Detection of Electrochemical Enzymatic Reactions by Surface

Anal. Chem. 2001, 73, 1595-1598

Detection of Electrochemical Enzymatic Reactions by Surface Plasmon Resonance Measurement Yuzuru Iwasaki,* Tsutomu Horiuchi, and Osamu Niwa

NTT Lifestyle and Environmental Technology Laboratories, 3-1, Morinosato Wakamiya, Atsugi, Kanagawa 243-0198, Japan

We describe the surface plasmon resonance (SPR) detection of an enzymatic turnover reaction and the measurement of glucose concentration using a multienzyme layer modified gold electrode. We constructed an osmium redox polymer mediated enzyme sensor on a gold thinfilm electrode and monitored electrochemical reaction by SPR measurement. Unlike the usual binding assay with SPR, here we used SPR to detect the redox state of an electron mediator that was the result of the electrontransfer reaction of sequential enzymatic reactions. Therefore, the degree of refractive index change was independent of the dielectric property of the substrate and enzymatic molecular recognition was converted to refractive index change with amplification. For the quantitative evaluation of glucose with this method, we used chronopotentiometry and a linear relation was obtained between the glucose concentration and the rate of refractive index change. Surface plasmon resonance (SPR) sensors are now widely used in the detection of biological molecules.1 The principle of this method is based on the detection of a small refractive index change on a gold surface modified with molecular recognition materials. The binding of the target molecule to the immobilized recognition molecule causes an increase in the surface concentration of the target molecule. The binding and association constants can both be derived by analyzing the SPR signal.2 SPR measurement uses near-field optics, which means that the detection field is limited to a few hundred nanometers from the gold surface. Molecules that do not bind to the modified layer do not provide a signal because the bulk and surface concentrations are same and they are washed away in a flow system. We also used SPR measurement in combination with electrochemical reactions. An electrochemical reaction is a heterogeneous surface reaction. Therefore, SPR can detect many types of electrochemically induced chemical changes in the surface.3-8 * Corresponding author. Fax: +81 46 240 4728. E-mail: [email protected]. (1) Homola, J.; Yee, S. S.; Gauglitz, G. Sens. Actuators 1999, B54, 3-15. (2) Schuck, P. Annu. Rev. Biophys. Biomol. Struct. 1997, 26, 541-566. (3) Kolb, D. M. Modern Problems in Condensed Matter Sciences, Surface Polaritons: Electromagnetic Waves at Surfaces and Interfaces; North-Holland Publishing Co.: Amsterdam, 1982; Chapter 8. (4) Jory, M. J.; Bradberry, G. W.; Cann, P. S.; Sambles, J. R. Sens. Actuators 1996, B35-36, 197-201. (5) Iwasaki, Y.; Horiuchi, T.; Morita, M.; Niwa, O. Electroanalysis 1997, 9, 12381241. (6) Kienle, S.; Ligler, S.; Krass, W.; Offenha¨usser, A.; Knoll, W.; Jung, G. Biosens. Bioelectron. 1997, 12, 779-786. 10.1021/ac0012851 CCC: $20.00 Published on Web 03/06/2001

