Inhibition of protein and cell attachment on materials generated from N -(2-hydroxypropyl) acrylamide

Article abstract of DOI:10.1021/bm500654q

Effective control over biointerfacial interactions is essential for a broad range of biomedical applications. At this point in time, only a relatively small range of radically polymerizable monomers have been described that are able to generate low fouling polymer materials and surfaces. The most important examples that have been successfully used in the context of the reduction of nonspecific protein adsorption and subsequent cell attachment include PEG-based monomers such as poly(ethylene glycol) methacrylate (PEGMA), zwitterionic monomers such as 2-methacryloyloxyethyl phosphorylcholine and noncharged monomers such as acrylamide and N-(2-hydroxypropyl) methacrylamide (HPMAm). However, issues such as oxidative degradation and poor polymerization characteristics limit the applicability of most of these candidates. Here we have synthesized the monomer N-(2-hydroxypropyl) acrylamide (HPAm), examined its polymerization kinetics and evaluated its suitability for RAFT mediated polymerization in comparison to HPMAm. We also synthesized hydrogels using HPMAm and HPAm and evaluated the ability of HPAm polymers to occlude protein adsorption and cell attachment. In RAFT-controlled polymerization, much faster (8×) polymerization was observed for HPAm relative to HPMAm and better control was achieved over the molecular weight distribution. The performance of hydrogels prepared from HPAm in the prevention of protein adsorption and cellular attachment was equivalent to or better than that observed for materials made from HPMAm and PEG. These results open the door for HPAm based polymers in applications where effective control over biointerfacial interactions is required.

Full text of DOI:10.1021/bm500654q

Article
pubs.acs.org/Biomac
Inhibition of Protein and Cell Attachment on Materials Generated
from N‑(2-Hydroxypropyl) Acrylamide
Benjamin D. Fairbanks,* Helmut Thissen, George Maurdev, Paul Pasic, Jacinta F. White,
and Laurence Meagher
CSIRO Manufacturing Flagship, Bayview Avenue, Clayton 3169 VIC, Australia
S
* Supporting Information
ABSTRACT: Effective control over biointerfacial interactions
is essential for a broad range of biomedical applications. At this
point in time, only a relatively small range of radically
polymerizable monomers have been described that are able to
generate low fouling polymer materials and surfaces. The most
important examples that have been successfully used in the
context of the reduction of nonspecific protein adsorption and
subsequent cell attachment include PEG-based monomers
such as poly(ethylene glycol) methacrylate (PEGMA),
zwitterionic monomers such as 2-methacryloyloxyethyl
phosphorylcholine and noncharged monomers such as
acrylamide and N-(2-hydroxypropyl) methacrylamide (HPMAm). However, issues such as oxidative degradation and poor
polymerization characteristics limit the applicability of most of these candidates. Here we have synthesized the monomer N-(2-
hydroxypropyl) acrylamide (HPAm), examined its polymerization kinetics and evaluated its suitability for RAFT mediated
polymerization in comparison to HPMAm. We also synthesized hydrogels using HPMAm and HPAm and evaluated the ability of
HPAm polymers to occlude protein adsorption and cell attachment. In RAFT-controlled polymerization, much faster (8×)
polymerization was observed for HPAm relative to HPMAm and better control was achieved over the molecular weight
distribution. The performance of hydrogels prepared from HPAm in the prevention of protein adsorption and cellular
attachment was equivalent to or better than that observed for materials made from HPMAm and PEG. These results open the
door for HPAm based polymers in applications where effective control over biointerfacial interactions is required.
intrinsically bioactive may cause a host of cascading biological
responses, many of which are undesirable. By employing
polymers which have low protein adsorption and cell
attachment properties, however, materials can be designed
that act as a nonreactive platform or “blank slate.” Discrete
biochemical signals and functional groups can subsequently be
added to elicit specific desired biological responses to the
material while preventing undesired responses that may
otherwise be generated as a result of nonspecific protein
adsorption.4−7
While a number of low protein and cell attachment polymers
have been identified, the stand-out performers all have
disadvantages which limit their broad application. Those
polymers belonging to the class of naturally derived materials,
including polypeptides and polysaccharides, lack long-term
stability under physiological conditions. Similarly, polymers
containing esters (e.g., polyesters, polyacrylates, and poly-
methacrylates) are subject to hydrolytic degradation, which has
been implicated in the clinical failure of hydroxyethyl acrylate
copolymer hydrogel implants.8,9 Poly(ethylene glycol) (PEG)
INTRODUCTION
■
The challenges in the development of effective biomaterials are
multifaceted. Not only must a functional biomedical material
possess the appropriate mechanical properties (strength and
flexibility so that the material does not break under
physiological stress, hardness so that it does not wear too
quickly, etc.), but it must also interact appropriately with the
biochemical environment in which it is placed. For a bone graft
this may mean a material that encourages cellular integration
while for an artificial heart valve or ocular lens, this may mean a
nonadhesive material to which cells fail to attach.
Cellular attachment to a material can be via specific
interactions, such as for materials that are synthesized to
present defined integrin binding sites to promote cellular
adhesion. More commonly, however, cellular attachment to
materials occurs as a result of nonspecific adsorption of serum
proteins on the surface of a material, presenting a myriad of cell
adhesion motifs.1−3 Minimizing the adsorption of proteins from
biological fluids therefore reduces cell attachment in many
cases.
