Nonfouling Behavior of Polycarboxybetaine ... - ACS Publications

Sep 12, 2008 - Zheng Zhang, Hana Vaisocherová, Gang Cheng, Wei Yang, Hong Xue, ... groups and the negative carboxyl groups and environmental factors ...
0 downloads 0 Views 841KB Size
Download PDF
2686

Biomacromolecules 2008, 9, 2686–2692

Articles Nonfouling Behavior of Polycarboxybetaine-Grafted Surfaces: Structural and Environmental Effects Zheng Zhang, Hana Vaisocherova´, Gang Cheng, Wei Yang, Hong Xue, and Shaoyi Jiang* Department of Chemical Engineering, University of Washington, Seattle, Washington 98195 Received January 21, 2008; Revised Manuscript Received July 19, 2008

Zwitterionic carboxybetaine (CB) has unique dual functionality for ligand immobilization on a nonfouling background. The properties of CB groups depend on their spacer groups between the positive quaternary amine groups and the negative carboxyl groups and environmental factors (e.g., ionic strengths and pH values). In this work, five polycarboxybetaines were prepared, including one polycarboxybetaine methacrylate (polyCBMA) and four polycarboxybetaine acrylamides (polyCBAAs) with different spacer groups. The polymers were grafted from a gold surface covered with initiators using surface-initiated atom transfer radical polymerization. Fibrinogen adsorption was measured as a function of ionic strengths and pH values using surface plasmon resonance sensors. The responsive protein adsorption on four polyCBAAs was mapped out. Results show that most of these surfaces exhibit high protein resistance in a wide range of ionic strengths and are more effective than zwitterionic selfassembled monolayers. Although protein adsorption tends to increase at low ionic strength and low pH value, it is still very low for polycarboxybetaines with a methylene, an ethylene, or a propylene spacer group but is more evident for polyCBAA with a longer spacer group (i.e., a pentene group). The response to ionic strengths and pH values can be attributed to the antipolyelectrolyte and protonation/deprotonation properties of polycarboxybetaines, respectively. Both of these properties are related to the spacer groups of CBs.

1. Introduction Biomimetic zwitterionic groups, such as functional groups of phosphorylcholine (PC), sulfobetaine (SB), and carboxybetaine (CB), can highly resist nonspecific protein adsorption.1-3 CB groups usually are more hydrophilic than SB groups and exhibit acid-base equilibrium.4,5 Moreover, CB groups have unique dual functionality.6 They not only highly resist nonspecific protein adsorption, but also have abundant functional groups for protein immobilization.6 The dual-functionality makes polycarboxybetaines very promising for many biomedical and engineering applications. While it is well-known that physical properties of zwitterionic polymers, such as chemical solubility, hydrophilicity, rheological properties, and mechanical properties, are usually responsive to ionic strength, temperature, and pH, the behavior of their interactions with biomolecules and microorganisms remains unknown. Zwitterionic polymers are characterized by their antipolyelectrolyte behavior, that is, polymeric chains tend to expand in aqueous solutions with the presence of ions.7-11 The antipolyelectrolyte behavior greatly affects the solubility and rheological properties of linear zwitterionic polymers and the swelling properties of zwitterionic hydrogels. In addition to ionic strengths, the solution properties of zwitterionic polymer are highly responsive to temperature.12-14 It has been shown that supercollapsed sulfobetaine brushes can change to a more hydrophilic form by increasing temperature.15 At low pH, CBbased compounds could be protonated, which makes them sensitive to pH. Many solution properties of both CB-based * To whom correspondence should be addressed. E-mail: sjiang@ u.washington.edu.

surfactants4 and polycarboxybetaines16-18 depend on both pH and ionic strength. For example, at low ionic strength, polycarboxybetaines may exhibit a dramatic increase in viscosity as pH is lowered. The addition of electrolytes will disrupt pHdependence due to electrostatic screening effects.16-18 With their unique responsive properties, polycarboxybetaine-grafted surfaces may exhibit different biological properties under different environmental conditions. For some CB-based compounds and polycarboxybetaines, the spacer groups between the positive quaternary amine groups and the negative carboxyl groups can affect the solution properties. For CB-based surfactants, the solution and foaming properties were changed by varying the number of methylene groups separating the charged groups.4,19 It was reported that when the spacer group changed from a methylene to a propylene of a copolymer containing CB, solution viscosity responds differently to pH and ionic strength.17 In our previous study, a carboxybetaine methacrylate (CBMA) with an ethylene spacer group between the charged groups was grafted from a gold surface to form a well-packed layer of polymer brushes, on which nonspecific protein adsorption was shown to be lower than 0.3 ng/cm2 from a buffer solution (pH 7.4, 150 mM).6,20 A copolymer containing CBMA with a methylene spacer group was also found with good blood compatibilities.21 However, protein interactions with polycarboxybetaines of other structures have not been investigated yet. In this work, five zwitterionic carboxybetaines were synthesized, including one CBMA and four carboxybetaine acrylamides (CBAAs; Scheme 1). The CBMA has an ethylene spacer group. Each of the four polyCBAA has a different spacer group, that is, a methylene, an ethylene, a propylene, and a pentene

