shown that this set-up
is essential for a reliable analysis of polymers and
specifically their end group conversion.[22,28] In particular,
this combination set-up allows distinguishing between
spectrally identical solutions of a true bioconjugate and a
mixture of components. Herein, we use this combination
of detectors to illustrate a successful removal of the
terminal Phth group. Indeed, the parent polymer exhibits
a pronounced absorbance at 285 nm which corresponds
to the Phth functionality (Figure 1A and B). In contrast,
UV-Vis spectrum of the elution volume corresponding to
the hydrazine-treated polymer bears no characteristic
signature of Phth and the solution has <10% absorbance
Scheme 1. Synthesis of PVP using O-ethyl-S-(phthalimidylmethyl) xanthate as a RAFT agent and subsequent removal of phthalimide
functionality to afford amine-functionalized polymer chains.
Macromol. Biosci. 2014, 14, 33–44
37
www.MaterialsViews.com
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund, A. N. Zelikin
www.mbs-journal.de
Figure 1. A,B) GPC characterization of RAFT-derived PVP before (A) and after (B) hydrazinoysis using a set-up with MALS and full-spectrum
1
UV-Vis detectors. C) Corresponding H NMR characterization of the polymers [before (a) and after (b) hydrozynolysis].
consume any species with defined UV, fluorescent, or other
spectroscopic properties to allow facile monitoring of the
process. Also, analysis of the product does not discriminate
between true conjugated species and impurities adsorbed
onto the polymer sample. Finally, this reaction was
conducted in organic solvent and not at physiological
conditions.
progression of reaction and in doing so significantly
increases the confidence in successful bioconjugation, i.e.,
synthesis of cholesterol derivatives. Finally, contributing to
the overall feasibility of this approach, thiol containing
analogue of cholesterol, thiocholesterol, is inexpensive and
readily available from commercial sources. Below, we
present our data on the use of thiocholesterol and thiol-
disulfide interconversion in cholesterol modification of
synthetic polymers.
Our prior experience in bioconjugation suggested that an
approach to circumvent the above discussed limitations
can be found in thiol-disulfide chemistry, specifically
interconversion facilitated by the use of activated thiols,
for example, mixed disulfides with Ellman’s reagent.[41,42]
This chemistry can be accomplished in a range of solvents
including physiological buffers thus, being attractive
for bioconjugation. Release of 5-thio-2-nitrobenzoic acid
(TNB) chromophore provides a facile means of monitoring
3.3. Cholesterol Modification Via Thiol-Disulfide
Exchange
Thiol-disulfide interconversion presents several opportu-
nities in terms of starting reagents to produce the final
disulfide linkage. First, we aimed to use commercial, non-
Scheme 2. Cholesterol conjugation to amine-functionalized PVP through carbamate linkage using cholesterol chloroformate.
Macromol. Biosci. 2014, 14, 33–44
38
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com

Cholesterol Modification of (Bio)Polymers Using . . .
www.mbs-journal.de
(Scheme 3). We have previously accomplished this on
RAFT derived, amine-containing PVA and showed that the
resulting polymer delivers versatile options in bioconju-
gation using this polymer.[21,28] Optimized reaction con-
ditions for this reaction were carbonate buffer (pH ¼ 8.0–
8.3) as reaction medium and 1:11:8 ratio of the polymer
to iminothiolane to Ellman’s reagent. Resulting polymer
was purified using a size-exclusion column (NAP-25) and
recovered via freeze-drying. Content of the terminal mixed
disulfide was estimated through a reaction with DTT, a
quantitative reaction which liberates TNB, a chromophore
with characteristic absorbance at 412 nm. Isolated poly-
mers revealed an apparent degree of thiolation of over
100%. To explain this, one has to consider that terminal
thioester group inherited with the RAFT agent affords a
thiol group upon hydrazynolysis[43] and is also detected
in the above employed assay. However, the thioester-
derived thiol undergoes spontaneous rearrangement in
aqueous solution of RAFT-derived PVP and polymer chains
gradually lose these thiol groups thus, leading to a varied
amount of R-group derived thiols[33] and thus being an
un-reliable site for conjugation.
1
Figure 2. A) H NMR and B) FT-IR characterization of cholesterol
conjugation to amine-functionalized PVP through carbamate
linkages.
For cholesterol modification, above obtained PVP
was reacted with thiocholesterol. In this reaction, we
exploit excellent solubility of PVP in organic solvents
which allowed performing this conversion in chloroform.
