Probing structure–antifouling activity relationships of polyacrylamides and polyacrylates

Biomaterials 34 (2013) 4714e4724

Contents lists available at SciVerse ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Probing structureeantifouling activity relationships of polyacrylamides and polyacrylates Chao Zhao a,1, Jun Zhao a,1, Xiaosi Li a, Jiang Wu a, c, Shenfu Chen c, Qiang Chen a, d, Qiuming Wang a, Xiong Gong b, Lingyan Li a, Jie Zheng a, * a

Department of Chemical and Biomolecular Engineering, The University of Akron, Akron, OH 44325, USA Department of Polymer Engineering, The University of Akron, Akron, OH 44325, USA State Key Laboratory of Chemical Engineering, Department of Chemical and Biological Engineering, Zhejiang University, Hangzhou 310027, China d School of Material Science and Engineering, Henan Polytechnic University, Jiaozuo 454003, China b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 January 2013 Accepted 11 March 2013 Available online 2 April 2013

We have synthesized two different polyacrylamide polymers with amide groups (polySBAA and polyHEAA) and two corresponding polyacrylate polymers without amide groups (polySBMA and polyHEA), with particular attention to the evaluation of the effect of amide group on the hydration and antifouling ability of these systems using both computational and experimental approaches. The influence of polymer architectures of brushes, hydrogels, and nanogels, prepared by different polymerization methods, on antifouling performance is also studied. SPR and ELISA data reveal that all polymers exhibit excellent antifouling ability to repel proteins from undiluted human blood serum/plasma, and such antifouling ability can be further enhanced by presenting amide groups in polySBAA and polyHEAA as compared to polySBMA and polyHEA. The antifouling performance is positively correlated with the hydration properties. Simulations confirm that four polymers indeed have different hydration characteristics, while all presenting a strong hydration overall. Integration of amide group with pendant hydroxyl or sulfobetaine group in polymer backbones is found to increase their surface hydration of polymer chains and thus to improve their antifouling ability. Importantly, we present a proof-of-concept experiment to synthesize polySBAA nanogels, which show a switchable property between antifouling and pH-responsive functions driven by acidebase conditions, while still maintaining high stability in undiluted fetal bovine serum and minimal toxicity to cultured cells. This work provides important structural insights into how very subtle structural changes in polymers can yield great improvement in biological activity, specifically the inclusion of amide group in polymer backbone/sidechain enables to obtain antifouling materials with better performance for biomedical applications. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Antifouling materials Polyacrylamide Polyacrylate Protein adsorption Cell adhesion Stimuli-responsive materials

1. Introduction Biofouling is defined as the spontaneous accumulation of undesirable proteins, cells, bacteria, and microorganisms on artificial surfaces of medical implants [1e3], drug delivery carriers [4,5], biosensors [6], and ship hulls [7e12]. In most cases, once biofouling occurs, it will irreversibly impair not only the function of biomedical devices through blood clot formation, tissue fibrosis, thrombosis coagulation, and bacterial infection, but also the performance of many industrial applications of nano/microfiltration, membrane separation, pipe corrosion, and ship navigation through biofilm * Corresponding author. E-mail address: [email protected] (J. Zheng). 1 These authors contributed equally to this work. 0142-9612/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biomaterials.2013.03.028

formation. Many polymeric materials including poly(ethylene glycol) (PEG) [13e19], poly(2-hydroxyethyl methacrylate) (polyHEMA) [20], poly(hydroxypropyl methacrylate) (polyHPMA) [21], tetraglyme [6], dextran [22], mannitol [23], glycerol dendron [24], poly(sulfobetaine methacrylate) (polySBMA) [25e27], poly(carboxybetaine methacrylate) (polyCBMA) [28e30], and poly(2-methacryloyloxyethyl phosphorylcholine) (polyMPC) [31e34] have been developed to resist protein adsorption, cell/bacterial adhesion, and biofilm formation. Although these antifouling polymeric materials have different chemical and structural characteristics (e.g. chemical structure, hydrophobicity, charge distribution, geometrical properties, molecular conformation/architecture/sequence/weight, etc.) in both monomeric and polymeric forms, they all possess certain degrees of antifouling capabilities in different biological media. However, the exact structuraleproperty relationship of these antifouling

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

materials still remains unclear, which leads to difficulties in fundamental understanding of the role of chemical and structural characteristics in antifouling properties and in practical design of new antifouling biomaterials. Several mechanisms have been proposed to interpret the resistance of materials to protein adsorption and microorganism attachment on the surface. Considering the fact that all existing antifouling materials contain either hydrophilic or zwitterionic moieties, the “water barrier” theory [14,35,36] suggests that a tightly bound water layer formed around the materials provides a physical and energetic barrier to prevent biomolecule adsorption on the surface. Expulsion of water molecules from the interfacial region between polymers and proteins requires strong surfacee protein interactions to compensate solvent entropy loss for protein adsorption. Such tightly bound hydration layer at the polymer interface can be achieved differently, i.e. hydrophilic polymers achieve surface hydration via hydrogen bonds, while zwitterionic polymers achieve hydration via ionic solvation. Due to flexible nature of polymer chains, the “steric repulsion” resulting from the compression of polymer chains as proteins approach the surface is also proposed to be responsible for prevention of protein adsorption [37]. Moreover, adsorption kinetic models highlight the importance of surface density of grafted polymers to resist protein adsorption, presumably because high surface coverage through increased polymer density reduces possible binding sites for protein adsorption on the supporting substrate, resulting in protein resistance [38]. In addition, polymer conformation and architecture were also found to be a key factor to control protein adsorption [39e42]. Despite of different antifouling mechanisms, it is not likely that a single mechanism is solely responsible for the onset of antifouling events, but rather a combination of many. More importantly, underlying intermolecular interactions among proteins, materials, and solvent at atomic level that are not well understood are the key determinant for the macroscopic antifouling

4715

performance of the materials. Thus, fundamental understanding of molecular interactions between materials and their surroundings is equally important for rational design of biomaterials. Recent studies from our and other works have shown that under optimal conditions, both zwitterionic polyCBAA and hydrophilic polyHEAA can achieve similar “zero” protein adsorption from undiluted human blood plasma and serum [28,43]. From a molecular structural point of view, although both CBAA and HEAA monomers have major structural differences in molecular size, surface hydrophobicity, and partial charge distributions, they both possess a common structural motif of an amide group in the backbone. The incorporation of the amide group into the hydroxyl group of HEAA or the carboxybetaine group of CBAA is expected to promote the formation of a hydration layer. In this study, we aim to elucidate the structureeantifouling activity relationship of polyacrylamide and polyacrylate with and without amide groups using combined experimental and computational approaches. The hydration and antifouling abilities in vitro of four different polymers of polySBAA and polyHEAA with amide groups and polySBMA and polyHEA without amide groups (Scheme 1) in polymer brush and hydrogel forms (Scheme 2) was characterized. Additionally, we synthesized polySBAA-based nanoparticles to test the control-release of R6G drugs upon pH-responsive changes by taking advantage of its integrated superlow fouling ability and zwitterionic nature (Scheme 3). 2. Materials and methods 2.1. Materials [3-(Methacryloylamino)propyl]dimethyl(3-sulfopropyl)ammonium hydroxide (SBAA, 96%), [2-(methacryloyloxy)ethyl]dimethyl-(3-sulfopropyl)ammonium hydroxide (SBMA, 97%), N-hydroxyethyl acrylamide (HEAA, 97%), 2-hydroxyethyl acrylate (HEA, 96%), 2-(methacryloyloxy) ethyl trimethyl ammonium (TM, 80 wt.% in H2O), 2,20 -bipyridyl (BPY, 99%), copper(I) bromide (99.999%), copper(I) chlorine (99.999%), N,N0 -methylene-bis-acrylamide (MBAA), 2-hydroxy-40 -(2-hydroxyethoxy)2-methylpropiophenone (98%), Span 80 (sorbitan monooleate), Tween 80

Scheme 1. Molecular structures of four monomers used in this work.

