Electrochemical Detection of Thiols with a Coenzyme Pyrroloquinoline

Financial support from the deArce Memorial Foundation at The University of Toledo and The Ohio Board of Regents Investment Fund for the development of...
0 downloads 0 Views 138KB Size
Download PDF
Anal. Chem. 2000, 72, 5755-5760

Articles

Electrochemical Detection of Thiols with a Coenzyme Pyrroloquinoline Quinone Modified Electrode Takayo Inoue and Jon R. Kirchhoff*

Department of Chemistry, University of Toledo, Toledo, Ohio 43606

A modified electrode sensor for the detection of thiols is described. The sensor was constructed by incorporation of the coenzyme pyrroloquinoline quinone (PQQ) into a polypyrrole (PPy) film on a glassy carbon electrode substrate by the electropolymerization of pyrrole in the presence of PQQ. The electrochemical properties of entrapped PQQ in the PPy film were influenced by the applied potential during electropolymerization and by film thickness, both of which were optimized to yield a stable and reproducible response for entrapped PQQ. The PQQ/ PPy sensor was utilized for the amperometric detection of cysteine, homocysteine, penicillamine, N-acetylcysteine, and glutathione. The response for each thiol in pH 8.42 borate buffer was found to be linear with detection limits (S/N ) 3) ranging from 13.2 µM for glutathione to 63.7 nM for cysteine with sensitivities of 0.023 nA/µM and 4.71 nA/µM, respectively. The response and detection limits were found to be sensitive to the nature of the thiol and the solution pH. Furthermore, in the presence of dopamine, ascorbic acid, or uric acid, the pH-dependent redox potential of the PQQ catalyst allows tuning of the detection potential to enhance the selectivity for thiols over these potential electroactive interferences. Thiols are important in biological systems due to their widespread occurrence in many proteins and in natural compounds such as glutathione, cysteine, coenzyme A, and lipoic acid.1 In particular, the reversible oxidation-reduction reactions between thiols and the corresponding disulfides (eq 1) are essential processes in many biological and chemical systems. In biological * Corresponding author: (phone) (419) 530-1515; (fax) (419) 530-4033; (email) [email protected]. (1) Friedman, M. The Chemistry and Biochemistry of the Sulfhydryl Group in Amino Acids, Peptides, and Proteins; Pergamon Press: Oxford, U.K., 1973. 10.1021/ac000716c CCC: $19.00 Published on Web 10/26/2000

© 2000 American Chemical Society

2RSH a RSSR + 2H+ + 2e-

(1)

systems, thiols are prominent marker molecules. For example, high levels of the disulfide cystine in the urine are indicative of kidney dysfunction,2 while elevated homocysteine levels have been implicated in coronary artery disease.3,4 In addition, the ratio of glutathione (GSH) to glutathione disulfide (GSSG) is a key marker for the redox status of cells and, thus, an indicator for cellular oxidative stress.5 Consequently, accurate determination of the changes in thiol levels provides critical insight into proper physiological functions or in the diagnosis of disease states. Numerous chemical and instrumental techniques for the determination of thiols have been reported. However, many suffer from difficulties with sample preparation, the need for derivatization, or the lack of sufficient sensitivity, all of which limit their utility. In contrast, electrochemical methods have been quite successful for the determination of free thiols. One initial challenge to the development of electrochemical methods for thiol detection was that direct oxidation of thiols at solid electrodes is slow and usually requires large overpotentials (g+1.0 V) to proceed.6,7 This problem is circumvented by indirect detection at mercury or mercury amalgam electrodes.8 Thiols interact with mercury-based electrodes to form stable mercury thiolate complexes. Detection at a less positive potential (+0.1 V vs Ag/AgCl) is then possible by the oxidation of Hg0 to Hg2+. Furthermore, when this ampero(2) Jocelyn, P. C. Biochemistry of the SH Group; Academic Press: London, 1972; p165. (3) McCully, K. S. The Homocysteine Revolution; Keats: New Canaan, CT, 1997. (4) McCully, K. S. Nat. Med. 1996, 2, 386-389. (5) Meister, A.; Anderson, M. E. Annu. Rev. Biochem. 1983, 52, 711-760. (6) Mefford, I.; Adams, R. N. Life Sci. 1978, 23, 1167-1174. (7) Kreuzig, F.; Frank, J. J. Chromatogr. 1981, 218, 615-620. (8) Rabenstein, D.; Saetre, R. Anal. Chem. 1977, 49, 1036-1039. (9) Saetre, R.; Rabenstein, D. Anal. Biochem. 1978, 90, 684-692. (10) Saetre, R.; Rabenstein, D. Anal. Chem. 1978, 50, 276-280.

