Langmuir 2004, 20, 9729-9735
9729
Influence of Surfactants and Antibody Immobilization Strategy on Reducing Nonspecific Protein Interactions for Molecular Recognition Force Microscopy Kathryn L. Brogan, Jae Ho Shin, and Mark H. Schoenfisch* Department of Chemistry, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-3290 Received June 24, 2004 Specific and nonspecific interactions between antibody-modified probes and substrate-immobilized proteins were monitored by atomic force microscopy (AFM). Probes were modified with anti-ovalbumin IgG antibodies immobilized in either an oriented or a random manner. The oriented immobilization of whole IgG was accomplished through the use of Protein A, and random immobilization was carried out with glutaraldehyde. Nonspecific interactions may lead to false detection of antibody-antigen binding events even when the antigen binding sites are properly positioned by an oriented immobilization strategy. Thus, nonionic and zwitterionic surfactants, including Tween 20, Tween 80, Triton X-100, and CHAPS, were evaluated to determine if nonspecific binding events could be reduced without compromising the desired specific antibody-antigen binding. Enzyme-linked immunosorbent assay and surface plasmon resonance assays were also employed to study antibody-antigen binding as a function of immobilization strategy and surfactant concentration. The data from these studies indicate that Protein A can be used to immobilize whole IgG onto AFM probes for force measurement experiments and that a surfactant is useful for improving the selectivity for such measurements.
Introduction Atomic force microscopy (AFM) has evolved into a useful analytical tool for studying antibody-antigen binding interactions.1-3 Intermolecular forces down to the piconewton level have been detected by measuring the binding-unbinding forces between an AFM probe (tip) modified with the protein of interest and a complementary molecule immobilized on a substrate.2 Combining the ability to measure such forces with the surface imaging capabilities of AFM has led to the development of a method for mapping adhesion forces between AFM tips and substrate-adsorbed molecules.1,3-5 Specific adhesion events in force measurements are often obscured by nonspecific interactions between the tip and substrate. While discrete nonspecific proteinprotein interactions (e.g., hydrophobic, electrostatic, van der Waals) are weak with respect to specific antibodyantigen binding, the accumulation of nonspecific interactions between proteins on the tip and proteins adsorbed to a substrate is often similar in magnitude to the forces required to break discrete antibody-antigen bonds.6 To date, few reports have specifically addressed the problem of nonspecific interactions in force measurement experiments. Poisson statistics have been used to calculate both single-molecule bond-rupture forces and the magnitude of nonspecific pull-off forces for chemical interactions * Author to whom correspondence should be addressed. E-mail:
[email protected]. (1) Raab, A.; Han, W.; Badt, D.; Smith-Gill, S. J.; Lindsay, S. M.; Schindler, H.; Hinterdorfer, P. Nat. Biotechnol. 1999, 17, 902-905. (2) Allen, S.; Chen, X.; Davies, J.; Davies, M. C.; Dawkes, A. C.; Edwards, J. C.; Roberts, C. J.; Sefton, J.; Tendler, S. J. B.; Williams, P. M. Biochemistry 1997, 36, 7457-7463. (3) Dammer, U.; Hegner, M.; Anselmetti, D.; Wagner, P.; Dreier, M.; Huber, W.; Guntherodt, H.-J. Biophys. J. 1996, 70, 2437-2441. (4) Radmacher, M.; Cleveland, J. P.; Fritz, M.; Hansma, H. G.; Hansma, P. K. Biophys. J. 1994, 66, 2159-2165. (5) Ludwig, M. D. W.; Gaub, H. E. Biophys. J. 1997, 72, 445-448. (6) Willemsen, O. H.; Cambi, A.; Greve, J.; deGrooth, B. G.; Figdor, C. G. Biophys. J. 2000, 79, 3267-3281.