4716

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724 (polyethyleneglycolsorbitan monooleate), Rhodamine 6G (R6G, 99%), ethylene glycol (99.5%), hexane (95%), phosphate citrate buffer (pH 5.0), hydrogen peroxide (30 wt.% in H2O), sulfuric acid (99.999%), o-phenylenediamine (98%), and ethanol (absolute 200 proof), human plasma fibrinogen (Fg), phosphate buffer saline (PBS, pH 7.4, 0.15 M, 138 mM NaCl, 2.7 mM KCl), methanol (99.5%), tetrahydrofuran (THF HPLC) and hydrochloric acid (HCl, 37%) were purchased from SigmaeAldrich (Milwaukee, WI). Horseradish peroxidase (HRP)-conjugated polyclonal goat anti-human fibrinogen was purchased from USBiological (Swampscott, MA). Tris[2-(dimethylamino)ethyl]amine (Me6TREN, 99%) was purchased from Alfa Aesar (Ward Hill, MA). Pooled human blood plasma and serum, and fetal bovine serum (FBS) were purchased from BioChemed Services (Winchester, VA). 2,20 -Azobis(4-methoxy-2.4-dimethyl valeronitrile) (V-70) was purchased from Wako Pure Chemical Industries (Osaka, Japan). Water used in these experiments was purified by a Millipore water purification system with a minimum resistivity of 18.0 MU cm. All other cell culture reagents were purchased from Invitrogen (Grand Island, NY). The ATRP initiator, u-mercaptoundecyl bromoisobutyrate, was synthesized through the reaction of bromoisobutyryl bromide and 11-mercapto-1-undecanol, using the method published previously [44].

SI-ATRP

Polymer Brush

n io at iz er m ly Po

Nanogel

o ot Ph

In ve rs Po e M ly ic m ro er e iz m a t ul io si n on

Acrylamide and Acrylate Monomers

Bulk Hydrogel

Scheme 2. Preparation of polymer brushes, nanogels, and hydrogels.

2.2. Surface-initiated atom transfer radical polymerization (SI-ATRP) on surface plasmon resonance (SPR) chips The SPR chips were prepared by coating an adhesion-promoting chromium layer (2 nm in thickness) and a plasmon active gold layer (48 nm) onto glass substrates using electron beam evaporation under vacuum. The SPR chips were rinsed with pure ethanol and water, a UV light treatment for 20 min, washed by water and pure ethanol again, and dried in a stream of dry air prior to use. The initiator selfassembling monolayers (SAMs) were formed by soaking pre-cleaned SPR chips into 0.1 mM ethanol solution of u-mercaptoundecyl bromoisobutyrate at room temperature overnight, and then the chips were washed sequentially with THF and ethanol to remove unbound initiators, and dried in a stream of dry air before use. PolySBAA and polyHEAA with amide groups were grafted onto SPR chips covered with initiators via SI-ATRP. Briefly, CuCl (16.8 mg, 0.170 mmol) and an SPR chip covered with initiator were placed in a reaction tube under nitrogen protection and sealed with rubber septum stoppers. Degassed methanol (2 mL) was first injected into the tube using a syringe to dissolve the catalysts. Then, a degassed solution (water and methanol in a 5:3 volume ratio, 8 mL) with SBAA or HEAA (3.42 mmol) and Me6TREN (40 mg, 0.174 mmol) was transferred to the tube using a

Scheme 3. Illustration of acid-triggered release of R6G from polySBAA nanogel.

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724 syringe under nitrogen protection. The mixture was protected by dry nitrogen until the reaction was completed. After reaction, the chip was taken out and washed by ethanol and water, rinsed with PBS buffer overnight to remove any unbound or weakly bound polymer. PolySBMA and polyHEA brushes were prepared using the same protocol as described above. Only difference to prepare these three brushes is to use CuBr (14.3 mg, 0.1 mmol) and BPY (31.2 mg, 0.2 mmol) as ATRP catalysts. The film thicknesses of all polymer brushes were maintained at 20 nm by carefully controlling ATRP reaction time as reported in our previous work [20,21,43]. The thicknesses of polymer brushes on SPR chips in air were confirmed by an a-SE Ellipsometer (J. A. Woollam Co. Lincoln, NE, USA) with a HeeNe laser (l ¼ 632.8 nm) and a fixed angle of incidence of 70 . A refractive index of 1.45 was assigned to the polymer brushes.

4717

to 100  C, and then re-heated back to 40  C at a rate of 5  C/min. under the protection of nitrogen gas at a flow rate of 25 mL/min. Weight ratio (Wnonfreezable) of nonfreezable water:polymer in the hydrogels is calculated by Wnonfreezable ¼ wnonfreezable/wpolymer ¼ (EWC  wfreezable)/wpolymer, where the wnonfreezable, wfreezable, and wpolymer represent the weight percentage of nonfreezable water, freezable water, and polymer in hydrogels, respectively. wfreezable can be experimentally obtained by wfreezable ¼ DHf/DHw  100%, where DHf is the enthalpy change associated with the melting of freezable water per weight of a hydrogel by DSC and DHw is the enthalpy change for the melting of bulk water by DSC [45,46]. Wnonfreezable ratio can also be converted to the number (Nw) of nonfreezable water molecules per repeating polymer unit using Nw ¼ Wnonfreezable  Mp/ Mw, where Mp is the molecular weight per repeating polymer unit and Mw is molecular weight of water.