Reaction was significantly accelerated in the presence
of triethylamine. Polymer purification was accomplished
via repetitive precipitation into diethyl ether. Polymer
characterization revealed cholesterol modification at 75–
modified thiocholesterol. To this end, PVP underwent a
one-pot conversion using Traut’s reagent (iminothiolane)
and Ellman’s reagent (Scheme 3). The first step of this
reaction affords thiol-containing polymer chains which
then immediately react with Ellman’s reagent. This yields
polymers which contain terminal mixed disulfide activated
toward further reactions with thiol-containing molecules
1
95%, as estimated by H NMR spectroscopy. This reaction
provides a proof-of-concept conjugation of polymers with
Scheme 3. Synthesis of disulfide-based conjugate of cholesterol with PVP. Reaction was conducted using PVP with terminal thiol groups,
activation of the latter towards thiol-disulfide exchange using Ellman’s reagent, and thiolcholesterol. Target conjugate was obtained via
thiol-disulfide reshuffling. Corresponding images demonstrate a colour change accompanying the progression of reaction.
Macromol. Biosci. 2014, 14, 33–44
39
www.MaterialsViews.com
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund, A. N. Zelikin
www.mbs-journal.de
Scheme 4. Syntheses of disulfide-based conjugate of cholesterol with PVP using thiocholesterol, Ellman’s reagent to activate
thiocholesterol toward disulfide reshuffling, and polymer chains equipped with a single terminal thiol group.
cholesterol via disulfide linkages but as described, fails to
facilitate monitoring the progress and quantification of the
product of this conversion. Also, above reaction can only be
performed in organic solvents thus, significantly limiting
its scope for biomedicine.
cholesterol, we hypothesize that it is more likely that
Ellman’s test underestimates the presence of TC. Specifi-
cally, it is plausible that thiol groups may undergo
oxidation and coupling into disulfide linkages during
preparation of the inclusion complex thus escaping
Ellman’s test. However, our efforts to obtain TC-CD and
quantify thiol groups under inert conditions did not
afford significantly higher values of thiol content.
In an attempt to prepare water-soluble derivative
of thiocholesterol, this compound was reacted with
Ellman’s reagent in THF in the presence of TEA to
produce thiocholesterol-5-thio-2-nitrobenzoic acid, TC-ER
(Scheme 4). Despite expectations, target cholesterol deriva-
tive did not exhibit solubility in water or physiological
buffers to accommodate aqueous phase bioconjugation.
Nevertheless, amine-containing PVP was treated with
Traut’s reagent to produce thiolated polymer chains.
Conjugation with TC-ER was conducted in mixed
solvents, aqueous tetrahydrofurane, and afforded yield of
cholesterol modification not exceeding 60%.
A more plausible explanation lies in that within TC-CD
inclusion complex, the thiol group (more specifically,
thiolate anion, the active species in a thiol-disulfide
exchange reaction) is positioned well within the cavity of
CD thus being sterically hindered and un-available for
Ellman’s test. Supporting this hypothesis, CD-based inclu-
sion complex with TC-ER successfully underwent a thiol-
disulfide conversion with DTT thus, revealing accessibility
of the disulfide linkage to further chemical modification
(Scheme 5). The adduct of cyclodextrin with TC-ER was
obtained via a protocol similar to that described above
for pristine TC and afforded a thiocholesterol loading
of 30 mg of TC-TNB per gram of compound, similar to
commercial preparations of cholesterol-CD adducts.
1H NMR characterization of this compound in deuterated
water revealed a well pronounced peak at d ¼ 5.5 charac-
teristicofcholesterol,andtwogroupsofpeaks,fromd ¼ 0.92
to 1.69 and from d ¼ 7.62 to 8.14, corresponding to the
methyl groups and aromatic protons of TC-ER, respectively
(Figure 3A). CD-based inclusion complex with TC-ER was
fully water soluble and presented a UV signature in its
aqueous solution matching that of Ellman’s reagent. To
verify that the adduct was indeed that between CD and
TC-ER, not un-reacted ER or impurities and derivatives
thereof, we further characterized the inclusion complex
using HPLC and a water-acetonitrile gradient as a mobile
phase (Figure 3B). This analysis revealed a UV-Vis signature
3.4. Water-Soluble Thiocholesterol and Aqueous
Thiol-Disulfide Exchange
To further increase solubility of thiocholesterol in aqueous
media, we turned to well-described inclusion complexes of
steroidal compounds with cyclodextrins (CD). Synthesis
of inclusion complexes using TC was accomplished
following Williams’s protocol[23] with slight modifications
and using methyl-b-cyclodextrin (M-b-CD) or hydroxyl-
propyl-b-cyclodextrin (HP-b-CD). Surprisingly, the CD-TC
adduct revealed only a minor content of thiol groups, as
verified using Ellman’s reagent. Apparent TC loading
was established at ꢃ3 mg per gram of inclusion complex,
which is only ꢃ10% of cholesterol content in commercial
preparations of CD-based ‘‘water-soluble cholesterol.’’
