Continuous On-Line Measurement of Cerebral Hydrogen

Anal. Chem. 2002, 74, 3684-3689

Continuous On-Line Measurement of Cerebral Hydrogen Peroxide Using Enzyme-Modified Ring-Disk Plastic Carbon Film Electrode Lanqun Mao,*,† Peter G. Osborne,‡,§ Katsunobu Yamamoto,‡ and Takeshi Kato|

Department of Electronic Chemistry, Interdisciplinary Graduate School of Science and Engineering, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8502, Japan, Department of Research and Development, BioelectroAnalytical Science Inc., 1-36-6 Oshiage, Sumida-ku, Tokyo 131-0045, Japan, and Laboratory of Molecular Recognition, Graduate School of Integrated Science, Yokohama City University, 22-2 Seto, Kanazawa-ku,Yokohama 236-0027, Japan

An amperometric method suitable for the continuous online measurement of cerebral hydrogen peroxide from a microdialysate has been successfully performed for the first time by using an enzyme-modified ring-disk plastic carbon film electrode (PCFE) in a thin-layer radial flow cell. PCFE consists of a ring electrode modified with horseradish peroxidase to detect H2O2 at 0.0 V (vs Ag/ AgCl) and a disk electrode coated with ascorbate oxidase (AOx) to preoxidize ascorbic acid (AA) and thus suppress interference via direct oxidation. Analytes in solution (brain dialysates or standards) are mixed on-line with a phosphate-buffered solution containing dissolved oxygen and chelating agent, EDTA. The buffered solution is used to provide the O2 necessary for the AOx catalytic reaction, stabilize the changes in dialysate pH that are associated with the in vivo formation of H2O2, and remove heavy metal ion impurities and thus suppress reactions between AA and H2O2. This procedure enables trace levels of H2O2 to be readily monitored, virtually interference-free from physiological levels of AA, uric acid, electroactive neurotransmitters and their principle metabolites, in a continuous-flow system. The development of a simple assay for reactive oxygen species (ROS), such as superoxide anion (O2•-), hydroxyl radical (•OH), and hydrogen peroxide (H2O2), has great practical potential because ROS are considered the mediators of the biochemistry of cellular pathology1 and increased levels have been found during ischemia and hypoxia2,3 and traumatic brain injury4,5 and may be involved in the etiology of aging6 and progressive neurodegen* Corresponding author: (fax) +81-45-924-5489; (e-mail) mao@echem. titech.ac.jp. † Tokyo Institute of Technology. ‡ BioelectroAnalytical Science Inc. § Present address: Department of Physiology 1, Asahikawa Medical University, Asahikawa 078-8510, Hokkaido, Japan. | Yokohama City University. (1) Halliwell, B. J. Neurochem. 1992, 59, 1609-1623. (2) Hall, E. D.; Braughler, J. M. Free Radical Biol. Med. 1989, 6, 303-313. (3) Vanella, A.; Di Giacomo, C.; Sorrenti, V.; Russo, A.; Castorina, C.; Campisi, A.; Renis, M.; Perez-Polo, J. R. Neurochem. Res. 1993, 18, 1337-1340. (4) Kontos, H. A.; Wei, E. P. J. Neurosurg. 1986, 64, 803-807. (5) Globus, M. Y.; Alonso, O.; Dietrich, W. A.; Busto, R.; Ginsberg, M. D. J. Neurochem. 1995, 65, 1704-1711.