Analytical Chemistry, Vol. 72, No. 23, December 1, 2000 5755

incorporates PQQ as a biocatalyst into a polypyrrole (PPy) matrix for the mediated detection of thiols is presented. The design and characterization of the sensor, as well as the amperometric detection of cysteine, homocysteine, penicillamine, N-acetylcysteine, and glutathione, are described. Figure 1. Redox reaction for PQQ.

metric detection scheme is coupled to a separation method such as liquid chromatography9-12 or capillary electrophoresis,13-15 the determination of thiols or the simultaneous determination of thiols and disulfides is possible in complex biological samples. One variation on this approach monitors the decrease in the reduction current for free Hg2+ at a tungsten wire electrode after complexation of Hg2+ to a thiol.16,17 Alternatively, several recent studies have used electrodes modified with inorganic electrocatalysts, such as cobalt phthalocyanine,18 prussian blue,19 ruthenium cyanide,20 aquocobalamin,21 copper hexacyanoferrate,22 and Bi doped in PbO2,23 to detect thiols and eliminate the need for a mercury-based electrode. The approach described here is the use of the redox cofactor pyrroloquinoline quinone (PQQ, 4,5-dihydro-4,5-dioxo-1H-pyrrolo [2,3-f] quinoline-2,7,9-tricarboxylic acid) as a biocatalyst for the detection of thiols. PQQ was discovered as a redox cofactor in bacterial alcohol dehydrogenases24 and more recently has been found widely distributed in eukaryotes. PQQ has been reported to function as an antioxidant,25 a tissue protective agent,26,27 and an essential nutrient.28 As part of its basic structure, PQQ contains an o-quinone moiety that exhibits an efficient, pH-dependent and reversible electron transfer between its oxidized and reduced forms as shown in Figure 1. Furthermore, PQQ catalyzes nonenzymatic reactions at moderate pH and temperature including the oxidation of thiols to disulfides.29,30 Thus, the unique redox properties of PQQ are such that a reduction in the overpotential for the oxidation of thiols is possible, leading to facile electrochemical detection. In this study, an electrochemical sensor that (11) Bergstrom, R. F.; Kay, D. R.; Wagner, J. G. Anal. Chem. 1978, 50, 21082112. (12) Allison, L. A.; Shoup, R. E. Anal. Chem. 1983, 55, 8-12. (13) O’Shea, T. J.; Lunte, S. M. Anal. Chem. 1993, 65, 247-250. (14) Lin, B. L.; Colon, L. A.; Zare, R. N. J. Chromatogr., A 1994, 680, 263-270. (15) Zhong, M.; Lunte, S. M. Anal. Chem. 1999, 71, 251-255. (16) Alexander, P. W.; Hidayat, A.; Hibbert, D. B. Electroanalysis 1994, 7, 290291. (17) Hidayat, A.; Hibbert, D. B.; Alexander, P. W. J. Chromatogr., B 1997, 693, 139-146. (18) Qi, X.; Baldwin, R. P. J. Electrochem. Soc. 1996, 143, 1283-1287. (19) How, W.; Wang, E. J. Electroanal. Chem. 1991, 316, 155-163. (20) Cox, J. A.; Gray, T. J. Electroanalysis 1990, 2, 107-111. (21) Li, H.; Li, T.; Wang, E. Talanta 1995, 42, 885-890. (22) Zhou, J.; Wang, E. Electroanalysis 1994, 6, 29-35. (23) Popovic, N. D.; Cox, J. A.; Johnson, D. C. J. Electroanal. Chem. 1998, 455, 153-156. (24) Salisbury, S. A.; Forrest, H. S.; Cruse, W. B. T.; Kennard, O. Nature 1979, 280, 843-844. (25) Hamagishi, Y.; Murata, S.; Kamei, H.; Oki, T.; Adachi, O.; Ameyama, M. J. Pharmacol. Exp. Ther. 1990, 255, 980-983. (26) Xu, F.; Mack, C. P.; Quandt, K. S.; Shlafer, M.; Massey, V.; Hultquist, D. E. Biochem. Biophys. Res. Commun. 1993, 193, 434-439. (27) Quandt, K. S.; Hultquist, D. E. Proc. Natl. Acad. Sci U.S.A. 1994, 91, 93229326. (28) Bishop, A.; Gallop, P. M.; Karnovsky, M. L. Nutr. Rev. 1998, 56, 287-293. (29) Itoh, S.; Kato, N.; Mure, M.; Ohshiro, Y. Bull. Chem. Soc. Jpn. 1987, 60, 420-422. (30) Park, J.; Churchich, J. E. BioFactors 1992, 3, 257-260.

