Anal. Chem. 2006, 78, 2549-2556
Grafting of Poly(ethylene glycol) onto Poly(acrylic acid)-Coated Glass for a Protein-Resistant Surface Tetsuichi Wazawa,*,† Yoshiko Ishizuka-Katsura,† So Nishikawa,‡ Atsuko Hikikoshi Iwane,§ and Shigeru Aoyama†,|
OMRON-Endowed Chair in Nano Optical Devices, and Soft Biosystem Laboratory, Graduate School of Frontier Biosciences, Osaka University, 1-3 Yamadaoka, Suita, Osaka 565-0871 Japan, CREST: the Formation of Soft Nanomachines, JST, 4-1-8, Honcho, Kawaguchi-shi, Saitama 332-0012 Japan, and Advanced Device Laboratories, OMRON Corporation, 9-1, Kizugawadai, Kizu-cho, Soraku-gun, Kyoto 619-0283, Japan
The surface of solid glass supports for samples in optical microscopy and for biosensors needs to be proteinresistant. A coating of a poly(ethylene glycol) monomethyl ether (mPEG) on the surface of the glass is one promising method for preventing the nonspecific adsorption of proteins. In this study, we have developed a novel technique for achieving an optimal coverage of a glass surface with mPEG to prevent protein adhesion. A clean glass substrate previously treated with (3-aminopropyl)dimethylethoxysilane (APDMES) was treated sequentially with poly(acrylic acid) and subsequently a primary amine derivative of mPEG in the presence of 1-ethyl-3-(3dimethylaminopropyl)carbodiimide. The resultant glass surface was demonstrated to be highly protein-resistant, and the adsorption of bovine serum albumin decreased to only a few percentage points of that on a glass surface treated with APDMES alone. Furthermore, to extend the present method, we also prepared a glass substrate on which biotinylated poly(ethylene glycol) was cografted with mPEG, and biotinylated myosin subfragment-1 (biotinS1) was subsequently immobilized on this substrate by biotin/avidin chemistry. Actin filaments were observed to glide on the biotin-S1-coated glass surface in the presence of ATP, and thus, the method is capable of immobilizing the protein specifically without any loss in its biological function. The solid glass support for samples is a key element in optical microscopy, DNA arrays, and waveguide and quartz crystal microbalance sensors. A receptor biomolecule can be specifically immobilized on the glass support of such an item of equipment to permit the measurement of the interactions between the biomolecule and its ligands. To measure the specific biomolecular interaction accurately, nonspecific adsorption of ligand molecules onto the glass support must be minimized, since such adsorption could cause measurement errors and ghost data. Many surface* To whom correspondence should be addressed. Present address: Laboratory of Physical Chemistry of Biomolecular Functions, Department of Materials Processing, Graduate School of Engineering, Tohoku University. Tel: 81-22-7957313. E-mail:
[email protected]. † OMRON-Endowed Chair in Nano Optical Devices, Osaka University. ‡ CREST: the Formation of Soft Nanomachines. § Soft Biosystem Laboratory, Osaka University. | OMRON Corp. 10.1021/ac052102j CCC: $33.50 Published on Web 03/02/2006
© 2006 American Chemical Society
treatment techniques for minimizing bioadhesion on glass surfaces have been extensively studied. These treatments include covalent grafting of surfactants, the physisorption of serum proteins and synthetic materials, and the dynamic coating of glass surface by adding, for example, tert-amines into the mobile phase for analysis.1 Among these techniques, the covalent grafting of synthetic surfactants may be convenient in terms of the reproducibility and durability of the coated layer. Poly(ethylene glycol) monomethyl ether (mPEG) is one of the most promising materials for minimizing protein adsorption on glass surfaces.2 For this purpose, mPEG has been conjugated to silanol groups on the glass surface by silane reagents containing mPEG moieties3,4 or a combination of functional mPEG derivatives and silane coupling reagents as primers.5,6 mPEG has also been conjugated to the silanol groups by chemisorption of its hydroxyl group on a H2O-plasma-treated glass surface.7 Among these, the (3-aminopropyl)trialkoxysilane-mediated grafting of mPEG onto glass surfaces shows promise as one of the most widely applicable methods. The alkoxy group of the silane reagent forms Si-O-Si linkages with a silanol group on glass surface, whereas its primary amino group serves as a primer for anchoring various types of functional groups such as mPEG, biotins, N-hydroxysuccinimides, maleimides, proteins, or DNA.8 However, the coupling reaction of the silane reagent with the glass surface can often be difficult to control, especially when preparing monomolecular organic layers, because trialkoxysilanes are reactive in the presence of only a trace amount of H2O in solvents and atmosphere, resulting in the hydrolysis of the alkoxy groups and subsequent generation of three reactive silanol groups, leading to self-polymerization.9-11 This chemical complexity of the trialkoxysilanes can cause a lack (1) Doherty, E. A. S.; Meagher, R. J.; Albarghouthi, M. N.; Barron, A. E. Electrophoresis 2003, 22, 34-54. (2) Kingshott, P.; Griesser, H. J. Curr. Opin. Solid State Mater. Sci. 1999, 4, 403-412. (3) Lee, S.-W.; Laibinis, P. E. Biomaterials 1998, 19, 1669-1675. (4) Jo, S.; Park, K. Biomaterials 2000, 21, 605-616. (5) Piehler, J.; Brecht, A.; Valiokas, R.; Liedberg, B.; Gauglitz, G. Biosens. Bioelectron. 2000, 15, 473-481. (6) Rasnik, I.; Myong, S.; Cheng, W.; Lohman, T. M.; Ha, T. J. Mol. Biol. 2004, 336, 395-408. (7) Alcantar, N. A.; Aydil, E. S.; Israelachvili, J. N. J. Biomed. Mater. Res. 2000, 51, 343-351. (8) Weetall, H. H. Appl. Biochem. Biotechnol. 1993, 41, 157-188. (9) Arkles, B. CHEMTECH 1997, 7, 766-778. (10) Plueddemann, E. P. Silane Coupling Agents, 2nd ed.; Plenum Press: New York, 1991. (11) White, L. D.; Tripp, C. P. J. Colloid Interface Sci. 2000, 232, 400-407.