3684 Analytical Chemistry, Vol. 74, No. 15, August 1, 2002

erative diseases, such as Parkinson’s disease.7 Although many strategies have been used for the detection of ROS,8-11 the direct measurement of these reactive species remains practically difficult because of the low endogenous concentrations and high reactivity. In biological environments, O2•- and •OH are extremely reactive and potentially the most biologically damaging of the ROS. Superoxide dismutase catalyzes the formation of H2O2 and O2 from O2•- 12 while H2O2, in the presence of Fe2+ or Cu2+, is a precursor for the production of •OH, by the Fenton reaction.13 As a consequence of this biochemistry, the physiological levels of H2O2 are closely associated with the degradation and formation of the reactive free radicals, O2•- and •OH, respectively. Unlike either O2•or •OH, H2O2 is not instantly degraded and is thus more amenable to direct analysis. In addition, H2O2 is, in itself, a toxic compound. In light of the above, a specific and technically undemanding method for reliable and durable measurement of H2O2 would be useful for investigations focusing on oxidative stress and lipid peroxidation.14 On-line analytical systems consisting of in vivo microdialysis sampling and direct analyte detection, without sample separation, are increasingly being applied to the continuous measurements of biological compounds, such as glucose,15,16 quinones,17,18 glutamate,19-21 acetylcholine,22 lactate,15 hypoxanthine,23 γ-aminobutyric acid,24 and nitric oxide.25 These assay systems combine (6) Benzi, G.; Moretti, A. Free Radical Biol. Med. 1995, 19, 77-101. (7) Maruyama, W.; Dostert, P.; Matsubara, K.; Naoi, M. Free Radical Biol. Med. 1995, 19, 67-75. (8) Li, B.; Gutierrez, P. L.; Blough, N. V. Anal. Chem. 1997, 69, 4295-4302. (9) Coudray, C.; Favier, A. Free Radical Biol. Med. 2000, 29, 1064-1070. (10) Chen, J.; Wollenberger, U.; Lisdat, F.; Ge, B.; Scheller, F. W. Sens. Actuators, B 2000, 70, 115-120. (11) Xue, J.; Ying, X.; Chen, J.; Xian, Y.; Jin, L.; Jin, J. Anal. Chem. 2000, 72, 5313-5321. (12) McCord, J. M.; Fridovich, I. J. Biol. Chem. 1969, 244, 6049-6055. (13) Theruvathu. J. A.; Aravindakumar, C. T.; Flyunt, R.; von Sonntag, J.; von Sonntag, C. J. Am. Chem. Soc. 2001, 123, 9007-9014. (14) Lei, B.; Adachi, N.; Arai, T. Brain Res. Protoc. 1998, 3, 33-36. (15) Osborne, P. G.; Niwa, O.; Yamamoto, K. Anal. Chem. 1998, 70, 17011706. (16) Mao, L.; Yamamoto, K. Talanta 2000, 51, 187-195. (17) Pravda, M.; Marvin, C. A.; Sarre, S.; Michotte, Y.; Kauffmann, J.-M. Anal. Chem. 1996, 68, 2447-2450. (18) Pravda, M.; Bogaert, L.; Sarre, S.; Ebinger, G.; Kauffmann, J.-M.; Michotte, Y. Anal. Chem. 1997, 69, 2354-2361. (19) Zilkha, E.; Obrenovitch, T. P.; Koshy, A.; Kusakabe, H.; Bennetto, H. P. J. Neurosci. Methods 1995, 60, 1-9. 10.1021/ac011261+ CCC: $22.00

