Mediated Bioelectrocatalytic Determination of Organophosphorus

Hawaii Biosensor Laboratory, Department of Chemistry, University of Hawaii at Manoa, 2545 The Mall,. Honolulu, Hawaii 96822. An electrochemical biosen...
0 downloads 0 Views 131KB Size
Anal. Chem. 1998, 70, 807-810

Mediated Bioelectrocatalytic Determination of Organophosphorus Pesticides with a Tyrosinase-Based Oxygen Biosensor W. Russell Everett† and Garry A. Rechnitz*

Hawaii Biosensor Laboratory, Department of Chemistry, University of Hawaii at Manoa, 2545 The Mall, Honolulu, Hawaii 96822

An electrochemical biosensor for the detection of organophosphorus pesticides is described. A tyrosinasebased oxygen sensor is constructed where enzymatic oxygen consumption is monitored electrochemically with the mediator 1,2-naphthoquinone-4-sulfonate. This bioelectrocatalytic system allows electrochemical initiation and termination of the enzymatic reaction. Amperometric and coulometric techniques are used to study inhibitory effects for pesticide determinations. This inhibition appears to be fully reversible, with full catalytic activity returning after removal of the pesticides. Detection limits are 5 µM for diazinon and 75 nM for dichlorvos. Enzymatic inhibition of 50% occurs at ∼1000 and ∼50 µM, respectively. In recent years, the use of organophosphorus and carbamate pesticides has become widespread. These pesticides are popular because they exhibit lower environmental persistence than organochlorates such as DDT.1 However, they do have a high acute toxicity due to prevention of neural impulse transmission by their inhibition of cholinesterase. Because of the possible health risks from contamination of water and food sources, research is continuing toward the development of a fast, accurate method for the determination of these types of pesticides. The use of electrochemical biosensor systems could possibly lead to future in-line and on-site measurements, which would be difficult to carry out with currently used analytical techniques. Several research groups have reported the development of enzyme-modified electrodes as a possible method of pesticide determination.2-15 * To whom correspondence should be addressed. † E-mail: [email protected]. (1) Jury, A. W.; Winer, A. M.; Spencer, W. F.; Focht, D. D. Rev. Environ. Contam. Toxicol. 1987, 99, 119-129. (2) Vlasov, Y.; Bratov, A.; Levichev, S.; Trantov, Y. Sens. Actuators B 1991, 4, 283-286. (3) Stein, K.; Schwedt, G. Anal. Chim. Acta 1993, 272, 73-81. (4) Fennouh, S.; Casimiri, V.; Burstein, C. Biosens. Bioelectron. 1997, 12, 97104. (5) Bernabei, M.; Cremisini, C.; Mascini, M.; Palleschi, G. Anal. Lett. 1991, 24, 1317-1331. (6) Wollenberger, U.; Seta, K.; Scheller, F. W.; Loffler, U.; Gopel, W.; Gruss, R. Sens. Actuators B 1991, 4, 257-260. (7) Palchetti, I.; Cagnini, A.; Del Carlo, M.; Coppi, C.; Mascini, M.; Turner, A. P. F. Anal. Chim. Acta 1997, 337, 315-321. (8) Marty, J. L.; Mionetto, N.; Noguer, T.; Ortega, F.; Roux, C. Biosens. Bioelectron. 1993, 8, 273-280. S0003-2700(97)00958-X CCC: $15.00 Published on Web 01/21/1998