10.1021/bm800407r CCC: $40.75  2008 American Chemical Society Published on Web 09/12/2008

Behavior of Polycarboxybetaine-Grafted Surfaces Scheme 1. Polycarboxybetaines Grafted on Gold Surfaces: PolyCBMA, PolyCBAA-1, PolyCBAA-2, PolyCBAA-3, and PolyCBAA-5

group. It is well-known that amide groups are more hydrolytically stable than ester groups due to their lower reactivity of carbonyl groups. Introduction of the acrylamide group to replace methacrylate might increase the hydrolytic stability of polycarboxybetaines. Both self-assembled monolayers (SAMs)22-26 and surface-grafted polymer brushes3,6,20,27 are convenient to form nonfouling zwitterionic surfaces. Zwitterionic SAMs such as PC-SAMs and SB-SAMs can resist nonspecific protein adsorption under certain ionic strengths. However, protein adsorption dramatically increases as ionic strength decreases, which was attributed to the dipole orientation on the SAM surfaces.23 Instead of SAMs, we use grafted polymer brushes to investigate the nonfouling behavior of CB-based materials. In this work, five different polycarboxybetaines were grafted from a layer of initiator SAMs using surface-initiated atom transfer radical polymerization (ATRP). Protein adsorption on the surfaces grafted with these five polycarboxybeteaines from buffers is compared. Fibrinogen adsorption on different polycarboxybetaine brushes was measured as a function of ionic strengths and pH values using surface plasmon resonance (SPR) sensors. The objective of this work is to study protein adsorption on different polycarboxybetaine-grafted surfaces as a function of ionic strengths and pH values so as to provide a fundamental understanding of how the chemical structures of carboxybetaines will affect their biological properties.

2. Experimental Methods 2.1. Materials. Human plasma fibrinogen was purchased from Sigma-Aldrich (Milwaukee, WI). N-(3-Dimethylaminopropyl) acrylamide (>98%) was from TCI America, Portland, OR. β-Propiolactone (95%), copper(I) bromide (99.999%), 2,2′-bipyridine (BPY, 99%), and tetrahydrofuran (THF, HPLC grade) were purchased from SigmaAldrich (Milwaukee, WI). Phosphate buffer saline (PBS, 0.01 M phosphate, 0.138 M sodium chloride, 0.0027 M potassium chloride, pH 7.4) was purchased from Sigma Chemical Co. Ethanol (absolute, 200 proof) was purchased from AAPER Alcohol and Chemical Co. Water used in experiments was purified using a Millipore water purification system with a minimum resistivity of 18.0 MΩ · cm. THF for reactions and washings was dried by sodium before use. ω-Mercaptoundecyl bromoisobutyrate was synthesized as described before.20,28 The buffers were prepared with the following compositions: a solution with an ionic strength of 10 mM (4.4 mM potassium phosphate, pH 7.4) and solutions with an ionic strength of 19, 50, 100, 150, and 200 mM (4.4 mM potassium phosphate with a NaCl concentration of 9, 40, 90, 140, or 190 mM, respectively, pH 7.4). Buffers were prepared in distilled, deionized water and filtered through 0.22 µm filters prior

