Application of Redox Enzymes for Probing the Antigen− Antibody

The application of biocatalysts for amperometric sensing of antigen−antibody .... Surface Plasmon Resonance Analyses: Development of Cholera Toxin S...
0 downloads 0 Views 241KB Size
Download PDF
Anal. Chem. 1996, 68, 3151-3157

Application of Redox Enzymes for Probing the Antigen-Antibody Association at Monolayer Interfaces: Development of Amperometric Immunosensor Electrodes Ron Blonder, Eugenii Katz, Yael Cohen, Norbert Itzhak, Azalia Riklin, and Itamar Willner*

Institute of Chemistry and Farkas Center for Light-Induced Processes, The Hebrew University of Jerusalem, Jerusalem 91904, Israel

Insulation of the electrical contact between a redox protein and an electrode surface upon association of an antibody to an antigen monolayer assembled on the electrode is used to develop immunosensor devices. In one configuration, a mixed monolayer consisting of the NE-(2,4dinitrophenyl)lysine antigen and ferrocene units acting as electron transfer mediators is applied to sense the dinitrophenyl antibody (DNP-Ab) in the presence of glucose oxidase (GOx) and glucose. In the absence of DNP-Ab, the mixed monolayer electrode stimulates the mediated electrocatalyzed oxidation of glucose that results in an amplified amperometric response. Association of the DNP-Ab to the modified electrode blocks the electrocatalytic transformation. The extent of the electrode insulation by the DNP-Ab is controlled by the Ab concentration in the sample. In the second configuration, a NE-(2,4dinitrophenyl)lysine antigen monolayer assembled on a Au electrode is applied to sense the DNP-Ab in the presence of a redox-modified GOx, exhibiting electrical communication with the electrode surface. Two kinds of redox-modified “electrically wired” GOx are applied: GOx modified by N-(ferrocenylmethyl)caproic acid, Fc-GOx, and a novel electrobiocatalyst generated by reconstitution of apo-GOx with a ferrocene-modified FAD semisynthetic cofactor. Electrocatalytic oxidation of glucose by the electrically wired biocatalysts proceeds in the presence of the antigen monolayer electrode, giving rise to an amplified amperometric signal. The electrocatalytic transformation is blocked upon association of the DNP-Ab to the monolayer electrode. The extent of electrode insulation toward the bioelectrocatalytic oxidation of glucose is controlled by the DNP-Ab concentrations in the samples. The application of biocatalysts for amperometric sensing of antigen-antibody interactions at the electrode surface makes the electrode insensitive to microscopic pinhole defects in the monolayer assembly. The antigen monolayer electrode is applied to sense the DNP-Ab in the concentration range 1-50 µg mL-1. The high specificity of antibody-antigen interactions and the possibility to elicit monoclonal antibodies for a variety of nonbiological materials or synthetic fragments of complex biological antigens open the way to develop immunosensor devices for clinical diagnosis, study of environmental pollutants, and food analysis.1,2 Radioisotopic labeling of antibodies and antigens has S0003-2700(96)00290-9 CCC: $12.00

