Surface Characterization and Protein Interactions of Segmented

ARTICLE pubs.acs.org/Langmuir

Surface Characterization and Protein Interactions of Segmented Polyisobutylene-Based Thermoplastic Polyurethanes David Cozzens, Arnold Luk, Umaprasana Ojha, Marina Ruths, and Rudolf Faust* Department of Chemistry, University of Massachusetts Lowell, One University Avenue, Lowell, Massachusetts 01854, United States

bS Supporting Information ABSTRACT: The surface properties and biocompatibility of a class of thermoplastic polyurethanes (TPUs) with applications in blood-contacting medical devices have been studied. Thin films of commercial TPUs and novel polyisobutylene (PIB)
poly(tetramethylene oxide) (PTMO) TPUs were characterized by contact angle measurements, X-ray photoelectron spectroscopy, and atomic force microscopy (AFM) imaging. PIB-PTMO TPU surfaces have significantly higher C/N ratios and lower amounts of oxygen than the theoretical bulk composition, which is attributed to surface enrichment of PIB. Greater differences in the C/N ratios were observed with the softer compositions due to their higher relative amounts of PIB. The contact angles were higher on PIB-PTMO TPUs than on commercial polyether TPUs, indicating lower surface energy. AFM imaging showed phase separation and increasing domain sizes with increasing hard segment content. The biocompatibility was investigated by quantifying the adsorption of fouling and passivating proteins, fibrinogen (Fg) and human serum albumin (HSA) respectively, onto thin TPU films spin coated onto the electrode of a quartz crystal microbalance with dissipation monitoring (QCM-D). Competitive adsorption experiments were performed with a mixture of Fg and albumin in physiological ratio followed by binding of GPIIb-IIIa, the platelet receptor ligand that selectively binds to Fg. The QCM-D results indicate similar adsorbed amounts of both Fg and HSA on PIB-PTMO TPUs and commercial TPUs. The strength of the protein interactions with the various TPU surfaces measured with AFM (colloidal probe) was similar among the various TPUs. These results suggest excellent biocompatibility of these novel PIB-PTMO TPUs, similar to that of polyether TPUs.

’ INTRODUCTION The properties of biomaterials surfaces are of utmost importance due to their role in thrombosis and inflammatory response. In long-term blood-contacting biomaterials in particular, thrombogenicity is a key concern.1,2 Thrombosis occurs through a complex biological pathway known as the coagulation cascade. The first step involves protein adsorption from blood serum onto the foreign substance, followed by platelet adhesion, and subsequent thrombus formation.3
6 The coagulation cascade is a healthy response of damaged tissues; however, thrombosis and occlusion in a synthetic biomaterial is a leading cause of failure, especially in application of synthetic polymers as vascular grafts. Protein adsorption, being the first step in thrombus formation, has been a crucial topic of investigation in order to understand how and why it occurs and, consequently, how to engineer thromboresistant surfaces. It is well-known that fibrinogen (Fg), a key blood serum protein, is responsible for platelet adhesion and the fibrin clot formation.4,5,7 For this reason, in vitro Fg adsorption is a commonly used predictor of in vivo biocompatibility. Another blood serum protein, human serum albumin (HSA), is known to have passivating properties when present on a surface.8 Fg is a rod-shaped protein with three domains, the α, β, and γ domains. The α domains are believed to be responsible for electrostatic binding effects through the amino groups of their r 2011 American Chemical Society

arginine and lysine residues.9 On the other hand, it has been widely reported that nonspecific protein adsorption (e.g., Fg, HSA, fibronectin) is preferred on hydrophobic over hydrophilic surfaces (e.g., octadecyl vs amino-terminated self-assembled monolayers) due to hydrophobic forces.6,10
15 Many studies have attempted to elucidate the chemical and physical processes that determine protein adsorption onto surfaces such as SiO2, TiO2, polyesters, polyurethanes, polystyrene (PS), polymethacrylates, polyethylene glycol (PEG), polycarbonates, maleic acid copolymers, polydimethylsiloxane (PDMS), and others.16
24 An array of techniques have been applied such as ellipsometry, quartz crystal microbalance with dissipation monitoring (QCM-D), surface plasmon resonance (SPR), and atomic force microscopy (AFM).16
24 As mentioned previously, it has been observed that hydrophobic surfaces are conducive to nonspecific protein adsorption; however, surfaces with adsorbed HSA appear to have reduced subsequent protein adsorption in addition to the hydrophilic systems above.16,17 For example, a model hydrophobic surface, plasma deposited hexamethyldisiloxane, showed Received: July 7, 2011 Revised: September 20, 2011 Published: October 24, 2011 14160

dx.doi.org/10.1021/la202586j | Langmuir 2011, 27, 14160–14168

Langmuir

ARTICLE

Table 1. Polymer Nomenclature and Composition soft segment

hard segment

SS:HS

shore

(SS)