2.3. Protein adsorption measurements by SPR Nonspecific protein adsorption on polymer brushes was evaluated with an inhouse four-channel SPR sensor, which measures change in the resonant wavelength at a fixed angle. The polymer-grafted SPR chip was first attached to the base of the prism. Optical contact between two surfaces was realized using a refractive index matching fluid (Cargille). A pre-adsorptive baseline was first established by flowing degassed PBS buffer through the sensor until the baseline became stable. 100% blood serum or 100% blood plasma was then delivered through the channels for 10 min using a peristaltic pump, followed by flushing each channel with PBS for 5 min to remove any unbound protein and to re-establish the post-adsorptive baseline. A flow rate of 0.05 mL/min was used for all experiments. Protein adsorption was finally quantified by measuring wavelength shift between the preadsorptive and post-adsorptive baselines and converting wavelength-shift difference to the amount of the adsorbed proteins. A 1-nm SPR wavelength shift at 750 nm corresponds to a protein surface coverage of 15 ng/cm2, as used in our previous works [20,21,43]. 2.4. Tapping mode atomic force microscopy (TM-AFM) To characterize the film roughness of polymer brushes prepared via SI-ATRP, a polymer-grafted SPR chip was rinsed in PBS buffer overnight to remove any unbound or weakly bound polymer, and dried with compressed N2 before AFM imaging. Tapping mode AFM imaging was performed in air using a Nanoscope IIId multimode scanning probe microscope (Veeco Corp., Santa Barbara) equipped with a 15 mm E scanner. Commercial Si cantilevers (NanoScience) with an elastic modulus of w40 N m1 were used. All images were acquired as 512  512 pixel images at a typical scan rate of 1.0e2.0 Hz with a vertical tip oscillation frequency of 250e350 kHz. Representative images of each sample were obtained by scanning at least 5 different locations. 2.5. Hydrogel preparation Four polymer hydrogels of polySBAA, polySBMA, polyHEAA, and polyHEA were prepared by adding “monomer solution” (4 mM monomer, 10 mg of photoinitiator, 2hydroxy-40 -(2-hydroxyethoxy)-2-methylpropiophenone, and MBAA as a crosslinker) to “mixed solvent” (0.375 mL ethanol, 0.565 mL ethylene glycol, and 0.565 mL H2O). The resulting solution was gently sonicated to well mix in an ice bath to prevent premature polymerization. To polymerize, the solution was transferred into a pair of glass plates separated by 1 mm-thick poly(tetrafluoroethylene) (PTFE) spacers. The photo-polymerization reaction was carried out at room temperature with 362 nm UV light for 1 h. The resulting hydrogels were removed from the plates and immersed in a large volume of PBS. The PBS was changed every 3 h daily for 5 days to remove un-reacted chemicals and excess salts before further use. Nonspecific protein interactions with these hydrogels were measured by enzyme-linked immunosorbent assay as described in Section 2.7. 2.6. Hydrogel characterization Equilibrium water content (EWC) measurement. The EWC of hydrogels (polySBAA, polySBMA, polyHEAA, and polyHEA) was assessed by weighting the mass of fully hydrated and dry hydrogel samples. First, hydrogel samples were incubated in PBS for 1 day for complete swelling. The fully wet and swollen samples were then punched into 8 mm disks (8 mm biopsy punch, Acuderm, FL) and weighed as mw. The wet samples were dehydrated at 65  C under vacuum for 72 h and re-weighted as md. The EWC of the samples was calculated by (mw  md)/mw  100%. All data were averaged by three repeated experiments. Differential scanning calorimetry (DSC) measurement. DSC is used to characterize the state of water (i.e. freezable and nonfreezable water) in polymer hydrogels. First, 4e6 mg hydrogels were placed in an aluminum pan, and the pan was hermetically sealed. Enthalpy change (DHf) associated with the melting of freezable water per weight of a hydrogel was measured using a DSC (Q2000, TA Instruments, DE) equipped with a liquid nitrogen cooling system (LNCS) for non-isothermal experiments. Temperature and enthalpy calibration were performed using an indium standard. Sapphire disks were used for heat capacity calibration. During the cooling and heating experiments, the samples were cooled down from room temperature

2.7. Fibrinogen adsorption on hydrogels by enzyme-linked immunosorbent assay (ELISA) The adsorption of fibrinogen (Fg) on polymer hydrogels was assessed using the ELISA method according to the standard protocol described below. First, the hydrogels were punched into 8 mm circular disks, and incubated in PBS at room temperature for 24 h. The 8-mm hydrogel disks were then placed into individual wells of a 24-well plate in triplicates, and incubated with 500 mL of 1 mg/mL Fg in PBS at 37  C for 90 min. To block the areas unoccupied by Fg, the disks were washed with PBS by 5 times, followed by the incubation in 500 mL of 1 mg/mL BSA at 37  C for additional 90 min. After washing with PBS 5 times, the disks were transferred to new wells and incubated in 500 mL of horseradish peroxidase (HRP)-conjugated anti-Fg (10 mg/mL buffer) at 37  C for 30 min, followed by 5 washes with PBS. For the color development, the disks were transferred to new wells, and 500 mL of 0.1 M citrate phosphate buffer (pH 5.0) containing 1 mg/mL o-phenylenediamine and 0.03% hydrogen peroxide was added to each well and left at 37  C for 20 min. Enzyme activity was stopped by adding an equal volume of 500 mL of 1 M sulfuric acid to each well. The light absorbance of the supernatant was measured at 492 nm by a microplate reader (Infinite M200, Tecan, Switzerland). The tissue culture polystyrene (TCPS) was used as a negative control experiment for comparison. The ELISA measurement was repeated using six independent disks (n ¼ 6) for each hydrogel substrates and the average results were reported. 2.8. Synthesis and characterization of nanogels PolySBAA nanogels were prepared using an inverse-microemulsion polymerization method. The continuous-phase solution contained 40 mL hexane, 1.4 g Tween 80 and 1.6 g Span 80 as co-emulsifiers, and 8 mg V-70 as an initiator, while the aqueous monomer stock solution contained 2 mL DI water, 876 mg (3 mmol) SBAA, and 18.4 mg (0.12 mmol) MBAA as a crosslinker. The two stock solutions of continuous-phase solution and aqueous stock solution were then mixed in a 100 mL flask with vigorous stirring for 2 min, followed by strong sonication to form the microemulsion. The reaction solution was purged with dry nitrogen at 0  C for 30 min to remove dissolved oxygen. During polymerization, the reaction was kept at 40  C with stirring and under the protection of nitrogen overnight. After reaction, the product solution was mixed with 30 mL THF and stirred for 5 h to remove the surfactants. Then, the mixture was centrifuged at 8000 rpm for 20 min, in which the supernatant was discarded using a 100 kD molecular-weight-cutoff Amicon Ultra centrifugal filter, while the precipitate was washed twice with 30 mL THF and twice with 30 mL DI water. The final precipitate was stored at 4  C for further characterization. The hydrodynamic diameters of the nanogels were analyzed at room temperature using a Zetasizer Nano ZS dynamic light scattering (DLS) instrument (Malvern, U.K.). The wavelength of 633 nm and the scattering angle of 173 were fixed. The dispersant refractive index was set to be 1.330 and the viscosity of water was set to be 0.8872 cP, respectively. The same experimental conditions and protocol were used to synthesize polySBAA nanogels loaded with R6G by adding additional 5 mg of R6G to the aqueous stock solution, and the R6G loading capacity in the polySBAA nanogels is 0.07%. 2.9. Release of encapsulated drug from nanogels The stock solution was prepared by re-suspending 20 mg purified polySBAA nanogels loaded with R6G into 20 mL buffer at a given pH value and left at 37  C. HCl was added into PBS buffer (pH ¼ 7.4) to adjust the pH to 3.6. At different time points, 0.5 mL sample was taken out from the stock solution, placed into a 3 kD molecular-weight-cutoff Amicon Ultra centrifugal filter device (Millipore, Burlington, MA, US), and centrifuged at 14,000 rpm for 30 min to collect filtrate. The fluorescence density of R6G in the filtrate at 555 nm was measured by a fluorescence spectrophotometer (LS 55 Fluorescence Spectrometer, PerkinElmer, MA, US) with an excitation wavelength of 526 nm at room temperature. The percentage of the released R6G from nanogels was defined as the ratio of the fluorescence density of filtrate at different time points to the fluorescence density of stock solution.