© 1996 American Chemical Society

been applied extensively in the last three decades in immunoassay systems.3 The disadvantages associated with radioactive labels led to the development of other antibody-antigen labels. Enzymelinked immunosorbent assays (ELISA) were extensively developed as a general immunoassay method, and the enzyme-catalyzed reaction provides a means to amplify the antigen-antibody interaction.4 Fluorescence,5 chemiluminescence,6 and electrode potential changes7 as a result of the enzyme reactions represent physical means that transduce the formation of the antibodyantigen complex. Other immunosensor devices use piezoelectric8,9 or surface plasmon resonance10 effects to transduce the formation of antibody-antigen complexes. Electrochemical detection of antibody-antigen interactions has been the subject of several research efforts. Monitoring of capacity changes at the electrode/electrolyte interface as a result of the antigen-antibody complex formation was applied to develop a series of capacitive affinity sensors.11 Amperometric detection12,13 of antigen-antibody interactions was accomplished by the application of redox-modified antigens or antibodies and their (1) (a) Gosling, J. P. Clin. Chem. 1990, 36 1408-1427. (b) BiosensorssA Practical Approach; Cass, A. E. G.; Ed.; IRL Press: New York, 1990. (2) (a) Van Emon, J. P.; Lopez Avila, V. Anal. Chem. 1992, 64, 79A-88A. (b) Wagner, G.; Guilbault, G. G. Food Biosensors Analysis; Marcel Dekker: New York, 1994. (3) Garvey, J. J.; Cremer, N. E.; Sussdorf, D. H. Methods in Immunology, 3rd ed.; W. A. Benjamin, Inc.: Reading, MA, 1977; pp 301-312. (4) Engvall, E. Methods Enzymol. 1980, 70, 419-439. (5) (a) Soin, E.; Lo ¨vgreen, T. In Nonisotopic Immunoassay; Ngo, T. T., Ed.; Plenum Press: New York, 1988; pp 231-243. (b) Diamands, E. P.; Christopoulos, T. K. Anal. Chem. 1990, 62, 1149A-1157A. (6) (a) Kohen, F.; Pazzagli, M.; Secio, M.; De Baever, J.; Vandekerchhove, D. In Alternative Immunoassay; Collins, W. P., Ed.; Wiley: London, 1985; pp 103-121. (b) Hage, D. S.; Kao, P. C. Anal. Chem. 1991, 63, 586-595. (7) Biotieux, J. L.; Thomas, D.; Desmet, G. Anal. Chim. Acta 1984, 163, 309313. (8) (a) Suleiman, A. A.; Guilbault, G. G. Analyst 1994, 119, 2279-2282. (b) Ko ¨sslinger, C.; Drost, S.; Aberl, F.; Wolf, H.; Koch, S.; Woias, P. Biosens. Bioelectron. 1992, 7, 397-404. (9) (a) Guilbault, G. G.; Hock, B.; Schmid, R. Biosens. Bioelectron. 1992, 7, 411-419. (b) Ko ¨nig, B.; Gra¨tzel, M. Anal. Chem. 1994, 66, 341-344. (10) (a) Liedberg, G.; Nylander, C.; Lundstro¨m, I. Biosens. Bioenerg. 1995, 10, i-ix. (b) Davies, J.; Roberts, C. J.; Dawkes, A. C.; Sefton, J.; Edward, J. C.; Glasbey, T. O.; Haymes, A. G.; Davies, M. C.; Jackson, D. E.; Lomas, M.; Shakesheff, K. M.; Tendler, S. J. B.; Wilkins, M. J.; Williams, P. M. Langmuir 1994, 10, 2654-2661. (c) Mrksich, M.; Grunwell, J. R.; Whitesides, G. M. J. Am. Chem. Soc. 1995, 117, 12009-12010. (11) (a) Bresler, H. S.; Lenkevich, M. J.; Murdock, J. F.; Newman, A. L.; Roblin, R. O. In Biosensor Design and Application; Mathewson, P. R., Finley, J. W., Eds.; ACS Symposium Series 511; American Chemical Society: Washington, DC, 1992; pp 89-104. (b) Newman, A. L. U.S. Patent 5,057,430, 1988; Chem. Abstr. 1990, 113, 2966v. (c) Newman, A. L. Canadian Patent 1,259,374, 1985; Chem. Abstr. 1988, 108, 183312u.

Analytical Chemistry, Vol. 68, No. 18, September 15, 1996 3151

competitive association to an electrode interface in the presence of the analyte substance (i.e., antigen or antibody, respectively). Enhanced sensitivities in the amperometric detection of the antibody or antigen was achieved by tailoring enzyme-linked electrochemical devices. Antigens and antibodies modified by redox enzymes allow the detection of the analyte antigen (or antibody) by competitive association to an electrode surface and amperometric transduction by the bioelectrocatalyzed transformation characteristic of the enzyme.14 Indirect electrochemical detection of the antibody-antigen complex formation was also achieved by coupling an enzyme to the antigen or antibody and electrochemically detecting an electroactive species formed by the biocatalyst.15,16 Amperometric or potentiometric detection of NADH, phenol, O2, H2O2, or NH3 generated by the enzyme-linked antigen or antibody label was used to probe antigen-antibody interactions. In a series of recent reports, we have addressed the use of redox enzyme layers assembled on electrodes for the amperometric detection of various substrates.17-20 We demonstrated general methods to assemble monolayers and multilayers on Au electrodes and means to establish electrical communication between the layered enzymes and electrode surfaces. We also reported the use of monolayer-modified electrodes or quartz crystals for electrochemical or piezoelectric sensing of the dynamics of guest-receptor interactions21,22 and the formation of antigen-antibody complexes22,23 at electrode surfaces. In these systems, electrodes or quartz crystals were modified with a monolayer of the molecular guest substrate, i.e., R-D-mannopyranose to sense concanavalin A, or by an antigen monolayer to detect the complementary antibody. Piezoelectric detection of the complementary protein that binds to the monolayer-modified crystal was based on the mass changes and the resulting frequency (12) (a) Weber, S. G.; Purdy, W. C. Anal. Lett. 1979, 12, 1-9. (b) Doyl, J. M.; Wehmeyer, K. R.; Heineman, W. R.; Halsall, H. B. In Electrochemical Sensors in Immunological Analysis; Ngo, T. T., Ed.; Plenum Press: New York, 1987; pp 87-102. (13) (a) Di Gleria, K.; Hill, H. A. O.; McNeil, C. J.; Green, M. J. Anal. Chem. 1986, 58, 1203-1205. (b) Di Gleria, K.; Hill, H. A. O.; Chambers, J. A. J. Electroanal. Chem. 1989, 267, 83-91. (c) Chambers, J. A.; Walton, N. J. J. Electroanal. Chem. 1988, 250, 417-425. (14) (a) Ho, W. O.; Athey, D.; McNeil, C. J. Biosens. Bioelectron. 1995, 10, 683691. (b) Rishpon, J.; Soussan, L.; Rosen-Margalit, I.; Hadas, E. Immunoassays 1992, 13, 231-235. (15) (a) Wright, D. S.; Halsall, H. G.; Heineman, W. R. In Electrochemical Sensors in Immunological Analysis; Ngo, T. T., Ed.; Plenum Press: New York, 1987; pp 117-130. (b) Tang, H. T.; Halsall, H. B.; Heineman, W. R. Clin. Chem. 1991, 37, 245-248. (c) Wehmeyer, K. R.; Halsall, H. B.; Heineman, W. R.; Volle, C. P.; Chen, C. Anal. Chem. 1986, 58, 135-139. (16) (a) Aizawa, M. In Electrochemical Sensors in Immunological Analysis; Ngo, T. T., Ed.; Plenum Press: New York, 1987; pp 269-291. (b) Franconi, C.; Bonori, M.; Orsega, E. F.; Scarpa, M.; Rigo, A. J. Pharm. Biomed. Anal. 1987, 5, 283-287. (c) Uditha de Alwis, W.; Wilson, G. S. Anal. Chem. 1985, 57, 2754-2756. (d) Tsuji, I.; Eguchi, H.; Yasukouchi, K.; Unoki, M.; Taniguchi, I. Biosens. Bioelectron. 1990, 5, 87-101. (e) Gebauer, C. R. In Electrochemical Sensors in Immunological Analysis; Ngo, T. T., Ed.; Plenum Press: New York, 1987; pp 239-255. (17) (a) Katz, E.; Riklin, A.; Willner, I. J. Electroanal. Chem. 1993, 354, 129144. (b) Willner, I.; Riklin, A.; Shoham, B.; Rivenzon, D.; Katz, E. Adv. Mater. 1993, 5, 912-915. (18) (a) Willner, I.; Katz, E.; Riklin, A.; Kasher, R. J. Am. Chem. Soc. 1992, 114, 10965-10966. (b) Willner, I.; Lapidot, N.; Riklin, A.; Kasher, R.; Zahavy, E.; Katz, E. J. Am. Chem. Soc. 1994, 116, 1428-1441. (19) (a) Willner, I.; Riklin, A. Anal. Chem. 1994, 66, 1535-1539. (b) Riklin, A.; Willner, I. Anal. Chem. 1995, 67, 4118-4126. (20) Shoham, B.; Migron, Y.; Riklin, A.; Willner, I.; Tartakovsky, B. Biosens. Bioelectron. 1995, 10, 341-352. (21) Cohen, Y.; Levi, S.; Rubin, S.; Willner, I. J. Electroanal. Chem., in press. (22) Willner, I.; Rubin, S.; Cohen, Y. J. Am. Chem. Soc. 1993, 115, 4937-4938. (23) Willner, I.; Blonder, R.; Dagan, A. J. Am. Chem. Soc. 1994, 116, 93659366.