5756

Analytical Chemistry, Vol. 72, No. 23, December 1, 2000

EXPERIMENTAL SECTION Apparatus. All electrochemical measurements were carried out in a conventional three-electrode electrochemical cell under anaerobic conditions with a Bioanalytical Systems (BAS, West Lafayette, IN) 100B electrochemical analyzer. Either a bare (3.0 mm in diameter) or modified glassy carbon (GC) electrode (BAS, MF-2012) was used as the working electrode. The bare GC electrode was polished with alumina (97% purity) was purchased from Fluka (Milwaukee, WI). Pyrrole (Aldrich, Milwaukee, WI) was distilled before use. The thiols L-cysteine, D,L-penicillamine, N-acetyl-Lcysteine, and glutathione were obtained from Sigma (St. Louis, MO), and D,L-homocysteine was obtained from Aldrich. Amino acids were purchased from several sources: D,L-alanine and glycine (Fisher, Pittsburgh, PA), L-phenylalanine, tyrosine, and L-tryptophan (Aldrich), D,L-methionine (Sigma), L-glutamine (Life Technolgies), and D,L-aspartic acid (Matheson, Coleman & Bell, Cincinnati, OH). Dopamine was obtained from Sigma. Uric acid and L-ascorbic acid were obtained from Fisher and J. T. Baker (Phillipsburg, NJ), respectively. All other chemicals were of analytical reagent grade and used as received. All solutions were prepared with distilled deionized water purified to a resistivity of at least 17 MΩ‚cm by a Barnstead B pure water purification system. Electropolymerization of Pyrrole on a GC Electrode. (i) Without PQQ. Electropolymerization of pyrrole was performed on a GC electrode according to the method of Shinohara et al.31 in a deoxygenated 0.1 M KCl aqueous solution containing 0.1 M pyrrole. Potentials of either +0.630 or +0.800 V were used to initiate polymerization. The thickness of the PPy film was controlled by monitoring the amount of charge passed during the polymerization reaction.31 (ii) With PQQ. A 1.55 mM PQQ solution was prepared by adding 0.51 mg of PQQ to 1.0 mL of a 0.1 M KCl aqueous solution. A few drops of 0.1 M KOH were added in order to dissolve PQQ completely. The PQQ solution was bubbled with oxygen for 30 min, followed by the addition of 7 µL of pyrrole. The electropolymerization of pyrrole on the GC electrode and entrapment of PQQ into the PPy film were achieved simultaneously by immersing a GC electrode into the PQQ/pyrrole solution and applying either +0.630 or +0.800 V. Deposition of PPy continued until the desired film thickness was achieved. Cyclic voltammetric measurements in 0.1 M phthalate buffer at pH 3.45 were conducted to evaluate the entrapment and stability of PQQ in the PPy film. (31) Shinohara, H.; Khan, G. F.; Ikariyama, Y.; Aizawa, M. J. Electroanal. Chem. 1991, 304, 75-84.