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

1 0

po

EA

25

2

H

20

3

ly

10 15 Time (min)

4

po

5

5

ly

0

6

A

0

7

EA

10

Plasma Serum

H

20

8

ly

30

9

po

40

b

M A

Plsama Serum

SB

50

ly

Wavelength shift (nm)

a

All polymer brushes of polySBAA, polySBMA, polyHEAA, and polyHEA were synthesized by SI-ATRP on gold deposited SPR sensor chips via immobilized thiol-functionalized initiators. Four polymer brushes were polymerized from four different monomers, respectively, including (1) SBAA containing sulfobetaine and amide groups, (2) HEAA containing hydroxyl and amide groups, (3) SBMA containing sulfobetaine group, and (4) HEA containing hydroxyl group (Scheme 1). Different chemical structures of these monomers enable us to classify them into different categories (e.g. acrylamide vs. acrylate and polymers with zwitterionic sulfobetaine group vs. polymers with hydrophilic hydroxyl group) for a more complete illustration of the structureeantifouling activity relationship. The polymerization conditions, catalyst, and solvent were carefully controlled to achieve similar film thickness of w20 nm for all polymer brushes, as confirmed by ellipsometer. The rms roughness of these polymer brushes, as determined by AFM, was ranged from 1.0 to 2.2 nm, which minimizes the dependence of film roughness on antifouling properties of the brushes. Protein resistance properties of the polymer brushes were evaluated by SPR. Our and other previous studies have clearly indicated that single protein solution or even diluted human blood serum and plasma are not sufficient to test antifouling properties of the polymer surfaces [4,20,43,49]. In this work, polymer brushes were challenged with undiluted human blood serum and plasma, which contain w500 highly concentrated proteins. Fig. 1a shows typical SPR spectra of protein adsorption from undiluted human serum and plasma onto the polySBAA brush. The average nonspecific adsorption from both 100% serum and 100% plasma was far below a detection limit of SPR sensor (<0.3 ng/cm2), as indicated by almost no wavelength shift between the baselines before and after human blood serum and plasma injection. Fig. 1b compared the adsorbed amount of proteins on four different brushes, which were converted from the SPR wavelength shift. The polySBAA, polySBMA, polyHEAA, and polyHEA brushes had adsorptions of 0, 3.4, 0,

A

Force field parameters for monomeric SBAA, SBMA, HEAA, and HEA molecules were developed using the ParamChem tool (https://www.paramchem.org/), which is compatible with the general CHARMM force field. After geometry optimization at MP2/6-31G* level by Gaussian95, the partial charges were derived by fitting to the gas-phase electrostatic potential using the restrained electrostatic potential (RESP) method. To validate the force field parameters of monomeric SBAA, SBMA, HEAA, and HEA, short 2-ns simulations in the explicit TIP3P water were conducted at 298 K. The bond lengths, bond angles, and torsion angles were well maintained in the MD simulations as compared to quantum mechanism structures. MD simulations of a 16-mer of polySBAA, polySBMA, polyHEAA, or polyHEA in a cubic TIP3P water box of 60  60  60 Å3 were performed using the NAMD program [47] with CHARMM22 force fields [48]. Each simulation system was first subject to 5000 steps of steepest decent minimization with position constraints on polymers, followed by additional 5000 steps of conjugate gradient minimization without any position constraint. After energy minimization, the system was then gradually heated from 50 K to 298 K in 200 ps and equilibrated at 298 K for 1 ns to adjust the system size and density under NPT ensemble with periodic boundary conditions in all three dimensions. Then, 20-ns production MD run was conducted to examine the mutual dynamics and interactions between polymers and water molecules. Shortrange VDW interactions were calculated by a switch function with a twin cutoff at 10 and 12 Å, while long-range electrostatic interactions were calculated by the particle-mesh Ewald method with a grid size of 1 Å and a real-space cutoff of 14 Å. The RATTLE algorithm was applied to constrain all covalent bonds involving

3.1. Antifouling performance of polymer brushes

po

2.11. MD simulations of polymers in solution

3. Results and discussion

A

The human neuroblastoma (SH-SY5Y) cells were cultured in a 75 cm2 T-flask (Corning) with sterile-filtered Eagle’s Minimum Essential Medium and Ham’s F-12 medium mixed at a 1:1 ratio containing 10% fetal bovine serum, and 1% penicillin/ streptomycin. Flasks were incubated in a humidified 37  C incubator with 5% CO2. Cells were cultured to allow 80e90% confluence, harvested using 0.25 mg/mL Trypsin/EDTA solution (Lonza), resuspended in Opti-MEM reduced serum medium and counted using a hemocytometer. Cells were then plated in a 24-well tissue culture plate with approximately 100,000 cells per well in 500 mL of medium, which allows cells to attach to the surface of tissue culture plate for 24 h inside the incubator. The cell viability of SH-SY5Y cells was tested by a typical live/dead cytotoxicity assay. A fixed amount of polySBAA or polyTM nanogels was added to each cultured cell well and pipetted up and down 5 times. The cells were then left for 24 h in a humidified 37  C incubator with 5% CO2, and then assessed for cell toxicity. Calcein AM was added to each well to achieve a final concentration of 2 mM for distinguishing the presence of live cells with a fluorescence excitation/emission of 494/517 nm, while ethidium homodimer-1 was added to each well to a final concentration of 5 mM for distinguishing the presence of dead cells with a fluorescence excitation/emission of 528/617 nm. The cells were incubated with dyes for 15 min in the live/dead assay to activate the fluorescent dyes. A Zeiss Axiovert 40 CFL inverted microscope fitted with filters at 510 nm and 600 nm was used to obtain fluorescence images of the live and dead cells. Fluorescence readings at 494/517 nm and 528/617 nm were detected using a Synergry H1 microplate reader (BioTek). Cell viability percentages were calculated using background subtraction.

hydrogen atoms, so that a time step of 2 fs was used in velocity Verlet integration. Each polymer system was independently run three times for validation with the different starting coordinates and velocities. MD trajectories were saved by every 2 ps for analysis.

SB

2.10. Cell culture and cell viability

Protein adsorption (ng/cm2)

4718

Fig. 1. (a) Typical SPR sensorgram of undiluted blood serum and plasma on polySBAA brush. (b) Nonspecific protein adsorption from undiluted human blood plasma and serum on polySBAA, polySBMA, polyHEAA, and polyHEA brushes at thickness of 20 nm.

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

and 4.9 ng/cm2 from plasma, as well as adsorptions of 0, 7.3, 0, and 5.0 ng/cm2 from serum. Both plasma and serum revealed a comparable trend. Although all of four brushes were highly resistant to protein adsorption of <8 ng/cm2 that is below the value of 9.2 ng/cm2 for PEG brushes [50], the dependence of protein adsorption on different brushes with and without amide groups was clearly observed. PolySBAA and polyHEAA with amide groups achieved the best antifouling performance with almost zero adsorbed proteins, while polySBMA and polyHEA brushes resulting from removal of amide groups from the corresponding polySBAA and polyHEAA led to the increase of nonspecific protein adsorption of 3.4e7.9 ng/cm2. This finding suggests that the incorporation of amide group, serving as both hydrogen-bonding donor and acceptors, into the backbone of polymers with an appropriate space distance from other pendant hydration groups (such as hydroxyl or sulfobetaine groups in our tested polymers) can greatly enhance the hydration ability of polymer brushes e probably through formation of a stronger hydration layer, which helps to further improve surface resistance to nonspecific protein adsorption. 3.2. Correlation between hydration ability and antifouling performance of polymer hydrogels To confirm whether the improved antifouling performance of polySBAA and polyHEAA is attributed to the enhanced hydration ability as hypothesized above, we prepared four different polymer hydrogels of SBAA, SBMA, HEAA, and HEA via photo-polymerization using MBAA as a crosslinker, followed by the evaluation of their hydration and protein-resistant abilities. Use of polymer hydrogels in contrast to polymer brushes not only provides distinct polymer architecture to re-assess the antifouling ability of coated surfaces but also allows to easily assess hydration properties (i.e.

a

85

95 90 EWC (%)