While it is plausible that TC loading into the structure
of CD is indeed significantly lower that for pristine
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
40
www.MaterialsViews.com

Cholesterol Modification of (Bio)Polymers Using . . .
www.mbs-journal.de
Scheme 5. Schematic illustration of thiol-disulfide exchange between cyclodextrin – based inclusion complexes of thiocholesterol (top)
and Ellman’s reagent – activated thiocholesterol (bottom) with Ellmans’ reagent and DTT, respectively.
for [CD/TC-ER] matching that of Ellman’s reagent and an
elution time matching that of TC-ER (Scheme 3B). This
result implies that in the presence of acetonitrile, TC-ER
is removed from the inclusion complex and also demon-
strates that inclusion complex indeed contains TC-ER
thus, substantiating successful incorporation of the thio-
cholesterol derivative into the structure of cyclodextrin.
To illustrate the utility of developed water soluble
thiocholesterol derivative for modification of biomolecules
at physiological conditions, we used a readily available,
natural thiol-containing tripeptide, GSH. This peptide is
found in mammalian cells at concentration of 5 ꢁ 10ꢀ3
M
and is typically used to probe degradation of disulfide-
stabilized drug carrier vehicles and also to probe novel
bioconjugation techniques relying on thiol chemis-
try.[27,41,44] The progress of reaction between GSH and
CD-complexed TC-ER was monitored via UV-Vis spectros-
copy monitoring absorbance of TNB chromophore, and
using DTT to set the 100% scale mark. This reaction proved
to be highly efficient and fast, achieving a near-complete
conversion in excess GSH within minutes, Figure 4A. For
modification of biomolecules, complete conversion of
target thiols is the prime aim, and with GSH, near-full
conversion of GSH into cholesterol containing disulfide
was achieved with as little as twofold excess of [CD/TC-ER]
over GSH. Taken together, results in Figure 4 present
the newly synthesized water-soluble thiocholesterol as
an effective tool for bioconjugation.
For PVP and PVA, aqueous solutions of amine-containing
polymers were chargedwith [CD/TC-ER] andiminothiolane
for an in situ conversion of amine terminal groups into
thiols and subsequent conjugation reaction (Scheme 6). In
optimizing reaction conditions we observed that solutions
buffered to pH ¼ 8 (e.g., carbonate buffer) favor spontane-
ous ring opening of the Traut’s reagent and reaction of
the products of this reaction with [CD/TC-ER]. This was
overcome by using phosphate buffers with pH ¼ 7.3. Lower
concentration of reagents also facilitated formation of
the desired polymer-containing product. Reaction mixtures
were incubated for 14 h, upon which solutions were first
Figure 3. A) 1H NMR characterization of an inclusion complex
between M-b-CD and TC-ER in deuterated water: i) CD, ii) [CD/TC-
ER]; B) overlaid HPLC elution traces for ER and [CD/TC-ER].
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
41
www.MaterialsViews.com

K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund, A. N. Zelikin
www.mbs-journal.de
Figure 4. A) Kinetic profile of bioconjugation in aqueous solutions of [CD/TC-ER] (inclusion complex between methyl-b-cyclodextrin and
Ellman’s reagent activated thiocholesterol) and glutathione (5 g Lꢀ1 GSH, carbonate buffer pH ¼ 8.3). B) Conversion of thiol groups within
GSH into disulfide linkages with thiocholesterol in aqueous solutions of GSH and [CD/TC-ER] as a function of the ratio of reagents.