3152

Analytical Chemistry, Vol. 68, No. 18, September 15, 1996

changes of the crystal accompanying the formation of guestreceptor or antigen-antibody complexes. Electrochemical detection of an antibody by the antigen monolayer electrode was based on the extent of electrode insulation toward a molecular redox probe solubilized in the electrolyte as a result of the formation of the antigen-antibody complex at the electrode surface. The molecular redox probe employed in these studies was the Fe(CN)63-/Fe(CN)64- couple. The changes in the amperometric responses of the molecular redox probe are, however, quite small in magnitude as a result of binding of the high-molecular-weight antibody to the electrode interface. Furthermore, it was found that the amperometric responses of the antigen monolayer electrodes in the presence of the antibodies are very sensitive to the quality of the antigen monolayer. That is, any slight defect in the antigen monolayer assembly or pinholes in the monolayer resulted in difficulties to probe the insulation of the electrode by the molecular redox probe. Thus, it is important to design systems where the effect of electrode insulation by the linked antibody is amplified and to apply redox probes that are insensitive to minor microscopic defects in the antigen monolayer assembly. Here we wish to report on the application of redox enzymes, i.e., glucose oxidase (GOx), as a redox probe that amplifies the antigen-antibody interactions at the electrode surface. We demonstrate that the high-molecular-weight biocatalyst is insensitive to pinholes in the antigen monolayer assembly. EXPERIMENTAL SECTION Materials. The N-(2-ferrocenylmethyl)caproic acid (2) was prepared according to the reported methods.20 [N-(2-methylferrocenyl)caproylamidoethyl]-FAD (4) was prepared by the reaction of N6-aminoethyl-FAD (10 mg, 1.1 × 10-2 mmol) with 2 (18 mg, 5.5 × 10-2 mmol) in the presence of hydroxysuccinimide (13.15 mg, 5.5 × 10-2 mmol) and 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride (EDC, 18 mg, 5.5 × 10-2 mmol) in 4 mL of 0.1 M HEPES buffer, pH 7.5. The solution was stirred for 3 h at room temperature. The crude reaction mixture was purified by chromatography on a Sephadex G-10 column (1 cm × 18 cm, elution with water). The product gives a single TLC spot (Rf ) 0.93; elution water-2-propanol 4:7). Ferrocene-modified GOx (FcGOx) was prepared by coupling of 2 to GOx. To a mixture of dry THF (1 mL) that included 2 (0.06 g) and N-hydroxysuccinimide (NHS, 0.017 g) was added a THF solution (1 mL) of dicyclohexylcarbodiimide (DCC, 0.031 g).24 The resulting mixture was stirred overnight at 0 °C and filtered. From the in situgenerated solution of the active ester of 2, 400 µL was then added to 1.6 mL of distilled water that included NaHCO3 (0.06 g) and GOx (100 mg). The resulting mixture was stirred for 24 h at 0 °C. The solution was dialyzed against phosphate buffer (0.01 M, pH 7.0), centrifuged, and lyophilized to yield the powder of the Fc-GOx. The loading of the enzyme by ferrocene was determined by using fluorescamine as probe.25 The loading of the resulting protein was 21. Reconstitution of glucose oxidase apoprotein with the synthetic FAD-ferrocene diad 4 was perfomed as described recently.24 All other materials, including N-(2,4-dinitrophenyl)lysine (1, Sigma), fluorescein isothiocyanate (3, isomer 1, Aldrich), dinitrophenyl antibody (monoclonal mouse IgE anti-DNP, Sigma), fluorescein antibody (FITC monoclonal antibody, Sigma), (24) Riklin, A.; Katz, E.; Willner, I.; Stocker, A.; Bu ¨ ckmann, A. F. Nature 1995, 376, 672-675. (25) Stein, S.; Bohlen, P.; Dairman, W. Science 1972, 178, 871-872.