Table 1. Electrochemical Data for PQQ as a Function of PPy-Membrane Thicknessa charge (mC/cm2)

thicknessb (nm)

ipa (µA)

ipc (µA)

E°′ (mV)

∆Ep (mV)

14.1 28.3 56.6 100.3 211.6

18 35 71 125 267

no peak unmeasurable 0.242 (0.005) 2.10 (0.05) 11.1 (0.5)

no peak unmeasurable 0.362 (0.022) 2.64 (0.10) 10.9 (0.8)

119 (1) 110 (2) 116 (4)

34 (4) 49 (2) 107 (2)

a From cyclic voltammetry measurements (n ) 3) at a scan rate of 2 mV/s. Standard deviations are given in parentheses. E°′ ) (E p,a + Ep,c)/2; ∆Ep ) Ep,a - Ep,c. b Reference 31.

Figure 2. Schematic sensor representation for the detection of thiols.

Amperometric Detection of Thiols. Amperometric detection of L-cysteine, D,L-homocysteine, D,L-penicillamine, N-acetyl-L-cysteine, and glutathione was conducted at the PQQ/PPy-modified GC electrode by measuring the current response at a fixed potential of +0.500 mV as a function of time under anaerobic and hydrodynamic conditions at pH 3.45 (0.1 M phthalate), 8.42 (0.1 M borate), and 10.0 (0.1 M borate). A range of concentrations was prepared for each thiol. The electrochemical cell initially contained 3 mL of buffer solution at the appropriate pH into which successive 25-µL aliquots of each solution were injected via a gastight syringe. The response for each thiol concentration was measured at least in triplicate. The current response (∆i) was plotted versus the analyte concentration, where ∆i is defined as the difference between the steady-state current before and after injection of analyte solution (see Figure 5). PPy-modified and unmodified GC electrodes were used for control experiments. The amperometric response to various nonthiol compounds was also measured and compared with that of cysteine in order to evaluate the selectivity of the sensor. Experiments were conducted at several different applied potentials and solution pH values to investigate these effects on the sensor selectivity. A dietary supplement, L-cysteine 500 (General Nutrition Corp., Pittsburgh, PA), was analyzed for cysteine content. A tablet was weighed (737.8 ( 12.3 mg, n ) 10) and then ground thoroughly. A 2.05-mg portion of the powdered sample was dissolved in 3 mL of 0.1 M phthalate (pH 3.45) buffer, filtered with a 0.2-µm filter to remove the insoluble cellulose, and finally analyzed with the PQQ/ PPy sensor. RESULTS AND DISCUSSION Sensor Design, Characterization, and Evaluation. A schematic representation of the sensor design is shown in Figure 2. Entrapment of PQQ during the electropolymerization of pyrrole in the presence of PQQ produces efficient loading of the biocatalyst into the PPy film. Biocatalyst loading and entrapment is aided by ion-pairing of negatively charged PQQ with the positively charged PPy film. PPy acts not only as a matrix for PQQ immobilization but also as an efficient conductive pathway between PQQ and the GC electrode surface.

Figure 3. Cyclic voltammograms of PQQ/PPy-modified electrodes in 0.1 M phthalate buffer, pH 3.45, as a function of film thickness. Scan rate, 2 mV/s. Film thickness: (A) 267, (B) 125, and (C) 71 nm.