purified on a desalting column followed by additional
dialysis for 48 h. Finally, the polymers were recovered via
freeze-drying and analyzed for cholesterol content using
NMR spectroscopy. Polymer loading with cholesterol
reached almost 100% for PVP and ꢃ80% for PVA thus,
illustrating successful modification of the polymers. We
note that overlapping peaks make direct NMR quantifica-
tion of remaining CD challenging, however, low affinity of
CD to solutes forming inclusion complexes suggests that
dialysis step removes an overall majority of CD from the
polymer termini.
developed novel tools and methods for micro-fabrication
using PVA hydrogels and their biomedical characterization
and functionalization.[22,25,28] Specifically, we presented
a novel method for conjugation of PVA with oligo-
peptide cargo wherein conjugation relies on thiol-disulfide
chemistry and is conducted at physiological conditions.[28]
Herein, we capitalize on these findings and use surface
adhered, micro-structured physical PVA hydrogels for
their functionalization with cholesterol. We hypothesized
that running bioconjugation on pre-formed hydrogels
offers opportunities in facile isolation of the product,
specifically through simple withdrawing the hydrogel
from the reaction medium. To illustrate this possibility,
amine-containing PVA was used to assemble the surface-
adhered hydrogel samples. Subsequently, hydrogels
were immersed in a buffered solution of iminothiolane
for 30 min to achieve an in situ thiolation of hydrogels.
Specimen were removed from the reaction milieu
and washed with copious amounts of fresh buffer.
Finally, to illustrate possibilities in cholesterol modifica-
tion of supramolecular materials, we tested [CD/TC-ER]
for modification of physical hydrogels based on PVA.
Despite decades of development, characterization, and
biomedical applications, these hydrogels present scant
opportunities in quantification of polymer within the
hydrogel matrix and only limited opportunities for
biochemical characterization. Over the past few years, we
Scheme 6. Schematic illustration of conjugation of thiocholesterol to PVP through a UV-Vis traceable chemistry and conducted in aqueous,
physiological conditions. Reaction was performed using methyl-b-cyclodextrin based inclusion complex with Ellman’s reagent activated
thiocholesterol and a RAFT-derived PVP with a single terminal amine group which was converted in situ into a thiol using iminothiolane.
For experimental details, see Experimental Section.
Macromol. Biosci. 2014, 14, 33–44
42
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.MaterialsViews.com

Cholesterol Modification of (Bio)Polymers Using . . .
www.mbs-journal.de
Figure 5. A) Differential interference contrast image of micro-structured, surface adhered physical hydrogel based on PVA (scale bar: 50 mm)
and schematic illustration of chemical modification of PVA chains within the structure of hydrogels using water-based chemical
transformations: thiolation using Traut’s reagent and subsequent bioconjugation with thiocholesterol using
a water-soluble
preparation, [CD/TC-ER] (inclusion complex between methyl-b-cyclodextrin and Ellman’s reagent activated thiocholesterol). B) UV-Vis
data on the reaction between the terminal thiol groups on PVA chains comprising physical hydrogels with Ellman’s reagent or [CD/TC-ER].
Subsequently, hydrogel samples were immersed in a
solution of Ellman’s reagent upon which supernatants
attained a characteristic yellow color. This indicated a
successful reaction of Ellman’s reagent with the thiol
groups immobilized within the hydrogel matrix, that is,
terminal groups of PVA chains, thus, demonstrating
successful thiolation of the hydrogel (Figure 5). Super-
natants drawn from above the hydrogel samples imme-
diately before this analysis revealed a significantly
lower absorbance which verifies that Ellman’s test
indeed registered thiols groups as part of hydrogel
biomaterials. When [CD/TC-ER] was introduced onto
thiolated hydrogels instead of Ellman’s reagent, resulting
absorbance was similar to that registered with ER only
indicating a highly efficient conversion of the thiol
groups within hydrogel matrix into cholesterol-containing
disulfides. We believe that a high degree of hydration of
the hydrogels greatly facilitates exchange of solutes
between the hydrogel and the solution bulk thus,
contributing to the feasibility of the reaction. We strongly
believe that this is the first described example of an in situ
thiolation of PVA physical hydrogels and also the first
report on cholesterol-functionalization of these bio-
materials, both goals achieved under physiological con-
ditions. We note that while no attempt was made in this
work to quantify the removal of cyclodextrin and ‘‘un-
masking’’ cholesterol, typical binding constants for
these inclusion complexes are in the 10ꢀ1 to 10ꢀ3 L molꢀ1
range[45] and CD removal is likely to proceed spontaneously
upon washing and/or storage of the hydrogel samples. We
are now investigating utility of cholesterol-containing
hydrogels for immobilization of liposomes as drug deposits
for surface-mediated drug delivery[46] and for biosensing
applications[47] as well as tools to achieve controllable
adhesion of mammalian cells towards the use of PVA
hydrogels as matrices in tissue engineering.
