Hemocompatible Mixed-Charge Copolymer Brushes of

Nov 30, 2009 - ACS Biomaterials Science & Engineering 2015 1 (11), 1183-1193. Abstract | Full Text HTML .... Yu-Ju Shih , Yung Chang , Andre Deratani ...
8 downloads 0 Views 5MB Size
Download PDF
pubs.acs.org/Langmuir © 2009 American Chemical Society

Hemocompatible Mixed-Charge Copolymer Brushes of Pseudozwitterionic Surfaces Resistant to Nonspecific Plasma Protein Fouling Yung Chang,*,† Shih-Hung Shu,‡ Yu-Ju Shih,† Chih-Wei Chu,§ Ruoh-Chyu Ruaan,‡ and Wen-Yih Chen*,‡ † R&D Center for Membrane Technology and Department of Chemical Engineering, Chung Yuan Christian University, Jhong-Li, Taoyuan 320, Taiwan, ‡Department of Chemical and Materials Engineering, National Central University, Jhong-Li, Taoyuan 320, Taiwan, and §Research Center for Applied Sciences, Academia Sinica 128 Sec. 2, Academia Road., Nankang, Taipei 11529, Taiwan

Received August 26, 2009. Revised Manuscript Received October 10, 2009 In this work, the hemocompatibility of a sulfobetaine-like copolymer brush resulting from a mixed-charge copolymerization of the positively charged 11-mercapto-N,N,N-trimethylammonium chloride (TMA) and negatively charged 11-mercaptoundecylsulfonic acid (SA) was studied. Mixed charge distribution in the prepared poly(TMA-co-SA) copolymer brushes was controlled by the regulation of the reaction rate of the surface-initiated atom transfer radical polymerization (ATRP). The adsorption behavior of plasma proteins on a surface grafted with poly(TMA-co-SA) was measured by a surface plasmon resonance (SPR) sensor. The effects of varying temperature, solution pH, and ionic strength on the antifouling characteristics of the mixed-charge copolymer brushes were systematically evaluated, and the protein-fouling resistance was discussed in detail, especially with respect to the effect of ionic strength on the intra- and intermolecular interactions of the poly(TMA-coSA) with proteins. The adhesion and activation of blood cells on the poly(TMA-co-SA)-grafted surface in contact with human whole blood was also demonstrated. The results suggest that mixed-charge copolymer brushes of poly(TMA-co-SA), which, like zwitterionic homopolymer brushes, have overall charge neutrality, can be used in similar applications for protein-fouling resistance and have excellent hemocompatibility with human whole blood at physiologic temperatures.

Introduction The development of nonfouling surfaces is critically important in biomedical applications for blood-contacting materials used in antithrombogenic implants, hemodialysis membranes, drug delivery carriers, and biosensors.1-7 However, only a very limited number of synthetic biomaterials are regarded as good hemocompatible candidates. For example, for those surfaces in contact with blood, even 10 ng/cm2 of adsorbed fibrinogen may induce a fullscale blood platelet adhesion, leading to thrombosis and embolism at the blood-contacting side of material surfaces in the bloodstream.8,9 A good nonspecific plasma protein-fouling resistance is one of the most important requirements for blood-contacting materials. When a plasma protein approaches an interface, electrical neutrality may be important in minimizing the electrostatic interactions, and the absence of hydrogen-bond donors may also be important for minimizing the hydrogen-bonding interactions. For many nonfouling materials or superlow-biofouling materials, the general properties of the functional groups on these surfaces are hydrophilic, electrically neutral, and hydrogen-bond acceptors *To whom correspondence should be addressed. E-mail: ychang@ cycu.edu.tw, [email protected]. (1) Hoffman, A. S. Advances in Chemistry Series; American Chemical Society: Washington, DC, 1982; p 3. (2) Ratner, B. D.; Hoffman, A. D.; Schoen, F. D.; Lemons, J. E. Biomaterials Science, an Introduction to Materials in Medicine, 2nd ed.; Elsevier: Amsterdam, 2004. (3) Chen, S.; Zheng, J.; Li, L.; Jiang, S. J. Am. Chem. Soc. 2005, 127, 14473. (4) Ratner, B. D. Biomaterials 2007, 28, 5144. (5) Chen, S. F.; Jiang, S. Y. Adv. Mater. 2008, 20, 335. (6) Krishnan, S.; Weinman, C. J.; Ober, C. K. J. Mater. Chem. 2008, 18, 3405. (7) Chiang, Y. C.; Chang, Y.; Higuchi, A.; Chen, W. Y.; Ruaan, R. C. J. Membr. Sci. 2009, 339, 151. (8) Shen, M. C.; Wagner, M. S.; Castner, D. G.; Ratner, B. D.; Horbett, T. A. Langmuir 2003, 19, 1692. (9) Kwak, D.; Wu, Y. G.; Horbett, T. A. J. Biomed. Mater., Res. Part A 2005, 74A, 69.