Biomacromolecules, Vol. 9, No. 10, 2008

2687

to use. Protein solutions were prepared by dissolving solid protein (10 mg) in the appropriate buffer (10 mL) at room temperature. The pH values may be adjusted with a small amount of HCl or NaOH. To study fibrinogen adsorption on polyCBAA-grafted sensors, a buffer with pH of 5 was prepared based on a 10 mM sodium acetate buffer and its ionic strength was adjusted with NaCl. 2.2. Monomer Synthesis. 2-Carboxy-N,N-dimethyl-N-(2′-methacryloyloxyethyl) ethanaminium inner salt (carboxybetaine methacrylate, CBMA) was synthesized as described in previous papers.3,6 The synthesis of (2-carboxymethyl)3-acrylamidopropyldimethylammonium bromide, methyl ester (CBAA-1-ester), (4-carboxypropyl)3-acrylamidopropyldimethylammonium bromide, ethyl ester (CBAA-3-ester), and (6-carboxypentyl)3-acrylamidopropyldimethyl ammonium bromide, ethyl ester (CBAA-5-ester) is described in the Supporting Information.29 2-Carboxy-N,N-dimethyl-N-(3′-acrylamidopropyl) ethanaminium inner salt (CBAA-2) was synthesized by the reaction of N-(3-dimethylaminopropyl) acrylamide and β-propiolactone. β-Propiolactone (2.88 g, 40 mmol) in 20 mL of dried acetone was added dropwise to a solution of N-(3-dimethylaminopropyl) acrylamide (3.90 g, 25 mmol) dissolved in 50 mL of dried acetone. The reaction mixture was stirred under nitrogen protection at 15 °C for ∼5 h. The white precipitate was washed with 100 mL anhydrous acetone and 150 mL anhydrous ether. The product was dried under reduced pressure to obtain the final CBAA monomer product (white powder, 98% yield). The monomer was kept at 2-8 °C before polymerization. 1H NMR (300 MHz, D2O): 2.01 (m, 2H, C-CH2-C), 2.62 (t, 2H, CH2-CdO), 3.05 (s, 6H, N+(CH3)2), 3.26-3.37 (4H, CH2-N+-CH2), 3.52 (t, 2H, N-CH2), 5.74 (m, 1H, CHdC-CON-trans), 6.17 (m, 1H, CHdC-CON-cis), 6.19 (m, 1H, dCH-CON-). 3-Carboxy-N,N-dimethyl-N-(3′-acrylamidopropyl) propanaminium inner salt (CBAA-3) was synthesized by hydrolysis from CBAA-3ester.4,18 The CBAA-3-ester was dissolved in water and passed through a column packed with Amberlite IRA-400 (hydroxide form) resin. After water was removed under reduced pressure, the product was then washed with anhydrous acetone and anhydrous ether. The solvent was then removed under reduced pressure and a colorless oil was obtained. The monomer was kept at 2-8 °C before polymerization. 1H NMR (300 MHz, D2O): 1.85 (m, 4H, C-CH2-C), 2.10 (t, 2H, CH2-CdO), 2.93 (s, 6H, N+(CH3)2), 3.11-3.25 (6H, CH2-N+-CH2 and N-CH2), 5.62 (m, 1H, CHdC-CON-trans), 6.07 (m, 1H, CHdC-CON-cis), 6.11 (m, 1H, dCH-CON-). 2.3. Surface-Initiated ATRP on Gold Surfaces. Five polycarboxybetaine brushes were grafted on gold surfaces using two different methods. PolyCBMA, polyCBAA-2, and polyCBAA-3 were grafted onto gold-coated SPR sensor chips or gold-coated silicon chips using surface-initiated ATRP. The preparation method of the polymer brushes was similar to that described in previous publications. 6,20 Briefly, CuBr (1 mmol) and a gold surface with a Br-terminated SAM prepared from mercaptoundecyl bromoisobutyrate were placed in a nitrogen-purged reaction tube. The degassed solution (pure water and methanol in a 1:1 volume ratio, 10 mL) with CBMA or CBAA (6.5 mmol) and BPY (2 mmol, in 5 mL of degassed methanol) was transferred to the tube using a syringe. After reaction for more than 1 h under nitrogen, the SPR chip or gold disk was removed and rinsed with ethanol, water, and PBS solution (Scheme 1). For preparing samples grafted with polyCBAA-1 or polyCBAA-5, a two-step reaction was performed. The monomer, CBAA-1-ester, or CBAA-5-ester, was first grafted onto gold-coated SPR sensor chips using surface-initiated ATRP as described above. The surfaces were then treated with 100 mM NaOH solution for 1-2 h to hydrolyze the ester groups. The samples were stored in PBS solutions before testing (Scheme 2). 2.4. Determination of Film Thickness and Surface Sensitivity. Due to the nature of surface plasmons, the sensitivity of a SPR sensor depends on the distance of the binding event from the SPR active (gold) surface.30 To calibrate the SPR surface sensitivity, the thickness and refractive index of a polymer layer must be found. A series of

2688

Biomacromolecules, Vol. 9, No. 10, 2008

Scheme 2. PolyCBAA-1-ester and PolyCBAA-5-ester were Grafted onto Gold Surfaces and Hydrolyzed into Zwitterionic PolyCBAA-1 and PolyCBAA-5

polyCBAA-2 layers in different thicknesses were first characterized by atomic force microscopy (AFM) for height measurements (described in the next paragraph) and SPR for wavelength shift measurements. These results were compared with a theoretical model,31 and the refractive index was determined to be 1.445 ( 0.006 RIU. Assuming that other polyCBAA polymers have similar refractive indices to polyCBAA-2, the SPR wavelength shifts were measured to determine the thicknesses of these polymers. The thicknesses for those polymer layers studied in this work were 15-20, 15-20, 15-25, and 25-32 nm for polyCBAA-1, polyCBAA-2, polyCBAA-3, and polyCBAA-5, respectively. To compensate the loss of the surface sensitivity due to the polymer layer, the sensor sensitivity was calculated for the surfaces with or without the polymer layer for their refractive indices (both adjusted to the operating wavelength of 750 nm). The calibration factors for the polymer chips used were 0.82, 0.77, 0.77, and 0.76 for polyCBAA-1, polyCBAA-2, polyCBAA-3, and polyCBAA-5, respectively. These values were used to calibrate the sensor response to fibrinogen adsorption under various binding conditions. For the measurements of film thicknesses by AFM, gold-coated Si wafers were patterned using standard photolithography techniques. Briefly, an ∼20 µm layer of P20 7/17 primer and an ∼100 µm layer of AZ 1512 photoresist were spin-coated onto the gold-coated wafers. The wafers were then placed under a photolithography mask that was patterned with 25 µm lines, spaced 25 µm apart. Following exposure to ultraviolet light, the Photoresist was developed using AZ 351 developer. Then, the underlying gold layer was etched away using THA gold etching solution. Following this etching step, the remaining photoresist was removed by rinsing the sample with isopropanol, followed by sonication in acetone. SAMs were formed on the patterned gold-grafted Si wafers and used for brush growth in the same polymerization solution as described above. Following the PBS soaking stage, the patterned samples were rinsed extensively with Milli-Q water and dried with filtered air. Contact mode AFM images were acquired using a Dimension 3100 AFM (Digital Instruments/Veeco, Woodbury, NY) operated in air. A commercial silicon nitride cantilever (DI) with an elastic module of 0.56 N/m was used. Different locations of each sample were measured to obtain an average thickness. The thicknesses of polyCBAA-2 layers were reproducibly in the range of 15-20 nm. 2.5. ATR-FTIR and FTIR Spectroscopy. The polymers were grafted on gold-coated silicon chips using surface-initiated ATRP as described above. Attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectra of polyCBMA-grafted golds were performed on the chips with a Bruker tensor 27 (Billerica, MA) in air with a relative humidity of about 30%. CBMA hydrogels were prepared as previously described.6 After CBMA hydrogels were immersed into aqueous solutions with different