EWC (%)

equilibrium water content and the state of water in the hydrogels). In addition, hydrogels with cross-linked polymer networks are more realistic platform for many antifouling applications ranging from tissue engineering to drug delivery. It is of interest to explore the water content of hydrogels because it is an important factor for antifouling properties. Fig. 2a shows the equilibrium water content (EWC) of the hydrogels as a function of crosslinker content (%), in which the EWC of the hydrogels was measured by weighting mass difference between fully hydrated and dry hydrogel samples. It can be seen clearly that an increase in crosslinker content resulted in a decrease of EWC for all hydrogels studied. Specifically, as the crosslinker content increased from 0.17 to 2.5%, the EWC decreased from 90.7 to 77.6% in polySBAA hydrogels and from 89.6 to 72.3% in polySBMA hydrogels. Consistently, an increase in crosslinker content from 0.17 to 7.5% led to a decrease of EWCs of polyHEAA hydrogels from 95.5 to 76.8% and of polyHEA hydrogels from 89.6 to 65.5%. We also evidently observed that polySBAA and polyHEAA hydrogels with the amide group always had higher EWCs than polySBMA and polyHEA hydrogels without the amide group overall ranges of crosslinker contents. This fact suggests that the presence of the amide groups in polymer chains is indeed able to accommodate more water molecules around polymer chains via the formation of hydrogen bonding. It is worth noting that hydration properties of the hydrogel are related to both quantity (EWC) and quality (bound water) of water molecules in the hydrogel [51]. On the basis of congelation characteristics, the water in polymer matrices is often classified into three types: (i) free water that freezes around 0  C as normal bulk water, (ii) freezable bound water that freezes at a temperature lower than 0  C, and (iii) nonfreezable bound water that never freeze in the present temperature range even at 100  C. In this work, the contents of freezable and nonfreezable water of interest

SBAA SBMA

90

80

4719

75

HEAA HEA

85 80 75 70 65

70 0.17

Crosslinker content (%) 30

25 Nw

0.17 0.83 2.5 5 7.5 Crosslinker content (%)

2.5 20 SBAA SBMA

HEAA HEA

15 Nw

b

0.83

20

15

10

5

10

0 0.17 0.83 2.5 Crosslinker content (%)

0.17 0.83 2.5 5 7.5 Crosslinker content (%)

Fig. 2. (a) Equilibrium water content (EWC) of hydrogels and (b) number (Nw) of nonfreezable water molecules per repeating polymer unit as a function of crosslinker content.

4720

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

on polySBMA, and 9.08% on polyHEA, and these results were consistent with protein-resistance results from polymer brushes. Although a variety of polymer hydrogels were shown to significantly reduce Fg adsorption compared to TCPS, comparing the EWC and Nw in Fig. 2 with the Fg adsorption in Fig. 3 further reveals a clear positive correlation between polymer hydration and antifouling properties, i.e. polyacrylamide hydrogels that possess a high degree of hydration properties exhibit less protein adsorption (better antifouling property) than polyacrylate hydrogels. 3.3. pH-responsive polySBAA nanogels Among four polymers as tested above, polySBAA with different architectures of hydrogels and polymer brushes achieves the best ultralow fouling to protein adsorption from both single protein solution and complex media. The unique structural characteristic that conjugates hydrophilic amide groups with zwitterionic sulfobetaine groups in the polyacrylamide backbones is attributed to its strong hydration e arising from both ionic solvation and hydrogen bonding between the zwitterionic/hydrophilic groups and water molecules. In addition to ultralow fouling property, sulfobetaine (SB) group also exhibits an acid/base equilibrium between a protonation (positively charged) state at lower pH and a deprotonated (neutral charged) state at higher pH (Scheme 3). The deprotonated SBAA form is ultralow fouling due to its zwitterionic nature, while the protonated SBAA form is active to release positively charged drugs due to the increased electrostatic repulsion. Driven by either acidic or basic condition, this unique reversibly switchable property makes the SBAA-based materials as a pH-responsive material for drug delivery. To support these claims above, we synthesized a series of crosslinked polySBAA nanogels loaded with positively charged R6G drugs using an inverse-microemulsion free-radical polymerization method. The controlled release of drugs from polySBAA nanogels was investigated in response to pH changes between a neutral pH of 7.4 in normal cells and acidic pH of 3.6 in cancer cells. Fig. 4 shows the cumulative release profiles of R6G from the polySBAA nanogels in PBS buffer at a physiological temperature of 37  C and different pH values (pH 3.6 and 7.4), respectively. The release rate of R6G from the polySBAA nanogels was relatively slow at pH 7.4, whereas it was accelerated under acidic condition of pH 3.6 (p < 0.05), clearly showing pH-dependent drug release

100 50 pH=3.6 pH=7.4

15 40 R6G release (%)

Fg adsorption relative to TCPS (%)

inside the hydrogels were quantified by DSC. The number (Nw) of nonfreezable water molecules normalized by repeating monomeric units enables direct comparison of hydration quality for different polymer hydrogels. Larger value of Nw indicates the stronger tightly bound water molecules around polymer chains [15,31,45,46,52]. As shown in Fig. 2b, over a range of crosslinker contents from 0.17 to 2.5%, the average Nw values monotonically decreased from 26.7 to 16.3 for polySBAA hydrogels and from 25.5 to 13.6 for polySBMA hydrogels, respectively. Similar trend of Nw was also observed for polyHEAA and polyHEA hydrogels. Over a range of crosslinker contents from 0.17 to 7.5%, Nw decreased from 16.0 to 7.5 for polyHEAA hydrogels and from 7.0 to 4.1 for polyHEA hydrogels, respectively. Comparison of hydration properties between polyacrylamide hydrogels (polySBAA and polyHEAA) and the corresponding polyacrylate hydrogels (polySBMA and polyHEA) confirms an improved hydration ability of polyacrylamides due to the presence of amide groups as compared to polyacrylates. It is also interesting to observe that the differences in EWC and Nw between polyHEAA and polyHEA hydrogels are rather larger than those differences between polySBAA and polySBMA. This fact suggests that the integration of amide groups with hydroxyl groups in the same polymer chain is expected to facilitate the formation of cooperative and bridging hydrogen bonds between two adjacent hydrophilic groups, and thus to promote hydration properties to relatively large extent. Upon characterization of hydration properties of the hydrogels, efforts were taken to evaluate the protein-resistant ability of these cross-linked hydrogels and to examine a structureehydratione antifouling relationship among the hydrogels. ELISA was carried out to assess protein resistance of the hydrogels. Fg was used here as the model protein to evaluate antifouling properties of the hydrogels, because Fg can easily adsorb to a wide range of material surfaces, and it is a coagulation protein involved in platelet aggregation and blot clot formation. TCPS was used as a control because it promotes protein adsorption and cell attachment. The massnormalized absorbance of the TCPS control at 490 nm was set to 100%. The absorbance of all other hydrogel samples was normalized to their mass to eliminate the impact of mass difference among samples. Fig. 3 showed that Fg adsorption relative to TCPS displayed an increased order of 5.41% on polySBAA, 6.88% on polyHEAA, 6.96%

10

5

PS

20 10 0

TC

po ly H EA

A

po ly H EA A

M po ly SB

po ly SB

A A

0

30

Fig. 3. Fibrinogen (Fg) adsorption on polySBAA, polySBMA, polyHEAA, and polyHEA hydrogels normalized by tissue culture polystyrene (TCPS). The results are averaged from three replicates.

-2

0

2

4

6 8 10 Time (day)

12

14

16

Fig. 4. In vitro pH-responsive release of Rhodamine 6G (R6G) from polySBAA nanogels, reported as the mean  SD (n ¼ 3).

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

180 160 140 120

0.1 mg/ml

80 60 40 20 0 polySBAA

polyTM

Fig. 6. Cytotoxicity of polySBAA and polyTM nanogels against SH-SY5Y cells. Data are averaged from three replicates.