While synthetic polymers that interact very little with
biological macromolecules and cells are understandably a
popular foundation for bioinert materials, they are also often a
crucial component of bioactive materials.4−7 A material that is
Received: May 6, 2014
Revised: August 14, 2014
Published: August 15, 2014
© 2014 American Chemical Society
3259
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
is by far the most popular low-fouling polymer employed in
academic research and industry, having very low protein
interactions.7,10−12 Another reason that PEG is so widely used
is the chemical versatility of the end groups and the regular
molecular weights of the polymers themselves; anionic
polymerization of ethylene oxide results in telechelic polymers
of extremely low dispersities. While PEG has been used (among
other purposes) as a polymer surface coating,12−15 a hydrogel
platform16,17 and attached to therapeutic molecules to improve
pharmacokinetic properties,18−20 the ether groups in the
polymer chains are observed to degrade oxidatively in
biologically relevant conditions,11,21,22 thus, limiting their use
in long-term applications. Moreover, concern has been raised in
recent years that PEG is more immunologically active than
previously understood.23
Recently, several poly zwitterionic polymers have shown
great promise in the development of low protein adsorption
materials. Phosphorylcholine methacrylate polymers, mimick-
ing the head groups of lipids composing the outer membrane of
red blood cells, have shown great resistance to protein
adsorption.24−26 However, the phosphoester group is suscep-
tible to hydrolysis27 and the hygroscopic character of the
compound makes it a difficult material to synthesize and
handle.28 Sulfobetaine-containing monomers are easier to
handle and more hydrolytically stable, but, in at least one
study, have not shown the same performance as the
phosphorylcholine monomers.29 Of the zwitterionic polymers,
carboxybetaine monomers appear now to have the best
combination of performance and stability.28,30 Most zwitter-
ionic monomers, however, fall into the categories of acrylates
and methacrylates, so long-term hydrolytic stability is an
ongoing issue.
mediated polymerization of HPMA, telechelic, or diend-
functional polymers of lower dispersities can be obtained,
making pHPMA competitive with PEGs where polymers with
more uniform, defined molecular weights, and reactive end-
groups are required.
Yet, for some applications, the HPMAm monomer has
limitations that preclude the effective application of its
otherwise desireable polymer. Characteristic of methacryla-
mides, HPMAm exhibits low polymerization rates relative to
(meth)acrylates and acrylamides. This shortcoming is partic-
ularly noted in nonaqueous solution polymerizations and limits
both the ultimate chain length and the speed with which
polymerizations can be performed.38,39 Additionally, for RAFT
mediated polymerization, the variety of appropriate RAFT
agents is limited relative to monomers of other polymerizable
groups. To address these issues, an acrylamide analogue of
HPMAm was synthesized, N-(2-hydroxypropyl) acrylamide or
HPAm (Figure 1). Despite its structural similarity, this
acrylamide analogue has, as of yet, not been explored in
many of those biomaterial applications for which HPMAm
excels. Relative rates of polymerization were determined as was
the molecular weight and dispersities of polymers resulting
from RAFT-mediated polymerization (see Figure 1 for
structure of RAFT agent). Potential for application in low
protein attachment and low cell attachment materials surfaces
was evaluated by measuring adsorption of europium labeled
fibronectin in fetal bovine serum and by seeding adherent cells
on hydrogels formed from HPAm and HPMAm compared
against hydrogels of PEGDMA as well as tissue culture
polystyrene and commercial low-attachment culture plates as
adsorption/attachment controls (Figure 1).
Pioneered by Jindrich Kopecek and his research group,
polymers made from N-(2-hydroxylpropyl) methacrylamide
(HPMAm), Figure 1, have been employed in biomaterial and
EXPERIMENTAL SECTION
Key compounds used in this work are shown in Figure 1. Unless
otherwise noted, all reactants and initiators were purchased from
Sigma-Aldrich and used as received.
■
Synthesis of HPMAm (1). HPMAm was synthesized by reacting
10 g (65 mmol) of methacrylic anhydride in diethyl ether (50 mL)
with an equimolar quantity of amino-2-propanol (4.9 g, Acros) in 25:1
mixture ether/dichloromethane (50 mL) at room temperature. In a
500 mL round-bottom flask, amino-2-propanol in diethyl ether was
stirred while dichloromethane was added dropwise until the complete
dissolution of the amine was obtained. The methacrylic anhydride
solution was added in a thin stream causing a vigorous reaction that
resulted in boiling of the solution in the reaction flask (for this reason
a large capacity flask was used). Following addition of the anhydride
solution the reaction was stirred for 1 h at room temperature. The
reaction mixture was transferred to a smaller flask and then cooled to
−20 °C overnight to facilitate crystallization of the product. Solid
HPMA was collected via filtration and recrystallized from 4:1 diethyl
ether/dichloromethane. The yield was 65%. This scheme was applied
in lieu of the reaction of methacryloyl chloride31 because of the less
stringent transport restrictions of methacrylic anhydride and because
of the ease of product recovery. Product identity and purity was
confirmed by HNMR and atmospheric pressure chemical ionization
mass spectroscopy (APCI MS; Figures S1 and S2 in Supporting
Information). Chemical shifts in deuterated chloroform: 1.16 d (3H),
1.92 m (3H), 3.12 m (1H), 3.38 br (1H), 3.45 m (1H), 3.90 m (1H),
5.30 m (1H), 5.68 m (1H), 6.49 br (1H).