Analytical Chemistry, Vol. 78, No. 8, April 15, 2006 2549
Figure 1. Reaction scheme for grafting of mPEG through PAA. Step 1: The silanol groups on the glass surface are allowed to react with APDMES. Step 2: Amide linkages are formed between PAA and the primary amino groups on the glass surface. Step 3: Amide linkages are formed between mPEG-NH2 and the carboxylic groups of PAA on the glass surface.
of reproducibility in preparing protein-resistant monomolecular organic layers. Furthermore, it has been reported that protein adsorption onto solid support surfaces is minimal when mPEG is grafted onto the surfaces at an optimal surface density.12 Thus, the direct conjugation of mPEG with glass surfaces simply treated with (3-aminopropyl)trialkoxysilane may not be a suitable method for preventing bioadhesion on the surface, because, in addition to the complicated chemistries of (aminopropyl)trialkoxysilanes with the glass surface, it is likely that it will be difficult to control the surface density of silanol group on the glass substrate. In this study, we have developed a novel method for grafting mPEG onto a glass substrate, a widely used material for the solid support of samples in biosensors and optical microscopy by using poly(acrylic acid) (PAA), a polyelectrolyte having one carboxylic group per monomer unit. A monolayer of propylamine is formed on a glass surface using (3-aminopropyl)dimethylethoxysilane (APDMES), which has only one ethoxy group per molecule and, therefore, cannot polymerize. We found that the glass surface became highly protein-resistant when an amine derivative of mPEG (mPEG-NH2) was grafted onto the APDMES-treated glass after the glass substrate had been treated with PAA. Moreover, as an application of the present method, biotinylated-PEG was cografted with mPEG on the glass surface and biotin-conjugated myosin was specifically immobilized by biotin/avidin interaction on the surface to perform an in vitro motility assay. By means of a motility study, we have examined the specific immobilization of a protein and the biocompatibility of the glass surface fabricated by the present technique. EXPERIMENTAL SECTION Materials. Poly(acrylic acid) sodium salts (mean molecular weights, 1200, 8000, and 15 000; code nos. 416010, 416029, and 416037, respectively), bovine serum albumin (BSA) (A-7030), (12) Unsworth, L. D.; Sheardown, H.; Brash, J. L. Langmuir 2005, 21, 10361041.