Synergry H1 microplate reader (Fig. 6) and the morphological damage of cells was qualitatively imaged using fluorescence microscopy (Fig. 7). In all the studied concentrations, polySBAA nanogels did not significantly induce cell death and maintained cell viability of 99.7%, compared with blank control cells under the same conditions (Fig. 6). Conversely, positively charged polyTM nanogels were highly toxic to cells and maintained w41.4% cell viability. Consistently, fluorescence microscopy showed no signs of morphological damage to the cells upon incubation with polySBAA nanogels (Fig. 7b), thereby demonstrating their minimal cytotoxicity, as compared to massive cell death induced by polyTM nanogels (Fig. 7c). The increase in cytotoxicity in the presence of positively charged polyTM nanogels could be attributed to the polymer aggregation on cell membranes, which may induce cell membrane disruption and interfere with critical intracellular process. Combining the established nonfouling property with pH-responsive characteristics, the polySBAA nanogels exhibit ultralow antifouling ability and excellent stability in FBS, and pH-responsive drug release, which hold a great potential for tumortargeted drug delivery. Despite of these encouraging in vitro results and a compelling biochemical rationale, future in vivo study of the

b

200

0.025 mg/ml 0.05 mg/ml

Hydrodynamic size (nm)

Hydrodynamic size (nm)

a

0 mg/ml

100

Cell Viability (%)

characteristics. When the R6G-loaded polySBAA nanogels were placed in a pH 7.4 buffer, the loaded R6G was slowly released by 3.3% at 2 days, 5.5% at 4 days, and 8.9% at 9 days, and then remained almost unchanged up to 14 days. In contrast, when the pH was decreased to 3.6, it took only 1 day to release the same amount of R6G (8.9%) from the same polySBAA nanogels at pH 7.4. Eventually, about 40% of the loaded R6G was released within 14 days. The significant retardation in drug release at different pH values indicates that the R6G drugs trapped in the networks of polySBAA nanogels can be released in a controlled manner. The interpretation for the pH-controllable drug release could be attributed to the pHresponsive protonation equilibrium of the sulfo groups of polySBAA, which affect both the intermolecular interaction between polymer chains and the drug R6G and the intramolecular interaction between polymer chains. First, the decrease in pH (e.g. <4.0) induces the protonation of the sulfo groups, leading to the change of sulfobetaine groups from neutral to positively charged (quaternary amine groups). The resulting protonated gel network further expels the positively charged R6G drugs from the nanogels to realize a fast drug release due to electrostatic repulsion between the positive polymer chains and the positive chemical groups of drugs. Second, pH reduction also induces the swelling of polySBAA nanogels with the leaky and porous structure due to the increased electrostatic repulsion between the positive polymer chains. As shown in Fig. 5a, the polySBAA nanogels increased their sizes by w17 nm as pH changed from 7.4 to 3.6. The acidebase equilibrium property makes poly(sulfobetaine) a good candidate for pH-stimuli drug release. Besides drug release, the long-term stability of polySBAA nanogels is another important parameter for drug efficacy. Fig. 5b shows the time-evolved hydrodynamic size of polySBAA nanogels in undiluted FBS measured by DLS. In FBS, the polySBAA nanogels were able to retain their initial hydrodynamic diameters at 140 nm up to 7 days. The barely changed size for polySBAA nanogels is mainly attributed to the effective prevention of nonspecific protein adsorption and nanoparticle aggregation, demonstrating its longterm stability and robust in FBS. This is very favorable for application in drug delivery. It has been reported that most of synthetic polymers were usually cytotoxic [53]. Therefore, we assess the cytotoxicity of polySBAA nanogels to the SH-SY5Y cells using a live/dead assay. For comparison, polyTM nanogels were prepared and used as a control. After 24-h incubation of SH-SY5Y cells with different nanogels, the quantity of live and dead cells was quantitatively measured using a

4721

180 160 140 120 100 80

100 0

1

2

3

4 pH

5

6

7

8

-1

0

1

2

3

4

5

6

7

8

Time (day)

Fig. 5. (a) Hydrodynamic size of polySBAA nanogel as a function of pH in PBS buffer and (b) stability of polySBAA nanogels in fetal bovine serum (FBS), reported as the mean  SD (n ¼ 3).

4722

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

Fig. 7. Representative fluorescence microscopy images of SH-SY5Y cells after 24 h incubation with (a) cell only, (b) polySBAA nanogels, and (c) polyTM nanogels at 0.05 mg/mL in medium. Live cells are stained as green while dead cells are stained as red. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

polySBAA nanogels is required to demonstrate the par pharmacologically active for future biological applications. 3.4. Atomic insights into polymer hydration To better understand the chemical structureeantifouling activity relationship of four polymers as experimentally tested above, we performed a series of long 20-ns MD simulations to examine the hydration of these polymers and then to correlate the hydration properties from MD simulations with their antifouling capacities from in vitro experiments. The radial distribution function (RDF) can be used to quantify the extent of the hydration of the solutes by quantitatively characterizing spatial distribution between water molecules and the solute of interest. Fig. 8a shows the RDFs of the oxygen atoms (Ow) of water molecules around all hydrogen-bonding atoms of HEAA and HEA 16-mers. It can be seen that both HEAA-Ow and HEA-Ow RDF profiles were almost identical to each other, both of which displayed the first hydration peaks at the position of w2.3 Å and the second hydration peaks at w5.0 Å. The almost identical peak positions for these two HEAA and HEA molecules suggest that water molecules form similar hydration shells around these two molecules. However, the subtle differences in peak heights by w2.8% also indicate that water densities within the first two hydration shells of HEAA are slightly higher than those of HEA, reflecting that the introduction of amide groups in HEAA can indeed induce different anisotropic wateresolution interactions, which further rearrange the spatial distribution of water molecules around the solutes differently. We also quantitatively account the coordination number of water molecules (N) in the first hydration shell around three different hydrogen-bonding donor/acceptor atoms of O13, N15, and O17 in HEAA or of O13, O16, and O15 in HEA. Comparing N numbers between HEAA and HEA molecules as shown in Scheme 4a, we found that overall N (3.04) of per HEAA molecule was slightly larger than that of HEA (2.86), consistent with subtle difference in the first peak heights of Ow-HEAA and Ow-HEA RDFs. Since both HEAA and HEA molecules have two identical hydrogen-bonding donor groups (i.e. O13 and O17 in HEAA vs. O13 and O15 in HEA), the relatively larger N of HEAA apparently comes from the amide group of eNH, which has a better hydration property than eO16 group in HEA. Along the similar line, Fig. 8b shows the comparison of watersolute RDFs for SBAA and SBMA molecules. Similar to the observation of RDFs for HEAA and HEA, both Ow-SBAA and Ow-SBMA RDFs had nearly identical shapes and peaks, indicating the similar solvation shell locations. However, the presence of the amide group of eNH in SBAA molecule slightly increased local water densities of solvation shells. Consistently, overall N values of SBAA and SBMA obtained from all hydrogen-bonding atoms were 5.66 and 5.31

(Scheme 4b), respectively, indicating that the SBAA had more water molecules in solvation shells than the SBMA, which is mainly contributed from the amide groups, despite different hydration capacities around different selected solute atoms. In additional, unlike three hydrogen-bonding acceptors and donors in HEAA and HEA molecules, SBAA and SBMA had six hydrogen-bonding acceptors and donors. Therefore, N numbers of SBAA (N ¼ 5.66) and SBMA (N ¼ 5.31) were considerably larger than those of HEAA (N ¼ 3.04) and HEA (N ¼ 2.86). The relatively larger N values in SBAA and SBMA interpret their better antifouling ability to resist nonspecific protein adsorption.

Fig. 8. Radial distribution function (RDF) between oxygen atoms (Ow) of water molecules and (a) HEAA/HEA and (b) SBAA/SBMA.

