Self-Assembled Layer of Thiolated Protein G as an Immunosensor

To achieve this, the amine groups from lysine residues in protein G were initially converted to thiol groups with 2-iminothiolane. The high affinity o...
0 downloads 0 Views 95KB Size
Anal. Chem. 2007, 79, 350-354

Self-Assembled Layer of Thiolated Protein G as an Immunosensor Scaffold Jeremy M. Fowler,† Margaret C. Stuart,‡ and Danny K. Y. Wong*,†

Department of Chemistry and Biomolecular Sciences and Department of Biological Sciences, Macquarie University, Sydney, NSW 2109, Australia

In the development of electrochemical biosensors, the first step often involves immobilization of a biological recognition component on a solid support. Electrochemical immunosensors are biosensors that are commonly designed with a two-antibody system to sandwich the analyte of interest between a capture antibody linked to an electrode surface and a signal antibody for generation of the analytical response. A key consideration is to immobilize the capture antibody in such a manner as to provide optimal orientation for binding to its analyte with minimal steric hindrance. The capture antibody may be immobilized via covalent attachment, physical adsorption, or electrostatic entrapment in a polymer matrix. This can be achieved using a variety of substances including Nafion, sol-gels, lipid membranes, conducting polymers. or self-assembled monolayers (SAMs).1-5 Although these

alternatives for antibody immobilization have been successfully used to construct electrochemical immunosensors, the majority offer limited or no control over the orientation of antibodies with respect to their antigen binding sites. Without the incorporation of a large excess of antibody, this uncontrolled orientation will result in a compromised dynamic range and sensitivity. The application of SAMs, particularly those based on alkanethiols, is an attractive method to address the orientation problem mentioned above. The high affinity of thiol groups for Au surfaces and the highly ordered nature of functionalized alkanethiol SAMs make these layers good candidates to assist in orientationcontrolled antibody immobilization. By chemically activating the exposed carboxylic acid terminals of a highly ordered ω-carboxylfunctionalized alkanethiol SAM, amine groups of a capture antibody can be covalently attached to the SAM.6,7 Despite the ease of immobilizing a SAM on an electrode surface, and the effectiveness of SAMs in achieving a more orientation-controlled antibody immobilization, several drawbacks exist with this method. First, the formation of the SAM, activation of its carboxyl terminals, and subsequent binding of the capture antibody may take up to 24 h to complete. More importantly, the binding of the capture antibody, with respect to its orientation, is still a somewhat random process. This is because the amine groups of the antibody that react with the activated SAM may be closely associated with the antigen binding sites, thus inhibiting the subsequent antibodyantigen interaction. To prevent antibodies from binding to the SAM through their antigen binding regions, a bacterial antibody binding protein (protein A or protein G) can be covalently bound to an activated alkanethiol SAM.8 Such proteins will only bind antibodies through their nonantigenic (Fc) regions. This leaves the antigen binding sites of the antibodies available to bind to their target antigen. There have been several reports on the use of these proteins in the field of immunosensing, particularly protein A, and only a limited number referring to the application of protein G. For example, Valat et al. have carried out a comparative study to determine the more suitable of these proteins in the development

* To whom correspondence should be addressed. E-mail: danny.wong@ mq.edu.au. Fax: +61-2-9850-8313. † Department of Chemistry and Biomolecular Sciences. ‡ Department of Biological Sciences. (1) Chetcuti, A. F.; Wong, D. K. Y.; Stuart, M. C. Anal. Chem. 1999, 71, 40884094. (2) Chen, J.; Yan, F.; Dai, Z.; Ju, H. Biosens. Bioelectron. 2005, 21, 330-336. (3) Nikolelis, D. P.; Petropoulou, S.-S. E.; Theoharis, G. Electrochim. Acta 2002, 47, 3457-3467.

(4) Jia, J.; Wang, B.; Wu, A.; Cheng, G.; Li, Z.; Dong, S. Anal. Chem. 2002, 74, 2217-2223. (5) Gaspar, S.; Zimmermann, H.; Gazaryan, I.; Csoregi, E.; Schuhmann, W. Electroanalysis 2001, 13, 284-288. (6) Duan, C.; Meyerhoff, M. E. Mikrochim. Acta 1995, 117, 195-206. (7) Susmel, S.; Guilbault, G. G.; O’Sullivan, C. K. Biosens. Bioelectron. 2003, 18, 881-889. (8) Akram, M.; Stuart, M. C.; Wong, D. K. Y. Anal. Chim. Acta 2004, 504, 243-251.