2550 Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
fibrinogen (F-8630), and lysozyme (L-6876) were purchased from Sigma (St. Louis, MO). Anti-rabbit immunoglobulin (IgG) (code no. 31216) and NeutrAvidin (code no. 31000) were obtained from Pierce (Rockford, IL). Poly(acrylic acid) (molecular weight, 25 000) was purchased from Wako Pure Chemicals (Osaka, Japan). 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide, hydrochloride salt (EDC) (code no. W001) was obtained from Dojindo Laboratories (Kumamoto, Japan) or Tokyo Chemical Industry (D1601; Tokyo, Japan). mPEG-NH2 and mPEG-COOH (mean molecular weight, 750; code nos. 12-750-2 and 12-750-3, respectively) were obtained from Rapp Polymere (Tu¨bingen, Germany). Biotin-PEG-NH2 (code no. 12141-0895) was obtained from PolyPure (Oslo, Norway). 2-Morpholinoethanesulfonic acid, monohydrate (MES) (code no. GB12) and 2-[4-(2-hydroxyethyl)1-piperazinyl]ethanesulfonic acid (HEPES) (code no. GB10) were obtained from Dojindo. APDMES (code no. SIA0603.0) was obtained from Gelest (Morrisville, PA). Cy3 monofunctional reactive dye (Cy3-OSu) (PA23001) was purchased from Amersham Biosciences (Buckinghamshire, U.K.). Grafting of mPEG onto Glass Substrates. The method for grafting mPEG onto glass surface is shown schematically in Figure 1. Borosilicate glass slides (S-2111, Matsunami Glass, Osaka, Japan) were ultrasonicated in an aqueous solution containing 10% (v/v) of a neutral detergent (Contami-Non N, Wako Pure Chemicals) for 20 min, rinsed exhaustively with ultrapure water, ultrasonicated again in ultrapure water for 20 min, rinsed with water, and dried in a vacuum at 150 °C for 2 h in a vacuum oven (DN-30S, Sato Vacuum Machinery, Tokyo, Japan). The slides were subsequently subjected to O2-plasma treatment with a 40-kHz plasma equipment (Pico UHP; Diener Electronic, Nagold, Germany) for 5 min at maximal generator power. The plasma-treated slides were immediately soaked in a solution of 1% (v/v) APDMES and 1.5% (v/v) acetic acid dissolved in a solvent containing 90% (v/v) ethanol (∞Pure Grade; Wako Pure Chemicals) and 10% (v/ v) H2O at 40 °C for 3 h, according to the method
reported by Shriver-lake13 with slight modifications. The slides were rinsed with ultrapure water and dried in a vacuum at 110 °C for 40 min to give the APDMES glasses. After cooling at room temperature, these were immersed in a 50 mM MES-NaOH buffer, pH 6.0, containing EDC and various concentrations (0.001-1 mM) of poly(acrylic acid) of molecular weights of 1200, 8000, 15 000, or 25 000 for 3 h at 40 °C. In the reaction, a 2-fold molar excess of EDC, based on the carboxylic group content of the PAA, was added to the reaction mixture. After the reaction, the slides were rinsed with ultrapure water, giving APDMES/ PAA glasses. The APDMES/PAA glasses were then treated with 1 mM mPEG-NH2 in the presence of 1 mM EDC dissolved in the 50 mM MES-NaOH buffer, pH 6.0, at 40 °C for 2 h, giving APDMES/PAA/mPEG glasses after rinsing with pure water. All the glass slides prepared by this method were stored in ultrapure water at 4 °C until they were required. For the APDMES/mPEG glass, the APDMES glass was immersed in an aqueous solution containing 1 mM mPEG-COOH, 1 mM EDC, and 50 mM of MES-NaOH, pH 6.0, causing a condensation reaction of the primary amine on the APDMES glass surface with the carboxylic group of mPEG-COOH. After incubation for 2 h at 40 °C, the glass slides were exhaustively rinsed with ultrapure water. To prepare a biotin-PEG grafted glass substrate (APDMES/ PAA/mPEG/biotin-PEG glass), the APDMES/PAA glass prepared with 0.1 mM PAA of molecular weight 25 000 (see above) was treated with an aqueous solution containing 1 mM mPEGNH2, 0.1 mM (or 0.01 mM) biotin-PEG-NH2, 1 mM EDC, and 50 mM MES-NaOH, pH 6.0, at 40 °C for 2 h. The glass was rinsed with ultrapure water. Fluorescence Modification of Proteins. BSA, lysozyme, fibrinogen, and NeutrAvidin were fluorescently labeled with Cy3OSu. Briefly, the proteins dissolved in buffers of pH 7.8-9.0 were mixed with a severalfold molar excess of Cy3-OSu to react with the proteins at room temperature. The mixtures were applied to NAP-10 columns (Amersham Biosciences) equilibrated with a 20 mM HEPES-NaOH buffer, pH 7.0, or an aqueous solution containing 0.15 M NaCl and 20 mM sodium phosphate, pH 7.2, and the fractions containing the proteins were dialyzed against the same buffer overnight at 4 °C. The dialyzed protein solutions were concentrated with YM centrifuge filters (Milli Pore, Billerica, MA) and then ultracentrifuged at 100 000 rpm for 10 min at 4 °C (rotor, TLA-120.2; ultracentrifuge, Optima MAX-E; BeckmanCoulter, Fullerton, CA). The supernatants were harvested, and the concentrations of the Cy3 dye and proteins were measured by absorbance spectrophotometry or by the Lowry method.14 We took the extinction coefficients as 150 000 and 12 000 cm-1 M-1 at 552 and 280 nm, respectively, for Cy3, 0.67 cm-1 mg-1 mL at 280 nm for BSA, and 2.6 cm-1 mg-1 mL at 280 nm for lysozyme, according to the manufacturers’ instructions. Measurement of Protein Adsorption. A prepared glass slide was dried under class 1000 clean air, and a sample cell was assembled by placing a clean coverslip (18 × 18 mm2, No. 1S; Matsunami Glass) on the substrate separated by two thin sheets of Teflon (∼2 × 20 mm2; thickness, 50 µm) (Aram, Osaka, Japan). (13) Shriver-lake, L. C. In Immobilized Biomolecules in Analysis; Cass, T., Ligler, F. S., Eds.; Oxford University Press: New York, 1998; pp 1-14. (14) Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. J. Biol. Chem. 1951, 193, 265-275.