(HS)

(wt:wt)

hardness31

polymer E2Aa

PDMS/PHMO

MDI-BDO

60:40

85A

P55Db

PTMO

MDI-BDO

45:55

55D

P80Ac

PTMO

MDI-BDO

56:44

80A

PTMO80A

PTMO

MDI-BDO

65:35

80A

PTMO60A

PTMO

MDI-BDO

80:20

60A

PIB60Ad

PIB/PTMO

MDI-BDO

79:21

60A

PIB80Ad PIB100Ad

PIB/PTMO PIB/PTMO

MDI-BDO MDI-BDO

65:35 60:40

80A 100A

SIBS

PIB

PS

70:30

60A

a

Elast-Eon 2A. b Pellethane 2363-55D. c Pellethane 2363-80A. d Novel PIB-PTMO TPUs.

∼40% reduction of adsorbed masses of Fg and IgG from mixtures with albumin.17 Block or graft copolymers have also been studied regarding protein adsorption. One well-known case is PEG-containing copolymers. PEG, which may be grafted or plasma polymerized onto various substrates, is readily solvated in aqueous media and forms brushlike layers.18 This makes it thermodynamically unfavorable for proteins to bind to the PEG surface. PEG-containing copolymers have consistently shown minimal protein adsorption (1.67 ng/cm2) in comparison with other polymeric surfaces (>200 ng/cm2 on various polyurethanes);6,19,20 however, in vivo cell studies have not always shown a good correlation between reduced protein adsorption and cell adhesion.21,22 Polyurethanes with hydrophilic segments showed minimal Fg adsorption (PEG-containing polyurethane showed 50 response units (RU) measured with SPR) compared to those with hydrophobic segments (PDMS-containing polyurethane showed 1390 RU).6,10 Furthermore, for block copolymers that show microphase separation, it has been observed that smaller domain sizes minimize protein adsorption on PDMS-containing block copolymers, especially when the domain sizes were ∼6
12 nm.23,24 Interaction forces between protein molecules and a surface have been studied with AFM-based techniques to measure the strengths of protein adhesion and obtain an understanding of the underlying mechanisms.25 In aqueous solution, both Fg and albumin show stronger adhesion forces in contact with hydrophobic surfaces than with hydrophilic ones (e.g., 2
3 times stronger on PS vs glass), which has been ascribed to hydrophobic forces.26
28 Increasing the time of interaction, as well as the loading force between protein and surface, increases the adhesion force by an exponential factor with time and up to two times with a doubling of the loading force.15,27,29 QCM-D has been employed in many adsorption studies due to its ability to precisely determine the various aspects of adsorption in real-time, including the kinetics of adsorption, the hydrated thickness or mass of adsorbed material, and viscoelastic properties of the adsorbed layers, in this case the proteins.30 Previously in our group a new series of thermoplastic polyurethanes (TPUs) were designed containing biostable polyisobutylene (PIB) in the soft segment (SS), intended to impart long-term oxidative stability.31 The physical-mechanical properties as well as the oxidative stability of these TPUs demonstrated superior in vitro biostability compared to commercial polyether TPUs.32 Oxidative degradation was greatly retarded in the PIB-PTMO TPUs as evidenced by reduced weight loss, greater retention of

ultimate tensile strength, and less marred surface morphology. In the present study, we characterize the surfaces of these novel TPUs using X-ray photoelectron spectroscopy (XPS), contact angle measurements, and AFM imaging, as well as study the interaction of proteins on these and commercial TPUs using QCM-D and the colloidal probe AFM technique. The results suggest good biocompatibility equal to current commercial biomedical TPUs.