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

4723

used to reflect the bind strength of water molecules to the solutes of interest, i.e. the longer ss of water molecules, the stronger binding affinity to the specified solutes. It can be seen in Fig. 9 that C(t) curves for HEAA and SBAA molecules with the amide groups decreased slower than the corresponding C(t) curves for HEA and SBMA, reflecting the dependence of water dynamics on the hydration shell and partial charge of the solute. We derived ss for four molecules by fitting the C(t) curves. The residence times were 13.5 ps for HEAA, 13.0 ps for HEA, 21.3 ps for SBAA, and 20.1 ps for SBMA, respectively. It is not surprised to observe that water molecules tend to stay longer around zwitterionic groups (SBAA and SBMA) than hydrophilic groups (HEA and HEAA). This finding also suggests that ionic solvation arising from a large dipolar headgroup in sulfobetaine groups is stronger than hydrogen-bonding solvation. 4. Conclusions

Scheme 4. Number of hydrogen bonds formed between all hydrogen bond donors/ acceptors of four monomers and the surrounding water molecules.

From a structural (static) point of view, the presence of amide groups in HEAA and SBAA molecules indeed slightly enhances the spatial distribution of water molecules near the solutes. Meanwhile, it is also important to characterize the dynamics of water molecules in the hydration shells, which enables to identify whether hydration water molecules are strongly or weakly bound to the solutes. We thus calculated the residence time (ss) of water in the first hydration shell of the solute by fitting a correlation funcP water PRj ð0ÞPRj ðtÞ=PRj ð0Þ2 using a single expotion CR ðtÞ ¼ 1=Nw N j¼1 nential function: CR ðtÞ ¼ Aexpðt=ss Þ; where PRj is a binary function that takes the value of 1 if the jth water molecule stays in the layer of thickness R for a time t without leaving the specified region during this interval and is equal to zero otherwise. ss can be

In this work, we synthesized and characterized different architecture forms of polymer brushes, polymer hydrogels, and polymer nanogels from the polyacrylamide polymers of polySBAA and polyHEAA and the corresponding polyacrylate polymers of polySBMA and polyHEA. Protein resistance to these surfaces was studied by SPR and ELISA. It was found that although all of polymer brushes were highly resistant to protein adsorption from single protein solution and undiluted human blood serum/plasma, the polyacrylamide brushes (<0.3 ng/cm2) exhibited far-better protein resistance over the polyacrylate brushes (3.5e7.5 ng/cm2). Further studies of polymer hydrogels reveal a positive correlation between the hydration properties of the polymers and their antifouling ability, i.e. polySBAA and polyHEAA hydrogels that contain more EWC and nonfreezable water than polySBMA and polyHEA hydrogels had better antifouling ability. MD simulations confirm that water molecules interact more strongly with polyacrylamide chains than with polyacrylate chains, as evidenced by higher spatial distribution and longer residence time of water molecules at polymer interface. Our results indicate that antifouling property of polymers can be readily enhanced by the conjugation of the amide group with other hydration groups such as hydroxyl or sulfobetaine groups in the same polymer backbone. In addition, since the SBAA molecular possesses a unique combination property of ultralow fouling and pH-simulative response that can be reversibly switched between acidic and basic condition, as a proof-of-concept, we synthesized the polySBAA nanogels with the loaded R6G drugs using an inverse-microemulsion free-radical polymerization method. The drug-loaded polySBAA nanogels exhibited an acidtriggered sustained-drug-release property, the ultralow fouling ability, high stability in undiluted FBS, and minimal cytotoxicity to SH-SY5Y cells, which make SBAA-based materials as a promising candidate in the application of drug and gene delivery. Overall, this work demonstrates that incorporation of the amide group with other hydrated pendant groups allows to obtain effective antifouling materials. Acknowledgment We thank for financial support from NSF CAREER Award CBET0952624 and NSF CBET-1158447. References

Fig. 9. Residence time (ss) of water molecules within the first hydration shell of 16-mers of SBAA, HEAA, SBMA, and HEA.

[1] Loiselle M, Anderson KW. The use of cellulase in inhibiting biofilm formation from organisms commonly found on medical implants. Biofouling 2003;19: 77e85.

4724

C. Zhao et al. / Biomaterials 34 (2013) 4714e4724

[2] Banerjee I, Pangule RC, Kane RS. Antifouling coatings: recent developments in the design of surfaces that prevent fouling by proteins, bacteria, and marine organisms. Adv Mater 2011;23:690e718. [3] Castro IB, Westphal E, Fillmann G. Third generation antifouling paints: new biocides in the aquatic environment. Quim Nova 2011;34:1021e31. [4] Zhao C, Chen Q, Patel K, Li L, Li X, Wang Q, et al. Synthesis and characterization of pH-sensitive poly(N-2-hydroxyethyl acrylamide)-acrylic acid (poly(HEAA/ AA)) nanogels with antifouling protection for controlled release. Soft Matter 2012;8:3848e57. [5] Ariga K, Ji Q, McShane MJ, Lvov YM, Vinu A, Hill JP. Inorganic nanoarchitectonics for biological applications. Chem Mater 2011;24:728e37. [6] Shen MC, Wagner MS, Castner DG, Ratner BD, Horbett TA. Multivariate surface analysis of plasma-deposited tetraglyme for reduction of protein adsorption and monocyte adhesion. Langmuir 2003;19:1692e9. [7] Kirschner CM, Brennan AB. Bio-inspired antifouling strategies. In: Clarke DR, editor. Annu rev mater res 2012;vol. 42. p. 211e29. [8] Cao S, Wang J, Chen H, Chen D. Progress of marine biofouling and antifouling technologies. Chin Sci Bull 2011;56:598e612. [9] Dafforn KA, Lewis JA, Johnston EL. Antifouling strategies: history and regulation, ecological impacts and mitigation. Mar Pollut Bull 2011;62:453e65. [10] Fusetani N. Antifouling marine natural products. Nat Prod Rep 2011;28:400e10. [11] Zoulalian V, Zuercher S, Tosatti S, Textor M, Monge S, Robin J-J. Self-assembly of poly(ethylene glycol)-poly(alkyl phosphonate) terpolymers on titanium oxide surfaces: synthesis, interface characterization, investigation of nonfouling properties, and long-term stability. Langmuir 2010;26:74e82. [12] Ranke J. Persistence of antifouling agents in the marine biosphere. Environ Sci Technol 2002;36:1539e45. [13] Yang J-C, Zhao C, Hsieh IF, Subramanian S, Liu L, Cheng G, et al. Strong resistance of poly (ethylene glycol) based L-tyrosine polyurethanes to protein adsorption and cell adhesion. Polym Int 2012;61:616e21. [14] Chen S, Li L, Zhao C, Zheng J. Surface hydration: principles and applications toward low-fouling/nonfouling biomaterials. Polymer 2010;51:5283e93. [15] Wu J, Chen S. Investigation of the hydration of nonfouling material poly(ethylene glycol) by low-field nuclear magnetic resonance. Langmuir 2012;28:2137e44. [16] Tugulu S, Klok H-A. Stability and nonfouling properties of poly(poly(ethylene glycol) methacrylate) brushes under cell culture conditions. Biomacromolecules 2008;9:906e12. [17] Dalsin JL, Hu B-H, Lee BP, Messersmith PB. Mussel adhesive protein mimetic polymers for the preparation of nonfouling surfaces. J Am Chem Soc 2003;125:4253e8. [18] Meyers SR, Grinstaff MW. Biocompatible and bioactive surface modifications for prolonged in vivo efficacy. Chem Rev 2011;112:1615e32. [19] Zhao C, Li L-Y, Guo M-M, Zheng J. Functional polymer thin films designed for antifouling materials and biosensors. Chem Pap 2012;66:323e39. [20] Zhao C, Li L, Wang Q, Yu Q, Zheng J. Effect of film thickness on the antifouling performance of poly(hydroxy-functional methacrylates) grafted surfaces. Langmuir 2011;27:4906e13. [21] Zhao C, Li L, Zheng J. Achieving highly effective nonfouling performance for surface-grafted poly(HPMA) via atom-transfer radical polymerization. Langmuir 2010;26:17375e82. [22] Holland NB, Qiu YX, Ruegsegger M, Marchant RE. Biomimetic engineering of non-adhesive glycocalyx-like surfaces using oligosaccharide surfactant polymers. Nature 1998;392:799e801. [23] Luk YY, Kato M, Mrksich M. Self-assembled monolayers of alkanethiolates presenting mannitol groups are inert to protein adsorption and cell attachment. Langmuir 2000;16:9604e8. [24] Wyszogrodzka M, Haag R. Synthesis and characterization of glycerol dendrons, self-assembled monolayers on gold: a detailed study of their protein resistance. Biomacromolecules 2009;10:1043e54. [25] Yang W, Chen S, Cheng G, Vaisocherová H, Xue H, Li W, et al. Film thickness dependence of protein adsorption from blood serum and plasma onto poly(sulfobetaine)-grafted surfaces. Langmuir 2008;24:9211e4. [26] Chen S-H, Chang Y, Lee K-R, Wei T-C, Higuchi A, Ho F-M, et al. Hemocompatible control of sulfobetaine-grafted polypropylene fibrous membranes in human whole blood via plasma-induced surface zwitterionization. Langmuir 2012;28:17733e42. [27] Shih Y-J, Chang Y. Tunable blood compatibility of polysulfobetaine from controllable molecular-weight dependence of zwitterionic nonfouling nature in aqueous solution. Langmuir 2010;26:17286e94. [28] Yang W, Xue H, Carr LR, Wang J, Jiang S. Zwitterionic poly(carboxybetaine) hydrogels for glucose biosensors in complex media. Biosens Bioelectron 2011;26:2454e9.

