Plastic Film Carbon Electrodes: Enzymatic Modification for On-Line

The effect of ascorbic acid was minimal and cross talk between partner .... Simultaneous microchip enzymatic measurements of blood lactate and glucose...
0 downloads 0 Views 85KB Size
Download PDF
Anal. Chem. 1998, 70, 1701-1706

Plastic Film Carbon Electrodes: Enzymatic Modification for On-Line, Continuous, and Simultaneous Measurement of Lactate and Glucose Using Microdialysis Sampling P. G. Osborne,*,† O. Niwa,‡ and K. Yamamoto†

Department of Research and Development, BioelectroAnalytical Science Inc., 36-6-1 Oshiage, Sumida-ku, Tokyo 131, and NTT Basic Research Laboratories, Morinosato, Wakamiya, Atsugi, Kanagawa, 243-01, Japan

Ring and split-disk plastic film carbon electrodes (PFCEs) were fabricated for use in thin-layer radial flow cells which were coupled to a microdialysis sampling system. PFCEs, were initially coated with osmium poly(vinylpyridine) redox polymer horseradish peroxidase (Os-gel-HRP). Then a second coat of oxidase enzyme was applied to produce enzyme bilayer (oxidase/Os-gel-HRP) PFCEs which were subsquently over-coated with cellulose acetate for use in the determination of glucose or lactate at 0 mV (vs Ag/AgCl). Split-disk electrode geometry enabled different oxidase enzymes to be immobilized on each half of a split-disk, Os-gel-HRP-coated, PFCE to facilitate the electrochemically independent yet continuous on-line determination of these two analytes from a single dialysate. In continuous-flow experiments, cellulose acetate overcoated oxidase/Os-gel-HRP cast-coated PFCEs were quick to stabilize background current and displayed linear and sensitive responses to substrates. The effect of ascorbic acid was minimal and cross talk between partner split-disk electrodes was demonstrated to be acceptable for in vivo applications. The utility of this analytical system is demonstrated by the quantitative on-line continuous assay of changes in dialysate striatal extracellular glucose and lactate from a conscious rat during (a) local stimulation of neurons by perfusion with the depolarizing agent, Veratridine, and (b) physical restraint. Amperometric enzyme biosensors for in vivo applications can be divided into two basic categories: those that are implanted directly into the tissue/medium to be assayed (such as fiber microelectrodes) and those that utilize an implanted sampling system (such as microdialysis), which carries the analyte to the distant enzymatic biosensor. Each technique has inherent advantages and disadvantages and as such neither technique is suited to all applications. Our laboratory has recently focused on optimization of the second of these amperometric enzyme biosensors. * Corresponding author: (fax) 3-36240940; (tel) 3-36240367; (e-mail) [email protected]. † BioelectroAnalytical Science, Inc. ‡ NTT Basic Research Laboratories. S0003-2700(97)00699-9 CCC: $15.00 Published on Web 03/27/1998