The inner space of the sample cell was filled with an aqueous solution containing 0.15 M NaCl and 20 mM HEPES-NaOH buffer (pH 7.0) (150-S buffer), and then 15 µL of a solution containing the fluorescently modified protein was allowed to flow into the cell twice to cause adsorption of the protein onto the glass surface. The cell was then incubated at room temperature for 2 min. The protein solutions used here were diluted to a concentration of 1 mg/mL for the measurement with the 150-S buffer, except for lysozyme, which was diluted to 0.3 mg/mL with buffer. The unadsorbed protein was flushed with 75 µL of 150-S buffer, and the aqueous medium in the sample cell was squeezed out after carefully removing the thin Teflon sheets from the sample cell assembly. The medium between the glass slide and the coverslip was sealed with a nail polish. The adsorption of fluorescently modified proteins onto the surface of glass slides was measured using a total internal reflection fluorescence microscope (TIRFM). The instrumentation of the TIRFM was reported previously.15,16 In the TIRFM, the fluorescence-excitation laser beam (wavelength, 532 nm) was incident on the glass/water interface, generating an evanescent wave, and thus, the fluorescent dyes adjacent to the interface were illuminated selectively. The fluorescence from the fluorescent protein adsorbed on the substrate was collected with a microscope objective (PlanApo/TIRFM, ×60, NA 1.45; Olympus, Tokyo, Japan) and the fluorescence was measured with a cooled chargecoupled device (CCD) camera (DU434-BV; Andor Technology, Belfast, Northern Ireland). X-ray Photoelectron Spectroscopy. The APDMES/PAA glass substrates that were prepared with 0.001-1 mM PAA (molecular weight, 25 000) were analyzed by X-ray photoelectron spectroscopy (XPS). The measurement was performed with an ESCA-5600Ci spectrometer (ULVAC-PHI, Kanagawa, Japan) with a monochromated Al source at 200 W (1486.6 eV). Emitted photoelectrons were detected at a takeoff angle of 60°. The peaks for C (1s), N (1s), O (1s), and Si (2p) were measured and the areas for the peaks were calculated from the spectra. In Vitro Motility Assay of Myosin Subfragment-1. Biotinylated myosin subfragment-1 (biotin-S1) was prepared according to Iwane et al.17 Fluorescent actin filament was prepared according to Harada et al.18 Briefly, actin (so-called G-actin) dissolved in an aqueous solution containing 2 mM HEPES-KOH, pH 7.8, and 0.2 mM ATP was polymerized by the addition of a final concentration of 0.1 M KCl at room temperature. The polymerized actin (F-actin) was treated with tetramethylrhodamine-B phalloidin (code no. 77418; Fluka, St. Louis, MO) to obtain TMRPH-actin filament. As in the protein adsorption experiment described above, a sample cell was assembled on a glass substrate (APDMES/PAA/ mPEG-, APDMES/PAA/mPEG/biotin-PEG-, or nitrocellulosecoated glass) with two strips of Teflon sheet and a coverslip: the nitrocellulose-coated glass slide was prepared according to Uyeda (15) Wazawa, T.; Ishii, Y.; Funatsu, T.; Yanagida, T. Biophys. J. 2000, 78, 15611569. (16) Wazawa, T.; Ueda, M. Adv. Biochem. Bioeng. Biotechnol. 2005, 95, 77106. (17) Iwane, A. H.; Kitamura, K.; Tokunaga, M.; Yanagida, T. Biochem. Biophys. Res. Commun. 1997, 230, 76-80. (18) Harada, Y.; Sakurada, K.; Aoki, T.; Thomas, D. D.; Yanagida, T. J. Mol. Biol. 1990, 216, 49-68.
Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
2551
Figure 2. Comparison of the fluorescence from Cy3-labeled proteins adsorbed on various glass surfaces. (a) Fluorescence micrographs of Cy3-BSA adsorbed on the APDMES, APDMES/PAA, APDMES/PAA/mPEG, APDMES/mPEG, and clean bare glasses. For comparison, all the micrographs are shown so that the white and black levels are the same. The glasses were contacted with 1 mg/mL Cy3-BSA solution for 2 min, and the unadsorbed Cy3-BSA was flushed with an aqueous buffer containing 0.15 M NaCl and 20 mM HEPES-NaOH, pH 7.0. Excitation wavelength, 532 nm. (b) Comparison of the protein adsorption for various types of glass substrates and proteins. The relative fluorescence intensity was calculated as F/FAPDMES,BSA × 100, where F is the fluorescence intensity of a Cy3-labeled protein adsorbed on a glass surface and FAPDMES,BSA is that of Cy3-BSA adsorbed on the APDMES glass. For adsorption, the glasses were contacted with 1 mg/mL BSA, fibrinogen, and IgG and 0.3 mg/mL lysozyme.