Figure 1. Key compounds used in this study: (1) N-(2-
hydroxylpropyl) methacrylamide (HPMAm) (2), N-(2-hydroxyprop-
yl) acrylamide (HPAm), (3) poly(ethylene glycol) dimethacrylate Mw
= 4000 (PEGDMA), and (4) cyano-4-[(ethylsulfanylthiocarbonyl)-
sulfanyl]pentanoic acid (RAFT agent).
biotechnological applications as an alternative to PEG.11,31,32
The amide bond of acrylamides and methacrylamides is
considerably more stable to hydrolysis than are the esters of
analogous (meth)acrylates. Rodriguez-Emmenegger et al.
report that the protein adsorption observed with pHPMAm
surfaces was below surface plasmon resonance (SPR) detection
limits, making the polymer a prime candidate for the generation
of low fouling biomaterial surfaces33,34 as well as for the
modification of therapeutic agents to improve their pharmaco-
kinetic properties.11,32,35−37 Additionally, by employing RAFT-
Synthesis of HPAm (2). Good yield was obtained by adding 5 g of
acryloyl chloride (55 mmol, Merck) dropwise in dry ethyl acetate (40
mL) to a stirred solution of 2-fold excess amino-2-propanol (8.3 g) in
dry ethyl acetate (50 mL) at a temperature of −10 °C under nitrogen.
By repeated synthesis attempts, it was determined that reduced
reaction time, relative to analogous protocols used for HPMAm
synthesis,31 improved yield. Following completion of the addition of
3260
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
the acryloyl chloride (30 min) the reaction mixture had separated into
two phases: a thin solution containing the majority of the recoverable
product and a viscous phase containing some product as well as
polymerized product and amine salt byproducts. The thin liquid phase
was passed through a 2 cm plug of silica. The viscous phase was
dissolved in water (50 mL) and extracted with ethyl acetate (50 mL).
The organic phase from the extraction was then passed through the
same silica plug which was then eluted with more ethyl acetate (200
mL). Ethyl acetate phases were combined and solvent removed via
rotary evaporation. The remaining viscous liquid product was placed in
a glass vial and stored at −20 °C whereupon it crystallized. The yield
was 50%. Product identity and purity was confirmed by 1H NMR and
APCI MS (Figures S3 and S4 in Supporting Information). Chemical
shifts in deuterated water: 1.06 d (3H), 3.14 m (1H), 3.21 br (1H),
3.84 m (1H), 5.66 dd (1H), 6.08 m (1H), 6.17 dd (1H).
Synthesis of PEGDMA. A total of 5 g of poly(ethylene glycol), Mw
= 4000, was reacted with 3 equiv methacryloyl anhydride in 50 mL of
anhydrous dichloromethane with 0.2 equiv dimethylaminopyridine as
catalyst. After 18 h, the PEGDMA was precipitated in cold diethyl
ether, redissolved in dichloromethane, washed with 5% sodium
bicarbonate solution, and brine and then reprecipitated in cold diethyl
ether. Residual solvent was removed via vacuum. Degree of
substitution was determined to be greater than 80% via HNMR in
deuterated chloroform.
Contact Angle Measurements. Static contact angle measure-
ments were performed by depositing 5 μL droplets of 0.2 um filtered
deionized water onto surfaces of hydrogels. Gels were placed on a glass
slide on top of moistened tissue in a closed container to maintain
humidity. Containers were opened only briefly for imaging with a KSV
Cam 200 tensiometer.
Europium Labeling and Protein Adsorption Quantification.
The labeling of fibronectin with europium and subsequent adsorption
assay were performed as described previously43 but with fibronectin
rather than human serum albumin. Briefly, 1 nmole fibronectin from
human serum (Sigma) in 100 μL of 0.1 M bicarbonate buffer (pH 9.3)
was treated with a 15-fold excess of Eu-labeling reagent (Delfia Eu−N1
ITC chelate, Perkin Elmer) at 4 °C for 18 h. Labeled fibronectin was
purified via dialysis.
Adsorption was performed with 10% FBS (fetal bovine serum) in
PBS (phosphate buffered saline) supplemented with 0.3 μg/mL Eu-
labeled fibronectin (approximately 1/100 of the endogenous
fibronectin component of FBS). Gels and surfaces were incubated
with supplemented FBS solution at 20 °C for 18 h then washed 7×
with PBS. After the last wash, Eu(III)-liberating Enhancement
Solution (Perkin Elmer) was added and allowed to cover surfaces
for 45 min. Following this short incubation, 0.1 mL of the Eu-complex
solution was removed from each surface and transferred to a white 96-
well plate for time-resolved fluorescence assay on a PHERAstar
instrument with excitation and measured emission at 337 and 620 nm,
respectively, as per Perkin Elmer product instructions. Concentrations
of Eu(III) were determined via comparison with standards. Raw
measurements were corrected for volume of enhancement solution
and surface area of substrates. Total fibronectin adsorption was
calculated based on the ratio of labeled to unlabeled fibronectin in 10%
FBS.
Cell Culture Experiments. Hydrogels were incubated with 2×
antibiotic antimycotic solution (anti−anti, Gibco) in PBS for 3 h at 37
°C to minimize risk of microbial and fungal contamination. After the
anti−anti solution was removed, L929 mouse fibroblast cells in media
supplemented with 10% fetal bovine serum (FBS), and 1× anti−anti,
were seeded at a density of 20000 cells/cm2. Photomicrographs of the
sample surfaces (i.e., attached and unattached cells) were taken after
24 h incubation. As controls, cells were also seeded on tissue culture
treated polystyrene (TCPS, Nunclon delta surface, Nunc) and ultra-
low attachment surface (ULA, Corning) under the same conditions,
except for the use of anti−anti.