glucose oxidase (GOx, from Aspergillus niger, EC 1.1.3.4., Sigma), and cystamine (2,2′-diaminodiethyl disulfide, Aldrich), were used as supplied without further purification. Ultrapure water from a Nanopure (Barnstead) source was used throughout this work. Electrode Characterization and Electrochemical Setup. A gold electrode (0.5 mm diameter Au wire of geometrical area ∼0.2 cm2) was used for all modifications and measurements. A cyclic voltammogram recorded in 0.5 M H2SO4 was used to determine the purity of the electrode surface just before modification. The real electrode surface area and the roughness coefficient (∼1.1) were estimated from the same cyclic voltammogram by integrating the cathodic peak for the electrochemical reduction of the oxide layer on the electrode surface.26 Electrochemical measurements were performed using a potentiostat (EG&G VersaStat) connected to a personal computer (EG&G research electrochemistry software, Model 270/250). All the measurements were carried out in a three-compartment electrochemical cell comprising the chemically modified electrode as a working electrode, a glassy carbon auxiliary electrode isolated by a glass frit, and a saturated calomel electrode (SCE) connected to the working volume with a Luggin capillary. All potentials are reported with respect to this reference electrode. Argon bubbling was used to remove oxygen from the solutions in the electrochemical cell. During the measurements the cell was thermostated (35 °C). Modification of Electrodes. (a) Electrode Pretreatment and Cystamine Monolayer Deposition. To remove the previous organic layer and to regenerate a bare metal surface, the electrodes were treated in a boiling 2 M solution of KOH for 1 h, rinsed with water, and stored in concentrated sulfuric acid. Immediately before modification, the electrodes were rinsed with water, soaked for 10 min in concentrated nitric acid, and then rinsed with water again. The clean bare gold electrode was soaked in a 0.02 M cystamine solution in water for 2 h. The electrode was then rinsed thoroughly with water to remove any physically adsorbed cystamine. (b) Preparation of Monolayers Containing an Antigen (NE(2,4,-Dinitrophenyl)lysine or Fluorescein) as a Single Component. The cystamine monolayer-modified Au electrode was soaked at room temperature for 3 h in a 0.01 M HEPES buffer solution (pH 7.3) that included N-(2,4-dinitrophenyl)lysine (3 mM) and EDC (10 mM, as a coupling reagent). Attachment of fluorescein units to the cystamine monolayer was accomplished by the reaction of the cystamine monolayer electrode with fluorescein isothiocyanate. A cystamine monolayer-modified Au electrode was soaked overnight in the 0.1 M phosphate buffer (pH 8.9) that included fluorescein isothiocyanate (5 mM). The resulting modified electrode was then thoroughly rinsed with water to remove any physically adsorbed components. (c) Preparation of a Mixed Monolayer Including NE-(2,4Dinitrophenyl)lysine as the Antigen and Ferrocene Units as an Electron Mediator on the Au electrode. The cystaminemodified Au electrode was soaked in a 0.01 M HEPES buffer soultion (pH 7.3) containing N-(2,4-dinitrophenyl)lysine (3 mM), and N-(2-ferrocenylmethyl)caproic acid (1 mM), and EDC (10 mM) for 3 h at room temperature. The modified electrode was then thoroughly rinsed with water and used immediately in the electrochemical experiments. (26) Woods, R. In Electroanalytical Chemistry; Bard, A. J., Ed.; Dekker: New York, 1976; Vol. 9, pp 119-125.

Scheme 1. Three Configurations for Antibody Analysisa

a (A) Electrochemical analysis of an antibody by a mixed monolayer composed of the complementary antigen and an electron mediator assembled on an electrode and a solubilized redox enzyme. (B) Amperometric analysis of an antibody by an antigen monolayer electrode using a solubilized, “electrically wired” enzyme modified with many relay groups. (C) Amperometric analysis of an antibody by an antigen monolayer electrode using a reconstituted electroenzyme as redox probe. Cyclic voltammetry responses are shown schematically.