© 1998 American Chemical Society

The microdialysis in vivo sampling technique, introduced in the early 1980s1 utilizes diffusion across a semipermeable membrane to sample the extracellular (EC) fluid of tissues. This sampling technique provides for versatility of procedure such that steady-state and quasi-steady-state conditions2-5 can be routinely obtained. The microdialysis technique lends itself to continuous sampling and analysis; however, less than 0.6% of microdialysis publications utilize real time, on-line, enzymatic biosensors to analyze the dialysate.6 Although the informational advantages to be gained from continuous analysis are considerable, the above-noted status quo attests, in part, to the following: (1) Off-line or on-line analysis of the dialysate by HPLC-ECD can be used to simultaneously measure a number of compounds. To date there are, to our knowledge, no publications that demonstrate this versatility in online, continuous, enzymatic analysis. (2) There are logistical difficulties in producing and commercially providing simple to use, enzyme-based electrodes to the neuroscience research community who’s researchers may be only transiently interested in an analyte as a means of elucidating a component of a physiological response under study and toward this end may be completely uninterested in allocating time and resources to the mastering of complex analytical procedures. (3) The technical difficulties in fabricating a specific, quantitative, interference-free, enzyme-based amperometric biosensor. Collectively, research into immobilization matrixes, coimmobilization of enzymes with peroxidases, incorporation of redox polymers to reduce the working potential and hence interference at the detector,7-9 refinement of cross-linking procedures,8 development of overcoating strategies,9 coupled with the employment (1) Ungerstedt, U. Measurement of neurotransmitter release by intracranial dialysis. In Measurement of Neurotransmitter Release; Marsden, C. A., Ed.; Wiley: New York, 1984; pp 81-107. (2) Lonnroth, P.; Janssen, P. A.; Smith. U. Am. J. Physiol. 1988, E228-E231. (3) Menacherry, S.; Hubert, W.; Justice, J. B. Anal. Chem. 1992, 64, 577-583. (4) Olson, R. J.; Justice, J. B. Anal. Chem. 1993, 65, 1017-1022. (5) Gerin, C.; Privat, A. J. Neurosci. Methods 1996, 66, 81-92. (6) CMA Bibliography, CMA Stockholm, 1995. (7) Ohara, T. J.; Rajagopalan, J.; Heller, A. Anal. Chem. 1994, 66, 2451-2457. (8) Lumley-Woodyear, T. d.; Rocca, P.; Lindsay, J.; Dror, Y.; Freeman, A.; Heller, A. Anal. Chem. 1995, 67, 1332-1338. (9) Zhang, Y.; Hu, Y.; Wison, G.; Moatti-Sirat, D.; Poitout, V.; Reach, G. Anal. Chem. 1994, 66, 1183-1188.

Analytical Chemistry, Vol. 70, No. 9, May 1, 1998 1701

of a preoxidizing electrode,10-12 and the optimization of perfusion and buffer configurations12-14 has resulted in the production of enzymatic biosensors suited to the measurement of analytes of high concentration that are regularly obtained in a microdialysate which largely fulfill the criteria of point 3. We have addressed the difficulties outlined in points 1 and 2 by focusing on the development of disposable, compound electrodes suited for use as enzyme-based amperometric biosensors for the on-line, continuous, and simultaneous analysis of glucose and lactate from microdialysis perfusate. We utilized ring and split-disk electrodes developed from photolithographically produced carbon film silica electrodes16,17 to produce carbon polymer film electrodes on a flexible plastic surface (PFCE). Computer-controlled, screen-printing procedures determine that PFCE of complex designs can be accurately fabricated, but unlike either silica film or glassy carbon electrodes, production costs are substantially reduced once the template has been fabricated. PFCE surfaces are suitable for enzymatic modification which enables the use of coatings of hydrogels of peroxidases and redox polymers to be cross-linked with oxidase enzymes for the quantification of the oxidation of the hydrogen peroxide produced by the reaction of substrate with a substratespecific oxidase at a potential of 0 mV, thereby significantly reducing the faradic influence of easily oxidizable interferents.4,15 In this paper, we report on the characteristics of a simple to use, disposable split-disk PFCE, coated with Os-gel-HRP polymer and either glucose or lactate oxidase and overcoated with cellulose acetate. The potential of this electrode type is demonstrated by in vivo, on-line, continuous, and simultaneous analysis of changes in EC brain glucose and lactate in the conscious rat in response to neural stimulation with Veratridine or the physiological response to restraint stressor. EXPERIMENTAL SECTION Apparatus. Cyclic voltammetry was performed using a 100 BW BAS (West Lafayette, IN). Amperometry was performed using two LC-4C amperometric controllers (BAS) and stored on a hard disk using a analog/digital converter DA-5 data acquisition interface (BAS). During amperometric studies solutions, were delivered to a radial flow cell (BAS) through FEP (CMA Microdialysis, Stockholm, Sweden) and polyethylene tubing by a CMA 100 (CMA Microdialysis) or BAS Micro LC pump. All experiments were performed at room temperature. Reagents. All chemicals used were reagent grade and purchased from Takara Chemical Co. (Osaka, Japan). Ascorbic acid (AA), bovine serum albumen (BSA), 2-butanone, cellulose acetate, dopamine (DA), glucose, gluteraldehyde, and hydrogen (10) Zilkha, E.; Koshy, A.; Obrenovitch, T. P.; Benetto, H. P.; Symon, L. Anal. Lett. 1994, 27, 453-473. (11) Osborne, P. G.; Niwa, O.; Kato, T.; Yamamoto, K. Curr. Sep. 1996, 15, 19-23. (12) Osborne, P. G.; Niwa, O.; Kato, T.; Yamamoto, K. J. Neurosci. Methods 1997, 77, 143-150. (13) Kuhr, W. G.; Korf, J. J. Cereb. Blood Flow Metab. 1988, 8, 130-137. (14) Fellows, L. K.; Boutelle, M. G.; Fillenz, M. J. Neurochem. 1993, 60, 12581263. (15) Yang, L.; Janle, E.; Huang, T.; Gitzen, J.; Kissinger, P.; Vreeke, M.; Heller, A. Anal. Chem. 1995, 67, 1326-1331. (16) Niwa, O.; Tabei, H. Anal. Chem. 1994, 66, 285-289. (17) Niwa, O.; Morita, O.; Solomon, B. P.; Kissinger, P. T. Electroanalysis 1996, 8, 427-433.