Zhang et al. pH values for more 1 h, the hydrogels were put into a large amount of acetone to remove water. Acetone was then removed by vacuum. The sample-KBr pellets were tested using a Perking Elmer 1720 Fourier transform infrared (FTIR) spectrophotometer with a wavenumber resolution of 0.5. 2.6. Surface Characterization by XPS and Ellipsometry. X-ray photoelectron spectroscopy (XPS) analysis was performed on polymer brushes using a Surface Science Instruments (SSI) S-Probe equipped with a monochromated Al KR X-ray source. Sample preparation is the same as in ATR-FTIR experiments. The energy of emitted electrons is measured with a hemispherical energy analyzer at pass energies ranging from 50 to 150 eV. Elemental composition present on the surface was identified from a survey scan. All data were collected at 55° from the surface normal takeoff angle. The binding energy (BE) scale is referenced by setting the peak maximum in the C1s spectrum to 285.0 eV. Three spots were measured for each sample. Ellipsometry was performed using a spectroscopic ellipsometer (Sentech SE-850, GmbH). Sample preparation is the same as in ATRFTIR experiments. Five separate spots were measured at three different angles of incidence (50, 60, and 70°) in the VIS region. The same batch of gold-coated chips was cleaned by UV-ozone cleaner for 20 min, washed with ethanol and Millipore water, and dried with nitrogen. The bare gold-coated chips were used as a reference. The film thicknesses of the grafted polymer brushes were determined using the Cauchy layer model with an assumed refractive index of 1.45. 2.7. Protein Adsorption from SPR. Fibrinogen solution of 1.0 mg/ mL in different ionic strengths and pH values was flowed over the sensor surfaces at a flow rate of 0.05 mL/min. A surface-sensitive SPR detector was used to monitor protein-surface interactions in real time. In this study, wavelength shift was used to measure the changes in surface concentration (mass per unit area).We took the amount of adsorbed fibrinogen from the buffer solution on HS(CH2)15CH3 SAM as a monolayer (ML). The wavelength shift induced due to protein adsorption on measured surfaces was normalized to be %ML by that on HS(CH2)15CH3 SAM.32

3. Results and Discussion 3.1. Preparation and Characterization of Carboxybetaine Polymer Brushes. In this work, five polycarboxybetaines were grafted on gold surfaces, including one polyCBMA and four polyCBAAs, with different spacer groups between the charged groups. A methylene, an ethylene, a propylene, and a pentene group were inserted between the quaternary amine groups and the negative carboxyl groups to form polyCBAA1, polyCBAA-2, polyCBAA-3, and polyCBAA-5, respectively (Scheme 1). For preparation of a polyCBMA-, polyCBAA-2-, or polyCBAA-3-grafted surface, the CBMA, CBAA-2, or CBAA-3 monomer was polymerized on a gold surface covered with initiator SAMs. CBAA-1 monomer was grafted onto a surface using the same method. However, very low polymer coverage was detected from XPS analysis. To prepare the polymer brushes with a similar thickness as polyCBMA and other polyCBAA, a two-step procedure was performed for preparing polyCBAA-1 brushes. The CBAA-1-ester was first grafted onto a surface using surface-initiated ATRP, and then the surface was treated with 100 mM NaOH solution to hydrolyze the ester group into the zwitterionic group (Scheme 2). For preparation of CBAA-5 monomer, CBAA-5-ester was dissolved in water and hydrolyzed via passing over Amberlite resin (hydroxide form). However, NMR showed that the monomer was partially polymerized after the hydrolysis and purification. Instead of using CBAA-5 monomer for surface-grafting, the two-step procedure was also applied for preparation of polyCBAA-5 from CBAA-5-ester monomer similar to how polyCBAA-1 was grafted described above (Scheme 2).