© 1998 American Chemical Society

The employment of enzyme electrodes for measuring pesticides is based on the inhibition of biocatalytic activity. Since the pesticides of interest are cholinesterase inhibitors, most researchers use acetyl- or butyrylcholinesterase as the biological component of their sensor.2-13 Amperometric determination of pesticides with cholinesterase systems is usually carried out using one of two inhibition methods. One method involves a two-enzyme cholinesterase/choline oxidase system, in which either an oxygen sensor4 or a hydrogen peroxide detector5-7 is used as an internal transducer. Alternatively, a single-enzyme system involving esterase and the substituted substrate thiocholine can be used, where analysis is carried out by direct electrochemical oxidation of the enzymatic reaction product.8,9 In the attempt to improve methods of pesticide determinations, researchers are also trying less traditional cholinesterase-based sensors. Recent reports have shown that the addition of electronic mediators to electrode composite mixtures lowers the potential for the electrocatalytic oxidation of thiocholine.10-13 The mediators cobalt phthalocyanine10,11 and 7,7,8,8,-tetracyanoquinodimethane12 have been used in systems for the detection of both carbamate and organophosphorus pesticides. La Rosa et al. have reported the use of 4-aminophenyl acetate as the enzyme substrate for a cholinesterase sensor for pesticide determination.13 This system allows for determination of esterase activities via oxidation of the enzymatic product 4-aminophenol rather than the typical thiocholine. These newer methods do seem to show slight improvements for pesticide determinations. The use of tyrosinase-modified electrodes to study inhibitory effects of possible environmental pollutants has also recently been reported.14-17 Besombes et al. used an electropolymerized pyrrole/tyrosinase coating for the detection of carbamate pesticides and reported a detection limit of 2 µM for chloroisopropylphenyl (9) Skladal, P.; Pavlik, M.; Fiala, M. Anal. Lett. 1994, 27, 29-40. (10) Skladal, P. Anal. Chim. Acta 1992, 269, 281-287. (11) Hartley, I. C.; Hart, J. P. Anal. Proc. 1994, 31, 333-337. (12) Martorell, D.; Cespedes, F.; Martinez-Fabregas, E.; Alegret, S. Anal. Chim. Acta 1997, 337, 305-313. (13) La Rosa, C.; Pariente, F.; Hernandez, L.; Lorenzo, E. Anal. Chim. Acta 1994, 295, 273-282. (14) Besombes, J.; Cosnier, S.; Labbe, P.; Reverdy, G. Anal. Chim. Acta 1995, 311, 255-263. (15) Wang, J.; Nascimento, V. B.; Kane, S. A.; Rogers, K.; Smyth, M. R.; Angnes, L. Talanta 1996, 43, 1903-1907. (16) Smit, M. H.; Rechnitz, G. A. Anal. Chem. 1993, 65, 380-385. (17) Smit, M. H.; Rechnitz, G. A. Electroanalysis 1993, 5, 747-751.

Analytical Chemistry, Vol. 70, No. 4, February 15, 1998 807

Figure 1. Strategy for using NQS as a bioelectrocatalysis for oxygen reduction. The enzymatic reaction is turned on by applying a reductive potential of -150 mV and switched off by applying a potential of +100 mV. The resulting current is monitored by chronoamperometry.

carbamate.14 Wang et al. reported the use of a tyrosinase-based screen-printed biosensor for the determination of carbamate pesticides15 and remarked on the fast response of the system without the preincubation periods required with cholinesterase sensors.11 Previous research in our laboratory has involved the use of mediated bioelectrocatalysis of tyrosinase for the determination of cyanide16 and benzoic acid.17 Tyrosinase catalyzes the fourelectron reduction of molecular oxygen to water at the expense of substrates. By replacing the natural substrates with redox shuttles to carry electrons to the active site of the enzyme, we create an on/off switch for the bioelectrocatalytic reduction of oxygen. Figure 1 shows a schematic of this process using the redox mediator 1,2 naphthoquinone-4-sulfonate (NQS). We hope to take advantage of the improvements made by both the use of tyrosinase enzyme and the substitution of substrates to make a new pesticide sensor. This paper reports the use of this mediated bioelectrocatalytic system with a tyrosinase-based oxygen sensor for the detection of organophosphorus pesticides. EXPERIMENTAL SECTION Reagents. Tyrosinase (EC 1.14.18.1, 2400 units/mg) and glutaraldehyde (30%) were purchased from Sigma Chemical Co. and used without further purification. NQS was purchased from Aldrich, and solutions were made fresh daily. Anhydrous sodium phosphate was used for the preparation of 0.05 M phosphate buffer, pH 7.0 (PBS). Electrolyte solutions consisted of 0.1 M sodium chloride in PBS. Pesticides were purchased from Chem Service and used as received. Absolute ethanol from Quantum Chemical Co. was used to make diazinon analyte solutions, due to its low water solubility. Dichlorvos analyte solutions were made up in water, which was distilled and deionized to 18 MΩ/cm for all solutions. Instrumentation. Electrochemical measurements were performed using a computer-interfaced Bioanalytical Systems 100B potentiostat. A three-electrode setup was used in all experiments, consisting of a modified glassy carbon working electrode (3-mm diameter), a platinum wire counter electrode, and a Ag/AgCl reference electrode (against which all potentials are quoted). All electrodes were purchased from Bioanalytical Systems. The electrochemical cell consisted of a 15-mL constant temperature 808