RESULTS AND DISCUSSION The concept of utilizing a redox enzyme to probe the association of an antibody to an antigen monolayer electrode is schematically outlined in Scheme 1. In configuration A, a mixed monolayer composed of the antigen and an electron relay unit is assembled on an electrode surface. Interaction of the mixed monolayer electrode with the appropriate redox enzyme and its substrate yields electrical communication between the electrode and the protein by the electron relay unit tethered onto the electrode surface. This results in the electrocatalyzed oxidation of the respective enzyme-substrate and the development of a current response. Interaction of the monolayer-modified electrode with the complementary antibody results in the antibody association to the monolayer. Formation of the antigen-antibody complex at the electrode surface introduces a barrier of electrical communication between the redox relay unit assembled in the monolayer and the redox enzyme. As a result, the electrobiocatalyzed oxidation of the substrate and the resulting amperometric signal are inhibited. The extent of the electrode coverage by the antibody is controlled by the antibody concentration in the bulk analyte sample and by the time of the electrode treatment with the antibody sample. Thus, the extent of insulation of the electrode surface toward the redox enzyme in solution relates to the bulk antibody concentration if the duration of the treatment is fixed. As a result, the decrease in the electrode amperometric response correlates with the antibody concentration in the solution. Analytical Chemistry, Vol. 68, No. 18, September 15, 1996

3153

Scheme 2. Assembly of a Mixed Monolayer of NE-(2,4-Dinitrophenyl)lysine (Antigen) and Ferrocene Units (Electron Mediator) on a Au Electrode for the Analysis of the DNP-Ab

In immunosensor electrode configuration B, an antigen monolayer is assembled on the electrode surface, and a redox relaymodified enzyme is used to probe the interactions of the antibody with the antigen monolayer electrode. Chemical modification of redox enzymes by electron relay units yields “electrically wired” enzymes that communicate with electrode surfaces.27 The redox relay units act as electron mediators for charge transport between the enzyme redox site and the electrode. In the presence of the antigen monolayer electrode, electrical communication between the modified enzyme and the electrode exists. This leads to the electrobiocatalyzed oxidation of the enzyme substrate and the formation of an electrical current by the electrode. Interaction of the monolayer-modified electrode with the antibody results in the formation of the antigen-antibody complex on the electrode surface. The bulky structure of the antibody perturbs the electrical communication between the redox protein and the electrode and inhibits the electrobiocatalyzed reaction. The extent of the electrode insulation by the antibody is controlled by its bulk concentration in the sample (and the time of incubation). Thus, the decrease in the amperometric signal of the electrode relates to the antibody concentration in the sample. An amperometric immunosensor electrode for the dinitrophenyl antibody (DNP-Ab), operating according to the configuration outlined in Scheme 1A, was assembled. A mixed monolayer, consisting of N-(2,4-dinitrophenyl)lysine, (1), acting as the DNPAb antigen, and the N-(2-ferrocenylmethyl)caproic acid (2), acting as an electron mediator unit, was assembled on a Au electrode as outlined in Scheme 2. The two components 1 and 2 were coupled to a preassembled cystamine monolayer on the Au surface. The surface density of the primary amine monolayer was determined by reacting the cystamine-modified monolayer electrode with pyrroloquinoline-quinone (PQQ) and coulometrically analyzing the covalent linked quinone units.28 The surface density is estimated to be ∼1 × 10-10 mol cm-2. This value should be considered as a lower limit value for the surface coverage with amino groups since complete modification of the monolayer by the quinone units is assumed. By coulometric analysis of the ferrocene oxidation wave, we estimate its surface coverage to be ∼2.5 × 10-11 mol cm-2. Figure 1 shows the cyclic voltammograms obtained by using the mixed N-(2,4-dinitrophenyl)lysine-ferrocene monolayer electrode with GOx and glucose (50 mM) in the absence and in the presence of the DNP-Ab. In the (27) (a) Degani, Y.; Heller, A. J. Am. Chem. Soc. 1988, 110, 2615-2620. (b) Schuhman, W.; Ohara, T. J.; Schmidt, H.-L.; Heller, A. J. Am. Chem. Soc. 1991, 113, 1394-1397. (28) Katz, E.; Schlereth, D. D.; Schmidt, H.-L. J. Electroanal. Chem. 1994, 367, 59-70.

3154 Analytical Chemistry, Vol. 68, No. 18, September 15, 1996

Figure 1. Cyclic voltammograms of the ferrocene/N-(2,4-dinitrophenyl)lysine mixed monolayer electrode. (a) In the background electrolyte composed of 0.1 M phosphate buffer, pH 7.0. (b) In the presence of the background electrolyte and added GOx (5 mg mL-1) and glucose (0.05 M). (c) After treatment of the electrode in an DNPAb solution (50 µg mL-1) for 6 min and recording of the electrochemical response in the background electrolyte that includes GOx (5 mg mL-1) and glucose (0.05 M). (d) Upon addition of ferrocenecarboxylic acid (5 × 10-4 M) to the DNP-Ab-treated electrode in the presence of GOx (50 mg mL-1) and glucose (0.05 M). For all experiments: scan rate, 2 mV s-1; temperature, 35 °C.