Acknowledgements: This work is financed by a grant from the
Lundbeck Foundation and Sapere Aude Starting Grant from
the Danish Council for Independent Research, Technology and
Production Sciences, Denmark.
Received: June 10, 2013; Revised: August 6, 2012; Published online:
September 17, 2013; DOI: 10.1002/mabi.201300286
Keywords: biomaterials; cholesterol; hydrogels; reversible addi-
tion fragmentation chain transfer (RAFT)
[1] L. Hosta-Rigau, Y. Zhang, B. M. Teo, A. Postma, B. Stadler,
Nanoscale 2013, 5, 89.
[2] H. Ishiwata, A. Vertutdoi, T. Hirose, K. Miyajima, Chem. Phar.
Bull. 1995, 43, 1005.
[3] S. Beugin, K. Edwards, G. Karlsson, M. Ollivon, S. Lesieur,
Biophys. J. 1998, 74, 3198.
[4] C. Carrion, J. C. Domingo, M. A. de Madariaga, Chem. Phys.
Lipids 2001, 113, 97.
[5] C. Lorenz, P. Hadwiger, M. John, H. P. Vornlocher, C. Unverzagt,
Bioorg. Med. Chem. Lett. 2004, 14, 4975.
[6] V. V. Ambardekar, H. Y. Han, M. L. Varney, S. V. Vinogradov,
R. K. Singh, J. A. Vetro, Biomaterials 2011, 32, 1404.
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
43
www.MaterialsViews.com

K. F. Rasmussen, A. A. A. Smith, P. Ruiz-Sanchis, K. Edlund, A. N. Zelikin
www.mbs-journal.de
[7] D. De Paula, M. V. L. B. Bentley, R. I. Mahato, RNA 2007, 13, 431.
[8] T. H. Senanayake, G. Warren, S. V. Vinogradov, Bioconjugate
Chem. 2011, 22, 1983.
[28] S.-F. Chong, A. A. A. Smith, A. N. Zelikin, Small 2013, 9, 942.
[29] J. T. Xu, L. Tao, C. Boyer, A. B. Lowe, T. P. Davis, Macromolecules
2011, 44, 299.
[9] B. Stadler, A. D. Price, R. Chandrawati, L. Hosta-Rigau, A. N.
Zelikin, F. Caruso, Nanoscale 2009, 1, 68.
[30] J. Xu, L. Tao, C. Boyer, A. B. Lowe, T. P. Davis, Macromolecules
2010, 44, 299.
[10] O. Shimoni, A. Postma, Y. Yan, A. M. Scott, J. K. Heath, E. C.
Nice, A. N. Zelikin, F. Caruso, ACS Nano 2012, 6, 1463.
[11] A. A. A. Smith, T. Hussmann, J. Elich, A. Postma, M.-H. Alves,
A. N. Zelikin, Polymer Chem. 2012, 3, 85.
[12] A. N. Zelikin, G. K. Such, A. Postma, F. Caruso, Biomacromo-
lecules 2007, 8, 2950.
[31] C. Boyer, V. Bulmus, T. P. Davis, V. Ladmiral, J. Q. Liu, S. Perrier,
Chem. Rev. 2009, 109, 5402.
[32] G. Moad, E. Rizzardo, S. H. Thang, Aust. J. Chem. 2012, 65, 985.
[33] G. Pound, J. M. McKenzie, R. F. M. Lange, B. Klumperman,
Chem. Commun. 2008, 3193.
[34] Y. Y. Tong, Y. Q. Dong, F. S. Du, Z. C. Li, J. Polym. Sci. Pol. Chem.
2009, 47, 1901.
[35] A. Postma, T. P. Davis, G. X. Li, G. Moad, M. S. O’Shea,
Macromolecules 2006, 39, 5307.
[13] M. A. Gauthier, M. I. Gibson, H. A. Klok, Angew. Chem. Int. Ed.
2009, 48, 48.
€
[14] J. U. A. Engstrom, L. J. Lindgren, B. Helgee, Macromol. Chem.
Phys. 2006, 207, 536.