Analytical Chemistry, Vol. 70, No. 4, February 15, 1998

cell held at 30 °C with a Fisher Scientific isotemp circulator, model 910. Experiments were run in 10 mL of PBS electrolyte with a NQS concentration of 50 µM. This concentration, based on previous work,17 was chosen so that the mediator, not dissolved oxygen, was the limiting reagent in the system. Analyte solutions were pipetted into the electrochemical cell, stirred for 15 s, and allowed to equilibrate for 1 min before application of the starting potential. Pipetted additions never exceeded 100 µL, 1% of PBS volume. Electrochemical data were imported into SigmaPlot (Jandel Scientific) spreadsheets for data analysis and plotting. Enzyme Electrode Construction. Enzyme electrode construction was based on previous work in our laboratory17 with the protocol as follows: glassy carbon electrodes were polished with 0.05-µm alumina and sonicated for 5 min. The method of Anjo et al.18 was used for activation of the electrode surface by holding the electrode at a potential of 1.5 V for 5 min in 1.0 M sodium hydroxide. Electrodes were then rinsed and air-dried, followed by the application of 20 µL of a 20 mg/mL tyrosinase solution. The enzyme layer was dried under vacuum for 15 min. Enzyme immobilization was carried out using a method described by Wilson et al.,19 where a 10-µL aliquot of 1% glutaraldehyde was added to the surface and allowed to react for 30 min. The resulting cross-linked enzyme was then again dried under vacuum for 15 min. Finally, enzyme-modified electrodes were stored overnight in PBS under refrigeration. Electrodes of this type are usually covered with a dialysis membrane; however, we found that this blocked access of analytes to the electrode surface and hindered reproducibility. Reproducibility was much higher, and stability of the electrodes did not appear to be affected without use of the membrane. RESULTS AND DISCUSSION Chronoamperometry. The mediator NQS is in its enzymatic unreactive state in bulk solution. Applying a reductive potential pulse of -150 mV converts the quinone to a reactive diol and facilitates its use as a bioelectrocatalyst. In turn, applying an oxidative pulse of 100 mV terminates the reaction (refer back to Figure 1 for the schematic of this process). The inhibitory effects of organophosphorus pesticides on enzyme activity are monitored using chronoamperometry by exploring the current decay during the 10-s reductive potential pulse. Figure 2 shows this inhibitory effect of diazinon in typical chronoamperometric profiles of NQS reduction at a tyrosinase-modified electrode. For clarity in showing the differences in the current decays at different diazinon concentrations, the inset in Figure 2 shows only the first 4 s of the 10-s potential pulse. In the presence of diazinon, the rate of enzymatic recycling of NQS is decreased due to the inhibitory effect of the pesticide. This is shown by the decrease in current with increasing concentration of pesticides. Control experiments involving electrochemical generation of the mediator without the presence of tyrosinase in the modifying layer showed no dependence on pesticide concentration. Inhibition in this system appears to be reversible, with the full catalytic activity returning (18) Anjo, D. M.; Kahr, M.; Khodabakhsh, M. M.; Nowinski, S.; Wanger, M. Anal. Chem. 1989, 61, 2603-2608. (19) Wilson, G. S.; The´venot, D. R. Unmediated amperometric enzyme electrodes. In Biosensors A Practical Approach; Cass, A. E. G., Rickwood, D., Hames, B. D., Eds.; The Practical Approach Series; Oxford University Press: New York, 1990; pp 1-16.

Figure 2. Inhibition effects of diazinon on the typical chronoamperometric current decay during the -150 mV potential pulse, with (solid curve) no diazinon and (dotted curve) 200 µM diazinon. (Inset) Effect of increasing concentration of diazinon, from top to bottom: 0, 50, 100, 200, 300, 500, and 1500 µM.

Figure 3. Effects of addition of ethanol on the typical chronoamperometric current decay during the -150 mV potential pulse, with (solid curve) no ethanol and (dotted curve) 200 µL of ethanol added to the electrochemical cell.

with removal of pesticides from the experimental cell and rinsing of the electrode with PBS. Data for dichlorvos were collected in the same manner and showed the same trends in enzymatic inhibition. Effects of Ethanol on Enzymatic Activity. Diazinon has very low aqueous solubility (40 mg/L),19 but it is miscible with ethanol. Therefore, concentrated analyte solutions of diazinon were made up in absolute ethanol and pipetted into the electrochemical cell for analysis. Since the presence of ethanol may have deleterious effects on enzymatic activity, control experiments were run to determine exactly what these effects might be on our sensor. For this control experiment, 200 µL of absolute ethanol was pipetted into the electrochemical cell. As mentioned in the Experimental Section, this injection was twice the concentration of ethanol injected during the study of analyte samples. Figure 3 shows chronoamperometric plots of the enzymatic reaction on the NQS mediator in the presence of different concentrations of ethanol. Steady-state current, measured in the last 5 s of the 10-s potential pulse, show values of 1.41 ( 0.06 µA without ethanol and 1.37 (

Figure 4. Inhibition effects of diazinon on the typical chronocoulometric data during the -150 mV potential pulse. Current was integrated from last 5 s of a 10-s potential pulse. Increasing concentration of diazinon, from top to bottom: 0, 50, 100, 200, 300, 500, and 1500 µM.