In this study, our goal was to produce a self-assembled layer on a gold electrode that would enable the capture of antibodies orientated for maximum binding to their specific antigen in an immunosensor. To achieve this, the amine groups from lysine residues in protein G were initially converted to thiol groups with 2-iminothiolane. The high affinity of thiols for a gold surface facilitates the direct formation of a self-assembled protein G layer. Following this, the coated gold electrode was exposed to a solution of capture antibody (mAb1) so that these antibodies could attach to the protein G layer through their nonantigenic regions, leaving antigen binding sites available with minimal steric hindrance for binding of target analyte. A comparative study between this method and the more conventional strategy of covalently attaching a layer of nonthiolated protein G on an alkanethiol selfassembled monolayer-coated gold electrode has been performed. Based on a reduced preparation time, and an enhanced capacity for immobilized capture antibody to bind its target analyte due to a more favorable orientation, the layer of thiolated protein G was found to be a more suitable backbone for an electrochemical immunosensor.

350 Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

10.1021/ac061175f CCC: $37.00

© 2007 American Chemical Society Published on Web 11/10/2006

of an electrochemical immunosensor.9 In their work, protein A or G was covalently bound to the surface of an activated graphitepolystyrene screen-printed electrode for the detection of IgG. Based on their findings, immobilized protein A exhibited a superior binding capacity for antibody and this protein was thus selected for further development of their immunosensor. In contrast, owing to the reversible nature of the protein G-antibody interaction, Yakovleva et al. determined protein G to be the more suitable antibody binding reagent for the development of a renewable microfluidic immunosensor.10 This highlights the importance of considering the relative merits of each protein when they are used in designing an immunosensor. More recently, the direct formation of self-assembled layers of proteins on a Au surface has been demonstrated.11,12 Amine groups within these proteins were initially modified to thiols. Protein G is a cell surface protein of group C and G Streptococci with three Fc binding domains located near its C-terminal and exhibits specificity for subclasses of antibodies from many species.13 Amine groups from lysine residues on the surface of protein G can be converted to thiol functional groups in a simple onestep reaction with 2-iminothiolane, giving the protein a high affinity for Au surfaces. This essentially facilitates the direct formation of a self-assembled layer of protein G, obviating the time-consuming steps of SAM formation, activation, and subsequent binding of protein G. Oh et al. have recently used this technique to demonstrate the self-assembly of protein G for development of a surface plasmon resonance-based immunosensor to achieve highly sensitive detection of a pathogenic microorganism down to 102 cfu mL-1.12 Nonetheless, there has hitherto been no systematic investigation to gain an understanding of the characteristics of this layer. In this work, we have investigated the feasibility of forming a self-assembled layer of thiolated protein G to provide a scaffold for controlled orientation of a capture antibody in an immunosensor. The characteristics of this layer have been investigated and compared to those of a layer constructed using a conventional SAM-unmodified protein G format. A long-term goal of our work is to develop electrochemical immunosensors using self-assembled layers of thiolated protein G. In this work, we have used the clinically important hormone, human chorionic gonadotrophin (hCG), as a model analyte. EXPERIMENTAL SECTION Materials and Instrumentation. Recombinant protein G (MW 17 000) was purchased from Amersham Biosciences Pty Ltd. (Sydney, Australia). Anti-hCG monoclonal antibodies (mAb1 and mAb2) were from Bioclone (Sydney, Australia). hCG of 12 500 IU mg-1 potency was purchased from Merck (Sydney, Australia). A bicinchoninic acid (BCA) QuantiPro Protein Assay Kit, 3-mercaptopropionic acid (MPA), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC), 2-iminothiolane, Chloramine (9) Valat, C.; Limoges, B.; Huet, D.; Romette, J. L. Anal. Chim. Acta 2000, 404, 187-194. (10) Yakovleva, J.; Davidsson, R.; Bengtsson, M.; Laurell, T.; Emneus, J. Biosens. Bioelectron. 2003, 19, 21-34. (11) Lee, W.; Oh, B.-K.; Bae, Y. M.; Paek, S.-H.; Lee, W. H.; Choi, J.-W. Biosens. Bioelectron. 2003, 19, 185-192. (12) Oh, B.-K.; Lee, W.; Kim, Y.-K.; Lee, Won, H.; Choi, J.-W. J. Biotechnol. 2004, 111, 1-8. (13) Bjoerck, L.; Kronvall, G. J. Immunol. 1984, 133, 969-974.