absence of the DNP-Ab (curve b), an electrocatalytic anodic current is observed at potentials more positive than the ferrocene units’ formal potential (E° ) 0.35 V). This implies that the ferrocene units associated with the electrode mediate the oxidation of GOx that further stimulates the electrobiocatalyzed oxidation of glucose, giving rise to the electrocatalytic anodic current. Interaction of the monolayer-modified electrode with the DNPAb (50 µg mL-1), with subsequent analysis of the electrochemical response of the electrode in the presence of GOx and glucose, is shown in Figure 1, curve c. The electrochemical response is completely blocked. This indicates that association of the DNPAb to the N-(dinitrophenyl)lysine antigen units which are part of the mixed monolayer perturbs the electrical communication between GOx and the electrode. Binding of the high-molecularweight DNP-Ab (MW ≈ 220 000) to the monolayer blocks the mediated electron transfer between the ferrocene units and the biocatalyst. This results in inhibition of the bioelectrocatalyzed oxidation of glucose, Scheme 1A. The cyclic voltammogram shown in Figure 1, curve d, is obtained upon addition of ferrocenecarboxylic acid, acting as diffusional electron mediator, to the mixed monolayer electrode that was treated with the DNP-Ab. An electrocatalytic anodic current is observed, and the current response is similar to that of the mixed monolayer electrode prior to its treatment with the DNP-Ab. These results reveal the importance of using a high-molecular-weight redox enzyme and the electron mediator as a part of the monolayer assembly as the redox pair to probe the formation of the antigen-antibody complex. The monolayer assembly composed of the antibodyantigen complex is imperfect and includes “pinholes” that allow the diffusion of low-molecular-weight components. Thus, in the presence of the ferrocenecarboxylic acid, its diffusion through the DNP-Ab-antigen complex monolayer enables the establishment of electrical communication between GOx and the electrode. In turn, the bulky enzyme, GOx, does not recognize the defects, and

Figure 2. Amperometric responses of the ferrocene/N-(2,4dinitrophenyl)lysine monolayer electrode at different DNP-Ab concentrations. All data were recorded by cyclic voltammetry, 2 mV s-1, in 0.1 M phosphate buffer, pH 7.0, that included GOx (5 mg mL-1) and glucose (0.05 M). Electrodes were inclubated with the DNP-Ab for 6 min.

its electrical interaction with the electron mediator which is a part of the monlayer is blocked by the antibody-antigen monolayer film. Figure 2 shows the amperometric responses of the ferrocene/N-(2,4-dinitrophenyl)lysine mixed monolayer electrode in the presence of different concentrations of DNP-Ab and with a fixed incubation time (6 min). The current responses decrease up to a DNP-Ab concentration of 20 µg mL-1. At this concentration, the electrode is entirely blocked by the antibody and fully insulated toward bioelectrocataylzed oxidation of glucose. Thus, the electrode configuration is applicable for quantitative electrochemical analysis of DNP-Ab within the concentration range ∼110 µg mL-1. Interaction of the ferrocene/N-(2,4-dinitrophenyl)lysine mixed monolayer with bovine serum albumin (BSA, 1 mg mL-1) prior to the treatment with DNP-Ab, or after treatment with the DNP-Ab, followed by rinsing of the electrode with phosphate buffer, did not affect the amperometric signal of the electrode. These results indicate that no nonspecific adsorption of BSA to the antigen monolayer or antigen-Ab monolayer takes place. The second electrode configuration, shown schematically in Scheme 1B, was examined for the development of monolayer electrodes for electrochemical sensing of the DNP-Ab or the antifluorescein antibody, anti-FITC. The redox biocatalyst employed to sense the formation of antigen-antibody complexes at the monolayer interface was glucose oxidase, modified by ferrocene electron mediator units, Fc-GOx. Previous studies indicated that covalent attachment of electron relay units, specifically, covalently tethered ferrocene units, to GOx yields a biocatalyst that communicates electrically with electrode surfaces.27 In these systems, electrochemical oxidation of redox units tethered to the external protein periphery allows electron transfer with intraprotein redox relay units and, ultimately, the oxidation of the proteinembedded flavin cofactor. This electron transfer path facilitates electrical contact between the biocatalyst redox center and the electrode and induces the bioelectrocatalyzed oxidation of glucose. GOx was modified by 2 (Scheme 3). The loading of the protein by the ferrocene units corresponds to 21. A N-(2,4-dinitrophenyl)lysine antigen monolayer was assembled on a Au electrode (Scheme 4A). Figure 3, curve b, shows the cyclic voltammogram of the monolayer electrode in the presence of Fc-GOx and glucose.

Figure 3. Cyclic voltammograms of N-(2,4-dinitrophenyl)lysine monolayer electrode. (a) In the presence of the background electrolyte composed of 0.1 M phosphate buffer, pH 7.0. (b) Upon addition of ferrocene-modified GOx (Fc-GOx, 5 mg mL-1) and glucose (0.05 M) to the electrolyte. (c) Upon treatment of the electrode with the DNP-Ab (50 µg mL-1) for 6 min in the presence of Fc-GOx and glucose. (d) With the DNP-Ab-treated electrode upon addition of ferrocenecarboxylic acid (5 × 10-4 M) to the electrolyte that includes GOx and glucose. For all experiments: scan rate, 2 mV s-1; temperature, 35 °C.