1702 Analytical Chemistry, Vol. 70, No. 9, May 1, 1998

Figure 1. Plastic film electrodes and flow cell: (A) disk electrode (3-mm diameter, area 28.2 mm2); (B) split-disk electrode (two semicircular electrodes 2.6-mm diameter with gap between electrodes of 0.8 mm; each electrode area 55.3 mm2); (C) thin-layer radial flow cell containing split-disk film electrode.

peroxide (30%) were purchased from Takara Chemical Co. (Osaka, Japan). Veratridine, lactate, L-lactate oxidase (from Pediococcus species), and L-glucose oxidase (from Aspergillus niger type VII) were purchased from Sigma Chemical Co. (St. Louis, MO). Osmium-poly(vinylpyrridine)-wired horseradish peroxidase gel polymer (Os-gel-HRP) was purchased from BAS. Pentobarbital sodium (Nembutal) was purchased from Abbott (Osaka, Japan). The buffer used was 0.15 M Na2HPO4/NaH2PO4 (pH 5.75) (phosphate buffer). Ringer’s solution for microdialysis was prepared as follows: [NaCl] ) 150 mM; [KCl] ) 3 mM; [MgCl2] ) 0.8 mM; [CaCl2] ) 1.2 mM. In all in vitro constant-flow experiments, phosphate buffer and Ringer’s solution were mixed on a 10:1 ratio (buffer/Ringer) to approximate conditions during in vivo experiments. Electrode Preparation. PFCE, disk (3 mm in diameter; area 28.2 mm2) and split-disk (two semicircular electrodes, 2.6 mm in diameter with a gap between electrodes of 0.8 mm; area of each 55.3 mm2) were prepared by screen printing carbon-based PVC glue on to the surface of PVC film. The printed electrode was heat cured and covered by an electrostatically charged plastic film until use. Figure 1 shows a schematic representation of (a) disk, (b) split-disk PFCE, and (c) thin-layer radial flow cell. Electrochemical Pretreatment. Prior to testing, all electrodes were polished using a sonicator with a solution of 0.05-µm aluminum particles (Buehler, Lake Bluff, IL) and then pretreated by potentiostating the electrode at 1.6 V for 3 min in 0.2 M sulfuric acid followed by cycling between -0.5 and 1.5 V at a sweep rate of 40 mV/s for 12 sweeps. Platinum wire was used as the auxiliary electrode. The reference electrode was saturated calomel (BAS, Tokyo, Japan). Os-gel-HRP Electrodes. The surface of the 3-mm disk and splitdisk PFCE were cast with 0.6 and 0.9 µL, respectively, of Os-gelHRP. These electrodes were dried for at least 15 h at room temperature before testing.

Figure 2. Schematic representation of analytical system for on-line, continuous determination of lactate and glucose concentrations from microdialysate or standard solutions: PFCE, plastic film carbon electrode; GE, generator electrode; RE, reference electrode; CE, counter electrode; J, plastic tube joiner.