0.05 µA in the presence of 200 µL of ethanol. Since these values are within experimental error, we assume there is no effect on the enzymatic reaction in our sensor. Thus, the use of small amounts of ethanol in analyte solutions should be acceptable. Chronocoulometry. Charging current is responsible for the majority of current during the reductive potential pulse. With this in mind, removal of the nonfaradaic portion of the current is carried out during data analysis. Integrating current in the last 5 s of the 10-s potential pulse achieves this by neglecting current due to double-layer charging and using only the steady-state current of the enzymatic reaction for data analysis. Figure 4 shows typical chronocoulometric analysis, plotting integrated steady-state current from the last 5 s of the 10-s, -150 mV potential pulse. As can be seen in Figure 2, increasing diazinon concentration decreases the enzymatic recycling of NQS, and the reductive charge decreases. Chronocoulometric data are used to make calibration plots for diazinon and dichlorvos, which are shown in Figures 5 and 6. Data are expressed as percent inhibition, which is calculated by dividing the total charge in the presence of inhibitor by the charge with no inhibitor, 100 × {1 - [Qinhib/Qno inhib]}. Inherent differences in current levels from one electrode to the next become averaged out during the integration method and percent inhibition calculation. This normalization corrects for differences in enzyme electrode manufacturing and allows for better reproducibility than if calibration curves were made with chronoamperometric data. Error bars represent the standard deviation of at least three samples. Error determinations using chronoamperometric data showed error 2-5 higher than those obtained when chronocoulometric data were used. The data for the two pesticides are very different. Detection limits were determined by the lowest analyte concentration at which a measurable electrochemical change took place. Corresponding values for our pesticides were 5 µM for diazinon, with an average percent inhibition of 1.35 ( 0.55, and 75 nM for dichlorvos, with an average percent inhibition of 0.95 ( 0.44. The concentrations of the two pesticides where 50% inhibition of the enzymatic activity occurs are ∼1000 µM for diazinon and ∼50 µM Analytical Chemistry, Vol. 70, No. 4, February 15, 1998

809

Figure 5. Normalized calibration curve for enzymatic inhibition as a function of diazinon concentration. Data are from coulometric analysis and expressed as percent inhibition, 100 × {1 - [Qinhib/ Qno inhib]}. Error bars represent standard deviation of at least three samples. (Inset) Narrow range data difficult to see on large scale.

for dichlorvos. These differences between the two pesticides may be explained by the their water solubility limits. As mentioned earlier, the solubility limit of diazinon is 40 mg/L, where as dichlorvos’s limit is 10 g/L.20 These solubility differences may greatly affect solvation characteristics of the pesticides and, in turn, their access to the enzyme. It is important to note that, in experiments where diazinon analyte solutions were above the solubility limits, we were working with suspensions rather than true solutions. The small amount of ethanol introduced with the injected analyte solutions was not enough to dissolve the diazinon in these circumstances. This may explain the observation that the linearity of the calibration plot for diazinon begins to deviate above 100 µM, where we are close to reaching the solubility limit. Some researchers are working with adding water-miscible solvents in order to improve the sensitivity of enzyme electrode systems to these types of hydrophobic analytes.4,7,21 Similar systems are being considered by us for future work. (20) Worthing, C. R., Hance, R. J., Eds. The Pesticide Manual, 9th ed.; British Crop Protection Council: Farnham, Surrey, UK, 1991. (21) Saini, S.; Hall, G. F.; Downs, M. E. A.; Turner, A. P. F. Anal. Chim. Acta 1991, 249, 1-15.

810 Analytical Chemistry, Vol. 70, No. 4, February 15, 1998

Figure 6. Normalized calibration curve for enzymatic inhibition as a function of dichlorvos concentration. Data are from coulometric analysis and expressed as percent inhibition, 100 × {1 - [Qinhib/ Qno inhib]}. Error bars represent standard deviation of at least three samples. (Inset) Narrow range data difficult to see on large scale.

CONCLUSIONS A tyrosinase-based enzyme oxygen sensor was constructed and used for inhibition studies of the organophosphorus pesticides, diazinon and dichlorvos. This inhibition presumably occurs through competitive interaction of the pesticide on the mediator side of the enzymatic cycle. Such an interaction stops the electron transfer between NQS and the copper center of tyrosinase, preventing its enzymatic recycling. Our sensor is much more sensitive to the more water soluble pesticide dichlorvos. Because of solubility limits, diazinon samples were made up in ethanol, and control experiments showed that the low concentrations of ethanol actually introduced into the cell had no affect on the enzymatic reactions. Detection limits for dichlorvos and diazinon are 75 nM and 5 µM, respectively. Further studies are underway to improve sensitivity and selectivity of this sensor. ACKNOWLEDGMENT Financial support of this work was provided by the National Science Foundation (Grant CHE-9216304). Timsey L. Everett is acknowledged for help in the preparation of this manuscript. Received for review September 2, 1997. December 10, 1997. AC970958L

Accepted