© 2002 American Chemical Society Published on Web 06/25/2002

short analytical time, high sensitivity, and specificity to provide near-real-time measurements. The serious specificity problems associated with the direct electrochemical measurement of H2O2 promoted the development of assays that used electron-transfer mediators or promoters, of which, horseradish peroxidase (HRP) is now the most widely employed.26,27 However, the cerebral environment poses unique and stringent demands upon the sensing technology, and thus, existing H2O2 sensors are not suited to adaptation for continuous on-line measurement of cerebral H2O2. A H2O2 sensor for use in the cerebral environment must overcome the problems associated with the low physiological concentration of H2O2, the great interference from other species, in particular ascorbic acid, and the changes in the endogenous levels of O2 and pH that are associated with the formation of H2O2 under pathophysiological conditions.28,29 In the present work, we prepared ring-disk plastic carbon film electrodes (PCFEs) (ring outer/disk inner) by modifying the disk electrode with ascorbate oxidase (AOx) and the ring electrode with HRP/polypyrrole (PPy)/polyphenol (PPh). Direct oxidation of ascorbic acid (AA) at the ring electrode is essentially eliminated by AOx-catalyzed preoxidation of AA at the disk electrode (O2 molecule is reduced to H2O rather than H2O2 in this enzymic reaction). HRP is immobilized on the ring electrode by entrapment in a PPy coating. The HRP/PPy-modified ring electrode was then overcoated with a thin film of PPh to limit the adsorption of the product of AA oxidation (dehydroascorbic acid, DAA) and reduce the chemical reaction between AA and H2O2 catalyzed by HRP. At the ring electrode, H2O2 is reduced, catalyzed by HRP with PPy as electron-transfer mediator or promoter, producing a reduction current proportional to the concentration of H2O2. The ring-disk electrode is maintained in a flow cell and irrigated by a phosphate buffer solution (PBS) that suppresses the homogeneous reaction between AA and H2O2 by removing heavy metal ion catalysts with the chelating agent, EDTA. The buffer also serves to deliver an excess of O2 for the AOx catalytic reaction and stabilizes the pH of the electrode environment. This analytical strategy enabled a repeatable and durable response to H2O2 to be recorded at the enzyme-modified PCFE maintained in a continuous-flow system. It is envisaged that this system may find suitable applications in the continuous, on-line measurement of cerebral H2O2 under pathophysiological conditions, such as hypoxia and ischemia. (20) Niwa, O.; Horiuchi, T.; Torimitsu, K. Biosens. Bioelectron. 1997, 12, 311319. (21) Berners, M. O. M.; Boutelle, M. G.; Fillenz, M. Anal. Chem. 1994, 66, 20172021. (22) Niwa, O.; Horiuchi, T.; Kurita, R.; Torimitsu, K. Anal. Chem. 1998, 70, 1126-1132. (23) Mao, L.; Yamamoto, K. Anal. Chim. Acta 2000, 415, 143-150. (24) Niwa, O.; Kurita, R.; Horiuchi, T.; Torimitsu, K. Anal. Chem. 1998, 70, 89-93. (25) Wang, J.; Lu, M.; Yang, F.; Zhang, X.; Baeyens, W. R. G.; Campao´a, A. M. G. Anal. Chim. Acta 2001, 428, 173-181. (26) Yang, L.; Janle, E.; Hiang, T.; Gitzen, J.; Kissinger, P. T.; Vreeke, M.; Heller, A. Anal. Chem. 1995, 67, 1326-1331. (27) Lumley-Woodyear, T. de.; Rocca, P.; Lindsay, J.; Dror, Y.; Freeman, A.; Heller, A. Anal. Chem. 1995, 67, 1332-1338. (28) Hyslop, P. A.; Zhang, Z.; Pearson, D. V.; Phebus, P. L. Brain Res. 1995, 671, 181-186. (29) Marzouk, S. A. M.; Ufer, S.; Buck, R. P.; Johnson, T. A.; Dunlap, L. A.; Cascio, W. E. Anal. Chem. 1998, 70, 5054-5061.