MTS Assays. To quantify the number of metabolically active cells
adhered to and removed from hydrogels’ surfaces, a colorimetric MTS
assay was performed using Promega’s CellTiter 96 aqueous non-
radioactive cell proliferation assay. This assay measures the enzymatic
conversion of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphen-
yl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS), which corresponds
to the number of metabolically active cells. At 24 h after cellular
seeding, the surface of each gel, as well as the ULA and TCPS surfaces,
was photographed and then lightly rinsed with the media already
present in its respective well. That media from each well
(approximately 750 μL) was transferred to a fresh well in a TCPS
plate, whereupon concentrated solutions of MTS and phenazine
methosulfate (PMS) were added to each well to meet the
recommended working concentrations (0.4 and 0.01 mg/mL
respectively). To the gels was then added 1 mL of fresh media
containing the assay’s constitutive compounds already at the
recommended concentrations. Following 3 h incubation at 37 °C,
the absorbance at 490 nm for each well was measured against media
containing assay components but no cells. Statistical significance was
determined by Student’s t test.
Synthesis of the RAFT Agent. The RAFT agent was synthesized
via a previously published protocol40,41 but with the substitution
mercaptoethane for 1-dodecanethiol and of potassium tert-butoxide for
sodium hydride.
Polymerization Kinetics. A total of 1 g of each monomer was
dissolved in deuterated DMSO to a concentration of 1.75 M. Small
volumes (small enough to not significantly dilute monomer) of
concentrated solutions of RAFT agent and azobis(isobutyronitrile)
(AIBN) in deuterated DMSO were added to a final concentration of
5.0 and 0.50 mM, respectively. Monomer solutions were deoxygenated
by three cycles of freeze−pump−thaw, whereupon the contents were
transferred, under nitrogen, to individual sealed vials with a volume of
200 μL each. All vials were placed in shaking incubator at 80 °C and
100 rpm and removed at various time points. These polymerizaiton
1
mixture samples were further diluted with deuterated DMSO and H
NMR spectra were obtained using a Bruker 400 MHz NMR
spectrometer. Monomer conversion to polymer was determined by
the integration of the alkene-associated proton peaks (at 5.5−6.5 ppm
for the acrylamide and 5.3−5.7 ppm for the methacrylamide) relative
to the total amide proton peak (at ∼6.5 ppm).
GPC-MALS. Molecular weights and dispersities were determined by
gel permeation chromatography and multiangle light scattering (GPC-
MALS) on a Shimadzu Chromatography System (LC-20AD) with
Ultrahydrogel 2000 and 120 columns connected in series. A column
temperature of 40 °C was maintained and 0.10 M sodium nitrate was
used as the mobile phase. The column outlet was connected to a Dawn
Heleos II light scattering detector and an Optilab REX refracteive
index detector (both from Wyatt Technology Coorporation) and data
was analyzed with ASTRA V software (Wyatt).
Hydrogel Formation. Hydrogels were made by the copolymeriza-
tion of monomers (concentration 2 M) with N,N-methylenebis-
(acrylamide) (40 mM) in acetate buffer (pH = 5.4). Polymerizations
were initated at 50 °C using 2,2′-azobis(2-methylpropionamidine)-
dihydrochloride at concentrations of 40 mM for 10 h for HPMAm gels
and at a concentration of 10 mM for 3 h for HPAm. Gels were formed
between untreated glass microscope slides sealed with a 1 mm thick
rubber gasket. Following polymerization, gels were placed in PBS to
remove any unreacted reagents and to equilibrate gels for biological
testing. After a minimum of 18 h, 10 mm discs were cut out using a
biopy punch for use in protein adsorption and cell culture experiments.
PEGDMA hydrogels were synthesized similarly to that as described
previously.42 Briefly, an aqueous solution of 20% PEGDMA with
0.02% photoinitiator LAP (synthesized according to previous
reports42) was irradiated between glass microscope slides for 5 min
at 10 mW/cm2, 365 nm UV light.
RESULTS AND DISCUSSION
■
As hypothesized, the polymerization of HPAm proceeded
much faster than that of HPMAm, as may be seen in Figure 2.
After 16 h, the conversion of HPAm exceeded 90%, while that
obtained for HPMAm remained just over 30%. To a large
extent, the low conversion of HPMAm after 16 h resulted from
3261
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
from the kinetics experiments. The results obtained are
summarized in Table 1. The concentrations of monomer,
initiator, and RAFT agent ideally would lead to a pHPMAm
with a Mn of 50 kg mol−1 and slightly lower for the pHPAm
because of the lower molecular weight of the monomer. The
target molecular weights for the polymers were calculated via
eq 2, assuming complete conversion.45
[M]o − [M]
M =
t mM + mT
̅
n
[T]o
(2)
Here [M]0 and [M]t are the initial and final molar
concentrations of monomer, respectively. [T]0 is the initial
concentration of RAFT agent and mM and mT are the molecular
weights of the monomer and RAFT agent, respectively.
The 50 kg mol−1 target was chosen as others had reported
difficulty in obtaining RAFT-controlled polymers of this
molecular wight with HPMA in organic solvents.38,39 After a
20 h reaction, the Mn for pHPMAm obtained was less than half
the target value, a consequence of low monomer conversion
(∼30%). The Mn of the pHPAm obtained, on the other hand,
was very close to the target Mn value. Additionally, the
molecular weight dispersity achieved for the pHPAm
synthesized was significantly closer to unity than that obtained
for pHPMAm. Theoretically, polymerization mediated with an
appropriate RAFT agent should result in lower or equivalent
dispersity at lower conversions than at higher conversions. The
exact reason for the higher dispersity obtained for the
pHPMAm is not known. It should be noted that conversion,
Mn and dispersity could possibly be improved for pHPMAm
with optimization of the polymerization conditions used
(solvent, initiator, RAFT agent, temperature). McCormick
and co-workers, for example, report obtaining high molecular
weights and low dispersities46 for RAFT-mediated polymer-
izations of HPMAm in acetate buffer. Similar conditions,
however, yielded inferior results in our experience. The use of
HPAm resulted in reactions that demonstrate desirable kinetics
while being amenable to RAFT control without time-
consuming optimization procedures.