An applied potential of +0.630 V was used to prepare films for all sensor applications. Electrodes prepared at electropolymerization potentials of +0.800 V yielded smaller and more ambiguous voltammetric peaks for PQQ, which is probably a result of the poor homogeneity of the PPy film. Cyclic voltammograms of PQQ/PPy-modified electrodes of various film thicknesses were obtained in phthalate buffer (0.1 M, pH 3.45) and are shown in Figure 3 with the corresponding electrochemical data summarized in Table 1. As the film thickness and hence the loading of PQQ into the PPy film increased, larger peak currents for the PQQ redox reaction were observed. The peak separation (∆Ep) between the oxidation and reduction peaks also increased due to the increased resistance (R) of the thicker film. Consequently, a larger overpotential is required to effect the redox reaction of PQQ with the increased film thickness. The peak current for the reduction of PQQ was also investigated as a function of time for each film thickness (Figure 4). Examination of Figure 4 clearly shows that the peak current decreased with time for all cases due to leaching of the PQQ biocatalyst. However, the initial rate of decrease was largest for the electrode modified with the thinnest film. For example, when the film was as thin as 54 nm, the peak current decreased to ∼50% of its initial value after 15 h. In contrast, the peak current with a Analytical Chemistry, Vol. 72, No. 23, December 1, 2000

5757

Table 2. Summary of the Sensitivity and Limits of Detection (S/N ) 3) for Each Thiol at pH 3.45 and 8.42a pH 3.45

a

pH 8.42

analyte

slope (nA/µM)

LOD (µM)

slope (nA/µM)

LOD (nM)

cysteine homocysteine penicillamine N-acetylcysteine glutathione

0.249 (0.005) 0.106 (0.003) 0.127(0.004) 0.077 (0.003) 0.030 (0.001)

1.42 (0.03) 3.26 (0.09) 2.74 (0.09) 5.54 (0.21) 11.4 (0.6)

4.71 (0.14) 0.67 (0.02) 1.51 (0.05) 0.114 (0.004) 0.023 (0.001)

63.7 (1.8) 447 (13) 198 (7) 2.62 (0.10) × 103 1.32 (0.06) ×104

Average values given for three electrodes. Standard deviations are given in parentheses.

Figure 4. Peak current of entrapped PQQ from cyclic voltammetry as a function of time. Film thickness: ([) 54, (9) 71, (2) 125, and (b) 267 nm.

Figure 5. Amperometric response for cysteine: (A) PPy-modified electrode with entrapped PQQ and (B) PPy-modified electrode without entrapped PQQ. Amount injected: 25 µL of a 4.95 mM cysteine solution. Electrode potential, +0.500 V.

267-nm film decreased only slightly and maintained ∼95% of the original response. The response essentially stabilized after 2 days in all cases. An effective sensor requires ample loading and minimal leaching of the catalyst in combination with efficient and rapid electron transfer between the catalyst and the electrode 5758