Bilayer Enzyme Electrodes. Bilayer enzyme electrodes were fabricated using a variation on published techniques.11,12 Briefly, for 3-mm disk or split-disk PFCEs, 0.6 or 1.3 µL, respectively, of glucose oxidase enzyme in 1% BSA in 0.15 M phosphate buffer (pH 5.7) (1 wt %) was cast coated onto the surface of the Os-gelHRP-coated electrode. Lactate oxidase was mixed with 1% BSA in 0.15 M phosphate buffer (pH 5.7) (1 µL per unit activity) and cast coated onto the surface of the partner Os-gel-HRP-coated electrode. These electrodes were dried in 25% gluteraldehyde and water vapor for 20 min and then dried for 5 h at room temperature 24 °C. Overcoating. Once dry, the bilayer electrodes were cast coated with 0.4 µL of 0.5% cellulose acetate in butanone and dried for at least 15 min before use. Greater sensitivity and baseline stability were obtained if the electrodes were soaked overnight at 4 °C in buffer/Ringer’s solution. Animal Preparation. Under Nembutal anesthesia (45 mg/ kg ip), a concentric design microdialysis probe (CMA 11, o.d. ) 0.28 mm), with a 2-mm dialyzing membrane at the tip, was permanently implanted into the left striatum (AP ) 0, L ) 3.0, D ) -6.0)18 of a male Wistar rat (350 gm) by previously published procedures.19 During and after surgery, local anesthetic (1% lignocain) was applied to the surface of the wound. Fluid Configuration for Continuous, On-Line Determination of Changes in EC Striatal Glucose and Lactate Concentrations. A schematic of the continuous, on-line, enzymatic/ amperometric assay system is presented in Figure 2. A variation of this system has been utilized previously for the in vivo analysis of EC striatal glucose using a glucose oxidase/Os-gel-HRP bilayer glassy carbon electrode.11,12 Briefly, in the present experiments, the perfusate, Ringer’s solution was pumped by a CMA 100 infusion pump (pump 2) at a rate of 1.5 µL/min through the microdialysis probe (channel A in diagram 2). The speed of the perfusate is chosen by the need to facilitate the desired recovery across the microdialysis mem(18) Paxinos, G.; Watson, C. The rat brain, 2nd ed.; Academic Press: Sydney, 1986. (19) Osborne, P. G.; O’Connor, W. T.; Kehr, J.; Ungerstedt, U. J. Neurosci. Methods 1991, 37, 93-97.

brane. A low-volume platinum generator electrode (i.d. 0.1 mm × 60 mm, internal volume less than 1 µL) maintained at 300 mV (vs Ag/AgCl-re) was positioned upstream of the detector to preoxidize ascorbic acid and provide diagnostic information about the flow of the dialysate. The dialysate was mixed, on-line, immediately prior to the cellulose acetate overcoated, split-disk, bilayer enzyme PFCE with 10 volumes of 0.15 M phosphate buffer (pH 5.7) pumped by another pump (BAS Micro LC) (pump 1) at 15 µL/min (channel B in Figure 2). The phosphate buffer and Ringer’s solutions were mixed when the tubings from channels A and B were joined with glue in a standard peek connector (From, Tokyo, Japan). The phosphate buffer served to stabilize the detector environment, inhibit the oxidation of ascorbic acid at the detector, and maximize the reaction kinetics of the glucose and lactate oxidases. The overcoated bilayer PFCE was positioned in a radial flow cell with a 50 µm gasket (BAS) and held at 0 mV (vs Ag/AgCl). The total flow rate of the mixture passing through the radial flow cell and across the surface of the PFCE was 16.5 µL/min. Glucose in the dialysate was enzymatically oxidized by glucose oxidase immobilized onto the surface of one of the split-disk electrodes, producing gluconolactone and H2O2. The hydrogen peroxide was then electrocatalytically reduced to H2O at 0 mV by the Os-gel-HRP complex.8,9 This current was recorded by one potentiostat. Lactate was measured on the partner split-disk PFCE using lactate oxidase to produce pyruvate and H2O2. The hydrogen peroxide was then electrocatalytically reduced to H2O at 0 mV by the Os-gel-HRP complex. This current was recorded by a second potentiostat. Quantification of dialysate glucose and lactate concentrations was possible by perfusing standard solutions through the flow cell prior to including the dialysate from the animal in the system. RESULTS AND DISCUSSION. Os-gel-HRP-Modified Disk PFCE. In agreement with previous reports for glassy carbon electrodes modified with Os-gelHRP,15 Os-gel-HRP-coated PFCE quickly stabilized background current (less than 1 h). Soaking the electrode in buffer at 4 °C overnight reduced this time to ∼30 min. Os-gel-HRP-coated disk Analytical Chemistry, Vol. 70, No. 9, May 1, 1998