[29] Cheng G, Xue H, Li G, Jiang S. Integrated antimicrobial and nonfouling hydrogels to inhibit the growth of planktonic bacterial cells and keep the surface clean. Langmuir 2010;26:10425e8. [30] Ladd J, Zhang Z, Chen S, Hower JC, Jiang S. Zwitterionic polymers exhibiting high resistance to nonspecific protein adsorption from human serum and plasma. Biomacromolecules 2008;9:1357e61. [31] Ishihara K, Ishikawa E, Iwasaki Y, Nakabayashi N. Inhibition of fibroblast cell adhesion on substrate by coating with 2-methacryloyloxyethyl phosphorylcholine polymers. J Biomater Sci Polym Ed 1999;10:1047e61. [32] Hasegawa T, Iwasaki Y, Ishihara K. Preparation of blood-compatible hollow fibers from a polymer alloy composed of polysulfone and 2-methacryloyloxyethyl phosphorylcholine polymer. J Biomed Mater Res 2002;63:333e41. [33] Chen X, Lawrence J, Parelkar S, Emrick T. Novel zwitterionic copolymers with dihydrolipoic acid: synthesis and preparation of nonfouling nanorods. Macromolecules 2012;46:119e27. [34] Kratz K, Breitenkamp K, Hule R, Pochan D, Emrick T. PC-polyolefins: synthesis and assembly behavior in water. Macromolecules 2009;42:3227e9. [35] Zheng J, Li L, Tsao H-K, Sheng Y-J, Chen S, Jiang S. Strong repulsive forces between protein and oligo (ethylene glycol) self-assembled monolayers: a molecular simulation study. Biophys J 2005;89:158e66. [36] Zheng J, Li LY, Chen SF, Jiang SY. Molecular simulation study of water interactions with oligo (ethylene glycol)-terminated alkanethiol self-assembled monolayers. Langmuir 2004;20:8931e8. [37] Jeon SI, Lee JH, Andrade JD, De Gennes PG. Proteinesurface interactions in the presence of polyethylene oxide: I. Simplified theory. J Colloid Interface Sci 1991;142:149e58. [38] Szleifer I. Protein adsorption on surfaces with grafted polymers: a theoretical approach. Biophys J 1997;72:595e612. [39] Wang RLC, Kreuzer HJ, Grunze M. Molecular conformation and solvation of oligo(ethylene glycol)-terminated self-assembled monolayers and their resistance to protein adsorption. J Phys Chem B 1997;101:9767e73. [40] Gunkel G, Weinhart M, Becherer T, Haag R, Huck WTS. Effect of polymer brush architecture on antibiofouling properties. Biomacromolecules 2011;12: 4169e72. [41] Reisch A, Voegel J-C, Gonthier E, Decher G, Senger B, Schaaf P, et al. Polyelectrolyte multilayers capped with polyelectrolytes bearing phosphorylcholine and triethylene glycol groups: parameters influencing antifouling properties. Langmuir 2009;25:3610e7. [42] Perrino C, Lee S, Choi SW, Maruyama A, Spencer ND. A biomimetic alternative to poly(ethylene glycol) as an antifouling coating: resistance to nonspecific protein adsorption of poly(l-lysine)-graft-dextran. Langmuir 2008;24:8850e6. [43] Zhao C, Zheng J. Synthesis and characterization of poly(N-hydroxyethylacrylamide) for long-term antifouling ability. Biomacromolecules 2011;12:4071e9. [44] Jones DM, Brown AA, Huck WTS. Surface-initiated polymerizations in aqueous media: effect of initiator density. Langmuir 2002;18:1265e9. [45] Morisaku T, Watanabe J, Konno T, Takai M, Ishihara K. Hydration of phosphorylcholine groups attached to highly swollen polymer hydrogels studied by thermal analysis. Polymer 2008;49:4652e7. [46] Ping ZH, Nguyen QT, Chen SM, Zhou JQ, Ding YD. States of water in different hydrophilic polymers e DSC and FTIR studies. Polymer 2001;42:8461e7. [47] Kale L, Skeel R, Bhandarkar M, Brunner R, Gursoy A, Krawetz N, et al. NAMD2: greater scalability for parallel molecular dynamics. J Comput Phys 1999;151: 283e312. [48] MacKerell AD, Bashford D, Bellott M, Dunbrack RL, Evanseck JD, Field MJ, et al. All-atom empirical potential for molecular modeling and dynamics studies of proteins. J Phys Chem B 1998;102:3586e616. [49] Jiang S, Cao Z. Ultralow-fouling, functionalizable, and hydrolyzable zwitterionic materials and their derivatives for biological applications. Adv Mater 2010;22:920e32. [50] Zhang Z, Zhang M, Chen S, Horbetta TA, Ratner BD, Jiang S. Blood compatibility of surfaces with superlow protein adsorption. Biomaterials 2008;29:4285e91. [51] Green RJ, Corneillie S, Davies J, Davies MC, Roberts CJ, Schacht E, et al. Investigation of the hydration kinetics of novel poly(ethylene oxide) containing polyurethanes. Langmuir 2000;16:2744e50. [52] Wu J, Lin W, Wang Z, Chen S, Chang Y. Investigation of the hydration of nonfouling material poly(sulfobetaine methacrylate) by low-field nuclear magnetic resonance. Langmuir 2012;28:7436e41. [53] Fischer D, Li YX, Ahlemeyer B, Krieglstein J, Kissel T. In vitro cytotoxicity testing of polycations: influence of polymer structure on cell viability and hemolysis. Biomaterials 2003;24:1121e31.