between organosilane-functionalized tips and surfaces.7 Unfortunately, the application of Poisson statistics to antibody-antigen systems is challenging due to the potential for multiple functional-group interactions within the antigen binding sites. Indeed, reports of using Poisson statistics to discern specific and nonspecific forces in biological systems have been limited to high-affinity monoclonal antibody-DNA aptamers8 and to the biotinavidin9 system. Imaging conditions have also been modified to address nonspecific interactions. Using an 8-nm flexible poly(ethylene glycol) (PEG) linker10 and a low force set point, Willemsen et al. conducted adhesion force imaging experiments in the repulsive regime to demonstrate the ability to detect specific binding events without interference from nonspecific interactions.11 Although promising, such sensitive force measurements required the use of a noncommercial AFM system.11 Surfactants have been widely employed in traditional immunoassay formats, including radiolabeling,12 solidphase immunofiltration,13 and enzyme-linked immunosorbent assays,14-16 to reduce background signal caused (7) Lo, Y.-S.; Simons, J.; Beebe, T. P. J. Phys. Chem. B 2002, 106, 9847-9852. (8) Jiang, Y.; Zhu, C.; Ling, L.; Wan, L.; Fang, X.; Bai, C. Anal. Chem. 2003, 75, 2112-2116. (9) Lo, Y.-S.; Huefner, N.; Chan, W.; Stevens, F.; Harris, J. M.; Beebe, T. P. Langmuir 1999, 15, 1373-1382. (10) Hinterdorfer, P.; Baumgartner, W.; Gruber, H. J.; Schilcher, K.; Schindler, H. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 3477-3481. (11) Willemsen, O. H.; Snel, M. M. E.; Kuipers, L.; Figdor, C. G.; Greve, J.; de Grooth, B. D. Biophys. J. 1999, 76, 716-724. (12) Vandenberg, E. T.; Krull, U. J. J. Biochem. Biophys. Methods 1991, 22, 269-277. (13) IJsselmuiden, O. E.; Herbrink, P.; Meddens, M. J. M.; Tank, B.; Stolz, E.; Eijk, R. V. W. V. J. Immunol. Methods 1989, 119, 35-43. (14) McCabe, J. P.; Fletcher, S. M.; Jones, M. N. J. Immunol. Methods 1988, 108, 129-135. (15) Rebeski, D. E.; Winger, E. M.; Shin, Y.-K.; Lelenta, M.; Robinson, M. M.; Varecka, R.; Crowther, J. R. J. Immunol. Methods 1999, 226, 85-92. (16) Kenna, J. G.; Major, G. N.; Williams, R. S. J. Immunol. Methods 1985, 85, 409-419.
10.1021/la048437y CCC: $27.50 © 2004 American Chemical Society Published on Web 09/29/2004
9730
Langmuir, Vol. 20, No. 22, 2004
by nonspecific protein-protein adsorption.12,16-19 Prior work suggests that nonspecific adsorption of proteinaceous material via primarily hydrophobic interactions is greatly reduced using nonionic surfactants.20 Indeed, several groups have reported that the addition of Tween 20, Triton X-100, and Tween 80 to antibody-antigen precipitation studies greatly reduced nonspecific adhesion.15,16,21-23 Such surfactants have also been shown to reduce nonspecific protein-protein interactions in other immunoassay formats. To date, the utility of surfactants for reducing nonspecific adhesion in the measurement of antibodyantigen interactions using AFM has not been studied. The antibody immobilization strategy is an equally critical parameter for measuring specific antibodyantigen adhesion forces. The orientation of immobilized antibodies has been shown to influence antigen binding for various immunoassays, including immunoaffinity chromatography,24,25 surface plasmon resonance spectroscopy26 and enzyme-linked immunoassays.27,28 The orientation of immobilized antibodies is also important for the measurement of specific adhesion forces with AFM.29,30 Probes coated with oriented antibodies have been prepared primarily with F(ab′) antibody fragments generated from the enzymatic digestion and subsequent reduction of whole IgG molecules.10,31 Because F(ab′) fragments contain a free thiolate group located opposite the antigen binding site, they can be covalently coupled to a variety of thiol-reactive linkers so that the antigen binding site is exposed to solution.10,11 The preparation of antibody fragments, however, is time-consuming. Alternatively, whole antibodies have been covalently coupled to AFM probes using linkers that react with the amino groups of lysine residues in the protein.32-34 Unfortunately, the surface of an antibody has numerous lysine groups that may lead to several attachment points,24 and thus, multiple antibody orientations. If the antibody is immobilized such that accessibility to the antigen is hindered or lysine groups near the antigen binding site are covalently bound to the surface, a reduction in the antigen (17) Butler, J. E. In Immunoassay; Diamandis, E. P., Christopoulos, T. K., Eds.; Academic Press: San Diego, 1996; pp 205-225. (18) Hoffman, W. L.; Jump, A. A. J. Immunol. Methods 1986, 94, 191-196. (19) Gardas, A.; Lewartowska, A. J. Immunol. Methods 1998, 106, 251-255. (20) Perrin, A.; Lanet, V.; Theretz, A. Langmuir 1997, 13, 25572563. (21) Crumpton, M. J.; Parkhouse, R. M. E. FEBS Lett. 1972, 22, 210-212. (22) Nishiyama, H.; Maeda, H. Biophys. Chem. 1992, 44, 199-208. (23) Ho, C. C.; Ch′ng, S. L. J. Colloid Interface Sci. 1988, 121, 564570. (24) Wimalasena, R. L.; Wilson, G. S. J. Chromatogr. 