EXPERIMENTAL SECTION Chemicals and Solutions. Horseradish peroxidase (EC 1.11.1), ascorbate oxidase (EC 1.10.3.3), dopamine (DA), norepinephrine (NE), 5-hydroxyindole-3-acetic acid (5-HIAA), 3,4-dihydroxyphenylacetic acid (DOPAC), 5-hydroxytryptamine (5-HT), 3-methoxy-4-hydroxyphenyl glycol (MHPG), and homovanillic acid (HVA) were all purchased from Sigma Chemical Co. (St. Louis, MO). Pyrrole, phenol, H2O2 (30% (w/v) solution), bovine serum albumin (BSA), glutaraldehyde, AA, UA, ethylenediaminetetraacetate acid (EDTA, disodium salt), and other chemicals were at least analytical grade and obtained from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). All chemicals were used as supplied. Ringer’s solution containing 147 mM NaCl, 4.0 mM KCl, 1.2 mM MgCl2, and 1.2 mM CaCl2 was prepared with doubly distilled water, filtered (filter pore size 0.22 µm, Millipore, Bedford, MA), and used as perfusion solution for on-line measurements. H2O2 solutions were prepared by dilution of the commercial 30% stock solution. H2O2 and AA solutions were prepared with Ringer’s solution immediately prior to use. Apparatus. Cyclic voltammetry was performed with a computercontrolled BAS 100B/W instrument (Bioanalytical Systems Inc. (BAS), West Lafayette, IN). The electrochemical cell consists of a three-electrode configuration; ring-disk PCFE (BAS, Tokyo, Japan) was used as working electrode, platinum wire as counter electrode, and Ag/AgCl electrode (3 M NaCl, BAS) as reference electrode. The electrochemical cell consists of a thin-layer radical flow block (BAS) with enzyme-modified ring-disk PCFE as working electrode, stainless steel as counter electrode, and Ag/ AgCl electrode (3 M NaCl) as reference electrode. The thickness of the gasket used (BAS) was 50 µm. The ring-disk PCFE in the electrochemical cell was electronically connected to an LC-4C amperometric detector (BAS) that was coupled to a DA-5 data acquisition interface (BAS). Stainless steel outer tubing and connectors were passivated with 6 M HNO3 to avoid leakage of iron ions, which are known to catalyze the reaction between H2O2 and AA and the decomposition of H2O2. Solutions or microdialysate samples were delivered from gas-impermeable syringes (Hamilton, Reno, NV), which were modified with silicon glass sleeves to be metal free, and pumped through tetrafluoroethylene-hexafluoropropene (FEP) tubing by microsyringe pumps (CMA 100, CMA Microdialysis AB, Stockholm, Sweden) to the radial flow cell. Preparation of Enzyme-Modified PCFEs. Ring-disk PCFE (3-mm disk diameter and 1-mm ring width with 0.50-mm gap between the disk and ring electrodes) was prepared by screenprinting carbon-based PVC glue on the surface of PVC film. Initially, the PCFEs were sonicated in distilled water containing 0.05-µm aluminum particles, rinsed with water, and then electrochemically pretreated in 0.50 M H2SO4 solution by cycling the potential from -0.50 to +1.30 V at a sweep rate of 100 mV s-1 until a stable cyclic voltammogram resulted (typically after 20 cycles). HRP Codeposition with PPy on the Ring Electrode. HRP was immobilized at the ring electrode by codeposition with PPy using potential-controlled amperometric polymerization at +0.90 V in a continuous-flow thin-layer flow cell. Initially, the ring electrode was polarized at +0.90 V and equilibrated to perfusion of 0.10 M PBS at 1 µL/min for 30 min. The polymerization solution Analytical Chemistry, Vol. 74, No. 15, August 1, 2002

3685

Figure 1. Schematic diagram of analytical system for on-line continuous measurement of H2O2 with HRP/PPy/PPh-modified ring and AOxmodified disk PCFE positioned in a thin-layer flow cell.

containing 25 mM PBS, 0.05 M pyrrole, and 1 mg/mL HRP was perfused for 50 min. A current-time curve was recorded for the electropolymerization to ensure continuity of the polymerization process between the electrodes. PPh Overcoating on HRP/PPy-Modified Ring Electrode. A PPh overcoat was applied to the surface of the HRP/PPy-modified ring electrode using the same flow cell procedure detailed above. The HRP/PPy-modified ring electrode was polarized at +0.90 V, and a polymerization solution containing 0.15 M phenol monomer in 0.10 M PBS solution was perfused through the flow cell at 1 µL/ min for 10 min. The modified PFCEs were rinsed thoroughly with distilled water to remove residual HRP and monomers of pyrrole and phenol. The electrodes were then air-dried for at least 30 min prior to AOx modification at the disk electrode. Hereafter, the electrodes modified with HRP, PPy, and PPh will be referred as HRP/PPy/PPh-modified ring electrodes. AOx Coating at the Disk Electrode. A 0.5% (w/w) BSA solution was prepared with distilled water and filtered (filter pore size, 0.22 µm). AOx solution (2 units/µL) was prepared with 25 mM PBS and mixed thoroughly with BSA solution (1:2 v/v). Two microliters of AOx-BSA solution was coated onto the disk electrode and crosslinked with 10 µL of 0.25% glutaraldehyde solution applied to the disk surface. The disk electrode was air-dried for 30 min. The enzyme-modified PCFEs were stored in a dry state at +4 °C while not in use. On-Line Measurements of H2O2. On-line responses of the modified PCFEs toward H2O2 were tested in the continuous-flow system. The modified PCFE was positioned in the flow cell in a wall jet configuration, and the HRP/PPy/PPh-modified ring electrode was polarized at 0.0 V for the measurement of H2O2. No potential was applied to the AOx-modified disk electrode. The experimental setup is schematically depicted in Figure 1. Conditions Relevant for On-Line Measurement of Cerebral H2O2. In vivo the formation of physiologically inappropriate levels of free radicals occurs in response to low blood flow, low oxygen levels, and low pH.19,29 These changes are transferred to the dialysate and adversely influence the electron transfer between HRP and the electrode and also the activity of the enzymes used. 3686 Analytical Chemistry, Vol. 74, No. 15, August 1, 2002