3522 DOI: 10.1021/la903172j

rather than hydrogen-bond donors. Two common types of polymeric segments, nonionic poly(ethylene glycol) (PEG) and zwitterionic betaine polymers, have been extensively employed for surface grafting to create nonfouling surfaces. PEG-based materials are the most commonly used synthetic surfaces for reducing protein adsorption.10,11 However, it has been long recognized that PEG decomposes in the presence of oxygen and transition metal ions, which are found in most biochemically relevant solutions.12-14 Thus, while PEG exhibits an excellent nonfouling capability, it faces problems of blood compatibility and stability.5,15 A zwitterionic structure contains both a positively and a negatively charged moiety within the same functional group while maintaining overall charge neutrality. In recent years, zwitterionic polymers containing the pendant groups phosphobetaine, sulfobetaine, and carboxybetaine have received growing attention for use in the new generation of blood-contacting materials because of their good plasma-protein-fouling resistance.3,5,12,16-22 (10) Harris, J. M. Poly(Ethylene Glycol) Chemistry: Biotechnical and Biomedical Applications, 1st ed.; Springer: 1992. (11) Li, L.; Chen, S.; Zheng, J.; Ratner, B. D.; Jiang, S. J. Phys. Chem. B 2005, 109, 2934. (12) Ostuni, E.; Chapman, R. G.; Holmlin, R. E.; Takayama, S.; Whitesides, G. M. Langmuir 2001, 17, 5605. (13) Luk, Y. Y.; Kato, M.; Mrksich, M. Langmuir 2000, 16, 9604. (14) Shen, M. C.; Martinson, L.; Wagner, M. S.; Castner, D. G.; Ratner, B. D.; Horbett, T. A. J. Biomater. Sci., Polym. Ed. 2002, 13, 367. (15) Li, L. Y.; Chen, S. F.; Jiang, S. Y. J. Biomater. Sci., Polym. Ed. 2007, 18, 1415. (16) Iwasaki, Y.; Ishihara, K. Anal. Bioanaly. Chem. 2005, 381, 534. (17) Chang, Y.; Chen, S. F.; Zhang, Z.; Jiang, S. Y. Langmuir 2006, 22, 2222. (18) Zhang, Z.; Chao, T.; Chen, S. F.; Jiang, S. Y. Langmuir 2006, 22, 10072. (19) Chang, Y.; Chen, S. F.; Yu, Q. M.; Zhang, Z.; Bernards, M.; Jiang, S. Y. Biomacromolecules 2007, 8, 122. (20) Chang, Y.; Liao, S. C.; Higuchi, A.; Ruaan, R. C.; Chu, C. W.; Chen, W. Y. Langmuir 2008, 24, 5453. (21) Zhang, Z.; Cheng, G.; Carr, L. R.; Vaisocherova, H.; Chen, S. F.; Jiang, S. Y. Biomaterials 2008, 29, 4719.

Published on Web 11/30/2009

Langmuir 2010, 26(5), 3522–3530

Chang et al.

Article

Jiang and co-workers showed that surfaces grafted with zwitterionic polymers are ideal for resisting nonspecific protein adsorption when the surface density and chain length of the zwitterionic groups are controlled, yielding excellent properties with respect to antithrombogenic response.17,18 Poly(sulfobetaine methacrylate) (polySBMA), with a methacrylate main chain and an analogue of the taurine betaine as the pendant group (CH2CH2Nþ(CH3)2CH2CH2CH2SO3-), has become the most widely studied zwitterionic polymer owing to its ease of synthetic preparation.17,18 Our previous work reported that zwitterionic polySBMA is an effective and stable nonfouling material, potentially providing a surface appropriate for use in human blood and implants.19,20,22 To further develop a potential nonfouling material for biomedical applications, especially in contact with human blood, we were inspired to study a polyampholyte system with a pseudozwitterionic structure bearing mixed positively and negatively charged moieties. Recently, it has been reported that nonfouling surfaces with homogeneous charge balances, prepared by mixing two oppositely charged compounds, are highly resistant to nonspecific single-protein adsorption.5,23,24 However, the study did not extend to the use or evaluation of these mixed-charge surfaces as human blood-contacting materials. In this work, a copolymer brush derived from the combination of the positively charged 11-mercapto-N,N,N-trimethylammonium chloride (TMA) and negatively charged 11-mercaptoundecylsulfonic acid (SA) was studied as an example of the pseudozwitterionic polyampholytes. For comparison, methyl- and oligo(ethylene glycol)-terminated self-assembled monolayers and a well-packed zwitterionic polymer brush formed by the surface polymerization of sulfobetaine methacrylate (SBMA) were similarly evaluated, especially with respect to blood-contacting properties. This study was aimed at addressing three important issues of mixed-charge surfaces grafted with poly(TMA-co-SA), namely: (i) distribution control of mixed-charge neutrality in the prepared copolymer brushes; (ii) systematic measurement of the protein adsorption characteristics due to the statistical charge balance of the copolymer brushes at different temperatures, solution pH, and ionic strengths; and (iii) in vitro evaluation of the blood compatibility of the charge-balanced poly(TMA-co-SA) surfaces using a human plasma solution and whole blood and comparison with the reference zwitterionic polySBMA surfaces.

Materials and Methods Materials. [2-(Methacryloyloxy)ethyl]dimethyl(3-sulfopropyl)ammonium hydroxide (sulfobetaine methacrylate, SBMA) was purchased from Monomer-Polymer & Dajac Laboratories, Inc., U.S. 11-Mercapto-N,N,N-trimethylammonium chloride (TMA) and 11-mercaptoundecylsulfonic acid (SA) were purchased from Sigma Chemical Co. Copper(I) bromide (99.999%), 2-bromoisobutyryl bromide (BIBB, 98%), pyridine (98%), 2-hydroxyethyl acrylate (97%), 2,20 -bipyridine (BPY, 99%), triethylamine (99%), tetrahydrofuran (THF, HPLC grade), and ethanol (absolute; 200 proof) were purchased from Sigma-Aldrich. 1-Undecanethiol (99þ%), (1-mercapto-11-undecyl)tetra(ethylene glycol) (99þ%), and 11-mercapto-1-undecanol (99þ%) were purchased from Asemblon, Inc. Fibrinogen (fraction I from human plasma), γ-globulin (Fraction II, III, 99%), and human serum albumin (HSA, 96-99%) were purchased from Sigma Chemical Co. Water used in the experiments was purified using a Millipore (22) Chang, Y.; Chen, W. Y.; Yandi, W.; Shih, Y. J.; Chu, W. L.; Liu, Y. L.; Chu, C. W.; Ruaan, R. C.; Higuchi, A. Biomacromolecules 2009, 10, 2092. (23) Holmlin, R. E.; Chen, X. X.; Chapman, R. G.; Takayama, S.; Whitesides, G. M. Langmuir 2001, 17, 2841. (24) Bernards, M. T.; Cheng, G.; Zhang, Z.; Chen, S. F.; Jiang, S. Y. Macromolecules 2008, 41, 4216.