’ MATERIALS AND METHODS Materials. All chemicals and proteins were used without further purification with the exception of the toluene and 1,4-butanediol used for TPU synthesis. Toluene (99.5+%), N,N-dimethylformamide (DMF, 99.9+%), 1-methyl-2-pyrrolidone (NMP, 99+%), pyridine (99+%), poly(tetramethylene oxide) (PTMO), 4,40 -methylenebis(phenylisocyanate) (MDI, 98%), sodium dodecyl sulfate (SDS), phosphate buffered saline (PBS), manganese(II) chloride (MnCl2, 97%), sodium chloride (NaCl, 98+%), Trizma hydrochloride (99+%), magnesium chloride (MgCl2, 98+%), human serum albumin (HSA, 97+%), and bovine serum albumin (BSA, 98+%) were purchased from SigmaAldrich. Calcium chloride (CaCl2, 96%) was purchased from Acros Organics/Fisher Scientific. Fibrinogen (Fg, 95+%) was purchased from Calbiochem and Glycoprotein IIb-IIIa (GPIIb-IIIa) from Enzyme Research Laboratories. Pellethane 2363-55D (P55D) and Pellethane 2363-80A (P80A) were obtained from Dow Chemical Co. Elast-Eon 2A (E2A) is a product of Aortech International plc. PBS buffer solution (pH 7.6) was prepared by dissolution of 1 salt packet (8 g of NaCl, 0.2 g of monobasic potassium phosphate, 1.15 g of dibasic sodium phosphate, and 0.2 g of potassium chloride) in 1 L of deionized water. Tris buffer (pH 7.4) was prepared by dissolution of 50 mM Trizma hydrochloride, 100 mM NaCl, 1 mM MgCl2, 1 mM CaCl2, and 1 mM MnCl2 in deionized water. Synthesis. The synthesis of one of the novel PIB-PTMO TPUs is described as a representative example as follows. Toluene was purified by refluxing over sodium and benzophenone and distilled. 1,4-Butanediol (BDO) was purified by refluxing over CaH2 and distilling under slight vacuum. Hydroxypropyl-telechelic PIB (Mn = 2000 g/mol, 4.0 g, 4.0 mmol) was synthesized as reported previously.33 PIB and PTMO (Mn = 1000 g/mol, 1.0 g, 1.0 mmol) were dissolved in dry toluene in a round-bottom flask equipped with a mechanical stirrer. The prepolymers were then azeotropically dried under vacuum at 50 °C. Dry toluene (27 mL) was added to dissolve the prepolymers, the temperature was raised to 100 °C, and the reaction system kept under N2. Then MDI (2.42 g, 9.47 mmol) was added to the reaction mixture, and the reaction was continued for 2 h at 100 °C. Then tin octoate (13.24 mg, 0.029 mmol) was added to the reaction mixture, followed by BDO (0.54 g, 6.0 mmol). The reaction was continued for 2 h at 100 °C under N2, after which the reaction was stopped and the mixture was cooled to 50 °C, poured into a Teflon mold, and left to dry at room temperature for 48 h. The resulting solid was cut into small pieces and vacuum-dried at 70 °C for 24 h before characterization. This synthesis has been slightly modified since initial publication where complete details on the synthesis and characterization can be found.31 These TPUs as well as the commercial TPUs under investigation are listed in Table 1. Spin-Coating. Thin films of the polymers listed in Table 1 were prepared by spin coating using a Spincoater P6700 (Specialty Coating Systems Inc.) instrument. The Pellethane samples and the synthesized PTMO TPUs were dissolved in DMF at a concentration of 0.5 wt %, E2A and SIBS were dissolved in toluene at 0.5 wt %, PIB60A was dissolved in pyridine at 0.5 wt %, and PIB80A and PIB100A were dissolved in NMP at 1 wt %. Films approximately 30
60 nm thick were spincoated from solution onto the top electrode of AT-cut quartz crystals as purchased from Q-Sense (QSX 301, Biolin Inc.). Preliminary evaluation 14161