Stabilization of the electrode environment was achieved by mixing the dialysate on-line with 0.10 M PBS containing 0.5 mM EDTA and dissolved O2. The microdialysate15 and exogenous phosphate buffer solution were perfused at 0.5 and 2.5 µL/min, respectively (see Figure 1). EDTA was used to chelate heavy metal ion impurities in the system and suppress the homogeneous reaction between AA and H2O2.30,31 O2 in the buffer acts as an electron acceptor of AOx for catalyzing AA oxidation at the disk electrode. The tubing and connections were made as short as possible to minimize the degradation reaction between H2O2 and endogenous quinones.18 All experiments were performed at room temperature. RESULTS AND DISCUSSION On-Line Responses of H2O2 at the Enzyme-Modified PCFE. The sensitivity and linearity of the enzyme-modified PCFE positioned in a continuous-flow system toward H2O2 in solution was evaluated using potential-controlled amperometry at 0.0 V. The sensitivity of the HRP/PPy/PPh-modified ring electrode to H2O2 was greatly dependent on the amount of HRP immobilized with PPy at the electrode as shown in Figure 2. Electrodes of different HRP content were prepared by varying the concentration of HRP in the polymerization solution (25 mM PBS containing 0.05 M pyrrole) and tested in the continuous-flow system at a perfusion speed of 3 µL/min. The sensitivity of the HRP/PPy/ PPh-modified ring electrode toward H2O2 increased with increasing HRP concentration in the solution and reached a maximum and a plateau when the concentration of HRP was 1 mg/mL. We optimized other electropolymerization conditions such as the applied potential and perfusion rate and determined that the optimum sensitivity was obtained when the electrodes were prepared as detailed in the Experimental Section. The response of the HRP/PPy/PPh-modified ring electrode toward H2O2 was greatly influenced by the operating potential as depicted in Figure 3. At a perfusion rate of 6 µL/min, the current increased as the operating potential shifted negatively and reached (30) Lowry, J. P.; McAteer, K.; Atrash, S. S. E.; Duff, A.; O’Neil, R. D. Anal. Chem. 1994, 66, 1754-1761. (31) Palmisano, F.; Zambonin, P. G. Anal. Chem. 1993, 65, 2690-2692.

Figure 2. Dependence of the sensitivity of the HRP/PPy/PPhmodified ring electrode in response to H2O2 on the concentration of HRP in the electropolymerization solution (25 mM PBS containing 0.05 M pyrrole). The modified PCFE was positioned in a thin-layer flow cell. The HRP/PPy/PPh-modified ring electrode was polarized at 0.0 V and no potential was applied to the AOx-modified disk electrode. Ringer’s solution was used as the perfusion solution. Flow rate, 3 µL/min. Data were presented as mean ( SEM. (n ) 5).

Figure 4. Representative on-line amperometric response of H2O2 standards at the HRP/PPy/PPh-modified ring electrode in a continuous-flow system. Flow rate, 3 µL/min. Other conditions are the same as in Figure 2.

Figure 4 shows a typical trace of on-line current versus time response of H2O2 at the HRP/PPy/PPh-modified ring electrode poised at 0.0 V. The electron transfer of HRP coimmobilized with PPy has been proposed to be mediated or promoted (not yet clarified) by pyrrole analogues, 32,33 which involves multiple steps as shown in the following scheme.

ferric-HRP + H2O2 f HRP compound (I) + H2O HRP compound (I) + e- + H+ f HRP compound (II) HRP compounds (I, II) + e + H+ f ferric-HRP + H2O

Figure 3. Potential dependence of the current response to 1 µM H2O2 recorded at the HRP/PPy/PPh-modified ring electrode in a continuous-flow system. Flow rate, 6 µL/min. Other conditions are the same as in Figure 2.

the maximum at 0.0 V versus Ag/AgCl. This low potential offers a high selectivity for H2O2 over other electrochemically oxidizable compounds present in the dialysate of extracellular fluid. At potentials more negative than -0.20 V, the background current changes and sensitivity declines, probably because PPy was reduced and HRP coimmobilized in the PPy film leaked from electrode surface.