Langmuir 2010, 26(5), 3522–3530

water-purification system with a minimum resistivity of 18.0 MΩ 3 m. THF for reactions and washing was dried with sodium before use. ω-Mercaptoundecyl bromoisobutyrate was synthesized through the reaction of BIBB with 11-mercapto-1undecanol using a method published previously.20,25 1H NMR (300 MHz, CDCl3) characteristics were as follows: 4.15 (t, J) 6.9 (2H, OCH2), 2.51(q, J) 7.5 (2H, SCH2), 1.92 (s, 6H, CH3), 1.57-1.72 (m, 4H, CH2), and 1.24-1.40 (m, 16H, CH2). Preparation of Self-Assembled Monolayers. In this study, three self-assembled monolayers (SAMs) were formed on the substrates: (1) ω-mercaptoundecyl bromoisobutyrate-initiated (Br), (2) methyl-terminated (CH3), and (3) oligo(ethylene glycol)-terminated (OEG) SAMs. BK-7 glass chips (SCHOTT Taiwan Ltd.) and P(100) Si wafers (Wafer Works Corporation Co., Inc.) were first coated with an adhesion-promoting chromium layer (thickness: 2 nm) and a surface-plasmon active gold layer (48 nm) by electron-beam evaporation under vacuum. Before SAM preparation, the gold-coated substrate was cleaned by washing with pure ethanol and deionized water in sequence, dried with N2, then left in a UV light cleaner for 20 min at a source power of 110 W, followed by rinsing with deionized water and ethanol, and finally dried with N2. For the preparation of an initiator SAM on a gold surface, the cleaned chip was soaked in a 2-mM ethanol solution of ω-mercaptoundecyl bromoisobutyrate for 24 h to form Br-SAMs on the gold surface, and the chip was rinsed with pure ethanol followed by THF and then dried in a stream of N2. For the preparation of CH3- or OEG-SAMs, the cleaned chip was soaked in a 2-mM ethanol solution of 1-undecanethiol or (1-mercapto-11-undecyl) tetra(ethylene glycol) thiols for 24 h to form SAMs on the gold surface, and the chip was rinsed in sequence with ethanol and deionized water, and then dried in a stream of N2. Preparation of Homopolymer Brushes. Three well-packed homopolymer brushes of polyTMA, polySA, and polySBMA on SPR glass chips and Si wafers were elaborated via the surfaceinitiated atom transfer radical polymerization (ATRP) method and were prepared by the following previously reported method.20,24,25 These homopolymer brushes were polymerized onto the gold substrates with immobilized Br-SAM initiators as in our previous reports. The reaction solutions of CuBr and BPY were first placed in a sealed glass reactor in a drybox under a nitrogen atmosphere. A 400 mM degassed solution (deionized water and methanol in a 1:3 volume ratio, 48 h) of monomers was transferred to the reactor, and the gold surface with immobilized initiators was then placed in the reactor under nitrogen. After polymerization, the substrate was removed and rinsed with ethanol and deionized water, and the samples were kept in deionized water overnight. The prepared substrates were rinsed with PBS buffer to remove unbound polymers before any experiments. The thickness of the substrates was measured by ellipsometry. The thickness of the polymer film was controlled to within a range of ∼30 to ∼45 nm. For further details on polySBMA film thickness as a function of polymerization time for different SBMA concentrations, we remand the reader to the previously reported work.25 Preparation of Mixed-Charge Copolymer Brushes. Surface-initiated ATRP was used to prepare mixed-charge poly(TMA-co-SA) on the gold substrates with immobilized Br-SAM initiators. The reaction solutions of CuBr and BPY were first placed in a sealed glass reactor in a drybox under nitrogen protection. A degassed solution (deionized water and methanol in a 1:3 volume ratio) of the mixed monomers in a 1:1 TMA:SA molar ratio was transferred to the reactor and the gold surface with immobilized initiators was then placed in the reactor under nitrogen. The poly(TMA-co-SA) layers produced were all of a similar thickness of 30-45 nm, which was controlled by adjusting total monomer concentration within a range of 0.4-1.2 M and (25) Zhang, Z.; Chen, S. F.; Chang, Y.; Jiang, S. Y. J. Phy. Chem. B 2006, 110, 10799.

DOI: 10.1021/la903172j

3523

Article

Chang et al. Table 1. Characteristic Data of Poly(TMA-co-SA) Copolymer Brushes co-monomer ratios (mol %)

sample ID

TMA

SA

reaction conditions a

concentration (M)

reaction time (h)

characterization of copolymer brushes N/S

film thickness (nm)c

contact angle (deg)

44.1 ( 4.5 54.7 ( 4.0 polyTMA 100 0 0.4 48 NA poly(TMA-co-SA) no. 1 50 50 0.4 36 1.08 36.8 ( 2.6 55.0 ( 1.1 poly(TMA-co-SA) no. 2 50 50 0.6 4 0.85 43.1 ( 3.5 49.9 ( 1.1 poly(TMA-co-SA) no. 3 50 50 0.8 3 0.96 36.2 ( 2.2 48.2 ( 1.0 poly(TMA-co-SA) no. 4 50 50 1.2 1 0.99 34.5 ( 2.8 36.7 ( 0.7 34.3 ( 5.7 13.3 ( 1.5 polySA 0 100 0.4 48 NAb a Reaction molar ratio of TMA and SA monomers was kept at 1:1 for the preparation of poly(TMA-co-SA). The ratio of Cu(I)Br/BPY was 1:2. b The ratio does not have a numerical value and was also characterized by XPS. c All polymer brush layers were elaborated with similar film thicknesses, ranging from 30 to 45 nm, as measured by ellipsometry. b