© 2001 American Chemical Society

Here we describe the SPR optical sensing of enzymatic reactions using our electrochemically activated multienzyme sensor via the redox state of the electron mediator. Basically, SPR measures the refractive index change in transparent material on a gold thin film. The sensitivity depends on the extent to which the refractive index changes in the binding event involving the surface-immobilized receptors and their ligands. Therefore, to improve the sensitivity we had to immobilize a high density of receptor molecules. When a ligand is smaller than the immobilized receptor, the reflection minimum angle (θSPR) shift will be small because the refractive index does not change greatly. In contrast, electrochemical reactions can be detected by SPR as θSPR changes around its redox potential in potential scanning experiments using the gold thin film as both the surface plasmon medium and the electrochemical working electrode. Here, freely diffusing charged molecules cause the refractive index change. If the redox-active species are immobilized in a thin film on the electrode, we can detect the dielectric change in the thin film induced by an electrochemical reaction. With this method, the application field of the SPR sensor can be enhanced since it can be used for detecting many substrates of oxidoreductases. We have reported on the possibility of detecting the electrochemical state of a mediator-type biosensor by SPR measurement.9 In this paper, we report the chemical conversion of an enzymatic reaction to a dielectric property change in a modified layer on the electrode (Figure 1). We used the enzymes as a molecular recognition element and an electron-transfer chain to transport electrons to the mediator. We used the mediator film to accumulate the charge and produce a large refractive index change in the surface by ion transport to balance the charge. As a first example, we describe the multienzyme electrochemical SPR detection of glucose. EXPERIMENTAL SECTION SPR and the Electrochemical Experiment. We measured the SP resonance angle using an SPR-20 (DKK, Tokyo), which is a Kretschmann configuration10 SPR measurement instrument. We used a cylindrical prism to couple the surface plasmon and the incident light (light-emitting diode, 860 nm). We carried out the electrochemical experiments by placing counter (Pt) and reference (7) Schlereth, D. D. J. Electroanal. Chem. 1999, 464, 198-207. (8) Boussaad, S.; Pean, J.; Tao, N. J. Anal. Chem. 2000, 72, 222-226. (9) Koide, S.; Iwasaki, Y.; Horiuchi, T.; Niwa, O.; Tamiya, E.; Yokoyama, K. Chem. Commun. 2000, 741-742. (10) Raether, H. ; Springer-Verlag: New York, 1988.

Analytical Chemistry, Vol. 73, No. 7, April 1, 2001 1595

Figure 2. Cyclic voltammograms and potential dependence of ∆θSPR for glucose oxidase and the Os-polymer-HRP modified gold electrode. The scan rate was 10 mV s-1. The glucose concentrations (0 M and 5 mM) were noted.

Figure 1. Substrate concentration detection scheme using the mediated enzyme electrode by SPR (above) and experimental setup (below).

(KCl saturated Ag/AgCl) electrodes in the instrument’s cell. The potentials we report here are relative to this reference electrode. The typical sample volume was 1 mL. The working electrodes were previously described gold films.9 We attached the goldsputtered cover glass to the cylindrical prism by using a refractive index matching liquid. The gold surface of the glass was covered with a silicone rubber sheet which had a hole in it (4 mm in diameter) for the electrolyte contact. We obtained cyclic voltammograms (CVs) using a HECS-972 potentiostat (FUSO, Kawasaki, Japan), a PARC-175 universal programmer (EG&G), and a PRO20 digital oscilloscope (Nicolet, WI). All the experiments were carried out at room temperature (20-21°C). Preparation of the Enzyme Film. Osmium-poly(vinylpyridine)-wired horseradish peroxidase (Os-polymer-HRP) films were formed by spotting 1-5 µL of a 1/10 dilute solution of Ospolymer-HRP solution (Bioanalytical Systems) on the gold electrodes and allowing them to dry in a refrigerator. The spot became a ring of Os-polymer-HRP film. Glucose oxidase from Aspargillus niger (Sigma, MO) was immobilized on the electrode by spin coating. The glucose oxidase was mixed in a 1:1 weight ratio with bovine serum albumin (BSA, Sigma) in a 2% water solution. First, we spread the enzyme solution on the gold electrode to cover the entire active surface and then we spin coated it at 500 rpm for 20 s and then 3000 rpm for 20 s. After coating the enzyme film, we exposed the electrode to glutaraldehyde vapor from a 25% solution (Sigma) for cross-linking. The spin-coating condition was critical for SPR measurement. When the film was not flat, it became difficult to determine the SPR reflection dip minimum angle (θSPR) because of dip broadening. The spin-coated enzyme film covered the entire active electrode area of 4 mm in diameter, whereas the spotted Os-polymerHRP film was smaller than the electrode area. We used 0.1 M phosphate buffer (pH 7) as the electrolyte. All the chemicals we used were of analytical grade, and the water was purified using a MilliQ system (Millipore, IL). 1596 Analytical Chemistry, Vol. 73, No. 7, April 1, 2001