dx.doi.org/10.1021/la202586j |Langmuir 2011, 27, 14160–14168

Langmuir of film uniformity was performed by visual inspection, followed by AFM imaging. Spin-coated samples were used for contact angle measurements, AFM imaging, colloidal probe force measurements, and QCM-D. Drop-cast samples were used for XPS. XPS. Elemental analysis of the PIB-PTMO TPU surfaces was performed using XPS. Samples were prepared by drop-casting the TPUs from solution onto silicon wafers. The solvent was slowly evaporated in air, followed by annealing at 125 °C for 12 h under nitrogen gas. Dropcasting was expected to result in the least frozen-in stress. However, XPS was also performed on extruded samples and the same trend and significant findings in C/N ratio were present. The samples were introduced into a VG ESCALAB MKII photoelectron spectrometer with a base pressure around 4  10
10 Torr. Mg Kα X-rays were used, and photoelectrons were detected at takeoff angles of 90°, 60°, and 30°. Contact Angle Measurements. Water contact angle measurements on spin-coated polymer films were performed on a Kr€uss DSA 100 drop shape analyzer using the sessile drop method. Both advancing and receding contact angle measurements were measured under nitrogen flow. Reported values are averages of six to nine drops of approximately 10 μL each. AFM Imaging. Spin-coated TPU surfaces were imaged using a Veeco Nanoscope IIIa Multimode atomic force microscope in tapping mode at ambient conditions and room temperature using an Applied Nanostructures Inc. ACCESS probe with a nominal spring constant of 50 N/m, and fundamental resonance frequency of 353.2 kHz. Height and phase images were recorded. Root mean square roughness values were calculated from three 1 μm  1 μm areas on each sample. QCM-D Protein Adsorption Measurements. Single Protein Adsorption. Quantitative measurements of protein adsorption were performed on an E4 QCM-D (Biolin Inc.) instrument. The adsorption of single proteins onto the various spin-coated polymer surfaces listed in Table 1 was studied at 37 °C at protein concentrations of 100 μg/mL. The solid lyophilized blood serum proteins, HSA and Fg, were dissolved in phosphate buffered saline (PBS) made with deionized (DI) water (Millipore). First, PBS was flowed through the QCM-D system to obtain a baseline signal, followed by protein solution until equilibrium adsorption was observed, indicated by a plateau in the signal. Then PBS was flowed again to rinse off any loosely coupled proteins. A Voigt model was used to calculate the adsorbed mass using QTools software (Biolin Inc.). Unlike the Sauerbrey equation which directly correlates the change in frequency with change in mass, the viscoelastic Voigt model considers the dissipative contribution as well, treating the adsorbed protein as a coupled oscillator.30 The data is fitted using a least-squares approach, and calculations are made iteratively until the fit of the data minimizes the chi square distribution.34 Fixed parameters and upper and lower limits of fitted parameters were set as has been validated previously for similar systems.35 The protein solution density was fixed at 1016 kg/m3 for Fg and 1011 kg/m3 for HSA as was measured from known masses and volumes. The viscosity of the adsorbed, extended protein layer in solution was typically around 2.4
3.6 cP, layer shear moduli were typically 1.7
2.3 MPa, and layer mass was allowed to vary between 0.01 to 100 μg/cm2 (data shown in Results section). Additionally, it was found that small changes to the protein solution viscosity (0.1
1 cP) could result in substantially lower chi square values. Therefore, this parameter was made a fitting variable instead of a fixed parameter. The resulting protein solution viscosity was in the range 1.0
1.07 cP, and the layer density values typically ranged from 1100 to 1400 kg/m3. Competitive Protein Adsorption. A QCM-D procedure used by Weber et al.20 incorporated the use of the soluble glycoprotein IIb-IIIa (GPIIb-IIIa), the platelet receptor which adheres to Fg and continues the coagulation cascade. Since GPIIb-IIIa selectively binds to Fg, the amount of Fg adsorbed from a competitive mixture can be quantified. We conducted experiments in a similar manner on select TPU surfaces. After adsorption of a mixture of both Fg and BSA, GPIIb-IIIa was

ARTICLE

Table 2. XPS Data: Surface Composition of Novel PIB-PTMO TPUs PIB-PTMO TPU PIB60A

takeoff angle (deg)

C (%)

O (%)

N (%)