The facilitated electron transfer by PPy between the electrode and the HRP compounds (I, II) yields a reduction current. At 3 µL/min, the current obtained at the HRP/PPy/PPhmodified ring electrode to H2O2 is linear within the concentration range from 0.2 to 15 µM with a sensitivity of 6.13 ( 0.2 nA/µM (n ) 5). The detection limit, based on a signal-to-noise ratio of 3 to 1, was calculated to be 0.06 ( 0.01 µM (n ) 5). The stability and reproducibility of the sensor were tested using both a continuous-flow system and flow injection analysis (FIA). Ringer’s solution was used as a carrier solution and pumped at 100 µL/ min for FIA measurements. The relative standard deviation of the current responses recorded by FIA to 50 injections of 2 µL of 2.5 µM H2O2 was calculated to be 1.3%. The electrode was also stable when used in a continuous-flow operation at 3 µL/min with no apparent reduction in sensitivity after the daily measurement of 1 µM H2O2 for 3-4 h on 6 consecutive days. (32) Tatsuma, T.; Gondaira, M.; Watanabe, T. Anal. Chem. 1992, 64, 11831187. (33) Wollenberger, U.; Bogdanovskaya, V.; Bobrin, S.; Scheller, F.; Tarasevich, M. Anal. Lett. 1990, 23, 1795-1808.

Analytical Chemistry, Vol. 74, No. 15, August 1, 2002

3687

Figure 5. Schematic depiction of electron-transfer process and chemical reactions of ascorbic acid in electrocatalytic reduction of hydrogen peroxide at a mediated HRP-based electrode.

Interference from Ascorbic Acid. In the mammalian brain, AA is present at high extracellular concentrations (200-400 µM34). The high levels, low oxidation potential of AA, and electrode fouling by the product of AA oxidation all contribute to the difficulties of ensuring analyte specificity with in vivo voltammetry35,36 and on-line analytical system coupled with in vivo microdialysis. Figure 5 schematically illustrates the variety of mechanisms37,38 through which AA interferes with the measurement of H2O2. These can be summarized as (I) direct oxidation at the electrode to produce DAA; (II) homogeneous reaction with H2O2 catalyzed by metal ions, e.g., Fe3+ and Cu2+;30,31,39 (III) reaction with H2O2 catalyzed by HRP;38 and (IV) reaction with some mediators, e.g., Os-gel.38 The use of PPy as a facilitator of electron transfer between HRP and the electrode rather than the use of other mediators, such as Os-gel,26 is advantageous for the present application because no chemical reaction occurs between PPy and AA and thus reaction IV should not contribute to the interference at the HRP/PPy/PPh-modified ring electrode. Interference from AA via Direct Oxidation. Cyclic voltammetry showed that the direct oxidation of AA at the HRP/PPy/PPhmodified ring electrode commences at ∼+0.10 V. In the continuous-flow system, at a perfusion rate of 3 µL/min, AA produced significant negative interference via reaction I when the HRP/ PPy/PPh-modified ring electrode was poised at 0.0 V. Under these conditions, when the disk electrode was not used to preoxidize AA, 0.2 mM AA produced ∼110 ( 7% (n ) 5) anodic current response relative to the cathodic current for 0.1 µM H2O2. This demonstrates that PPh overcoating and the low operating potential of the electrode (0.0 V) are insufficient to prevent interference from physiologically relevant levels of AA. AA interference at the HRP/PPy/PPh-modified ring electrode was significantly reduced by coating the disk electrode with AOx as demonstrated in Figure 6. The AOx coating at the disk electrode is very durable and stable. No interference responses were recorded at the HRP/PPy/PPhmodified ring electrode to prolonged application of 0.2 mM AA. The efficiency of the AOx disk electrode to catalyze the oxidation (34) Schenk, J. O.; Miller, E.; Gaddis, R.; Admas, R. N. Brain Res. 1982, 253, 353-356. (35) Ciszewski, A.; Milczarek, G. Anal. Chem. 1999, 71, 1055-1061. (36) Hu, Y.; Mitchell, K. M.; Albahadily, F. N.; Michaelis, E. K.; Wilson, G. S. Brain Res. 1994, 659, 117-125. (37) Garguilo, M. G.; Michael, A. C. J. Am. Chem. Soc. 1993, 115, 12218-12219. (38) Maidan, R.; Heller, A. Anal. Chem. 1992, 64, 2889-2896. (39) Lowry, J. P.; O’Neil, R. D. Anal. Chem. 1992, 64, 453-456.