reaction time from 1 to 36 h, as listed in Table 1. After polymerization, the substrates were removed from the reactor, rinsed with ethanol and deionized water, and immersed in phosphatebuffered saline (PBS) solution at 45 °C overnight to remove unreacted monomers and unbound copolymers before analysis. Surface Characterization of Polymer Brushes. The surface compositions of the grafted polymer brushes were also characterized by X-ray photoelectron spectroscopy (XPS). XPS analysis was performed using a PHI Quantera SXM/Auger spectrometer with a monochromated Al KR X-ray source (1486.6 eV photons). The energy of emitted electrons was measured with a hemispherical energy analyzer at pass energies ranging from 50 to 150 eV. All data were collected at the photoelectron takeoff angle of 45° with respect to the sample surface. The binding energy (BE) scale is referenced by setting the peak maximum in the C 1s spectrum to 284.6 eV. The high-resolution C 1s spectrum was fitted using a Shirley background subtraction and a series of Gaussian peaks. Data analysis software was from Service Physics, Inc. Ellipsometry was performed using a spectroscopic ellipsometer (VASE, J.A. Woollam Co., Inc.). Five separate spots of the sample were measured at three different angles of incidence (50°, 60°, and 70°) in the visible region. The same batch of gold-coated substrates was cleaned in a UV-ozone cleaner for 30 min, washed with ethanol and deionized water, and dried with nitrogen. The bare gold-coated chips were used as a reference. The thicknesses of the studied films were determined using the Cauchy layer model with an assumed refractive index of 1.45. Water contact angles were measured with an automatic contact angle meter (model CA-VP, Kyowa Interface Science Co., Ltd. Japan) at 23 °C. The deionized water was dropped on the sample surface at 10 different sites. Plasma Protein Adsorption. In this work, a custom-built surface-plasmon resonance (SPR) biosensor based on wavelength interrogation with a four-channel Teflon flow cell was used to monitor protein adsorption on the coated substrate. An SPR sensor chip was attached to the base of the prism and optical contact was established using refractive-index matching fluid (Cargille). Protein solutions of 1.0 mg/mL human fibrinogen, γ-globulin, or HSA in phosphate buffered saline (PBS, 0.15 M, pH 7.4) were delivered to the surfaces at a flow rate of 0.05 mL/ min. In this study, platelet poor plasma (PPP) solution containing plasma proteins was also tested on the coated substrate. Blood was obtained from a healthy human volunteer. PPP was prepared by centrifugation of the blood at 3000 rpm for 10 min. A diluted solution containing 20% (v/v, in PBS) plasma proteins from PPP was used to reduce the effects of plasma viscosity in the SPR flow channel and to prevent the formation of small bubbles. A surfacesensitive SPR detector was used to monitor protein-surface interactions in real time. The wavelength shift was used to measure the change in surface-adsorbed protein (mass per unit area). The SPR measurement was repeated using three independent chips (n = 3 in total) for each modified substrate, and the result was reported in average. The calibration of the wavelength shift from SPR data associated with the amount of adsorbed (26) Jung, L. S.; Campbell, C. T.; Chinowsky, T. M.; Mar, M. N.; Yee, S. S. Langmuir 1998, 14, 5636.

3524 DOI: 10.1021/la903172j

protein was calculated on the basis of the quantitative mathematical formalism established by Campbell and co-workers.26 The calibration followed the standard calculation for the same custom-built SPR system, resulting in a 1-nm wavelength shift in the SPR response being equivalent to ∼15 ng/cm2 of adsorbed protein.3,26 Human Blood Cell Adhesion. The gold-coated Si wafers with polymer brushes (polyTMA, polySA, and polySBMA), copolymer brushes (poly(TMA-co-SA)), and SAMs (CH3- and OEGSAMs), each with a 1-cm2 surface area, were placed in individual wells of a 24-well tissue culture plate, and each well was equilibrated with 1000 μL of PBS for 2 h at 23 °C. Blood was obtained from a healthy human volunteer. A 200 μL portion of the whole blood, first recalcified by the addition of calcium (1 M CaCl2, 5 μL), was placed on the substrate surface in each well of the tissue culture plate and incubated for 120 min at 37 °C. After the chips were rinsed twice with 1000 μL of PBS, they were immersed in 2.5% glutaraldehyde in PBS for 48 h at 4 °C to fix the activated platelets and other adhering blood cells, then rinsed twice with 1000 μL of PBS and gradient-dried with ethanol in 75% v/v PBS, 50% v/v PBS, 25% v/v PBS, 5% v/v PBS, and 0% v/v PBS, successively, for 20 min each and dried in air. Finally, the samples were sputter-coated with gold prior to observation under a JEOL JSM-5410 SEM operating at 7 keV. The morphology of blood cells adhering to the substrates was observed in scanning electron microscopy (SEM) images at a 1000 magnification at five different places on each chip.

Results and Discussion In this present work, four well-packed grafted surfaces with homopolymer brushes (polyTMA, polySA, and polySBMA) and mixed-charge copolymer brushes (poly(TMA-co-SA)) were prepared via surface-initiated ATRP, and two monolayer surfaces with CH3- and OEG-SAMs were achieved by self-assembly, as shown in Figure 1. To achieve similar grafting qualities with respect to chain length and surface density, all polymer brushes were elaborated with similar film thicknesses in the range of 3045 nm as measured by ellipsometry. The chemical composition of the poly(TMA-co-SA) copolymer brushes was determined by XPS and is shown in Table 1. SPR was used to estimate the nonfouling properties, such as resistance to plasma protein adsorption, using CH3- and OEG-SAMs as references for comparison. When new protein-resistant surfaces are judged, it is generally accepted that CH3-SAMs presenting hydrophobic methyl groups will induce large amounts of protein adsorption, while a SAM surfaceterminated with OEG ((EG)nOH, n = 4) groups is used as a standard for low protein adsorption. The zwitterionic surface grafted with polySBMA has also been described as an ideal low protein-adsorption material in previous reports.17,20,22,25 Effects of Mixed Charge Distribution in the Poly(TMAco-SA) on Nonspecific Protein Adsorption. An important characteristic of zwitterionic polymer brushes for resisting nonspecific protein adsorption is the overall charge neutrality within Langmuir 2010, 26(5), 3522–3530

Chang et al.