RESULTS AND DISCUSSION Potentiostatic Measurement. Figure 2 shows cyclic voltammograms of the gold electrode modified with Os-polymerHRP and glucose oxidase and simultaneously measured ∆θSPR (relative to the θSPR at a starting potenial of 0.45 V). Glucose is known to be oxidized on gold electrodes.11 Since we used a gold electrode for the SPR measurements, we set the highest potential at 0.45 V to avoid the nonenzymatic oxidation of glucose. At potentials lower than this value, the direct oxidation current of the glucose was negligible. Also, BSA-enzyme modified film prevented this electrochemical reaction. In the absence of the glucose, the CV showed mediator reduction and oxidation and the θSPR showed a bimodal change around the CV peak potential. The ∆θSPR direction can be either positive or negative with respect to the potential, depending on the ion exchange capability of the surface film. In the current case, the osmium redox polymer and HRP complex provided the overall anion exchange capability and the observed θSPR was larger when the polymer was reduced than when the polymer was oxidized.12-14 As the charge of the center metal of the mediator polymer is changed by the electrochemical reaction, the swelling status of the polymer may change and influence the refractive index observed by SPR. We expect the swelling to be more affected in weak electrolytes; however, the amplitude of ∆θSPR during electrochemical reactions was the same even if the electrolyte concentration was 8 mM. Therefore, the observed ∆θSPR was most likely due to the ion transport that occurred to balance the charge neutrality of the film. In the presence of glucose, the CV showed a constant reduction current at potentials below 0.2 V. This is a typical catalytic current. The glucose was oxidized by glucose oxidase, the hydrogen peroxide produced by the glucose oxidase was reduced by HRP, and the HRP was reduced by the mediator.15,16 By contrast, ∆θSPR showed the same potential dependence as in the absence of (11) Hsiao, M. W.; Adzi×a6, R. R.; Yeager, E. B. J. Electrochem. Soc. 1996, 143, 759-767. (12) Varineau, P. T.; Buttry, D. A. J. Phys. Chem. 1987, 91, 1292-1295. (13) Sharp, M.;.Aberg, S. J. Electroanal. Chem. 1998, 449, 137-151. (14) Tja¨rnhage, T.; Sharp, M. Electrochim. Acta 1994, 39, 623-628. (15) Pishko, M. V.; Michael, A. C.; Heller, A. Anal. Chem. 1991, 63, 22682272. (16) Vreeke, M.; Maidan, R.; Heller, A. Anal. Chem. 1992, 64, 3084-3090.

Figure 3. Time course of ∆θSPR in the chronopotentiometry of the glucose oxidase and Os-polymer-HRP modified gold electrode. The potential was set at 0 V, and the potentiostat was set in the electrometer mode at time zero.

glucose. This was because the electrode potential ruled the surface concentration ratio of the two redox states of the mediator.17,18 The bulk glucose, hydrogen peroxide concentration, and other molecules themselves will induce a refractive index change and cause a shift in θSPR; however, the difference between the θSPR values of the two potentials was rather large and canceled out this θSPR shift. Under controlled potential conditions, The surface concentration ratio of the two redox states of the mediator was governed by the electrode potential. As the redox state of the mediator changes, the rearrangement of the concentration profile of the mobile ion will occur to keep the charge neutrality. This rearrangement of the ions (ion transport) causes the refractive index change, and the result of this transport was directly reflected in the θSPR. This can be used for the determination of the substrate concentration using SPR measurement in a chronopotentiometric experiment. Chronopotentiometry. The facile electron transport of osmium redox polymer enabled us to determine the substrate concentration quantitatively by the chemical oxidation of the osmium ions combined with an enzymatic electron-transfer chain.18 We conducted the experiment as follows. The substrate was added to the electrochemical cell, and the electrode potential was held at 0 V. Then, the potentiostat was set in the electrometer mode (denoted as 0 s) and the potential change and θSPR were recorded. We set the time for reducing the osmium ions to 1 min for experimental convenience, but this time can be shortened by using the automatic mode change of the potentiostat. Figure 3 shows the result of this experiment. At time 0, the mediator was fully reduced by the electrochemical reaction because the potential was set far lower than the redox potential of the mediator. In the presence of glucose, the θSPR decreased because the reduced osmium was oxidized by sequential reactions of glucose oxidase and HRP. The slope of ∆θSPR (∆θSPR/∆t) was steeper at higher glucose concentrations. The ∆θSPR was not changed in the absence of glucose (0 M). The amplitude of ∆θSPR between two redox states was always the (17) Uhe, B.; Schuhmann, W.; Janker, G.; Schmidt, H.-L. Sens. Actuators 1992, B7, 389-392. (18) Vering, T.; Schumann, W.; Seiwald, D.; Schmidt, H.-L.; Speiser, B.; Ye, L. J. Electroanal. Chem. 1994, 364, 277-279.