Scheme 3. Chemical Modification of GOx by Ferrocene Units

Scheme 4. Assembly of (A) a NE-(2,4-Dinitrophenyl)lysine Monolayer and (B) a Fluorescein Monolayer on a Au Electrode

An electrocatalytic anodic current is observed, indicating that the modified enzyme exhibits electrical communication with the antigen monolayer electrode. This induces the electrobiocatalyzed oxidation of glucose. Figure 3, curve c, shows the cyclic voltammogram of the antigen monolayer electrode in the presence Analytical Chemistry, Vol. 68, No. 18, September 15, 1996

3155

Figure 4. Amperometric responses of the N-(2,4-dinitrophenyl)lysine monolayer electrode at different DNP-Ab concentrations. All data were recorded by chronoamperometry (4.5 s after potential step from 0 to 0.5 V) in 0.1 M phosphate buffer, pH 7.0, in the presence of Fc-GOx (5 mg mL-1) and glucose (0.05 M). Electrodes were incubated with the DNP-Ab for 6 min.

of Fc-GOx and glucose, after treatment with the DNP-Ab (50 µg mL-1, 6 min). The electrical communication between the redoxmodified enzyme and the electode is entirely blocked, and only the background response of the electrolyte solution is observed. This is attributed to the association of the bulky DNP-Ab to the monolayer electrode, which prevents the electrical contact between the Fc-GOx and the electrode surface. Figure 3, curve d, shows the cyclic voltammogram of the antigen monolayer electrode after treatment with DNP-Ab and in the presence of FcGOx/glucose, upon addition of the diffusional electron mediator, ferrocenecarboxylic acid. An electrocatalytic anodic current is observed that is only slightly lower in magnitude than that of the freshly prepared antigen monolayer electrode. Similarly to the former configuration, this result is attributed to pinhole defects in the resulting antigen-antibody monolayer. In the presence of the low-molecular-weight mediator, diffusion of ferrocenecarboxylic acid through the monolayer pinholes provides a route to establish electrical communication between the redox enzyme and the electrode. This mechanism yields the electrocatalytic anodic current even in the presence of the antigen-antibody complex monolayer. Figure 4 shows the decrease in the amperometric responses of the antigen monolayer electrode at different DNPAb concentrations using Fc-GOx and glucose as the redox indicator to probe the formation of the antigen-Ab complex. The current response decreases as the concentration of DNP-Ab increases, and the amperometric signal is totally blocked at a DNPAb concentration of 50 µg mL-1 (incubation time, 6 min). The antigen monolayer electrode is thus an active interface for electrochemical detection of the DNP-Ab in the concentration region ∼2-50 µg mL-1. Similar results are observed with a fluorescein antigen monolayer electrode for the analysis of the anti-fluorescein-Ab. Scheme 4B outlines the method for the preparation of the fluorescein monolayer electrode. A cystamine monolayer assembled on a Au electrode was reacted with fluorescein isothiocyanate (3) to yield a thiourea-linked fluorescein antigen on the monolayer. Figure 5 shows the cyclic voltammograms of the fluorescein monolayer electrode in the absence (curve b) and in the presence (curve c) of the antifluorescein-Ab using Fc-GOx and glucose as redox probe for the antigen-antibody binding interactions. With the fluores3156 Analytical Chemistry, Vol. 68, No. 18, September 15, 1996

Figure 5. Cyclic voltammograms of the fluorescein monolayer electrode. (a) In the presence of the background electrolyte composed of 0.1 M phosphate buffer, pH 7.0. (b) Upon addition of FcGOx (5 mg mL-1) and glucose (50 mM). (c) Upon treatment of the electrode with the fluorescein-Ab (50 µg mL-1) for 6 min in the presence of Fc-GOx and glucose. (d) With the fluorescein-Ab-treated electrode with ferrocenecarboxylic acid (5 × 10-4 M), Fc-GOx (5 mg mL1-1), and glucose (50 mM). For all experiments: scan rate, 2 mV s-1; temperature, 35 °C. The electrode used is a Au disk of 1 mm diameter.

cein monolayer electrode, an electrocatalytic anodic current is observed (curve b), indicating electrical communication between Fc-GOx and the electrode that yields the electrobiocatalyzed oxidation of glucose and the resulting anodic current. Formation of the antigen-antibody complex blocks the electrical contact between Fc-GOx and the electrode, and the biocatalyst is insulated for oxidation of glucose (Figure 5, curve c). As in previous examples, addition of the diffusional electron mediator ferrocenecarboxylic acid to the antigen monolayer electrode that includes the associated fluorescein antibody restores the electrocatalytic anodic current to a value slightly lower than that of the amperometric response to the fluorescein monolayer electrode by itself (Figure 5, curve d). This is due to the diffusional penetration of ferrocenecarboxylic acid through pinholes in the antigen-Ab monolayer, which establishes the electrical contact between the redox enzyme and the electrode, as described above. The N-(2,4-dinitrophenyl)lysine monolayer and fluorescein monolayer electrodes reveal high specificity for the respective antibodies. Interaction of the monolayer-modified electrodes with BSA (1 mg mL-1) for 30 min, followed by rinsing of the electrode, did not affect the electrocatalytic current of the electrodes in the presence of Fc-GOx/glucose. Also, treatment of the respective electrodes after the formation of the antigen-Ab complex monolayer under similar conditions did not influence the transduced currents. These results clearly imply that no nonspecific adsorption of BSA to the monolayer electrodes or the antigen-Ab complex monolayers occurs. A further means to establish electrical contact between redox enzymes and electrode surfaces was recently demonstrated by us and involves the reconstitution of flavoenzymes with a semisynthetic relay-modified FAD diad.24 Reconstitution of apo-GOx with a ferrocene-modified FAD diad, 4 (Scheme 5), generated a biocatalyst that revealed electrical communication with electrodes. The resulting “electroenzyme” stimulated the electrobiocatalyzed oxidation of glucose. This electroenzyme could be similarly