One feature of RAFT-mediated polymerization is that not
every RAFT agent will provide the desired control over every
monomer. For example, some RAFT agents are more suitable
for methacrylate monomers. The degree of control that a
particular RAFT agent facilitates is determined by the
substituents, the “Z” and “R” groups, of the central
thiocarbonylthio group. In this study we explicitly chose a
RAFT agent, a trithiocarbonate with a cyanovaleric acid
substituent, which is compatible with and has been used
previously for the controlled polymerization of HPMAm.47−49
It should be noted, however, that in addition to increased
polymerization rates, one advantage of acrylamide monomers
relative to methacrylamide monomers is the greater variety of
RAFT agents that can provide control of polymerization.
Figure 2. Conversion plotted against time of HPMAm (solid line) and
HPAm (dashed line) under identical polymerization conditions: 1.75
M monomer, 5.0 mM RAFT agent, and 0.50 mM AIBN in DMSO at
80 °C. Measured conversions indicated by standard error bars. Lines
are drawn to guide the eye. Error bars represent standard deviation at
each time point.
the low residual concentration of AIBN, which at this point
would be almost entirely exhausted. But examining just the first
2 h of polymerization, where the depletion of AIBN was not
limiting and the consumption of monomers was linear, the
polymerization rate of HPAm exceeded that of HPMAm by a
factor of 8. The rate of polymerization is described by eq 1,
where kp is the polymerization rate constant, [M] is the
concentration of monomer, [I] is the concentration of initiator,
kd is the disassociation rate constant of the initiator, kt is the
rate constant of termination, and f is the initiator efficiency.44
1/2
⎛
⎜
⎝
⎞
⎟
⎠
fkd[I]
kt
R = k [M]
p
p
(1)
Polymerization conditions were chosen so as to keep as many
variables in eq 1 equivalent among polymerizations, and
therefore, based on the relative rates of monomer consumption
observed in these experiments, kp k−t 1/2 = 8kp kt−1/2, where
Am Am
MAm Am
kp and kt are the polymerization and termination rate
Am
Am
constants, respectively, for HPAm and kp and kt are the
Am
Am
corresponding rate constants for the HPMAm.
Many applications in the development of biomedical
materials require polymers of defined and narrowly dispersed
molecular weights that are of a particular architecture. For this
reason, the amenability of HPAm to RAFT-controlled
polymerization was evaluated against HPMAm under identical
reaction conditions. GPC-MALS was used to determine the
molecular weights and dispersities of the polymers resulting
a
Table 1. Summary of GPC-MALS Results for Polymers Resulting from Kinetics Experiment
Mn (target) kg mol−1
Mn (theoretical) kg mol−1
conversion %
Mn (measured) kg mol−1
PDI
pHPMAm
pHPAm
50
45
15
41
29
91
1
1
21 2.1
41 2.8
1.4 0.10
1.1 0.06
a
Based on the concentration of initiator and RAFT agent, target Mn is the value that would be obtained if reactions proceeded without nonidealities
and conversions were complete. Theoretical Mn refers to the value that is calculated based on the concentration of initiator and RAFT agent and the
observed, rather than absolute, conversion. Measured Mn and PDI were determined via GPC-MALS.
3262
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
Polymers of HPMAm have demonstrated outstanding
characteristics in biomedical research but have been limited
by relatively low molecular weights and slow polymerization.
The improvement of polymerization kinetics and control with
HPAm relative to HPMAm allow the monomer greater
versatility in its application. Radical damage during grafting to
therapeutic proteins, for example, may be mitigated by a lower
initiator concentration. Growing polymers off of RAFT-
functionalized surfaces may be performed in less time and
with a greater polymer surface density. Polymers with pHPAm
blocks of high degrees of polymerization may be generated to
obtain otherwise elusive desired properties. It must be
demonstrated, however, that pHPAm shares the same advanta-
geous qualities of pHPMAm. To that end, materials made with
HPAm were compared directly with those made from HPMAm
to determine whether the biological interfacial reactions were
similar.
hypothesized that materials made from HPAm would show
similarly low protein adsorption properties. Protein adsorption
to surfaces was measured using a time-resolved fluorescence
europium assay. Surfaces were incubated for 18 h with 10%
fetal bovine serum in PBS supplemented with 0.30 μg/mL
europium-labeled fibronectin (approximately 1% relative to the
endogenous fibronectin in FBS). These conditions were chosen
to approximate physiological environments. Following adsorp-
tion and subsequent washing, europium(III) was liberated from
adsorbed protein via a dissociation “enhancement” solution.
From the 620 nm wavelength emission of the released
europium, surface-bound fibronectin was calculated. The
sensitivity of this technique (detection of femtomole of Eu-
labeled protein) is similar to that of 125I radiolabeling. As Figure
3 indicates, total fibronectin adsorption on TCPS was
One area in which pHPMAm has shown great utility is in the
generation of surfaces which have low protein adsorption
properties. Such surfaces ostensibly allow the fabrication of so-
called, “stealth” biomaterials that can be implanted with
potentially little foreign body response. To explore the
suitability of pHPAm for the preparation of similar materials,
hydrogels of both pHPMAm and pHPAm, cross-linked via
N,N-methylenebis(acrylamide), were synthesized and the
surface wetability and protein adsorption to the surface of
each assessed. The effect of the cross-linker on these properties
was assumed to be negligible because of the small relative
proportion in composition (approximately 2% by weight) and
the chemical group similarity with HPAm and HPMAm. Each
hydrogel was compared against PEGDMA hydrogels and
ultralow attachment coated polystyrene (Corning; ULA) and
tissue culture treated polystyrene (TCPS) in cell attachment
assays.