3688 Analytical Chemistry, Vol. 74, No. 15, August 1, 2002

Figure 6. Typical time vs current responses to H2O2 and AA at the HRP/PPy/PPh-modified ring electrode in a continuous-flow system. Flow rate, 3 µL/min. Other conditions are the same as in Figure 2.

of AA was dependent upon the perfusion rate. At perfusion rates of less than 4 µL/min, no observable current was recorded for 0.2 mM AA even at HRP/PPy-modified ring electrodes not overcoated with PPh. At 4 µL/min, 0.2 mM AA produced ∼8.5% anodic current response relative to the cathodic current for 0.1 µM H2O2 at the HRP/PPy-modified ring electrode. These perfusion rates are at the limit of usefulness for on-line analysis of microdialysates,40 and greater utility of this assay would be achieved if enhanced specificity against AA at higher perfusion rates could be engineered. To improve the selectivity, and thus utility, of the HRP/PPy-modified ring electrode, we coated the HRP/PPy-modified ring electrode with a thin film of PPh. This increased the perfusion speed to which no measurable response to 0.2 mM AA was recorded at the HRP/PPy/PPh-modified ring electrode to 8 µL/min. The PPh layer slows the diffusion of H2O2, and a 20% decrease was recorded in the sensor sensitivity toward H2O2 while the dynamic linear range was extended up to 15 µM. Interference from AA via Homogeneous Reaction with H2O2. The reaction of AA with H2O2 is catalyzed by heavy metal ion impurities, e.g., Fe3+, Cu2+ (reaction II), or peroxidase enzymes, e.g., HRP (reaction III). In the present case, interference from reaction III between AA and H2O2 catalyzed by HRP can be considered negligible because this reaction does not proceed as fast as the electrode reaction of HRP compounds when HRP is electrically “wired” to the electrode while the PPh film limits the diffusion of AA to the HRP inner layer. Reaction II is not expected to occur in the absence of added catalysts. It is assumed to be accelerated by the presence of impurities in solutions.39 To circumvent this potential problem, we introduced a chelating agent, EDTA, into the buffer that was mixed with the sample to chelate heavy metal ions and thus suppress possible reactions before the sample reached the detector. The contribution of a homogeneous reaction between 0.2 mM AA and 0.1 µM H2O2 (reaction II) at the HRP/PPy/PPhmodified ring electrode in the continuous-flow system was measured. H2O2 standard solution was perfused from pump 2 and mixed with the PBS perfused from pump 1 (see Figure 1). The PBS was varied to contained (i) EDTA, (ii) AA, or (iii) EDTA + (40) Robinson, T. E.; Justice, J. B., Jr. Microdialysis in the Neurosciences; Elsevier Science Publisher BV: Amsterdam, 1991.