1703

PFCEs were sensitive and stable, being usable for up to three days of experimentation (∼4 h/day). In characterization experiments, the reduction current of a 3-mm Os-gel-HRP PFCE poised at 0 mV (vs Ag/AgCl) to 2.5 µM H2O2 in 0.15 M buffer/Ringer was determined using a 50-µm gasket in the flow cell when the flow of the buffer/Ringer’s solution was varied between 5 and 50 µL/min. The reduction current increased linearly for buffer/ Ringer flow rates from 5 to 25 µL/min. Thereafter current increase tended toward saturation as flow increased to 50 µL/ min. At a constant flow rate of 20 µL/min, the current was linear (nA ) 12 [H2O2 µM] +17.5, r ) 0.997) for buffer/Ringer’s solutions of 0.1-200 µM H2O2. At concentrations of 250 µM H2O2, the relation was exponential (data not presented). Overcoated Bilayer Disk PFCEs. The current response of a 3-mm disk cellulose acetate overcoated glucose oxidase/Osgel-HRP PFCE was determined using a 50-µm gasket in the flow cell when the flow rate of the buffer/Ringer containing 10 µM glucose was varied between 5 and 30 µL/min. The reduction current generated across this range of flow rates exhibited a bellshaped response. the maximum current being generated when buffer/Ringer flow rate was 17.5 µL/min. At flow rates ranging from 10 to 20 µL/min, the current generated by enzyme-modified PFCEs varied by less than 15% of the maximum. At flow rates of 5 and 30 µL/min, current was reduced to ∼60% of maximum. At a constant perfusion speed of 20 µL/min, the current was linear (nA ) 1.96 [glucose µM] + 16.3, r ) 0.997) for buffer/Ringer’s solutions of 1-400 µM glucose. Under in vivo conditions and accounting for dilution of dialysate by phosphate buffer, this range would correspond to a dialysate glucose concentration of 10 µM to 4 mM and exceed the range of concentrations likely to be encountered in conventional microdialysis experiments. Cross Talk between Split-Disk PFCE. The distance between the two partner electrodes on the split-disk PFCE is 800 µm. Utilizing conditions of continuous flow, 0.15 M buffer/ Ringer’s solution (pH 5.75) containing 500 nM dopamine was pumped into the flow cell containing a 50-µm gasket at flow rates varying between 5 and 35 µL/min. Cross talk between the two partner electrodes of the split-disk PFCE at each perfusion speed was determined by calculating the percentage of the reduction current measured at the electrode maintained at 50 mV to the oxidation current measured at the partner electrode maintained at 750 mV (vs Ag/AgCl). At these perfusion speeds, cross talk was less than 1% (data not presented). This demonstrates that when this electrode and fluid configuration is utilized, the two partner electrodes are effectively independent. This result is consistent with previous experiments utilizing photolithograhically produced silicon wafer, split-disk electrodes with an interelectrode distance of 250 µm where cross talk between electrodes positioned in a radial flow cell was less than 2.5% at all speeds tested and decreased as perfusion speed increased.17 Cross Talk between Bilayer Glucose/Lactate Split-Disk PFCE. A split-disk electrode was fabricated such that half was a glucose bilayer electrode while the partner electrode was a bilayer lactate electrode. Using the fluid configuration employed during a microdialysis experiment (Figure 2), the buffer was pumped through channel B at 15 µL/min. The concentration of glucose and lactate in Ringer’s solution (channel A pumped at 1.5 µL/ min) was individually varied among 2.5, 10, 100, 500, 1000, 1500, 1704 Analytical Chemistry, Vol. 70, No. 9, May 1, 1998

Figure 3. (Left-hand axis) Reduction current (nA) vs dialysate concentration (µM) relation for glucose (solid circle) and lactate (solid triangles) generated by overcoated, oxidase/Os-gel-HRP, split-disk PFCE (seven-point standard curve). (Right-hand axis) Cross talk (percent of current at each electrode derived from cross reaction of substrates) vs dialysate concentration. Glucose cross talk (open circles); lactate cross talk (open triangles); mean ( SEM (n ) 5).