Behavior of Polycarboxybetaine-Grafted Surfaces

Biomacromolecules, Vol. 9, No. 10, 2008

2689

Figure 1. (a) ATR-FTIR spectrum of polyCBMA brushes grafted on a gold surface covered with initiators (characteristic peaks: 1727 cm-1, CdO stretching; 1592 cm-1, COO- asymmetric stretching; and 1380 cm-1, COO- symmetric stretching). (b) FTIR spectra of polyCBMA hydrogels under different pH values.

The prepared surfaces were characterized using ATR-FTIR, XPS, and ellipsometry. For polyCBMA brushes, ATR-FTIR spectrum shows the characteristic peaks of polyCBMA (1727 cm-1, CdO stretching; 1592 cm-1, COO- asymmetric stretching; and 1380 cm-1, COO- symmetric stretching). The ATRFTIR spectrum of polymer brushes is very similar to the FTIR spectrum of a piece of polyCBMA hydrogel, both of which the deprotonated carboxyl groups (COO-) can be observed (Figure 1b, pH ) 7). The thickness of grafted polyCBMA is about 10-15 nm measured from ellipsometry. XPS survey scans were

applied to analyze the elemental compositions of five polycarboxybeatine-grafted surfaces. Most atom % of C, N, and O measured from XPS is close to the theoretical atom % ratio of corresponding polymers (Table 1). One oxygen composition (polyCBAA-2) is higher than the theoretical value, likely due to surface contamination. Oxygen is commonly a contaminant on XPS samples and it can come from silicon oxide, metal oxides, silicone, or other organic contamination. 3.2. Protein Adsorption on PolyCBMA-Grafted Surfaces: Effect of Ionic Strength. Adsorption of human fibrinogen from

2690

Biomacromolecules, Vol. 9, No. 10, 2008

Zhang et al.

Table 1. Thickness and Surface Chemical Compositiona of PolyCBMA, PolyCBAA-1, PolyCBAA-2, PolyCBAA-3, and PolyCBAA-5 Grafted onto Gold Surfaces

thickness (nm) C% O% N% b

CBMA

CBAA-1

CBAA-2

CBAA-3

CBAA-5

10-15b 66.7 ( 3.2 (68.8) 28.2 ( 2.8 (25.0) 5.1 ( 1.7 (6.2)

15-20c 71.1 ( 4.5 (66.7) 19.3 ( 3.2 (20) 9.6 ( 4.4 (13.3)

15-20c 65.4 ( 2.3 (68.8) 24.3 ( 2.6 (18.7) 10.3 ( 0.3 (12.5)

15-25c 73.5 ( 3.9 (70.6) 16.1 ( 3.1 (17.6) 10.4 ( 1.5 (11.8)

25-32c 72.0 ( 0.7 (73.7) 18.4 ( 0.9 (15.8) 9.5 ( 0.2 (10.5)

a The atomic percentages of the polymers are calculated from XPS spectra (the values in the parentheses are the theoretical atomic percentages). The thickness was measured from ellipsometry. c The thicknesses were evaluated from the SPR wavelength shift.

Scheme 3. Deprotonation and Protonation of a Carboxybetaine Unita

a Corresponding to a transition from a nonfouling polymer to a fouling polymer.

Figure 2. The pH and ionic strength dependence of human fibrinogen adsorption on polyCBMA-grafted sensors from SPR. The wavelength shift induced due to protein adsorption on measured surfaces was normalized to be % monolayer (ML) by that on HS(CH2)15CH3 SAM at the same ionic strength and pH value. These reference values for protein adsorption on HS(CH2)15CH3 SAM are provided in Supporting Information.

buffers with different ionic strengths and pH values onto polyCBMA-grafted SPR sensors is presented in Figure 2. It is shown that fibrinogen adsorption on the polySBMA brushes is very low in a wide range of ionic strengths. The protein adsorption behavior of polyCBMA-grafted surfaces is different from that of zwitterionic SB or PC SAMs on which protein adsorption is much higher at low ionic strengths. It was reported that fibrinogen adsorption on SB-based SAMs and PC-based SAMs is about 88 ML% and 48 ML% at an ionic strength of 10 mM, respectively,23 which is much higher than 4.3 ML% on polyCBMA-grafted surfaces (Figure 2). For zwitterionic SB or PC SAMs, high protein adsorption at low ionic strengths is attributed to the dipole vectors, which have a perpendicular component to the surface and correspond to a net electric field associated with the dipoles at the SAM surfaces.23,26 At higher ionic strengths, the net electric field can be reduced due to electrostatic screening effects. Furthermore, polymer brushes provide a much thicker zwitterionic layer than that of SAMs, and long flexible polymer brushes also provide some flexibility to resist nonspecific protein adsorption.33,34 Unlike zwitterionic SAMs, the dipole vectors on polymer pendant groups are flexible to be oriented in different directions and the net electric field is reduced. Because of all these factors, protein adsorption on the polymer brushes is much lower that that on zwitterionic SAMs at low ionic strengths. At low ionic strengths, fibrinogen adsorption on polyCBMA brushes still slightly increases with the decrease of ionic strength. At a pH value of 7.4, the surfaces are nonfouling when ionic strength is higher than 100 mM, while fibrinogen adsorption is 1.5 ML% at the ionic strength of 50 mM and 4.3 ML% at the ionic strength of 10 mM. This can be explained by the antipolyelectrolyte behavior of polyCBMA. Different from other hydrophilic polymers, zwitterionic polymers are characterized