Analytical Chemistry, Vol. 72, No. 23, December 1, 2000

substrate. From these results, a compromise in film thickness is necessary, not only to ensure the availability of sufficient catalyst but also to provide a rapid response for detection at the lowest possible applied potential. Therefore, a film thickness of 125 nm was chosen for further development of the thiol sensor. Amperometric Detection of Thiols by the PQQ/PPy Electrode. Electrochemical detection of thiols is mediated through the use of the PQQ biocatalyst. Initially, immobilized PQQ chemically oxidizes a thiol to its disulfide while being reduced to PQQH2 (Figure 2). Thiols are then quantitatively detected by monitoring the amperometric current resulting from the regeneration of PQQ by the oxidation of PQQH2 at the GC electrode. A typical amperometric response for thiol detection at a PQQ/PPymodified electrode is shown in Figure 5 for cysteine at pH 3.45. Immediately upon injection of cysteine into the electrochemical cell, the current response increased rapidly and reached a steady state (Figure 5A). In contrast, no response was detected for cysteine at a PPy-modified electrode prepared in the absence of PQQ (Figure 5B), which clearly demonstrates immobilized PQQ is a critical component for the catalytic oxidation and detection of thiols. Similar results were obtained for homocysteine, Nacetylcysteine, glutathione, and penicillamine. The percent relative standard deviation of the current response for repetitive injections (n g 3) of each thiol was found to be between 2.3 and 4.8%. The sensor response is dependent on the pH of the analyte solution. In all cases, calibration plots for the detection of cysteine, homocysteine, N-acetylcysteine, glutathione, and penicillamine at pH 3.45 and 8.42 exhibited a linear relationship (R2 > 0.99) between ∆i and thiol concentration. The comparative response as measured by the slope of the calibration plot for each thiol and pH, as well as the concentration detection limits (S/N ) 3), is summarized in Table 2. The relative sensitivity of the sensor follows the general trend cysteine > penicillamine > homocysteine > N-acetylcysteine > glutathione with detection limits in the micromolar to nanomolar range at pH 3.45 and 8.42; the lowest detection limit was 63.7 nM for cysteine at pH 8.42. These data clearly suggest that the solution pH relative to the thiol pKa is an important factor governing the current response of the sensor due to the ease of oxidation of RS- relative to RSH.32 The enhancement of the current response at pH 8.42 is greatest for cysteine and penicillamine with pKa values of 8.3 and 7.9, respectively. Homocysteine (pKa ) 8.9) and N-acetylcysteine (pKa ) 9.5) also showed enhancement at pH 8.42, although less than that observed for cysteine and penicillamine. The response for glutathione (pKa ) 8.8) essentially remained the same. The response of the PQQ/ PPy sensor for thiol detection was also examined at pH 10. The

current responses for the sensor toward thiol detection, as well as for the cyclic voltammograms of entrapped PQQ, were highly variable and in some cases rapidly decreased with time. Yamamoto et al.33 reported that rapid and irreversible overoxidation of PPy occurs at potentials above 1.0 V and consequently results in loss of conductivity. Slower, irreversible oxidation also occurs at lower potentials and is enhanced at higher pH values. These factors presumably account for the sensor instability at pH 10. Further complicating detection at pH 10 is the general loss of reversibility for the PQQ redox reaction under basic conditions. At pH 8.42, a decrease in conductivity for the PPy film was gradual and occurred to only a minimal extent, while no significant loss of conductivity was detected at pH 3.45. The observation that the response for glutathione is unaffected by pH indicates a complicated interplay of specific thiol characteristics and properties such as size, structure, and redox potential may also influence the sensor response to varying extents. The relatively large size presumably restricts glutathione from entering the polymer film, which has a small pore size.34 Thus, only PQQ near the surface can participate in the redox reaction, resulting in a comparatively small response. On the other hand, smaller thiols such as cysteine may enter the PPy film and maximize the PQQ-thiol interaction resulting in a larger relative current response. Selectivity. A high degree of selectivity is critical for the development and application of sensor systems. In order for the PQQ/PPy sensor to be effective for the selective detection of thiols, the response of potential interferences commonly found in biological samples needs to be examined. These include species that either undergo a direct electrochemical reaction at the modified electrode or simply react with PQQ to indirectly produce a response due to the regeneration of PQQ by the electrode. For example, PQQ is reported to be involved in many enzymatic reactions, e.g., alcohol dehydrogenases, and therefore could potentially react with simple alcohol substrates.24 Similarly, PQQ could also react with nucleophillic substances found in biological media such as amines and amino acids.35 At pH 3.45 and a detection potential of +0.500 V, the PQQ/PPy sensor had either no response or a minimal response to the following group of potential interfering agents in comparison to the response of an injection of the buffer alone and at concentrations equal to or greater than used for thiol detection: methanol, ethanol, glycine, alanine, aspartic acid, phenylalanine, tyrosine, glutamine, and tryptophan. Response toward the thioether methionine was similar to that observed for the amino acids and alcohols. In contrast, the electroactive species ascorbic acid and uric acid, which are notorious interferences for analyses in biological media, produced a large response and would constitute a major interference for thiol detection under conditions where either species is present. Dopamine also yielded a large response at the PQQ/PPy electrode. To counter this problem, a unique selectivity advantage of the PQQ/PPy electrode is derived from the pHdependent electrochemical response of PQQ and, thus, the ability (32) Burner, U.; Jantschko, W.; Obinger, C. FEBS Lett. 1999, 443, 290-296. (33) Yamamoto, H.; Ohwa, M.; Wernet, W. J. Electroanal. Chem. 1995, 397, 163-170. (34) Wang, J.; Chen, S. P.; Lin, M. S. J. Electroanal. Chem. 1989, 273, 231242. (35) Ohshiro, Y.; Itoh, S. In Principles and Applications of Quinoproteins; Davidson, V. L., Ed.; Dekker: New York, 1993; Chapter 12.