and 2500 µM and the reduction current was recorded at each splitdisk bilayer electrode. Figure 3 demonstrates the mean relation between reduction current measured at the bilayer glucose and lactate split-disk PFCE (n ) 5 electrode pairs) in response to varying the concentration of glucose (2.5-2500 µM) and lactate (2.5-2500 µM) in channel A. Cross talk, as determined by the percentage of current generated at one split-disk bilayer PFCE from equimolar substrates (glucose and lactate), was greatest for the lactate film bilayer electrode. For both electrodes, cross talk was greatest at very low dialysate concentrations (less than 10 µM) and decreased as the dialysate concentration increased, becoming less than 7% for both compounds at dialysate concentrations of 100 µM. This is well suited to measurements of glucose and lactate within the EC environment using microdialysis since these dialysate concentrations are easily achieved, even using the smallest commercial microdialysis probes,12 and where decreases of glucose are invariably associated with increases of lactate. Since each bare electrode was essentially independent of cross talk, the increase in cross talk exhibited by bilayer enzyme PFCE is consistent with a reduction of enzyme specificity consequence of cross-linking or immobilization during fabrication. This phenomenon has been reported previously for other enzymes.20 The electrodes exhibited a durable response. After calibration over the above-mentioned range (2.5-2500 µM using 14 individual concentration determinations, taking a total of 2 h) and with perfusion for the subsequent 85 min with a dialysate containing 225 µM lactate and glucose, the reduction current at each splitdisk electrode was measured to decrease 6-7%. This time course is sufficient to perform routine behavioral and pharmacological manipulations. Preliminary storage tests demonstrate that enzymemodified PFCEs have negligible loss of strength when stored for 2 weeks at 4 °C. Interference to Ascorbic Acid. Using the fluid configuration employed during a microdialysis experiment (Figure 2), phosphate (20) Gohbadi, S.; Csoregi, G.; Marko-Varga, G.; Gorton, L. Curr. Sep. 1996, 14, 94-102.

buffer was pumped through channel B at 15 µL/min. In thoroughly wetted or soaked, overcoated, bilayer enzyme PFCEs, inclusion of 100 µM AA in Ringer’s solution into the dialysate (channel A pumped at 1.5 µL/min) produced a small oxidation current at each overcoated, bilayer enzyme, split-disk PFCE (12% of equimolar glucose (n ) 5), 2-3% of equimolar lactate (n ) 5)). The current was larger at nonovercoated, bilayer enzyme, split-disk PFCEs (5.0 ( 0.4% (n ) 5) of equimolar glucose, 5.8 ( 0.4% of equimolar lactate (n ) 5)). These oxidation currents were reduced if a generator poised at 300 mV (vs Ag/AgCl) was employed. Interestingly, if the overcoated, bilayer enzyme PFCE was insufficiently soaked in buffer/Ringer’s solution, perfusion with 100 µM AA produced a small reduction current at the overcoated, lactate/Os-gel-HRP split-disk PFCE. This was usually less than 5% of the current generated to equimolar lactate. This reduction current increased with age of the electrode, being larger on each consecutive day. A reduction current to readily oxidizable interference, urate and acetaminophen, has been previously observed using osmium-coated electrodes, cross-linked with dimethyl suberimidate.8 Using the overcoated, bilayer lactate/Os-gel-HRP PFCE, the reduction current to AA was strongly potentiated by saturating the phosphate buffer with oxygen and completely abolished by reducing the partial pressure of oxygen by bubbling with N2 for 15 min. Since glucose oxidase is an O2-dependent enzyme, N2saturated phosphate buffer completely inhibited glucose enzymatic reactions, whereas O2-saturated buffer increased glucose oxidase activity by up to 100%. The oxygen dependency of the current is unlikely to influence in vivo determinations of glucose or lactate as conditions such as ischemia or hypoxia, which influence the oxygen tension of the brain, will be reflected as much attenuated decreases in the oxygen tension of the dialysate. This is because at these perfusion rates recovery across the dialysis membrane is unlikely to exceed 15-20%. Any effect will be further attenuated because the flow of the phosphate buffer is in 10-fold excess of the dialysate and the partial pressure of O2 in the phosphate buffer should remain stable over the duration of an experiment. Continuous, On-Line, Simultaneous in Vivo Determination of EC Striatal Glucose and Lactate in a Conscious Rat. In vivo experiments were performed on the first (24 h) and second (48 h) day after implantation of the microdialysis probe.19 At the beginning of each experiment, three-point standard curves were generated for the relationship between diaylsate glucose and lactate concentration and the current generated at each overcoated, bilayer enzyme PFCE and the effect of Veratridine on each overcoated, bilayer enzyme PFCE. All standard curves were linear over the range of concentrations measured during the in vivo experiments. Linear regression of all standard curves gave a fit better than r ) 0.997. On each experimental day, the rat was connected and the microdialysis probe was perfused with Ringer’s solution for at least 90 min at 1.5µL/min before the dialysate was switched into the analytical circuit. On day 1, once a stable baseline was obtained (usually ∼15 min after inclusion of the dialysate into the analytical circuit), the microdialysis probe was perfused for 7.5 min with 50 µM Veratridine, a neuron-depolarizing compound, after which time the probe was again perfused with Ringer’s solution. EC striatal