by their antipolyelectrolyte behavior, that is, polymeric chains tend to expand in aqueous solutions in the presence of ions, while tend to shrink at low ionic strengths.7-10 The polyCBMA chains collapse because of the strong intramolecular and intermolecular attractions of zwitterionic pendant groups at low ionic strengths. This may decrease the hydration ability of polyCBMA and increase their protein adsorption. However, this effect of ionic strength is smaller on polyCBMA than on zwitterionic SAMs. Thus, the polyCBMA surfaces are still highly protein-resistant in a wide range of ionic strengths. 3.3. Protein Adsorption on PolyCBMA-Grafted Surfaces: Effect of pH. At lower ionic strengths, protein adsorption on polyCBMA-grafted surfaces also presented a slight pHdependence (Figure 2). For example, for a protein solution with an ionic strength of 50 mM, the polyCBMA-grafted surface is nonfouling at pH 8.7, but protein adsorption is 1.5 ML% at pH 7.4. For a fibrinogen solution with an ionic strength of 200 mM, the surface presents a protein adsorption of 2.4 ML% at pH 5, while nonfouling continues at both pH 7.4 and 8.7. CB groups can exhibit acid-base equilibria at different pH values. At lower pH, due to the protonation of the carboxyl groups, the surface can be positively charged and tends to adsorb fibrinogen via electrostatic interactions. When the pH increases, the carboxyl groups are deprotonated and the surface becomes neutral and protein-resistant (Scheme 3). To observe the protonation/deprotonation of CB groups, a bulk hydrogel containing 94 mol % polyCBMA was prepared and the FTIR spectra of the hydrogel at different pH values were measured (Figure 1b). For a polyCBMA hydrogel stored at an aqueous solution, pH 7, both a peak at 1727 cm-1, which is attributed to stretching of CdO groups, including COOH groups, and a peak at 1592 cm-1, which is attributed to asymmetric stretching of COO- groups, were present. However, when the pH value of the solution decreases to 2, only one peak at 1727 cm-1 is present, indicating the carboxyl groups on polyCBMA are all protonated. It should be pointed out that the total protonation only occurs at very low pH in aqueous solutions. It was reported previously that the pKa of a CB-2 surfactant in an aqueous solution is about 3.25, much lower than acetic acid (ca. 4.8).4 The presence of electrolyte may decrease the protonation of carboxybetaine groups.18 Usually, the polymer brushes can be

Behavior of Polycarboxybetaine-Grafted Surfaces

Biomacromolecules, Vol. 9, No. 10, 2008

2691

Figure 3. Fibrinogen adsorption as a function of ionic strengths and pH values from SPR on various surfaces grafted with (a) polyCBAA-1, (b) polyCBAA-2, (c) polyCBAA-3, and (d) polyCBAA-5. The wavelength shift of 1 nm in SPR is equivalent to 0.15 mg/m2 adsorbed proteins.

weakly charged from a buffer solution at low pH, leading to a slight pH-dependence of protein adsorption on polyCBMA grafted, as shown in Figure 2. 3.4. Protein Adsorption on Four PolyCBAA with Different Spacer Groups. Four polyCBAA polymers were grafted on the surfaces with an acrylamide backbone and different spacer groups between the charged groups (Scheme 1). It was reported that the thickness of polymer brushes grafted via ATRP is a function of the degree of polymerization rather than grafting density. The grafting density is fixed and determined by the density of the initiator density before the polymerization.28 The thicknesses of polyCBAA-1, polyCBAA-2, polyCBAA-3, and polyCBAA-5 were 15-20, 15-20, 15-25, and 25-32 nm, respectively. We prepared polyCBAA-2 of different thicknesses and tested their fibrinogen adsorption. For the film thicknesses of 10-15, 23-26, 25-30, 35-40, and 45-50 nm, fibrinogen adsorption was 0.3, 0.5, 0.1, 0, and 0.8 ng/cm2, much lower than 1 ML% (pH 7.4, 150 mM). Thus, we attribute the changes in protein adsorption to different CB structures or environmental factors. Fibrinogen adsorption from SPR as a function of ionic strengths and pH values on the four different surfaces is presented in Figure 3. It is shown that polyCBAA-grafted surfaces can also present very low fibrinogen adsorption in a wide range of ionic strengths and pH values. Protein adsorption