C/N

90

90.0

8.2

1.8

50.0

60 30

91.8 93.3

6.7 5.4

1.5 1.3

61.2 71.8

90.3

7.3

2.4

37.6

extrap. 0 bulk theoretical PIB80A

90

87.3

9.9

2.8

31.2

60

88.6

9.0

2.4

36.9

30

90.3

7.8

1.9

47.5

3.0

54.8 29.5

extrap. 0 bulk theoretical PIB100A

82.8

88.6

8.4

90

86.4

10.3

3.3

26.2

60

87.5

9.5

2.9

30.2

30

89.2

8.4

2.4

extrap. 0 bulk theoretical

37.2 42.2

87.0

9.5

3.5

24.9

adsorbed onto the Fg. In the competitive adsorption experiments, the protein solution contained 50 μg/mL Fg and 833 μg/mL BSA in PBS. The ratio 50/833 corresponds to the approximate ratio of Fg to albumin in blood serum.36 The coated QCM crystals were equilibrated in PBS until stable baselines were reached. The Fg/BSA protein solution was then flowed through the instrument chambers for 1 h. After adsorption, the chambers were rinsed with Tris buffer for 30 min and then 10% w/v BSA in Tris was flowed for 30 min to fill remaining adsorption sites on the surface. Another Tris rinse was then done for 30 min, followed by a solution of 30 μg/mL GPIIb-IIIa in Tris for 2 h. After GPIIb-IIIa adsorption, another Tris rinse was performed for 1 h. Colloidal Probe Force Measurements. Measurements of Fg adhesion forces were performed with a colloidal probe, a technique based on AFM.37 Probes that had uniform, smooth gold beads (see Supporting Information section for details on preparation and characterization of colloidal probes) were soaked in a 100 μg/mL Fg solution for 10 min to allow Fg to adsorb onto the gold. Subsequently, the probes were gently rinsed by dipping in PBS and mounted in the fluid cell attachment for the AFM. Force measurements were performed with the Fg-coated colloidal probe and the spin-coated polymer substrates immersed in PBS. As the loading force and residence time are known to give variable results for protein adhesion,15,27,29 the loading force was controlled using the software trigger function, with maximum load around 8 nN, and the approach and separation rate was kept at 0.5 Hz. The probe was cleaned for reuse by immersion in 1% SDS solution in water, followed by rinsing in DI water, followed by drying and UV/ozone irradiation for about 2 min. Several probes with radii in the range 2.4
8.8 μm and spring constants 0.30
0.35 N/m were prepared and used to obtain the data shown in the Results section. Statistical Analysis. All results below were measured with at least three replicates, with the exception of the XPS analysis, which was performed on one sample of each polymer. Data were compared using a two-tailed Student’s t test. A p value of 65°, yet there are slight differences in adsorbed amounts (Figure 2) and strengths (Figure 5), depending on the SS when comparing E2A to the others, and also when comparing the PIB-PTMO TPUs and the commercial polyether TPUs, indicating that hydrophobic forces are not solely responsible for protein interactions with nonpolar polymer surfaces. When Fg is in direct contact with the surface, the hydrophobic nature of the TPUs dictates the strength of adhesion which is 14166

dx.doi.org/10.1021/la202586j |Langmuir 2011, 27, 14160–14168

Langmuir similar among the polyether- and PIB-PTMO TPUs. E2A exhibits reduced adhesion strength and a lower amount of adsorbed Fg in the QCM-D studies (∼1.3 μg/cm2 compared to 1.7
2.0 μg/cm2), although the adsorbed amount is not as low as that reported on PEG-containing surfaces (0.092 μg/cm2).15 It has been shown that PDMS-containing TPUs adsorb more than 20 times more Fg than PEG-containing TPUs, and thus the low surface energy of the PDMS segment does not confer protein repellency.10 A recent study suggests that decreased adsorption may be due to (polar) HS enrichment upon hydration.44 Xu et al. reported a 2/3 to 3/4 decrease in adhered platelets after 1 day of hydration of PDMS and PTMO based TPUs, respectively. If this HS enrichment mechanism is correct, then our results suggest that although hydration was not performed over similar time scales (about 1 h for QCM-D experiments), the HS enrichment is highest and fastest for E2A since the PIB-PTMO TPUs do not show reduced protein adsorption as does E2A. This indicates that the PDMS SS of E2A is fundamentally different from that of PIB in its interactions with proteins. XPS (Table 2) showed that the low surface energy PIB is enriched at the surface in vacuum/air, as has also been observed in polymers with PDMS SS.23 Therefore, chain flexibility alone, present in both E2A and PIB, is not sufficient to give HS enrichment. This different mechanism of reduced protein adsorption for E2A then explains why Fg on SIBS has a similar low adsorption amount, but high adhesion force. The HS enrichment theory will not be applicable since the PS HS is completely nonpolar. It can be understood that Fg has a high affinity for the SIBS surface as we see directly from the force measurements, but this high affinity presents itself as adsorbing to the SIBS surface in a more “flat” manner which reduces the total adsorbed amount. Regardless of surface, albumin shows an initial steric repulsion from the surface, whereas Fg shows monotonic attraction upon approach. Considering the results in Table 4, it is seen that Fg adsorbs similarly onto the PIB-PTMO TPU and the polyether TPU surfaces in a competitive environment. In this competitive environment, the HS enrichment theory will not be adequate to explain the differences between E2A and the other TPUs. The conclusion that can be drawn is that BSA binds more strongly to E2A than Fg does, resulting in a reduced amount of adsorbed Fg and subsequent high GP/Fg ratio (5 molecules of GP per Fg).