Article

Figure 1. Chemical structures of gold surfaces with (a) polyTMA, polySA, and polySBMA polymer brushes, (b) mixed-charge poly(TMAco-SA), copolymer brushes, and (c) CH3- and OEG-SAMs.

the functional groups containing positively and negatively charged moieties. Therefore, it is a critical consideration to control the charge distribution in the mixed-charge poly(TMAco-SA) copolymer brushes composed of two oppositely charged monomer components. To better understand the protein-foulingresistance properties of the polyampholyte brushes, the compositions and thicknesses of the poly(TMA-co-SA) copolymer brushes were characterized by XPS and ellipsometry. During the surface-initiated ATRP reaction, a monomer ratio of 1:1 was used to form a 1:1 statistical copolymer brush with homogeneous mixed-charge components. The quantification of the charge ratio of TMA and SA in the copolymer brushes was determined by the ratio of the atomic percentages of nitrogen and sulfur (N/S), as summarized in Table 1. From the compositional analysis of the 1:0 TMA/SA sample (polyTMA) and the 0:1 TMA/SA sample (polySA), nitrogen can only be found in the polyTMA and sulfur can only be found in the polySA. Therefore, the N/S ratio in the poly(TMA-co-SA) is a good indicator for the determination of reaction conditions that resulted in the formation of a homogeneous mixed-charge distribution in the copolymer brushes. As shown in Table 1, it can be seen that the N/S ratio approached 1.0 as the total monomer concentration of TMA and SA was increased from 0.4 to 1.2 M, resulting in the formation of a nearly Langmuir 2010, 26(5), 3522–3530

1:1 mixture of the two monomers based on the calculated N/S ratio of 0.99. The quantitation of single-protein adsorption was used to simply describe the molecular arrangement of mixed charged groups on the prepared brush surface. It has been suggested that neutral surfaces composed of positively and negatively charged pairs present protein-resistant properties.23 Figure 2 shows the typical sensor responses for single-protein adsorption on the CH3-SAM and poly(TMA-co-SA) surfaces. The net responses of the SPR signal after the surfaces were in contact with the protein solution were proportional to the amount of protein adsorbed on the surface. In general, CH3-SAMs showed an approximately monolayer adsorption of fibrinogen, with an SPR shift of 21 ( 1 nm (corresponding to 315 ( 15 ng/cm2) at 23 °C. As shown in Figure 2, it was found that the reduction of protein adsorption on the poly(TMA-co-SA) surfaces was associated with the charge ratio, which depended on the amounts of TMA and SA in the copolymer brush, which in turn depended on the polymerization conditions. The poly(TMA-co-SA) copolymer brushes formed under all of the reaction conditions in Table 1 resulted in a similar thickness of 30-45 nm. On the basis of the previous study, it was not expected that thickness differences of (15 nm in the copolymer brush layers would influence the DOI: 10.1021/la903172j

3525

Article

Figure 2. The SPR sensor response to a fibrinogen solution (1 mg/mL) on CH3-SAMs, poly(TMA-co-SA) no. 1, and poly(TMA-co-SA) no. 4 grafted surfaces at 23 °C. The wavelength shift of 1 nm in SPR is equivalent to 0.15 mg/m2 adsorbed proteins.

resistance to protein adsorption.27 The adsorptions of HSA, γ-globulin, and fibrinogen were measured to characterize nonspecific protein adsorption from single-protein solutions onto the poly(TMA-co-SA), and the results are summarized in Figure 3. We observed significant reductions in the adsorption of proteins on all prepared poly(TMA-co-SA) copolymer brushes compared to that on surfaces of CH3-SAMs. However, there was still significant γ-globulin and fibrinogen adsorption, and limited HSA adsorption on the copolymer brushes of poly(TMA-coSA) samples no. 1-3. On the basis of the XPS analysis in Table 1, it can be seen that the reaction conditions yielding samples 1, 2, and 3 (monomer concentrations of 0.4, 0.6, and 0.8 M, respectively) resulted in the formation of poly(TMA-co-SA)s having N/S ratios of 1.08, 0.85, and 0.96, respectively. The results indicate that even a slight surface charge can induce electrostatic interactions between proteins and surfaces, leading to surface adsorption. Concerning the reaction kinetics, the reaction rate of a polymerization is usually proportional to the monomer concentration in the reaction solution. In Figure 3, it can be seen that the copolymer surface of poly(TMA-co-SA) no. 4, with an N/S ratio of 0.99, prepared from a high monomer concentration of 1.2 M and a short reaction time of 1 h, reduced the single-protein adsorption to a level comparable with the adsorption on the surfaces with OEG-SAMs and polySBMA homopolymer brushes. The adsorbed amounts of each tested protein on poly(TMA-co-SA) no. 4 surfaces were found to be less than 2.5 ng/cm2 at 23 °C. The increased total amount of TMA and SA monomers in the reaction solution promoted the charge neutrality of poly(TMA-co-SA) in the prepared copolymer brushes by an equal distribution of mixed charged groups. These results suggest that the increase of polymerization rate might have reduced reaction selectivity for different charged monomers and readily formed a nonfouling statistical copolymer brush of poly(TMA-co-SA) from TMA and SA monomers. It was also found that the increased charge neutrality of poly(TMA-co-SA) made its copolymer surface more hydrophilic as evidenced by the decrease in water contact angle, indicating the formation of a strong hydration layer from the zwitterionic pairs of positively charged TMA and negatively charged SA in a 1:1 ratio. These results strongly (27) Feng, W.; Zhu, S.; Ishihara, K.; Brash, J. L. Langmuir 2005, 21, 5980.

3526 DOI: 10.1021/la903172j

Chang et al.