AA. PBS containing AA alone produced 87.5 ( 5.6% (n ) 3) of the response of perfusion with the other two solutions. This 12% reduction in current magnitude is not due to reactions I, III, or IV and is attributed to the homogeneous reaction between AA and H2O2 (reaction II). In addition to the degradation reaction demonstrated for H2O2 with endogenous quinines,18 a degradation reaction between H2O2 with AA is probably inherent to most biological environments during pathophysiology because high levels of AA are present endogenously and trace levels of metals may be derived from biodegradation of endogenous heme containing proteins. We conclude that the homogeneous interference whereby AA reacts directly with H2O2 can be eliminated when the microdialysate is mixed with a buffered solution containing EDTA. Interference from Other Species. The interference from physiological levels of other electroactive endogenous species and compounds, such as O2•- and hypoxanthine, probably coformed with H2O2 under the pathophysiological conditions was tested. Although the formal potential of the O2/O2•- redox couple is -0.31 V versus Ag/AgCl,41 the short half-life-time of O2•- is incompatible with the time scale of microdialysis sampling and thus O2•- should not contribute to the interference. Hypoxanthine, which has been found to accumulate during ischemia/reperfusion,42 does not interfere with H2O2 measurement because of its poor electrochemical property and the low operating potential of this electrode. At 3 µL/min, no current response was recorded to physiological levels43,44 of monoamine transmitters, e.g., DA (0.1 µM), 5-HT (10 nM), or NE (10 nM). Metabolites, e.g., DOPAC (10 µM), HVA (10 µM), 5-HIAA (10 µM), MHPG (10 µM), or uric acid (50 µM) all produced less than 1.2% current response relative to that produced by 0.1 µM H2O2. Electrode Fouling. Palmisano and Zambonin31 attributed current depression at an H2O2-detecting enzyme electrode to adsorption of DAA, the product of oxidation of AA. Fouling of the HRP/PPy/PPh-modified ring/AOx disk electrode was determined using the continuous-flow system in response to a mixture of 0.2 mM AA and 0.1 µM H2O2. AA and H2O2 were perfused into the system in a manner as demonstrated in Figure 1. The electrode response during 30-min perfusion in the continuous-flow system was very stable. This indicates that DAA produced from AOxcatalyzed reaction does not foul the modified ring electrode, probably because the PPh film coated onto the ring electrode inhibits the adsorption of DAA. On-Line Measurement of H2O2 in Microdialysate. Figure 7 is a typical recording from the modified PCFE using the continuous-flow system to a brain microdialysate. The small current obtained for the microdialysate shows that the basal concentration of H2O2 is less than the linear range of the present system after on-line mixing and dilution. Numerous reports have demonstrated basal levels of H2O2 to be below the limit of detection that increase markedly during experimental manipulation to approximately 20-100 µM.28 To further test the applicability of the system for on-line measurement of cerebral H2O2, the microdialysate was mixed with H2O2 standard and tested. The (41) Matsumoto, F.; Tokuda, K.; Ohsaka, T. Electroanal. 1996, 8, 648-653. (42) Mao, L.; Xu, F.; Xu, Q.; Jin, L. Anal. Biochem. 2001, 292, 94-101. (43) Miele, M.; Fillenz, M. J. Neurosci. Methods 1996, 70, 15. (44) Zetterstrom, T.; Vernet, L.; Ungerstedt, U.; Jonzon, T. B.; Fredholm, B. B. Neurosci. Lett. 1982, 29, 111.

Figure 7. On-line current responses recorded to brain dialysate, H2O2 standard + brain dialysate, and H2O2 standard at the HRP/ PPy/PPh-modified ring electrode in the continuous-flow system as per Figure 1. The samples (brain dialysate, H2O2 standard, and the mixture) were perfused from pump 2 at 0.5 µL/min and were mixed on-line at a T-joint with 0.10 M PBS containing 0.50 mM EDTA that was pumped at 2.5 µL/min from pump 1. The concentration of H2O2 standard was 4.5 µM, and the final concentration of H2O2 added in the brain dialysate was 4.5 µM. Other conditions are the same as in Figure 2.

same H2O2 standard was also tested for comparison as shown in Figure 7. Consistent with the assay of a pure dialysate, the current response to the mixed sample was a little higher than that for H2O2 standard. These demonstrate that this enzyme-modified PCFE used in a continuous-flow system can be utilized for continuous on-line measurement of cerebral H2O2 under pathophysiological conditions associated with enhanced levels of formation of H2O2. CONCLUSION HRP/PPy/PPh- and AOx-modified ring-disk PFCEs were coupled with a continuous-flow system. Electrode fabrication and perfusion conditions were optimized to enable the selective and sensitive on-line measurement of cerebral H2O2 for use with in vivo microdialysis. The interference from AA was efficiently eliminated by AOx-catalyzed preoxidation at the disk electrode and the use of PPy to facilitate the electron transfer between HRP and the electrode and the use of PPh overcoat on the ring electrode. The on-line mixing of the dialysate with a buffer containing EDTA suppressed the homogeneous reaction between AA and H2O2 and stabilized the O2 concentration and pH at the electrodes. These strategies enable the durable and reliable online measurements of H2O2 well suited for measurements of biological tissue sampled by microdialysis. ACKNOWLEDGMENT The authors thank Professor Takeo Ohsaka in Tokyo Institute of Technology for the helpful discussion.

Received for review December 14, 2001. Accepted May 7, 2002. AC011261+ Analytical Chemistry, Vol. 74, No. 15, August 1, 2002

3689