Figure 4. Effect of perfusion of the striatum with Ringer’s solution or 50 µM Veratridine in Ringer’s solution for 7.5 min (boxed V) on the EC striatal dialysate levels (µM) of lactate (upper darker trace) and glucose (middle trace). Generator electrode current is lower trace. Fluctuations in signal result from pump noise.

Figure 5. Effect of restraint (boxed restraint) on the EC striatal dialysate levels (µM) of lactate (darker trace) and glucose. Generator electrode current is lower trace. Fluctuations in signal result from pump noise.

glucose and lactate were monitored for an additional 90 min after switching back to Ringer’s solution. Local perfusion of the microdialysis probe with Veratridine (50 µM) caused a profound, rapid, and long-lasting decrease in striatal glucose and a temporally delayed long-lasting increase in EC striatal lactate (Figure 4). Veratridine was without effect upon the basal current at either electrode but was observed to be oxidized by the generator electrode, thus providing a useful indicator of the timing of the drug treatment. The magnitude and duration of the effect of Veratridine on glucose is consistent with previous studies.12,21 The increase in lactate is consistent with the induction of a cellular environment where oxidative energy demand exceeds supply. On day 2, the effect of restraint upon ECF glucose and lactate was determined. Once a stable baseline had been obtained, the rat, while still in the perfusion bowl, was restrained in the experimenter’s hand for 7.5 min. EC striatal glucose and lactate were monitored for an additional 35 min. Restraint of the rat was associated with a large, rapid increase of EC striatal lactate and a more delayed increase in EC glucose. Lactate returned to basal levels before glucose, which remained elevated 30 min after application of the stressor (Figure 5). The (21) Fellows, L. K.; Boutelle, M. G.; Fillenz, M. J. Neurochem. 1992, 59, 21412147.

Analytical Chemistry, Vol. 70, No. 9, May 1, 1998

1705

magnitude and duration of these responses are consistent with previous studies.12,21,22 CONCLUSIONS PFCEs are well suited to the production of a range of complex electrode geometries and may be enzymatically modified and overcoated to produce linear, sensitive, and robust amperometric responses to glucose and lactate substrates in solution. The current generated by enzyme-modified PFCEs to a constant concentration of substrate varied by less than 15% of the maximum for flow speeds ranging from 10 to 20 µL/min which makes these electrodes well suited for use with sampling techniques that utilize dilution of a microdialysate with an appropriate buffer. In the simultaneous determination of glucose and lactate at concentrations routinely obtained from microdialysis dialysates, the over(22) van der Kuil, J. H.; Korf, J. J. Neurochem. 1991, 57, 648-654.

1706 Analytical Chemistry, Vol. 70, No. 9, May 1, 1998

coated, enzyme bilayer split-disk PFCE demonstrated negligible cross talk. It is possible that the use of bilayer enzymatic, split-disk PFCEs for the simultaneous determination of changes in brain EC glucose and lactate will facilitate understanding of the complex mechanisms that regulate the interactions between cerebral EC glucose and lactate. It is envisaged that other oxidase enzymes may be used in fabrication of bilayer enzyme electrodes and that this development in combination with the production of segment PFCEs will ultimately enable the inexpensive, on-line measurement of numerous analytes simultaneously from a single dialysate.

Received for review July 2, 1997. Accepted February 6, 1998. AC9706990