Figure 3. Adsorption of 1 mg/mL HSA, γ-globulin, and fibrinogen, respectively in PBS buffer and water contact angle on poly(TMA-co-SA) no. 1, poly(TMA-co-SA) no. 2, poly(TMA-co-SA) no. 3, and poly(TMA-co-SA) no. 4 grafted surfaces at 23 °C. The wavelength shift of 1 nm in SPR is equivalent to 0.15 mg/m2 of adsorbed protein.

support the hypothesis that hydrated surfaces with overall charge neutrality formed from mixed-charge components can present nonfouling characteristics for resisting nonspecific protein adsorption. Protein Adsorption Characteristics for Statistical Copolymer Brushes of Poly(TMA-co-SA) with Overall Charge Neutrality. In this study, the effects of the temperature, solution pH, and ionic strength of the protein solution on the amount of protein adsorbed on the poly(TMA-co-SA) polymer brushes were further tested to elucidate the antifouling surface characteristics. All polymer brushes were attached to the base of the SPR prism followed by the in situ evaluation of single-protein adsorption on the surfaces by SPR measurements. Figure 4 shows the adsorption of various proteins on prepared surfaces with five polymer brushes in PBS buffer (0.15 M and pH 7.4) at room temperature (23 °C) and at human body temperature (37 °C); for the latter, ionic strength and pH were held similar to that of the human bloodstream. Hydrophobic CH3-SAMs and zwitterionic polySBMA brushes were used as references. As indicated by SPR measurement, the polySBMA surfaces were almost completely resistant to the adsorption of HSA, γ-globulin, and fibrinogen. It is known that CH3-SAMs presenting hydrophobic methyl groups usually induce large amounts of protein adsorption,20,22,25 which was also observed here (see Figure 4). It was found that poly(TMA-co-SA) brushes with overall charge neutrality are highly resistant to nonspecific adsorption of various proteins at both 23 and 37 °C, while positively charged polyTMA and negatively charged polySA brushes showed high protein adsorption. In general, the increase of temperature usually enhances the exposure of hydrophobic patches of protein in buffer medium and therefore promotes protein adsorption to material surfaces by hydrophobic interaction, which is consistent with the results shown in Figure 4. Concerning thermodynamics of protein adsorption onto the surfaces grafted with mixed-charge copolymer brushes, the so-called “entropy of hydrophobic hydration” (ΔS) dominates the overall free energy of hydrophobic hydration. In general, increasing temperature expands the self-associated network of water, creating more space for hydrophobic moieties to interact between hydrocarbons of polymer brushes and hydrophobic patches of proteins, and the ΔS becomes more positive. From the consideration of thermodynamic principles governing Langmuir 2010, 26(5), 3522–3530

Chang et al.

Article

Figure 5. Adsorption of 1 mg/mL HSA, γ-globulin, and fibrinoFigure 4. Adsorption of 1 mg/mL HSA, γ-globulin, and fibrinogen in PBS buffer on CH3-SAMs, polyTMA, polySA, poly(TMAco-SA), and polySBMA grafted surfaces at 23 and 37 °C.

the adsorption of proteins onto surfaces, the major factors favoring protein adsorption will be the entropy gain of releasing water molecules from surfaces associated with the increase of temperature.1,28 Figure 5 compares the adsorption of three human proteins (HSA, γ-globulin, and fibrinogen) onto a poly(TMA-co-SA) surface in buffers with pH values from 3 to 11 at 23 °C, while ionic strength was kept low (0.01 M NaCl). These results show that the variation of solution pH may affect single-protein adsorption on a poly(TMA-co-SA) surface. An increase in the amount of proteins adsorbed onto the poly(TMA-co-SA) surface was observed at pH below 5.0, which is similar to the results for polySBMA brushes reported in our previous work.20 It was shown that proteins would have denatured structures and cause protein-protein interactions resulting in the formation of a denser packing or multilayer of protein adsorption at lower pH.29-31 From the contact angle measurements with pH values from 7.4 to 3.0, it was found that the poly(TMA-co-SA) surface displayed a low contact angle, from 25.6° ( 1.1° to 13.4° ( 1°; the poly(TMAco-SA) brushes displayed a highly hydrophilic surface at low pH. This indicates that the protein adsorption at low pH might not be due to hydrophobic interactions between proteins and the poly(TMA-co-SA) surface. These results suggest that electrostatic dipoles at the protein-poly(TMA-co-SA) interface might be the major contributors to the significant increase in protein adsorption. The three proteins with different isoelectric points (IEP) are 4.7-4.9, 5.2-5.5, 5.5 for HAS, γ-globulin, and figrinogen, respectively. The significant adsorption of proteins onto the poly(TMA-co-SA) surface at pH below 5.0 was mainly due to the electrostatic interaction between the positively charged proteins and the positively charged moieties of poly(TMA-co-SA) brushes. However, the adsorption of the proteins onto the poly(TMA-co-SA) surface does not solely depend on the electrostatic interactions between the proteins and the poly(TMA-co-SA) surface. Conformational changes of denatured proteins in acid media are commonly observed; these are induced by the electrostatic repulsion of positively charged patches inside the proteins at (28) Norde, W.; Lyklema, J. J. Colloid Interface. Sci. 1979, 71, 350. (29) Muzammil, S.; Kumar, Y.; Tayyab, S. Eur. J. Biochem. 1999, 266, 26. (30) Kumar, Y.; Tayyab, S.; Muzammil, S. Arch. Biochem. Biophys. 2004, 426, 3. (31) Norde, W.; Macritchie, F.; Nowicka, G.; Lyklema, J. J. Colloid Interface Sci. 1986, 112, 447.