’ CONCLUSIONS AFM and QCM-D were used to investigate protein interactions with novel PIB-PTMO TPUs and commercial TPUs. Single protein adsorption experiments have been shown to be an inadequate predictor of biocompatibility in comparison with competitive adsorption experiments. Competitive adsorption experiments showed that the PDMS-based TPU, E2A, binds less Fg when present in a mixture with albumin, compared to polyether or PIB-PTMO TPUs. Commercial polyether and novel PIB-PTMO TPUs showed similar behavior in a competitive protein environment. Force measurements showed similar interactions between Fg and polyether and PIB-PTMO TPUs and lower adhesion strength between Fg and E2A. The slightly reduced adhesion strength and adsorbed amounts of Fg on E2A are attributed to the distinctive properties conferred by the PDMS co-soft-segment. The protein interactions with PIB-PTMO TPUs are thus similar to those with commercial polyether TPUs, suggesting good biocompatibility and great promise for use in long-term blood-contacting biomaterials.

ARTICLE

’ ASSOCIATED CONTENT

bS

Supporting Information. Calculation of colloidal probe contact area, preparation and analysis of colloidal probes, force measurements with bare gold probe, control experiment demonstrating selective GPIIb-IIIa binding to Fg, and correlation of GPIIb-IIIa/Fg to %Fg adsorbed. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Fax: (978) 934-3013. E-mail: [email protected].

’ ACKNOWLEDGMENT The authors thank Dr. Jagdeep Singh for assistance in acquiring XPS data and AFM images. Financial support by the Massachusetts Life Sciences Center and Boston Scientific Corp. (through funding the Boston Scientific Biomaterials Laboratory) is gratefully acknowledged. M.R. acknowledges support through NSF CAREER Award #NSF-CMMI 0645065. ’ REFERENCES (1) Schmedlen, R. H.; Elbjeirami, W. M.; Gobin, A. S.; West, J. L. Clin. Plast. Surg. 2003, 30, 507–517. (2) Teebken, O. E.; Haverich, A. Eur. J. Endovasc. Surg. 2002, 23, 475–485. (3) Balasubramanian, V.; Grusin, N. K.; Bucher, R. W.; Turitto, V. T.; Slack, S. M. J. Biomed. Mater. Res. 1999, 44, 253–260. (4) Anderson, J. M.; Bonfield, T. L.; Ziats, N. P. Int. J. Artif. Organs 1990, 13, 375–382. (5) Farrell, D. H.; Thiagarajan, P.; Chung, D. W.; Davie, E. W. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 10729–10732. (6) Wu, Y. G.; Simonovsky, F. I.; Ratner, B. D.; Horbett, T. A. J. Biomed. Mater. Res. 2005, 74A, 722–738. (7) Nagai, H.; Handa, M.; Kawai, Y.; Watanabe, K.; Ikeda, Y. Thromb. Res. 1993, 71, 467–477. (8) Kottke-Marchant, K.; Anderson, J. M.; Umemura, Y.; Marchant, R. E. Biomaterials 1989, 10, 147–155. (9) Jung, S. Y.; Lim, S. M.; Albertorio, F.; Kim, G.; Gurau, M. C.; Yang, R. D.; Holden, M. A.; Cremer, P. S. J. Am. Chem. Soc. 2003, 125, 12782–12786. (10) Ma, C. F.; Hou, Y.; Liu, S.; Zhang, G. Z. Langmuir 2009, 25, 9467–9472. (11) Sagvolden, G.; Giaever, I.; Feder, J. Langmuir 1998, 14, 5984–5987. (12) Renner, L.; Pompe, T.; Salchert, K.; Werner, C. Langmuir 2005, 21, 4571–4577. (13) Osaki, T.; Renner, L.; Herklotz, M.; Werner, C. J. Phys. Chem. B 2006, 110, 12119–12124. (14) Ishizaki, T.; Saito, N.; Sato, Y.; Takai, O. Surf. Sci. 2007, 601, 3861–3865. (15) Bremmell, K. E.; Kingshott, B.; Ademovic, Z.; Winther-Jensen, B.; Griesser, H. J. Langmuir 2006, 22, 313–318. (16) Welle, A.; Kroger, M.; Doring, M.; Niederer, K.; Pindel, E.; Chronakis, I. S. Biomaterials 2007, 28, 2211–2219. (17) Lassen, B.; Malmsten, M. J. Colloid Interface Sci. 1996, 180, 339–349. (18) Morra, M. J. Biomater. Sci., Polym. Ed. 2000, 11, 547–569. (19) Leckband, D.; Sheth, S.; Halperin, A. J. Biomater. Sci., Polym. Ed. 1999, 10, 1125–1147. (20) Weber, N.; Wendel, H. P.; Kohn, J. J. Biomed. Mater. Res. 2005, 72A, 420–427. (21) Bretagnol, F.; Lejeune, M.; Papadopoulou-Bouraoui, A.; Hasiwa, M.; Rauscher, H.; Ceccone, G.; Colpo, P.; Rossi, F. Acta Biomater. 2006, 2, 165–172. (22) Jordan, S. W.; Chaikof, E. L. J. Vasc. Surg. 2007, 45, 104A–115A. 14167