Figure 6. Cyclic voltammograms of the N-(2,4-dinitrophenyl)lysine monolayer electrode. (a) In the presence of the background electrolyte composed of 0.1 M phosphate buffer, pH 7.0. (b) Upon addition of Fc-FAD-reconstituted GOx (5 mg mL-1) and glucose (0.05 M). (c) After treatment of the electrode with the DNP-Ab (50 µg mL-1) for 6 min in the presence of Fc-FAD-reconstituted GOx. (d) With the DNP-Ab-treated electrode after addition of ferrocenecarboxylic acid (5 × 10-4 M) to the electrolyte that includes Fc-FAD-reconstituted GOx and glucose. For all experiments: scan rate, 2 mV s-1; temperature, 35 °C.

Scheme 5. Transformation of GOx into an “Electroenzyme” via Reconstituion

(4)

employed as a redox probe to follow the formation of the antigenantibody complex at the monolayer electrode. Figure 6 shows the cyclic voltammograms of the N-(2,4-dinitrophenyl)lysine monolayer electrode in the presence of the ferrocene-FAD reconstituted GOx and glucose (curve b) and the electrode current response after treatment with the DNP-Ab (curve c). In the presence of the antigen monolayer electrode, electrical communication between the reconstituted electroenzyme and the electrode is attained, and electrobiocatalyzed oxidation of glucose proceeds. Formation of the antigen-antibody complex at the electrode interface blocks the electrical contact of the biocatalyst with the electrode surface (Scheme 1C). Further addition of the diffusional electron mediator ferrocenecarboxylic acid restores the electrocatalytic oxidation of glucose (Figure 6, curve d). As in all previous systems, this low-molecular-weight electron mediator establishes electrical communication between the electroenzyme and the electrode by a diffusional route.

CONCLUSIONS We have demonstrated two antigen monolayer electrode configurations for the amperometric analysis of antigen-antibody interactions. In both configurations, a redox enzyme, GOx, was employed to probe the formation of antigen-antibody complexes at the monolayer interface. The principle for the detection of the formation of the antigen-antibody pair in the two configurations is similar in both cases and involves the control of electrical interaction between the enzyme and the electrode by means of the insulating antigen-antibody layer. The two configurations represent advantages over prior electrochemical methods to sense antigen-antibody interactions, where redox labeling of the antigen (or antibody) was employed for the electrochemical analysis of the complementary component, or the indirect electrochemical sensing of an enzyme-linked antigen (or antibody) product (i.e., H2O2, NADPH, O2, etc.) was used. The new method does not require the chemical modification of the antigen (or antibody) and allows the application of the same probing redox enzyme for any antigen-antibody pair. The described methods also reveal the advantages of the application of a redox enzyme over molecular redox probes to follow the antigen-antibody interactions: (i) The bioelectrocatalyzed oxidation of the substrate, i.e., glucose, provides a means for amplification of the amperometic signal of the electrode. Thus, insulation of the electrode by the antigen-antibody complex results in a high-magnitude change in the amperometric signal and enables the amplification of the blocking effect by the antibody association to the antigen monolayer. (ii) The antigen-antibody complex monolayer includes pinhole defects that prevent sensitive detection of the insulating effect by a diffusional low-molecular-weight redox probe. The bulky redox enzyme is insensitive to these pinholes, and it recognizes the blocking effect of the antigen-Ab layer assembled on the electrode. The two configurations shown in Scheme 1A and B reveal comparable sensitivities. The sensitivities are, however, controlled by the time of incubation of the electrodes with the respective antibody, and enhanced sensitivities are observed at longer incubation times. We recommend the second configuration, where a single-component antigen monolayer is assembled on the electrode for practical applications, due to the simpler fabrication of the electrode surface. The use of small molecular diffusional electroactive substrates to probe the antigen-Ab interacting at the electrode surface23 suffers from serious disadvantages. The presence of microscopic defects in the monolayer assembly leads to poor reproducibilities in the extent of insulation of the electrode, and the use of molecular electroactive probes for quantitative analyses of the antibodies is rather limited. ACKNOWLEDGMENT This study is supported in part by the Szold Foundation, The Hebrew University of Jerusalem, and Savyon Diagnostics, Ltd., Ashdod, Israel.

Received for review March 25, 1996. Accepted July 1, 1996.X AC960290V X

Abstract published in Advance ACS Abstracts, August 15, 1996.

Analytical Chemistry, Vol. 68, No. 18, September 15, 1996

3157