et al.19 The sample cell was first filled with an aqueous solution containing 25 mM KCl, 20 mM HEPES-KOH, pH 7.8, and 5 mM MgCl2 (25-P buffer) and then 1 mg/mL NeutrAvidin solution was injected into the cell. After incubation for 5 min, the unadsorbed NeutrAvidin was washed out with 25-P buffer. Biotin-S1 was subsequently allowed to flow into the sample cell to react with the NeutrAvidin adsorbed on the glass surface, and then the unadsorbed biotin-S1 was washed out with 25-P buffer. TMRPHactin filament suspended in an aqueous solution containing 25-P buffer with 2 mM ATP was subsequently injected into the sample cell, and the motility of the actin filament was observed using an inverted epifluorescence microscope (IX-71; Olympus, Tokyo, Japan) at 20 °C. The microscope image was recorded with an electron-multiplying CCD camera (DV887-ECS-DV, Andor Technology) at a video rate of 10 Hz. RESULTS AND DISCUSSION Surface Treatment Dependence of Protein Adsorption. Cy3-BSA was allowed to adsorb onto five types of glass substrate. BSA was chosen to investigate protein adsorption on the substrates, because BSA is well known for its strong adsorption property onto solid supports and has been widely used as a reference protein in adsorption studies. The substrates used for this experiment were APDMES, APDMES/PAA, APDMES/PAA/ mPEG, and APDMES/mPEG glasses, as well as a clean bare glass slide. After the glass plates had been allowed to contact a 1 mg/ mL solution of Cy3-BSA, the unadsorbed protein was washed out with 150-S buffer and the Cy3-BSA adsorbed on the glass slides was observed by using the TIRFM (see Experimental (19) Uyeda, T. Q.; Kron, S. J.; Spudich, J. A. J. Mol. Biol. 1990, 214, 699-710.
2552
Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
Section). The fluorescence micrographs taken from the glasses are shown in Figure 2a. It was observed that the specimen area illuminated by a laser beam emitted fluorescence arising from the Cy3 fluorophore conjugated to BSA and the fluorescence intensity was dependent on the type of glass substrate, indicating that the fluorescence intensity must be related to the amount of adsorbed Cy3-BSA. Thus, the fluorescence intensity was taken as a measure of protein adsorption. In addition, the APDMES glass was taken as a reference substrate, because (3-aminopropyl)alkoxysilane-treated glass substrates are often starting materials for the fabrication of functionalized glass supports. Therefore, the fluorescence intensities of Cy3-BSA adsorbed on the glass slides were normalized by that on the APDMES glass. In Figure 2b, the fluorescence intensities of adsorbed Cy3BSA are shown for the five types of glass slide. The results in the top panel show that the fluorescence intensity of adsorbed Cy3BSA decreased to 4 and 1.4% for the APDMES/PAA and APDMES/PAA/mPEG glasses, respectively, relative to the APDMES glass. Thus, it is shown that a coating of PAA on the APDMES glass surface exhibits protein resistance, but further grafting with mPEG is necessary. Moreover, note that the APDMES/mPEG glass in which mPEG-COOH was directly conjugated to the propylamine on the APDMES glass had no protein resistance, and in fact, the relative fluorescence intensity of Cy3-BSA adsorbed on the substrate was 94%. It has been postulated that a high surface density of grafted PEG is necessary for protein resistance,2,12 and thus, on the APDMES/mPEG glass, the surface density of grafted mPEG chain may not be high enough to prevent protein adhesion. In addition, the adsorption of Cy3-BSA onto a clean bare glass surface was examined. The amount of adsorbed
protein on the bare glass can vary depending on its cleanness, for example, the duration of its exposure to ambient atmosphere after O2-plasma treatment, the cleanness of the surrounding air, and so forth, and thus, the protein adsorption step was carried out as soon as possible after the O2-plasma treatment to maximize the reproducibility of the experiment. The relative fluorescence intensity of Cy3-BSA adsorbed on the bare glass was 48% and not significantly different from that of APDMES glass. Hen egg lysozyme, a highly basic protein with an isoelectric point of 11.1,20 was also investigated for the adsorption study. Like BSA, Cy3-lysozyme was allowed to adsorb on the glass substrates, and the fluorescence intensity of Cy3-lysozyme adsorbed on the glasses was measured by using the TIRFM. It is noteworthy that the fluorescence intensity of adsorbed Cy3-lysozyme on the APDMES/PAA was as high as 35% relative to that of Cy3lysozyme adsorbed on the APDMES (Figure 2b). These results indicate that positively charged lysozyme (at pH 7.0) would interact electrostatically with negatively charged carboxylic groups of PAA, causing physisorption of lysozyme onto the APDMES/ PAA glass, in contrast to BSA, which has an isoelectric point of 4.85.20 Thus, these differences in the adsorption properties should further rationalize the present chemical modification processes for glass surfaces. IgG and fibrinogen were also examined for the adsorption experiment. IgG was chosen because it is widely used for and immobilized on biosensors and thereby the nonspecific adsorption on the sensor surface must be a critical issue. In addition, fibrinogen was chosen for another model protein, because it is well known to be a sticky protein with a strong tendency to adsorb onto solid surfaces.21 Both proteins exhibited adsorption results similar to that of BSA (Figure 2b), and thus, the results demonstrate that the APDMES/PAA/mPEG treatment markedly prevents protein adhesion. PAA Conjugation-Condition Dependence in Preventing Protein Adsorption. To minimize protein adhesion on the APDMES/PAA/mPEG glass surface, the conditions for the condensation reaction between PAA and the APDMES glass had to be examined. On one hand, the reactions of the APDMES with the clean bare glass and mPEG-NH2 with the APDMES/PAA glass should be performed at as a high efficiency as possible, since mPEG should be grafted in a high surface density. On the other hand, the molecular mass and concentration of PAA at reaction could be adjusted to achieve an optimal coverage of the substrate surface with PAA (Figure 1). The APDMES glasses were allowed to react with PAAs of molecular weight of 1200, 8000, 15 000, and 25 000 and at concentrations of 0.001, 0.01, 0.1, and 1 mM. The resulting APDMES/PAA glasses were treated with mPEG-NH2, and thus, 16 different samples of APDMES/PAA/mPEG glass were obtained. Figure 3 shows the fluorescence intensities of Cy3-BSA adsorbed on the glass slides relative to that onto the APDMES glass. The results show that BSA adsorption decreased as the molecular weight of PAA increased and was minimal when the APDMES/PAA/mPEG glass was prepared at a concentration of 0.01 or 0.1 mM PAA for each molecular weight of PAA. Therefore, in this study, a concentration of 0.1 mM PAA with a (20) Sapsford, K. E.; Ligler, F. S. Biosens. Bioelectron. 2004, 19, 1045-1055. (21) Hemmersam, A. G.; Foss, M.; Chevallier, J.; Besenbacher, F. Colloids Surf., B 2005, 43, 208-215.
Figure 3. Comparison of the fluorescence intensity of Cy3-BSA adsorbed on the APDMES/PAA/mPEG glasses that were prepared under different conditions. The APDMES glasses were treated with PAAs of molecular weight of 1200, 8000, 15 000, and 25 000 at concentrations of 0.001, 0.01, 0.1, and 1 mM in the presence of EDC. The relative fluorescence intensity was calculated as F/FAPDMES × 100, where F is the fluorescence intensity of Cy3-BSA adsorbed on the APDMES/PAA/mPEG glass surface and FAPDMES is that adsorbed on the APDMES glass. In the adsorption experiment, the glasses were contacted with 1 mg/mL Cy3-BSA. Table 1. Surface Elemental Composition Relative to Silicon Analyzed by XPS
bare glass APDMES APDMES/PAA (0.001 mM)a APDMES/PAA (0.01 mM)a APDMES/PAA (0.1 mM)a APDMES/PAA (1 mM)a a
C/Si
N/Si
O/Si
0.10 0.36 0.37 0.95 1.06 1.21
0.00 0.03 0.03 0.07 0.10 0.12
2.19 2.09 2.09 2.17 2.06 2.04
The glasses were prepared with PAA of molecular weight of 25 000.
molecular weight of 25 000 was chosen as the reaction conditions for preparing the APDMES/PAA/mPEG glass. In addition, the elemental surface compositions of the APDMES, APDMES/PAA, and bare glasses were analyzed by XPS, and the composition values relative to silicon are summarized in Table 1. The results show that the C/Si ratio for the APDMES/ PAA (0.001 mM) glass was very similar to that for the APDMES glass. In contrast, the APDMES/PAA glasses that were prepared with 0.01-1 mM PAA exhibited severalfold higher C/Si values, and this ratio slightly increased with the reaction concentration of PAA. The N/Si ratio also increased with the reaction concentration of PAA (Table 1), and this may be responsible for some residual products formed from the reaction between EDC and PAA. Thus, these results suggest that little coverage of PAA is formed on the APDMES glass when the glass is treated with 0.001 mM PAA, whereas a much higher coverage of PAA is obtained Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
2553
Figure 4. Concentration profile of protein adsorption. Upper panel: The fluorescence intensity of Cy3-BSA adsorbed on the APDMES and APDMES/PAA/mPEG glasses is plotted as a function of the concentration of Cy3-BSA that was in contact with the glass surfaces. The intensities are relative values (%) compared with that of 1 mg/ mL BSA on APDMES glass. Lower panel: The ratio of the fluorescence intensity for the APDMES/PAA/mPEG glass to that for APDMES glass is plotted as a function of the Cy3-BSA concentration.