1991, 572, 85102. (25) Egodage, K. L.; Wilson, G. S. In Immobilized Biomolecules in Analysis: A practical approach; Cass, T., Ligler, F. S., Eds.; Oxford University Press: New York, 1998; pp 35-53. (26) Catimel, B.; Nerrie, M.; Lee, F. T.; Scott, A. M.; Ritter, G.; Welt, S.; Old, L. J.; Burgess, A. W.; Nice, E. C. J. Chromatogr., A 1997, 776, 15-30. (27) Spitznagel, T. M.; Jacobs, J. W.; Clark, D. S. Enzyme Microb. Technol. 1993, 15, 916-921. (28) Lu, B.; Smyth, M. R.; O’Kennedy, R. Analyst 1996, 121, 29R32R. (29) Harada, Y.; Kuroda, M.; Ishida, A. Langmuir 2000, 16, 708715. (30) Castner, D. G.; Ratner, B. D. Surf. Sci. 2002, 28-60. (31) Hinterdorfer, P.; Gruber, H. J.; Kienberger, F.; Kada, G.; Riener, C.; Borken, C.; Schindler, H. Colloids Surf., B 2002, 23, 115-123. (32) Stevens, M. M.; Allen, S.; Davies, M. C.; Roberts, C. J.; Schacht, E.; Tendler, S. J. B.; VanSteenkiste, S.; Williams, P. M. Langmuir 2002, 18, 6659-6665. (33) Wagner, P.; Hegner, M.; Kernen, P.; Zaugg, F.; Semenza, G. Biophys. J. 1996, 70, 2052-2066. (34) Chowdhury, P. B.; Luckham, P. Colloids Surf., A 1998, 143, 53-57.
Brogan et al.
binding activity may be observed. Whole IgG has been immobilized to surfaces in an oriented manner using the carbohydrate groups in the Fc region of the antibody.25 The chemical or enzymatic oxidation of the carbohydrate groups result in reactive aldehydes that can be covalently bonded to surface hydrazide groups.35 However, intermolecular polypeptide chain cross-linking of the reactive aldehyde groups can occur, leading to polymerization of the proteins that can result in a loss of activity.36 Protein A, a bacterial coat protein produced by Staphylococcus aureus that contains four high-affinity sites for the Fc portion of IgG,37 has been shown to yield oriented immunosurfaces with higher antigen-binding activities than corresponding surfaces prepared with a random coupling strategy.26,38 Since Protein A binds to the native Fc region, chemical or enzymatic treatment of the antibody is not required. Thus, Protein A may provide a simpler method for functionalizing AFM probes with whole IgG in an oriented manner. Herein, we report on the specific and nonspecific interactions that occur between anti-ovalbumin IgGmodified AFM probes and both antigen (ovalbumin) and nonantigen (bovine serum albumin, BSA) proteins adsorbed to a substrate surface. We demonstrate that the antibody-immobilization strategy influences the magnitude of the specific and nonspecific adhesion forces measured. Additionally, surfactants commonly employed in solid-phase immunoassays are proven useful for reducing nonspecific adhesion forces and differentiating specific and nonspecific interactions in force recognition measurements. Systematic studies of the magnitude of specific and nonspecific adhesion as a function of antibodyimmobilization strategy, surfactant type, and surfactant concentration using enzyme-linked immunosorbent assay (ELISA), surface plasmon resonance (SPR), and AFM are described. Materials and Methods Reagents. Rabbit anti-ovalbumin IgG (RAO-IgG) and horseradish peroxidase-labeled rabbit anti-ovalbumin IgG (HRPRAO-IgG) were purchased from Rockland Immunochemical (Gilbertsville, PA). Bovine serum albumin (free of IgG and protease) was purchased from Jackson Immunoresearch (West Grove, PA). Albumin, chicken egg (98%), 11-mercaptoundecanoic acid, Tween 20, Tween 80, Triton X-100, and 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) were purchased from Sigma. Protein A was purchased from American Qualex (San Clemente, CA). Glutaraldehyde and ethanolamine hydrochloride were purchased from Fisher Scientific. Aminopropyltrimethoxysilane (APTMS) was purchased from Gelest (Tulleytown, PA). 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), N-hydroxysuccinimide (NHS), o-phenylenediamine (OPD), and dimethyl pimelimidate (DMP) were purchased from Pierce (Rockford, IL). The above chemicals were used without further purification. Water was purified using a Milli-Q UV Gradient A10 System (Millipore Corp., Billerica, MA) to a final resistivity of 18.2 MΩ/cm, and a total organic content of