T, bovine serum albumin (BSA), and human IgG were purchased from Sigma-Aldrich (Sydney, Australia). A PD-10 column was purchased from Pharmacia (Uppsala, Sweden). Centrifugal filtration devices with 10 000 MW cutoff were purchased from Millipore (Sydney, Australia). Sodium metabisulfite was from BDH (Victoria, Australia). Iodine-125 used for radiolabeling mAb1 and hCG was received as Na125I in NaOH from Australian Nuclear Science and Technology Organisation (Sydney, Australia). Sheep anti-mouse serum was from Chemicon (Melbourne, Australia). Alumina powder of 0.05-µm diameter was purchased from LECO (Sydney, Australia). The following buffers were prepared in Milli-Q water and adjusted to appropriate pH levels with NaOH or HCl: phosphatebuffered saline (PBS), pH 7.6, containing 10 mM phosphate (Na2HPO4/KH2PO4) and 0.15 M NaCl; BSA-PBS, pH 7.6, containing 0.5% BSA. A three-electrode system, consisting of a 3-mm gold disk working electrode, a Ag|AgCl reference electrode (both purchased from Bioanalytical Systems), and a platinum auxiliary electrode, was accommodated in a 1.5-mL electrochemical cell. Electrochemical experiments were performed using a MacLab Potentiostat (eDAQ Pty Ltd., Sydney, Australia) interfaced with a PC via EChem v2.0 (eDAQ). Prior to all voltammetric experiments, the electrolyte solution was sparged with nitrogen for 10 min and a blanket of nitrogen was maintained over the solution throughout the experiment. Formation of Self-Assembled Thiolated Protein G Layer. In the direct formation of a protein G layer on a gold electrode surface, the amine groups associated with a series of lysine residues in the protein G molecule were first converted to thiol groups using 2-iminothiolane.12 A 10-fold molar excess of 2-iminothiolane prepared in degassed PBS was reacted with protein G that had been dissolved in degassed PBS for 30 min at 4 °C. Excess 2-iminothiolane was immediately removed by centrifugal filtration, and the protein was concentrated to 1 µg µL-1. Prior to formation of the protein layer, the gold electrode surface was polished to a mirror finish using 0.05-µm alumina slurry. A 10-µL aliquot of thiolated protein G was applied to the electrode surface for 1 h at room temperature. Following protein immobilization, the electrode was washed thoroughly in 1 mL of PBS for 5 min with occasional shaking. The quantity of thiolated protein G bound to the electrode surface was determined after estimating the residual protein content of the wash solution using the BCA-based QuantiPro Protein Assay Kit. Briefly, the assay working reagent was prepared by combining 1 part of 4% CuSO4 with a premixed solution containing 25 parts of QuantiPro Buffer QA and 25 parts of QuantiPro BCA QB. This reagent was mixed in equal quantities with a blank solution, BSA protein standards, and electrode wash solutions, respectively. The reaction proceeded for 1 h at 60 °C before being quenched in iced water. The absorbance of the violet solutions at 562 nm was measured against the blank. Formation of Nonthiolated Protein G Layer Covalently Bound to an MPA SAM. In these experiments involving a nonthiolated protein G layer, clean gold electrodes were initially coated with an MPA SAM. This was achieved by applying MPA (20 µL; 50 mM) to the electrode surface for 20 min at room temperature, before rinsing off any unbound MPA. To facilitate binding of protein G to the freshly formed SAM, the electrode Analytical Chemistry, Vol. 79, No. 1, January 1, 2007

351

was immersed in a 1% (w/v) solution of EDC in acetonitrile for 5 h. After rinsing thoroughly with acetonitrile, the electrode was dried in a nitrogen atmosphere. Nonthiolated protein G (10 µL; 1 mg mL-1) in PBS was then applied to the electrode surface and allowed to bind overnight at 4 °C. Any unbound protein G was removed by washing the electrode thoroughly with PBS. The quantity of protein G immobilized on the SAM was determined using the QuantiPro Protein Assay Kit in the manner described previously for thiolated protein G. Characteristics of Thiolated and Nonthiolated Protein G Layers. The layer of thiolated protein G and the layer of nonthiolated protein G on an MPA SAM were characterized by investigating their respective degrees of binding to 125I-labeled antihCG monoclonal antibody (mAb1), the antibody selected as the capture component of a two-site sandwich immunoassay. Additionally, the orientation of this capture antibody was investigated by determining its capacity for binding 125I-labeled analyte hCG. A modified Chloramine T method was employed to label mAb1 and our model analyte, hCG.14 In these experiments, 10 µg of the respective proteins was diluted to 1 mg mL-1 in PBS (pH 7.6) and 10 µL of Na125I solution (∼37 MBq) was added. After the addition of aqueous Chloramine T (5 µL; 1 mg mL-1), the reaction proceeded for 40 s on a vortex mixer and was then quenched with sodium metabisulfite (30 µL; 2 mg mL-1) before separation of the radiolabeled protein from unbound iodide by gel filtration on a PD-10 column using 0.5% BSA-PBS, pH 7.6. The radioactivity of eluted fractions was counted to determine the recovery and activity of the radiolabeled proteins. The specific activities of 125ImAb1 and 125I-hCG were 5.6 × 106 and 2.0 × 107 cpm µg-1, respectively. The ability of the protein G layers to bind mAb1 was investigated by applying 20 ng of 125I-mAb1 in PBS for 1 h at room temperature. After washing the electrodes in four separate solutions of PBS, radioactivity associated with the washes and electrodes was counted and used to determine the extent of mAb1 binding. In the fourth wash, the radioactivity count was very low (