with g0.01 mM PAA. However, when the APDMES glass was treated with PAA at a concentration as high as 1 mM, the large number of PAA molecules presumably competed for the primary amino group on APDMES glass, resulting in a defective PAA coverage on the glass surface. Thus, the PAA coverage on the APDMES glass would be maximal when the glass is treated with PAA at a concentration of 0.01-0.1 mM; this would permit the optimal grafting of mPEG-NH2 onto the substrate. Protein Concentration Dependence of Protein Adhesion. The degree of protein adsorption was also examined as a function of the protein concentration. Cy3-BSA (0.001-1 mg/mL) was allowed to adsorb onto the APDMES and APDMES/PAA/mPEG glasses for 2 min, and the fluorescence intensities of Cy3-BSA adsorbed on the glasses were measured after the unadsorbed protein was washed out with the flow of 150-S buffer. The fluorescence of adsorbed Cy3-BSA increased with its concentration for both APDMES and APDMES/PAA/mPEG glasses, but the fluorescence intensity for the APDMES glass was much higher than for the APDMES/PAA/mPEG glass over the concentration range (top panel, Figure 4). As a measure of protein adsorption, we calculated the ratio of the fluorescence intensity of Cy3-BSA adsorbed on the APDMES/PAA/mPEG glass to that on the APDMES glass for each Cy3-BSA concentration. The results demonstrate that the ratio of the fluorescence is unchanged and the APDMES/PAA/mPEG glass surface shows only a few percentage points of protein adhesion ratio over a wide range of protein concentrations (0.001-1 mM Cy3-BSA) (bottom panel, Figure 4). Although the experimental process for measuring the BSA adsorption measurement may be too crude to permit discussion of the adsorption isotherm, the results suggest that the BSA adsorption onto both types of substrates may follow a Langmuir mechanism over a wide range of protein concentrations. 2554
Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
In Vitro Motility Assay of Myosin Subfragment-1. To demonstrate the immobilization of a specific protein as an application of the present technique, biotinylated PEG and mPEG were cografted onto the APDMES/PAA glasses. The APDMES/ PAA glasses were treated with the mixtures of mPEG-NH2 and biotin-PEG-NH2 at molar ratios of 10:1 and 100:1 to give APDMES/PAA/mPEG/biotin-PEG glasses (see Experimental Section). The resultant glasses have been shown to have a high specificity for immobilizing NeutrAvidin, a biotin binding protein. Cy3-NeutrAvidin (0.1 mg/mL) was allowed to adsorb onto the APDMES/PAA/mPEG and APDMES/PAA/mPEG/biotin-PEG glasses, and the fluorescence of the adsorbed Cy3-NeutrAvidin was measured by the same procedure as used in the protein adsorption experiment described as above. Figure 5a shows the fluorescence intensities of Cy3-NeutrAvidin adsorbed on the glasses. The data demonstrate that the fluorescence of immobilized Cy3-NeutrAvidin on the APDMES/PAA/mPEG glass is 200-fold lower than those of the APDMES/PAA/mPEG/biotinPEG glasses, indicating highly specific binding of NeutrAvidin to the biotin moiety grafted on the APDMES/PAA/mPEG/biotinPEG glass substrates. Therefore, the present method was expanded to permit specific immobilization of a biotinlyated protein by means of biotin/avidin chemistry. An in vitro motility assay of myosin was performed with the APDMES/PAA/mPEG/biotin-PEG glass. Biotinylated myosin subfragment-1 was immobilized onto the APDMES/PAA/mPEG/ biotin-PEG glass (mPEG-NH2/biotin-PEG-NH2 ratio at reaction, 10:1) by using biotin/avidin chemistry within the sample cell, which was assembled on the glass slide (Figure 5b). TMRPHactin filaments suspended in an assay solution (25-P buffer with 2 mM ATP) were allowed to flow into the sample cell. Epifluorescence microscopy showed that TMRPH-actin filaments were attached onto the surface of the APDMES/PAA/mPEG/biotinPEG glass on which biotin-S1 was immobilized (Figure 5c). In contrast, TMRPH-actin filaments did not attach onto the surface of the APDMES/PAA/mPEG glass that was treated with NeutrAvidin and biotin-S1, indicating that biotin-S1 was not immobilized onto this glass surface (data not shown). Furthermore, we observed the gliding of the TMRPH-actin filaments on the biotin-S1-treated APDMES/PAA/mPEG/biotin-PEG glass. Thus, the results show that biotin-S1 was successfully immobilized on the APDMES/PAA/mPEG/biotin-PEG glass surface through biotin/avidin chemistry without any loss in the motor function of myosin and actin, as was expected. By video analysis, the sliding velocity of actin filaments was measured to be 1.7 ( 0.5 µm/s (mean ( SD; n ) 50). For comparison, biotin-S1 was immobilized on a nitrocellulose-coated glass, a conventional method for performing in vitro motility assays for myosin.19 The sliding speed of TMRPH-actin filament on the glass was measured to be 1.2 ( 0.3 µm/s (n ) 50), significantly slower than that on the APDMES/ PAA/mPEG/biotin-PEG glass (Figure 5d). The faster sliding speed for the APDMES/PAA/mPEG/biotin-PEG glass suggests that biotin-S1 would be more stable on this glass surface than on the nitrocellulose-coated glass so that many fewer myosin S1 molecules on the APDMES/PAA/mPEG/biotin-PEG glass are denatured upon immobilization, leading to less friction for the sliding of actin filament.