Figure 4. Relation between the glucose concentration and the slope of ∆θSPR in Figure 3. The inset showed the catalytic current at 0 V.

same (Figure 2). Therefore, the sensitivity of the substrate concentration did not depend on the refractive index of the substrate. In binding assays using SPR, molecular recognition molecules such as antibodies and DNA are useful because they trap the target molecules and hold them until the detaching condition is applied. Enzymes are also molecular recognition molecules, but they are catalysts and the concentrations of enzyme-substrate intermediates are usually very low or shortlived and may not be easy to detect by SPR binding assays. This technique provides one example of a detection scheme for turnover-type molecular recognition reactions. Figure 4 shows the relationship between glucose concentration and electrochemical detection and SPR detection in the experiment shown in Figure 3. The slope of the linear part in Figure 3 is plotted here. Because we measured the oxidation rate of the prereduced mediator, the factors affecting sensitivity of this sensor are the reaction rate of HRP, the amount of mediator, and the rate of nonenzymatic oxidation of the mediator. One of the reasons for the lower sensitivity to our previous hydrogen peroxide sensor9 came from the thinness of the glucose oxidase layer. These can be controlled by choosing suitable spin-coating conditions for the desired detection range. Although the sensitivity and the dynamic range can be optimized by adjusting the amount of redox polymer and glucose oxidase, Figure 4 shows that quantitative determination is possible by the SPR measurement of the mediator-type electrochemical biosensor. A class of biosensors using Ospolymer-HRP and oxidoreductases including for glutamate,19 GABA,20 acetylcholine,21 lactate,22 NADH,23 and histamine24 have been reported. These substrates will be measured using the SPR enzyme sensor because all enzymes can be used by the same scheme as shown in Figure 1. In the current film preparation method, the mediator film covers a smaller area than the glucose oxidase film because of (19) Niwa, O.; Torimitsu, K.; Morita, M.; Osborne, P.; Yamamoto, K. Anal. Chem. 1996, 68, 1865-1870. (20) Niwa, O.; Kurita, R.; Horiuchi, T.; Torimitsu, K. Anal. Chem. 1998, 70, 89-93. (21) Niwa, O.; Horiuchi, T.; Kurita, R.; Torimitsu, K. Anal. Chem. 1998, 70, 1126-1132 (22) Osborne, P. G.; Niwa, O.; Yamamoto, K. Anal. Chem. 1998, 70, 17011706. (23) Liu, Z.; Niwa, O.; Horiuchi, T.; Kurita, R.; Torimitsu, K. Biosens. Bioelectron. 1999, 14, 631-638. (24) Niwa, O.; Kurita, R.; Hayashi, K.; Horiuchi, T.; Torimitsu, K.; Maeyama, K.; Tanizawa, K. Sens. Actuators 2000, 67, 43-51.

Analytical Chemistry, Vol. 73, No. 7, April 1, 2001

1597

the availability of the Os-polymer-HRP. The sensitivity and reproducibility of the sensor can be improved by spin coating the mediator to cover the larger area of the enzyme film. CONCLUSION We described the electrochemical SPR measurement detection of glucose using a mediated enzyme electrode. Instead of measuring the catalytic current, we used the θSPR change induced by a chemical oxidation of the mediator. A larger signal change can be obtained in SPR measurement than in the direct measurement

1598

Analytical Chemistry, Vol. 73, No. 7, April 1, 2001

of the refractive index change caused by substrate accumulation itself. This method opens the way to the integration of multiple electrochemical biosensors with a two-dimensional SPR system on one electrode by spotting different enzymes.

Received for refiew October 30, 2000. Accepted January 17, 2001. AC0012851