Certain property requirements are generally considered
important in polymer coatings used in low adhesion
applications: very hydrophilic and composed of materials
containing no hydrogen bond donors.10,50 Surfaces grafted
with HPMAm polymers, however, not only present hydrogen
bond donor sites but also exhibit only moderate wetability. This
was observed in the fabricated hydrogels where sessile contact
angles at room temperature were measured to be 45°, Table 2.
Figure 3. Binding of fibronectin to TCPS, ULA, pHPAm, pHPMAm,
and PEGDMA hydrogels, n = 3. Error bars indicate confidence
intervals (α = 0.01).
approximately 160 ng/cm2. Adsorption on pHPAm hydrogels
was very low-less than 1% of that measured for the fibronectin
adsorption on TCPS and roughly equal to that of the ULA-
treated polystyrene. Fibronectin adsorption was also very low
on the PEG hydrogels, approximately 3% of that of TCPS.
While adsorption was low on pHPMAm hydrogels, it was
substantially higher than on the other hydrogels, roughly 10%
relative to the TCPS control.
Table 2. Contact Angles and Representative Images for
Hydrogels
Typically, low protein adsorption on materials translates to
low cell attachment as bound endogenous proteins are not
present to provide adhesion sites for the cells. It is also notable
that soluble polymers with low protein interaction are often
conjugated to therapeutic agents to tune pharmacokinetic
properties. Low interaction of proteins at pHPAm surfaces
suggests the possible application of pHPAm as an alternative to
PEGylation of bioactive compounds in a similar fashion to that
demonstrated for pHPMAm.11,32,35−37
Presented in Figure 4 are the results obtained for cell
attachment experiments on synthesized hydrogels as well as the
positive and negative controls (TCPS and ULA surfaces,
respectively). As can be seen in Figure 4a, TCPS supported
substantial L929 cellular adhesion. No cells were observed
floating freely when the plate was rocked gently. Moreover,
many of the attached cells were flattened with a spread
morphology and protrusions, suggesting significant integrin-
mediated binding.
Hydrogels of HPAm exhibited statistically similar contact
angles (44°), indicating similar hydophilicity on the surface of
the materials, while PEGDMA hydrogels were determined to
be only slightly more hydrophilic (Table 2). These results are
consistent with measurements made by authors of another
study on brush-polymer coatings of pHPMAm and pPEGMA33
and support their observations that the mechanisms for the
suppression of protein adsorption are poorly understood.
Because surfaces coated with pHPMAm brushes have been
reported to show very low protein adsorption, it was
3263
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
TCPS samples resulted in no observable change in cellular
density, attachment, or spreading.
To quantitatively verify observed cellular attachment, the
metabolic activity of the adherent cells was determined via a
MTS assay. This assay measures the enzymatic conversion of
MTS to a colorimentric product, roughly proportional to the
number of metabolically active cells present in or on a sample.
Presented in Figure 5a is the absorbance measured from cells
attached to TCPS, ULA, pHPAm, pHPMAm, and PEGDMA
hydrogel surfaces following a gentle rinse to remove non-
adherent cells. The absorbances values, proportional to the
number of remaining cells, were normalized to that of the
TCPS sample (the positive control). The total metabolic
activity from cells attached to pHPAm, pHPMAm, and
PEGDMA hydrogels was approximately 1% of that obtained
for the TCPS surface. These results represent a statically robust
difference (p < 10−7) for both pHPMAm and pHPAm surfaces
compared to TCPS). There was no statistically significant
difference among the pHPMAm, pHPAm, PEGDMA, and ULA
surfaces. To verify that the low cell adhesion observed was a
result of occlusion of cellular binding at the surface and not
because of cell morbidity, metabolic assays were performed on
the reserved media with which the plate wells and hydrogels
were rinsed. These measurements were normalized to that
obtained for the ULA plate. As can be seen in Figure 5b, the
media from TCPS exhibited extremely low metabolic activity,
because the viable cells remained adhered and few were
removed by the gentle rinsing procedure. The reserved media
from the ULA plate was, however, metabolically active as the
rinsing procedure collected most cells from the surface. The
metabolic assay confirms the viability of those cells as well.
Similarly, cells removed from the pHPAm, pHPMAm, and
PEGDMA hydrogels remained viable and metabolically active.
There was no statistical difference between the reserved ULA
media and the media from the hydrogels (α = 0.01), indicating
little if any cell morbidity associated therewith.
Figure 4. Cellular attachment of L292s after 24 h on (a) tissue culture
polystyrene, (b) ultralow adhesion polystyrene, (c) HPAm hydrogels,
(d) HPMAm hydrogels, and (e) PEG hydrogels.
Cells seeded on ULA, on the other hand, bound together in
cellular clumps rather than to the surface, Figure 4b. Cells
maintained a rounded morphology and floated freely in the
media. A small minority of cells remain unattached to larger
clumps, but they too exhibited a rounded morphology and
failed to adhere to the plate.