usually slightly increases with the decrease of ionic strengths and pH values, showing the similar nonfouling behavior for all CB-based polymer brushes. At lower ionic strength, polyCBAA-2 exhibited a better protein resistance than polyCBMA, both of which have an ethylene spacer between two charged groups. For instance, on the polyCBMA-grafted surfaces, protein adsorption induced a wavelength shift of 0.05 nm (100 mM), 0.66 nm (50 mM), and 1.71 (10 mM) at pH 7. On the polyCBAA-2-grafted surfaces, the wavelength shift is 0 nm (100 mM) and 0.04 nm(19 mM), lower than those on the polyCBMAgrafted surfaces. The hydrogen in acrylamide may serve as a hydrogen bond donor, which may alter the nonfouling properties of grafted polyCBAA polymers.35 However, the zwitterionic pendant groups are large enough to eliminate any possible negative effects from the acrylamide backbone, and the hydrophilic backbone might improve the nonfouling performance at lower ionic strength. Compared with polyCBMA, polyCBAA might be suitable for a long-term application in a complex environment with a more hydrolytically stable acrylamide backbone. Figure 3 also shows that the spacer groups have effects on their nonfouling properties, especially at low ionic strengths and low pH values. For a polycarboxybetaine with a higher spacer length, such as CBAA-3 and CBAA-5, protein adsorption increases at low pH values. The polyCBAA-5 with a pentene

2692

Biomacromolecules, Vol. 9, No. 10, 2008

spacer group has much higher protein adsorption at ionic strengths lower than 150 mM. The introduction of long spacer groups tends to make the CB group more hydrophobic. Due to its antipolyelectrolyte behavior, low ionic strengths will make the CB group even more hydrophobic. As previously indicated, a higher pKa makes polyCBAA-5 much easier to be protonated at lower pH values. All these factors make polyCBAA-5 brushes exhibit less protein-resistance than other polyCBAA polymers with smaller space groups at low pH values and low ionic strengths. At higher pH values and higher ionic strengths, zwitterionic polymer brushes of polyCBAA-5 are highly resistant to protein adsorption (Figure 3d). From the data reported previously from CB-based surfactants, CB-1, CB-2, CB-3, and CB-5 have a pKa of 1.8, 3.25, 3.96, and 5.12 in water at room temperature, respectively.4 This indicates that the polyCBAA with a longer spacer is easier to be protonated at lower pH values. It was found that a CB-copolymer with a propylene spacer has a higher solution viscosity at low pH than a CB-copolymer with a methylene group.17 At low pH, polyCBAA with a longer spacer group such as CBAA-3 or CBAA-5 tends to be more protonated and to have higher protein adsorption. On the other hand, with the increase of the spacer groups, the distance between the charged groups also increases. For zwitterionic (trimethylammonio)alkanoate molecules, (CH2)3+N(CH2)nCOO-, it is estimated that CB-1, CB-2, CB-3, and CB-5 (corresponding to n ) 1, 2, 3, and 5) have an intercharge distance from the quaternary amine to the carboxyl groups of 2.48, 3.81, 4.50, and 6.15 Å in D2O, respectively.36 With the increase of the distance, the dipole moment increases accordingly, which tends to make the molecule more hydrophilic. Because polymer brushes are different from surfactants, their intercharge distances would be different form those values measured from smaller surfactant molecules in dilute solutions, particularly those with larger separations between two charged sites. However, for more rigid CB-1 or CB-2, the difference in intercharge distance between polymer brushes and surfactants should not be large.36 The results from CB-based surfactants showed that CB-2 are more hydrophilic than CB-1.4 Thus, for polyCBAA brushes, polyCBAA-2 may be more hydrophilic than polyCBAA-1, which makes polyCBAA-2 to be potentially more protein-resistant under certain conditions than polyCBAA-1.

4. Conclusions In this work, five polycarboxybetaines, including one CBMA, and four polyCBAA with different spacer groups were grafted from gold surfaces covered with initiators via surface-initiated ATRP. Most of these surfaces exhibit high protein resistance in a wide range of ionic strengths and pH values, and protein adsorption tends to increase at low ionic strengths and low pH value. Fibrinogen adsorption on polycarboxybetaine brushes is much lower than that on zwitterionic SB or PC SAMs at low ionic strengths. Fibrinogen adsorption slightly increases with the decrease of ionic strengths, which can be attributed to the antipolyelectrolyte behavior of zwitterionic polymers. PolyCBMA was found to be protonated at low pH values, leading to net positive charge and protein adsorption. Due to their higher hydrophobicity and higher pKa, polycarboxybetaines with longer space groups, such as polyCBAA-5, show higher protein adsorption than polyCBAA with shorter spacer groups at low ionic strengths and pH values.

Zhang et al.

Acknowledgment. This work has been supported by the Defense Threat Reduction Agency (HDTRA1-07-1-0033) and the National Science Foundation (DMR-0705907). Supporting Information Available. Synthesis of CBAAesters and human fibrinogen adsorption on HS(CH2)15CH3 SAMs used for calculating %ML in Figure 2 are provided. This material is available free of charge via the Internet at http:// pubs.acs.org.