[36] D-a. Wang, A. S. Narang, M. Kotb, A. O. Gaber, D. D. Miller,
S. W. Kim, R. I. Mahato, Biomacromolecules 2002, 3, 1197.
[37] S. O. Han, R. I. Mahato, S. W. Kim, Bioconjugate Chem. 2001,
12, 337.
[38] J. H. Kim, Y. Li, M. S. Kim, S. W. Kang, J. H. Jeong, D. S. Lee, Int. J.
Pharm. 2012, 427, 435.
[15] M. G. Tardajos, M. Nash, Y. Rochev, H. Reinecke, C. Elvira,
A. Gallardo, Macromol. Chem. Phys. 2012, 213, 529.
[16] T. Trellenkamp, H. Ritter, Macromolecules 2010, 43, 5538.
[17] C. Boyer, J. Liu, L. Wong, M. Tippett, V. Bulmus, T. P. Davis, J.
Polym. Sci. Part A: Polym. Chem. 2008, 46, 7207.
[18] J. Q. Liu, H. Y. Liu, C. Boyer, V. Bulmus, T. P. Davis, J. Polym. Sci.
Part A: Polym. Chem. 2009, 47, 899.
[39] M. Oliver, M. R. Jorgensen, A. Miller, Tetrahedron Lett. 2004,
45, 3105.
[19] L. J. Wong, C. Boyer, Z. F. Jia, H. M. Zareie, T. P. Davis, V. Bulmus,
Biomacromolecules 2008, 9, 1934.
[40] P. J. Roth, C. Boyer, A. B. Lowe, T. P. Davis, Macromol. Rapid
Comm. 2011, 32, 1123.
[20] M. H. Alves, B. E. B. Jensen, A. A. A. Smith, A. N. Zelikin,
Macromol. Biosci. 2011, 11, 1293.
[41] S. F. Chong, A. Sexton, R. De Rose, S. J. Kent, A. N. Zelikin,
F. Caruso, Biomaterials 2009, 30, 5178.
[21] B. Fejerskov, B. E. B. Jensen, N. B. S. Jensen, S.-F. Chong, A. N.
Zelikin, ACS Appl. Mater. Interfaces 2012, 4, 4981.
[22] B. Fejerskov, A. A. A. Smith, B. E. B. Jensen, T. Hussmann, A. N.
Zelikin, Langmuir 2012, 29, 344.
[23] B. Fejerskov, A. N. Zelikin, PLOS One 2012, 7, e49619.
[24] L. Hosta-Rigau, B. E. B. Jensen, K. S. Fjeldsꢀ, A. Postma, G. Li, K. N.
[42] S. F. Chong, R. Chandrawati, B. Stadler, J. Park, J. H. Cho, Y. J.
Wang, Z. F. Jia, V. Bulmus, T. P. Davis, A. N. Zelikin, F. Caruso,
Small 2009, 5, 2601.
[43] W. Shen, Q. Qiu, Y. Wang, M. Miao, B. Li, T. Zhang, A. Cao, Z. An,
Macromol. Rapid Commun. 2010, 31, 1444.
€
[44] S. Bauhuber, C. Hozsa, M. Breunig, A. Gopferich, Adv. Mater.
€
Goldie, F. Albericio, A. N. Zelikin, B. Stadler, Adv. Healthcare
2009, 21, 3286.
Mater. 2012, 1, 791.
[25] B. E. B. Jensen, M.-H. Alves, B. Fejerskov, B. Stadler, A. N.
Zelikin, Soft Matter 2012, 8, 4625.
[45] V. J. Stella, Q. R. He, Toxicol. Pathol. 2008, 36, 30.
[46] M. E. Lynge, M. Baekgaard Laursen, L. Hosta-Rigau, B. E. B.
€
Jensen, R. Ogaki, A. A. A. Smith, A. N. Zelikin, B. Stadler, ACS
[26] G. Moad, E. Rizzardo, S. H. Thang, Polym. Int. 2011, 60, 9.
[27] G. Saito, J. A. Swanson, K. D. Lee, Adv. Drug Delivery Rev. 2003,
55, 199.
Appl. Mater. Interfaces 2013, 5, 2967.
[47] M. Bally, K. Bailey, K. Sugihara, D. Grieshaber, J. Voros, B. Stadler,
€ €
€
Small 2010, 6, 2481.
Macromol. Biosci. 2014, 14, 33–44
ß 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
44
www.MaterialsViews.com