Like those cells seeded on the ULA surface, L929 fibroblasts
seeded on the pHPAm, pHPMAm, and PEGDMA hydrogel
surfaces (Figure 4c−e) formed clumps that floated freely in the
media and were not attached. The clumps formed by cells over
the pHPAm hydrogel surface appeared, on average, to be larger,
and fewer individual cells or clumps of 2−3 cells were observed.
Following gentle rinsing, cells were removed from the ULA
plate as well as from the pHPAm, pHPMAm, and PEGDMA
gel surfaces almost entirely. The same rinsing procedure on the
CONCLUSION
■
While HPMAm has demonstrated efficacy as a polymer for the
control of biointerfacial interactions, its slow polymerization
kinetics limits its utility. Here we have demonstrated that
HPAm provides a promising substitute for HPMAm for
applications in low protein and cell attachment materials
featuring equivalent performance, but improved polymerization
Figure 5. (a) Absorbance from MTS metabolic assay of cells remaining on surfaces following gentle rinsing. Values normalized to that of TCPS. (b)
Absorbance from MTS metabolic assay of media collected from rinsing of cell-seeded surfaces. Values normalized to that of media removed from the
ULA plate. Taken together, these data indicate that cells exhibited neglible mortality on and failed to adhere to HPAm, HPMAm and PEGDMA
hydrogels similar to ultralow adhesion culture plates. Error bars represent confidence intervals (α = 0.01).
3264
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
(13) Bazile, D.; Prudhomme, C.; Bassoullet, M. T.; Marlard, M.;
Spenlehauer, G.; Veillard, M. J. Pharm. Sci. 1995, 84, 493.
(14) Kenausis, G. L.; Voros, J.; Elbert, D. L.; Huang, N. P.; Hofer, R.;
Ruiz-Taylor, L.; Textor, M.; Hubbell, J. A.; Spencer, N. D. J. Phys.
Chem. B 2000, 104, 3298.
and control. The polymerization in DMSO was determined to
be faster by almost an order of magnitude. In addition, we have
demonstrated that higher molecular weights can be obtained
and that more uniform dispersities can be achieved via
controlled processes such as RAFT-mediated polymerization.
Hydrogels formed from HPAm exhibited at least equivalent
prevention of cell attachment compared to HPMAm-based
materials, underlining the suitability of this monomer in
applications where low biointerfacial interactions are required.
Additionally, the low interaction of proteins with the
polyHPAm material may portend its broader suitability in the
development of polymer-conjugated therapeutics. HPAm is
structurally identical to HPMAm, with the exception of a single
methyl group. This small distinction translates to a remarkable
difference in polymerization characteristics, but with no
apparent deficit in properties, which are important in
controlling biological interaction. This study establishes the
monomer HPAm as one of relatively few monomers that can be
used for the manufacture of low cell attachment materials and
surfaces that are expected to be more stable under physiological
conditions than (meth)acrylate polymers and other materials
containing hydrolabile moieties.
(15) Kingshott, P.; Thissen, H.; Griesser, H. J. Biomaterials 2002, 23,
2043.
(16) Fairbanks, B. D.; Schwartz, M. P.; Halevi, A. E.; Nuttelman, C.
R.; Bowman, C. N.; Anseth, K. S. Adv. Mater. 2009, 21, 5005.
(17) Cruise, G. M.; Scharp, D. S.; Hubbell, J. A. Biomaterials 1998,
19, 1287.
(18) Allen, T. M.; Cullis, P. R. Science 2004, 303, 1818.
(19) Moghimi, S. M.; Hunter, A. C.; Murray, J. C. FASEB J. 2005, 19,
311.
(20) Tanaka, Y.; Nishida, N.; Sugiyama, M.; Kurosaki, M.; Matsuura,
K.; Sakamoto, N.; Nakagawa, M.; Korenaga, M.; Hino, K.; Hige, S.;
Ito, Y.; Mita, E.; Tanaka, E.; Mochida, S.; Murawaki, Y.; Honda, M.;
Sakai, A.; Hiasa, Y.; Nishiguchi, S.; Koike, A.; Sakaida, I.; Imamura, M.;
Ito, K.; Yano, K.; Masaki, N.; Sugauchi, F.; Izumi, N.; Tokunaga, K.;
Mizokami, M. Nat. Genet. 2009, 41, 1105.
(21) Zhang, M. Q.; Desai, T.; Ferrari, M. Biomaterials 1998, 19, 953.
(22) Branch, D. W.; Wheeler, B. C.; Brewer, G. J.; Leckband, D. E.
Biomaterials 2001, 22, 1035.
(23) Hamad, I.; Hunter, A. C.; Szebeni, J.; Moghimi, S. M. Mol.
Immunol. 2008, 46, 225.
(24) Feng, W.; Zhu, S. P.; Ishihara, K.; Brash, J. L. Langmuir 2005,
21, 5980.
(25) Lewis, A. L. Colloids Surf., B 2000, 18, 261.
(26) Nakabayashi, N.; Williams, D. F. Biomaterials 2003, 24, 2431.
(27) Wang, D. A.; Williams, C. G.; Li, Q. A.; Sharma, B.; Elisseeff, J.
H. Biomaterials 2003, 24, 3969.
ASSOCIATED CONTENT
* Supporting Information
The NMR and MS spectra of synthesized monomers. This
material is available free of charge via the Internet at http://
pubs.acs.org.
■
S
(28) Zhang, Z.; Chao, T.; Chen, S.; Jiang, S. Langmuir 2006, 22,
10072.
AUTHOR INFORMATION
Corresponding Author
*E-mail: ben.fairbanks@csiro.au.
Notes
■
(29) West, S. L.; Salvage, J. P.; Lobb, E. J.; Armes, S. P.; Billingham,
N. C.; Lewis, A. L.; Hanlon, G. W.; Lloyd, A. W. Biomaterials 2004, 25,
1195.