References and Notes (1) Chapman, R. G.; Ostuni, E.; Takayama, S.; Holmlin, R. E.; Yan, L.; Whitesides, G. M. J. Am. Chem. Soc. 2000, 122, 8303–4. (2) Ishihara, K.; Ziats, N. P.; Tierney, B. P.; Nakabayashi, N.; Anderson, J. M. J. Biomed. Mater. Res. 1991, 25, 1397–407. (3) Zhang, Z.; Chao, T.; Chen, S.; Jiang, S. Langmuir 2006, 22, 10072–7. (4) Weers, J. G.; Rathman, J. F.; Axe, F. U.; Crichlow, C. A.; Foland, L. D.; Scheuing, D. R.; Wiersema, R. J.; Zielske, A. G. Langmuir 1991, 7, 854–67. (5) Bonte, N.; Laschewsky, A. Polymer 1996, 37, 2011–9. (6) Zhang, Z.; Chen, S.; Jiang, S. Biomacromolecules 2006, 7, 3311–5. (7) Lowe, A. B.; McCormick, C. L. Chem. ReV. 2002, 102, 4177–89. (8) Huglin, M. B.; Radwan, M. A. Makromol. Chem. 1991, 192, 2433– 45. (9) Huglin, M. B.; Rego, J. M. Macromolecules 1991, 24, 2556–63. (10) Huglin, M. B.; Rego, J. M. Colloid Polym. Sci. 1992, 270, 234–42. (11) Ivanov, I. A.; Kamenova, I. P.; Georgieva, V. T.; Kamenska, E. B.; Georgiev, G. S. Colloid Surf., A 2006, 282&283, 129–33. (12) Huglin, M. B.; Rego, J. M. Polymer 1991, 32, 3354–8. (13) Georgiev, G. S.; Mincheva, Z. P.; Georgieva, V. T. Macromol. Symp. 2001, 164, 301–11. (14) Arotcarena, M.; Heise, B.; Ishaya, S.; Laschewsky, A. J. Am. Chem. Soc. 2002, 124, 3787–93. (15) Azzaroni, O.; Brown, A. A.; Huck, W. T. S. Angew. Chem., Int. Ed. 2006, 45, 1770–4. (16) Kathmann, E. E.; McCormick, C. L. J. Polym. Sci., Part A: Polym. Chem. 1997, 35, 243–53. (17) Kathmann, E. E.; White, L. A.; McCormick, C. L. Polymer 1997, 38, 879–86. (18) Kathmann, E. E.; White, L. A.; McCormick, C. L. Polymer 1997, 38, 871–8. (19) Chevalier, Y.; Storet, Y.; Pourchet, S.; Leperchec, P. Langmuir 1991, 7, 848–53. (20) Zhang, Z.; Chen, S.; Chang, Y.; Jiang, S. J. Phys. Chem. B 2006, 110, 10799–804. (21) Kitano, H.; Tada, S.; Mori, T.; Takaha, K.; Gemmei-Ide, M.; Tanaka, M.; Fukuda, M.; Yokoyama, Y. Langmuir 2005, 21, 11932–40. (22) Ostuni, E.; Chapman, R. G.; Holmlin, R. E.; Takayama, S.; Whitesides, G. M. Langmuir 2001, 17, 5605–20. (23) Holmlin, R. E.; Chen, X.; Chapman, R. G.; Takayama, S.; Whitesides, G. M. Langmuir 2001, 17, 2841–50. (24) Chen, S.; Liu, L.; Jiang, S. Langmuir 2006, 22, 2418–21. (25) Tegoulia, V. A.; Rao, W. S.; Kalambur, A. T.; Rabolt, J. R.; Cooper, S. L. Langmuir 2001, 17, 4396–404. (26) Chen, S.; Zheng, J.; Li, L.; Jiang, S. J. Am. Chem. Soc. 2005, 127, 14473–14478. (27) Feng, W.; Brash, J.; Zhu, S. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 2931–42. (28) Jones, D. M.; Brown, A. A.; Huck, W. T. S. Langmuir 2002, 18, 1265–9. (29) Zhang, Z. Ph.D. Dissertation, University of Washington, Seattle, WA, 2008. (30) Homola, J. Chem. ReV. 2008, 108, 462–93. (31) Homola, J. Surface Plasmon Resonance Based Sensors;SpringerVerlag: Berlin-Heidelberg-New York, 2006; Vol. 4. (32) Bergstroem, K.; Holmberg, K.; Safranj, A.; Hoffman, A. S.; Edgell, M. J.; Kozlowski, A.; Hovanes, B. A.; Harris, J. M. J. Biomed. Mater. Res. 1992, 26, 779–90. (33) Milner, S. T. Science 1991, 251, 905–14. (34) Lee, J. H.; Lee, H. B.; Andrade, J. D. Prog. Polym. Sci. 1995, 20, 1043–79. (35) Ostuni, E.; Chapman, R. G.; Liang, M. N.; Meluleni, G.; Pier, G.; Ingber, D. E.; Whitesides, G. M. Langmuir 2001, 17, 6336–43. (36) Chevalier, Y.; Leperchec, P. J. Phys. Chem. 1990, 94, 1768–74.

BM800407R