Langmuir 2010, 26(5), 3522–3530

gen in buffer containing 0.01 M NaCl on poly(TMA-co-SA) grafted surfaces as a function of solution pH at 23 °C from SPR measurements.

lower pH.32 These denatured protein structures might enhance the electrostatic interaction with the poly(TMA-co-SA) surface. It is important to note that the net charge of proteins becomes less positive at higher pH. Therefore, proteins will tend to be less electrostatically attracted to a poly(TMA-co-SA) surface as the pH increases, as with the results shown in Figure 5. Three different ionic salts (NaCl, KCl, and NH4Cl) were used in this work, and the ionic strengths of the buffer media were adjusted by dissolving the salts in phosphate buffer (with an ionic strength of 10 mM; potassium phosphate 4.4 mM and pH 7.4) over a salt concentration range of 0.0 to 1.0 M. The effects of varying the ionic strength of the protein solution on the amount of protein adsorbed onto the surfaces with neutral mixed-charge poly(TMA-co-SA) copolymer brushes and zwitterionic polySBMA homopolymer brushes were investigated using SPR sensors. The adsorption of 1 mg/mL of human fibrinogen from buffers with different ionic strengths onto polymer-grafted SPR sensors is presented in Figure 6 for poly(TMA-co-SA)-grafted surfaces having an N/S ratio of 0.99 and a film thickness of ∼34.5 ( 2.8 nm. Interesting phenomena were observed for the protein adsorption amounts for all the salts. The protein adsorption was increased as the salt concentration increased from 10 to 60 mM, whereas the protein adsorption was decreased as the salt concentration increased from 60 mM to 1.0 M. There appeared to be an inflection point in protein adsorption at an ionic strength of 60 mM for each ionic salt, as shown in Figure 6a. For polySBMA, it was shown that fibrinogen adsorption on the polySBMAgrafted surfaces with film thicknesses in the range of 30-45 nm was ultralow at a wide range of ionic strengths; see Figure 6b. A possible explanation could be based on interchain and/or intrachain interactions, as in the simplified proposed model illustrated in Figure 7. The dependence on ionic strength of the protein adsorption behavior of both polymer-grafted surfaces can be explained by the different possible conformational states among the polymer brushes, including inter- and intrachain associations and nonassociations of the polymer brushes. At low ionic strengths (below 60 mM), fibrinogen adsorption on poly(TMAco-SA) brushes slightly increased with the increase of ionic strength. In general, it is a typical observation that proteins tend to adsorb onto hydrophobic surface due to the entropy gain from (32) Kondo, A.; Murakami, F.; Higashitani, K. Biotechnol. Bioeng. 1992, 40, 889.

DOI: 10.1021/la903172j

3527

Article

Figure 6. Human fibrinogen adsorption as a function of ionic strengths from SPR on surfaces grafted with (a) poly(TMA-coSA) and (b) polySBMA. The wavelength shift of 1 nm in SPR is equivalent to 0.15 mg/m2 of adsorbed protein. The standard deviation for all SPR data is below 5% (n = 3).

the excluded water molecules released into the ionic medium. This suggests that the partial exposure of hydrophobic chain segments bearing methyl groups is provided by the inter- and intrachain associations of poly(TMA-co-SA) brushes at lower ionic strengths, which is attributed to the associations between positive and negative charges in the polymer side chains. Thus, the abstraction of water molecules from the protein surface resulting in partial protein adsorption onto the poly(TMA-co-SA) surface becomes spontaneously driven by hydrophobic interactions. At higher ionic strengths (above 60 mM), it was found that protein adsorption decreased with increasing ionic strength for a given film thickness of poly(TMA-co-SA) brushes and exhibited a minimum level comparable to polySBMA brushes. These results indicate the increase in ionic strength can decrease internal charge interactions between polymer brushes, resulting in an obvious reduction of protein adsorption. It also indicates that polymeric chains of poly(TMA-co-SA) tend to expand in aqueous solutions with high ionic strengths due to strong electrostatic screening effects, as well as antipolyelectrolyte behavior, resulting in the formation of a mixed-charge conformational state and nonassociations, as 3528 DOI: 10.1021/la903172j

Chang et al.

illustrated in Figure 7a. The protein adsorption behavior of zwitterionic polySBMA-grafted surfaces was different from that of the mixed-charge poly(TMA-co-SA)-grafted surfaces, on which protein adsorption was much higher at low ionic strengths. The zwitterionic structure of polySBMA has positive and negative charges in the same side chain and can also assume different conformational states in inter- and intrachain associations, as shown in Figure 7b, which was proposed by Huck and coworkers.33 However, Figure 6b shows that a nonfouling surface of polySBMA brushes can be achieved when the solution contains a small amount of ionic salt above 10 mM. This indicates that the zwitterionic charge groups of SBMA are strongly hydrated through ionic solvation, resulting in the formation of a zwitterionic conformational state in nonassociations, even at the low ionic strength of 20 mM. Thus, the proposed model in Figure 7 depicts a significant molecular insight that the distance between two charged groups has effects on their inter- and intramolecular interactions between polymer brushes associated with the nonfouling properties. At high ionic strength, poly(TMA-co-SA), with a homogeneous mixture of balanced charge groups, exhibited a nonfouling, protein-resistant surface. These results demonstrate that the spacing between two oppositely charged groups at the nanometer scale might be similar to that of the ethylene spacer between two charged groups in the polySBMA side chains, resulting in the nonassociations of poly(TMA-co-SA) brushes. Plasma Protein Adsorption and Human Blood Cell Adhesion. The extent of plasma protein adsorption was evaluated by the inspection of blood compatibility of the surface grafted with poly(TMA-co-SA) brushes. Real-time adsorption of human plasma proteins onto poly(TMA-co-SA)-grafted surfaces was monitored using SPR at 37 °C. A dilute solution containing 20% plasma proteins (v/v, in PBS) from platelet poor plasma was used to reduce the effects of plasma viscosity in the SPR flow channel and to prevent the formation of small bubbles. It should be noted that the response of the SPR with broader sense of adsorption from PPP solution is not only to major components of plasma proteins but also to other small biomolecules, such as lipids and polysaccharides. However, it is accepted that the major change of SPR signal is attributed mainly to the higher molar mass of plasma proteins than other small biomolecules in PPP solution.20-22 Figure 8 shows the adsorption of plasma proteins on CH3-SAMs, OEG-SAMs, poly(TMA-co-SA), and polySBMA surfaces at 37 °C. It is known that CH3-SAMs presenting hydrophobic methyl groups usually induce large amounts of plasma protein adsorption,20,22,25 which can be observed in Figure 8. We observed significant decreases in adsorption of plasma proteins on OEG-SAMs as compared to CH3-SAMs. Similar to reductions in the aforementioned single protein adsorption, poly(TMA-co-SA) and polySBMA surfaces also greatly resisted plasma protein adsorption. The amount of plasma proteins adsorbed was 38.25 ng/cm2 on OEG-SAMs, 7.65 ng/cm2 on the poly(TMA-co-SA) surface, and 1.65 ng/cm2 on the polySBMA surface from the dilute plasma solution. Horbett et al. showed that the adhesion and activation of platelets from the bloodstream might be correlated with the adsorption of proteins on surfaces.8,9 Even a small amount of plasma protein on a surface can lead to the thrombogenic reaction. Figure 9 shows SEM images, at a magnification of 1000, of the prepared substrates in contact with recalcified human whole blood for 120 min at 37 °C in vitro. The SEM results showed the formation of thrombosis on CH3-SAMs, with full-scale platelet (33) Cheng, N.; Brown, A. A.; Azzaroni, O.; Huck, W. T. S. Macromolecules 2008, 41, 6317.