(30) Zhang, Z.; Chen, S.; Jiang, S. Biomacromolecules 2006, 7, 3311.
(31) Kopecek, J.; Bazilova, H. Eur. Polym. J. 1973, 9, 7.
(32) Kopecek, J.; Kopeckova, P. Adv. Drug Delivery Rev. 2010, 62,
122.
(33) Rodriguez-Emmenegger, C.; Brynda, E.; Riedel, T.; Houska, M.;
Subr, V.; Alles, A. B.; Hasan, E.; Gautrot, J. E.; Huck, W. T. S.
Macromol. Rapid Commun. 2011, 32, 952.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The authors would like to acknowledge John Bentley who
performed the europium labeling of fibronectin.
■
(34) Kamei, S.; Kopecek, J. Pharm. Res. 1995, 12, 663.
(35) Sugiyama, K.; Mitsuno, S.; Shiraishi, K. J. Polym. Sci., Polym.
Chem. 1997, 35, 3349.
(36) Etrych, T.; Jelinkova, M.; Rihova, B.; Ulbrich, K. J. Controlled
Release 2001, 73, 89.
(37) Omelyanenko, V.; Kopeckova, P.; Gentry, C.; Kopecek, J. J.
Controlled Release 1998, 53, 25.
(38) Roy, D.; Ghosn, B.; Song, E.-H.; Ratner, D. M.; Stayton, P. S.
Polym. Chem. 2013, 4, 1153.
(39) Valade, D.; Boyer, C.; Davis, T. P.; Bulmus, V. Aust. J. Chem.
2009, 62, 1344.
(40) Moad, G.; Chong, Y. K.; Postma, A.; Rizzardo, E.; Thang, S. H.
Polymer 2005, 46, 8458.
(41) Convertine, A. J.; Benoit, D. S. W.; Duvall, C. L.; Hoffman, A.
S.; Stayton, P. S. J. Controlled Release 2009, 133, 221.
(42) Fairbanks, B. D.; Schwartz, M. P.; Bowman, C. N.; Anseth, K. S.
Biomaterials 2009, 30, 6702.
REFERENCES
■
(1) Absolom, D. R.; Zingg, W.; Neumann, A. W. J. Biomed. Mater.
Res. 1987, 21, 161.
(2) Amiji, M.; Park, K. J. Colloid Interface Sci. 1993, 155, 251.
(3) Dewez, J. L.; Berger, V.; Schneider, Y. J.; Rouxhet, P. G. J. Colloid
Interface Sci. 1997, 191, 1.
(4) Banerjee, P.; Irvine, D. J.; Mayes, A. M.; Griffith, L. G. J. Biomed.
Mater. Res. 2000, 50, 331.
(5) Hubbell, J. A. Nat. Biotechnol. 1995, 13, 565.
(6) Ameringer, T.; Fransen, P.; Bean, P.; Johnson, G.; Pereira, S.;
Evans, R. A.; Thissen, H.; Meagher, L. J. Biomed. Mater. Res., Part A
2012, 100, 370.
(7) Yu, Q.; Zhang, Y.; Wang, H.; Brash, J.; Chen, H. Acta Biomater.
2011, 7, 1550.
(8) Roldan-Pallares, M.; Hernandez-Montero, J.; Llanes, F.;
Fernandez-Rubio, J. E.; Ortega, F. Arch. Ophthalmol. (Chicago, IL, U.
S.) 2007, 125, 511.
(43) Coad, B. R.; Lu, Y.; Meagher, L. Acta Biomater. 2012, 8, 608.
(44) Odian, G. Principles of Polymerization; John Wiley and Sons,
Inc.: New York, 1991.
(9) Roldan-Pallares, M.; Sanz, J. L. D.; Awad-El Susi, S.; Refojo, M. F.
Arch. Ophthalmol. (Chicago, IL, U. S.) 1999, 117, 197.
(10) Chen, H.; Yuan, L.; Song, W.; Wu, Z.; Li, D. Prog. Polym. Sci.
2008, 33, 1059.
(45) Moad, G.; Solomon, D. The Chemistry of Radical Polymerization;
Elsevier Ltd.: Oxford, 2006.
(11) Knop, K.; Hoogenboom, R.; Fischer, D.; Schubert, U. S. Angew.
(46) Scales, C. W.; Vasilieva, Y. A.; Convertine, A. J.; Lowe, A. B.;
McCormick, C. L. Biomacromolecules 2005, 6, 1846.
(47) Hong, C. Y.; Pan, C. Y. Macromolecules 2006, 39, 3517.
Chem., Int. Ed. 2010, 49, 6288.
(12) Thissen, H.; Gengenbach, T.; du Toit, R.; Sweeney, D. F.;
Kingshott, P.; Griesser, H. J.; Meagher, L. Biomaterials 2010, 31, 5510.
3265
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266

Biomacromolecules
Article
(48) Luo, K.; Yang, J.; Kopeckova, P.; Kopecek, J. Macromolecules
2011, 44, 2481.
(49) Duvall, C. L.; Convertine, A. J.; Benoit, D. S. W.; Hoffman, A.
S.; Stayton, P. S. Mol. Pharmaceutics 2010, 7, 468.
(50) Ostuni, E.; Chapman, R. G.; Holmlin, R. E.; Takayama, S.;
Whitesides, G. M. Langmuir 2001, 17, 5605.
3266
dx.doi.org/10.1021/bm500654q | Biomacromolecules 2014, 15, 3259−3266