dx.doi.org/10.1021/la202586j |Langmuir 2011, 27, 14160–14168

Langmuir

ARTICLE

(23) Childs, M. A.; Matlock, D. D.; Dorgan, J. R. Biomacromolecules 2001, 2, 526–537. (24) Fang, H. X.; Feng, L. B.; Wu, L. M. Chin. J. Polym. Sci. 2009, 27, 327–334. (25) Averett, L. E.; Schoenfisch, M. H. Analyst 2010, 135, 1201– 1209. (26) Sagvolden, G.; Giaever, I.; Feder, J. Langmuir 1998, 14, 5984– 5987. (27) Xu, L. C.; Siedlecki, C. A. Langmuir 2009, 25, 3675–3681. (28) Xu, L. C.; Siedlecki, C. A. J. Biomed. Mater. Res. 2010, 92A, 126–136. (29) Xu, L. C.; Vadillo-Rodriguez, V.; Logan, B. E. Langmuir 2005, 21, 7491–7500. (30) Voinova, M. V.; Rodahl, M.; Jonson, M.; Kasemo, B. Phys. Scr. 1999, 59, 391–396. (31) Ojha, U.; Kulkarni, P.; Faust, R. Polymer 2009, 50, 3448–3457. (32) Cozzens, D.; Ojha, U.; Kulkarni, P.; Faust, R.; Desai, S. J. Biomed. Mater Res. 2010, 95A, 774–782. (33) Bela, I.; Kennedy, J. P. Polym. Mater. Sci. Eng. 1988, 58, 869–872. (34) Nelder, J. A.; Mead, R. Comp. J. 1965, 7, 308–313. (35) Johnson, P. A.; Luk, A.; Demtchouk, A.; Patel, H.; Sung, H. J.; Treiser, M. D.; Gordonov, S.; Sheihet, L.; Bolikal, D.; Kohn, J.; Moghe, P. V. J. Biomed. Mater. Res. 2010, 93A, 505–514. (36) Horbett, T. A. Protein Adsorption on Biomaterials. In Biomaterials: Interfacial Phenomena and Applications; Cooper, S.L., et al. , Eds.; American Chemical Society: Washington, D.C., 1982; pp 233
244. (37) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Langmuir 1992, 8, 1831–1836. (38) Luk, A.; Cozzens, D.; Faust, R. To be submitted. (39) Kim, J. T.; Weber, N.; Shin, G. H.; Huang, Q.; Liu, S. X. J. Food Sci. 2007, 72, E214–E221. (40) Messina, G. M. L.; Satriano, C.; Marletta, G. Colloids Surf., B. 2009, 70, 76–83. (41) Reedy, E. D., Jr. J. Mater. Res. 2006, 21, 2660–2668. (42) Sethuraman, A.; Han, M.; Kane, R. S.; Belfort, G.. Langmuir 2004, 20, 7779–7788. (43) Hayashi, T.; Tanaka, M.; Yamamoto, S.; Shimomura, M.; Hara, M. Biointerphases 2007, 2, 119–125. (44) Xu, L. C.; Runt, J.; Siedlecki, C. A. Acta Biomater. 2010, 6, 1938
1947.

14168

dx.doi.org/10.1021/la202586j |Langmuir 2011, 27, 14160–14168