Figure 5. In vitro motility assay of myosin subfragment-1 on the APDMES/PAA/mPEG/biotin-PEG glass surface. (a) The fluorescence intensities of Cy3-NeutrAvidin bound to the surfaces of the APDMES/PAA/mPEG and the APDMES/PAA/mPEG/biotin-PEG glasses. To prepare the APDMES/PAA/mPEG/biotin-PEG glasses, the APDMES was treated with 0.1 mM PAA (molecular weight, 25 000) to obtain the APDMES/PAA glasses, and they were treated with the mixtures of mPEG-NH2 and biotin-PEG-NH2 at molar ratios of 100:1 and 10:1. The glasses were allowed to contact 0.1 mg/mL Cy3-NeutrAvidin for 5 min and free Cy3-NeutrAvidin was washed out with 150-S buffer. The fluorescence intensities are values (%) relative to that of Cy3-NeutrAvidin bound to the APDMES/PAA/mPEG/biotin-PEG glass that was prepared with a mPEG-NH2/biotin-PEG-NH2 ratio of 10:1. (b) Schematic drawing of in vitro motility assay of biotin-S1. (c) Fluorescence micrograph of TMRPHactin filaments attached to biotin-S1-treated APDMES/PAA/mPEG/biotin-PEG glass (mPEG-NH2/biotin-PEG-NH2 ratio, 10:1). (d) Histograms of sliding velocity of TMRPH-actin filament on the biotin-S1-treated APDMES/PAA/mPEG/biotin-PEG and nitrocellulose-coated glasses.
Versatility of mPEG Grafting Using Poly(acrylic acid). Substantial adsorption of BSA was observed on the APDMES/ mPEG glass, whereas much less adsorption occurred on the APDMES/PAA/mPEG glass (Figure 2). This suggests that the surface density of mPEG on the APDMES/mPEG glass is too low to prevent BSA adsorption onto the surface, whereas that on the APDMES/PAA/mPEG glass is high enough to reduce protein adsorption. Hence, many more binding sites for the mPEG derivatives are available on the APDMES/PAA glass than on the
APDMES glass as a result of the PAA treatment forming a network of polycarboxylate backbones on the surface. It is also suggested that the surface density of carboxylic groups of PAA remaining after the condensation reaction with APDMES glass could be optimized by adjusting the molecular weight and reaction concentration of PAA. Thus, the use of PAA in our method is a versatile process for grafting mPEG onto glass surfaces. Moreover, because the chemical procedure is independent of the nature of the solid support if a primary amino group is introduced or present Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
2555
on the surface, the technique developed in this study could be applicable to other solid surfaces, such as plastics, metals, or metal oxides. CONCLUSIONS In this study, we have developed a novel method for grafting mPEG onto glass surfaces treated sequentially with APDMES and PAA. The APDMES/PAA/mPEG glass substrates prepared by this method have been shown to be highly protein-resistant. This property could be due to the conjugation of mPEG-NH2 with PAA that has previously reacted with primary amino groups on the APDMES glass, possibly giving an optimal coverage of mPEG. The present method was also shown to be capable of functionalizing the glass surface with biotin-PEG chains, permitting specific immobilization of biotinylated proteins by biotin/avidin chemistry. In fact, biotin-S1 was successfully immobilized on the APDMES/ PAA/mPEG/biotin-PEG glass without any loss in motility. The present method will be useful for preparing highly sensitive
2556
Analytical Chemistry, Vol. 78, No. 8, April 15, 2006
biosensor surfaces and microscopy glass slides that require very low protein adsorption. ACKNOWLEDGMENT The authors appreciate the extensive support in the myosin motility study from Prof. Toshio Yanagida at Graduate School of Frontier Biosciences, Osaka University. We thank Ms. Kyoko Yamauchi and Mr. Takashi Morimoto at the Nippon Sheet Glass Techno-Research Co. (Hyogo, Japan) for the XPS analysis. We also thank Dr. Yuko Takeuchi at CREST, JST for critical reading of the manuscript. This work was partly supported by a grant from the Inter-Field Fusion Research Project under Osaka University 21st Century COE Program, Dynamics of Biological Systems, and grant 16201030 from the Japan Society for the Promotion of Science. Received for review November 30, 2005. Accepted February 3, 2006. AC052102J