Langmuir 2010, 26(5), 3522–3530

Chang et al.

Article

Figure 7. Simplified model of the possible conformational states as a function of ionic strengths: (a) mixed-charge copolymer brushes of poly(TMA-co-SA) and (b) zwitterionic homopolymer brushes of polySBMA.

Figure 8. Adsorption of plasma proteins on CH3-SAMs, OEGSAMs, poly(TMA-co-SA), and polySBMA grafted surfaces at 37 °C from human blood plasma. Langmuir 2010, 26(5), 3522–3530

adhesion and activation at the blood-contact site. However, there also clearly appeared an obvious amount of slightly activated platelets and other adhered blood cells on OEGSAMs, while here the adsorbed amount of plasma proteins was only 38.25 ng/cm2. On the basis of previous reports from Horbett et al., it is believed that reducing plasma protein adsorption levels to below 10 ng/cm2 on biomaterial surfaces can effectively prevent the adhesion and activation of platelets from the bloodstream.8,9 It was shown that no blood cells adhered to the poly(TMA-co-SA) and polySBMA surfaces. These results confirmed the previous hypothesis that even a small amount of proteins adsorbed on a surface can lead to the adhesion and activation of platelets from the bloodstream. From the results of both the protein adsorption measurements and the blood cell adhesion in vitro, this study demonstrated for the first time that mixed-charge copolymer brushes of poly(TMA-co-SA) with overall charge neutrality can be used to achieve high blood compatibility with human whole blood. We concluded that a nanometer-scale homogeneous, balanced mixture of charged surfaces from either mixed-charge or zwitterionic groups can provide excellent blood compatibility at human body temperature. DOI: 10.1021/la903172j

3529

Article

Chang et al.

Figure 9. SEM photographs of human blood cells adhered onto the surface of (a) CH3-SAMs, (b) OEG-SAMs, (c) poly(TMA-co-SA), and (d) polySBMA surface. All images at a magnification of 1000.

In this work, it is demonstrated that copolymer brushes with homogeneous charge balance, prepared from two oppositely charged compounds, are highly resistant to nonspecific protein adsorption. However, it was found that the nonfouling surfaces prepared from the zwitterionic structure of polySBMA polymer brushes are better than that the mixed-charge structure of poly(TMA-co-SA) copolymer brushes, under all conditions tested. It could be attributed to the distance between two oppositely charged groups associated with the hydrated-promoting effects on the inter- and intramolecular interactions between copolymer brushes, as shown in the simplified model in Figure 7. Therefore, the general concept for preparing new nonfouling materials from common oppositely charged groups could be performed in the ease of synthesis, but it should concern the issue of separations between two charged sites even with overall charge neutrality of copolymer brushes. The proposed principle will have a strong impact on a broad range of biomedical applications.

Conclusions In this work, a mixed charge distribution with overall charge neutrality was obtained in the prepared poly(TMA-co-SA) copolymer brushes by surface-initiated ATRP at a high reaction rate. We utilized SPR measurements to demonstrate that an antifouling surface of neutral poly(TMA-co-SA) brushes with nonspecific protein-binding resistance in PBS buffer at both 23 and 37 °C can be achieved when the surface contains a balanced mixture of charge groups from the two oppositely charged components, TMA and SA. The results show that the increased charge

3530 DOI: 10.1021/la903172j

neutrality of poly(TMA-co-SA) resulting from the high initial concentrations of TMA and SA and a short reaction time made the copolymer surface more hydrophilic, indicating the formation of a strong hydration layer due to the pseudozwitterionic pairs of TMA and SA in a 1:1 ratio. We also found that the presence of salts at different ionic strengths had different antipolyelectrolyte effects on the prepared copolymer brushes. The results yielded significant molecular insight into poly(TMA-co-SA) association states. The distance between two charged groups in a polymer chain can have effects on their inter- and intramolecular interactions, which can promote or inhibit hydration of the charged groups. In the interactions between copolymer brushes, this hydration is associated with the nonfouling properties of the surface. Furthermore, the results also showed a remarkable reduction in the adsorption of plasma proteins and adhesion of blood cells on the poly(TMA-co-SA)-grafted surface compared to the others. The results suggest that the hemocompatible nature of poly(TMA-co-SA) copolymer brushes with an equal distribution of charge groups in the side chains gives it great potential in antithrombogenic surface coatings. Acknowledgment. The authors express their sincere gratitude to the Center-of-Excellence (COE) Program on Membrane Technology from the Ministry of Education (MOE), R.O.C., to the project Toward Sustainable Green Technology in the Chung Yuan Christian University, Taiwan (CYCU-98-CR-CE), and to the National Science Council (NSC 97-2221-E-033-006 and NSC 95-2221-E-008-172-MY3) for their financial support.

Langmuir 2010, 26(5), 3522–3530