Figure 6. Comparison of the pH-dependent oxidation potentials from Osteryoung square-wave voltammetry for the (9) PQQ/PPy electrode, and (2) uric acid, ([) dopamine, and (b) ascorbic acid at a GC working electrode.

Figure 7. Amperometric current response at the PQQ/PPy-modified electrode for (9) ascorbic acid, (2) dopamine, and (b) uric acid relative to cysteine at pH 8.42 as a function of applied potential. Each point represents the response from 25-µL injections of 0.17 mM stock solutions of cysteine, dopamine, ascorbic acid, and uric acid.

to tune its redox potential with pH. This feature provides flexibility in the applied detection potential of the sensor and allows the response of potential electroactive interferences to be minimized. Figures 6 and 7 illustrate this point. Figure 6 depicts the pH-dependent oxidation potential as determined by Osteryoung square wave voltammetry for the PQQ/PPy electrode, and for dopamine, ascorbic acid, and uric acid at a GC electrode. By carefully selecting a pH for the analysis, a detection potential in the range between the oxidation potential of PQQ and the interference can be used such that excellent selectivity is achieved for thiols. Figure 7 shows a plot of the ratio of the current response from dopamine, ascorbic acid, and uric acid to the current response from cysteine at pH 8.42 and confirms this point. As the detection potential is moved more negative, the effect of the interference decreases. Detection potentials of e -0.1 V for ascorbic acid and e +0.1 V for uric acid and dopamine provide Analytical Chemistry, Vol. 72, No. 23, December 1, 2000

5759

analyses but also much lower detection limits.36

Figure 8. Amperometric response at the PQQ/PPy-modified electrode for L-cysteine in a dietary supplement tablet. Amount injected: 25 µL of a 2.05-mg tablet/3 mL of phthalate buffer pH 3.45. Electrode potential, +0.500 V.

the best selectivity for thiol determination in the presence of these compounds. Determination of Cysteine in a Dietary Supplement. The PQQ/PPy sensor was used for the detection and quantitation of L-cysteine in a common over-the-counter dietary supplement tablet. An excellent response was observed for cysteine in pH 3.45 phthalate buffer at an applied potential of +0.500 V (Figure 8). The amount of cysteine in the L-cysteine 500 tablets was determined to be 477 ( 14 mg (n ) 6) with a coefficient of variation of 3.0%. This result demonstrates the utility of the PQQ/PPy sensor for quantitation in a single-component analysis. Because the sensor responds to a class of molecules rather than a specific analyte, additional selectivity is needed to achieve the level of specificity required for analysis of thiols in a complex mixture such as the various biological media. Similar to the mercury film electrode,9-15 the PQQ/PPy sensor can therefore be used as a detector for a separation method. Along these lines, we are currently miniaturizing the sensor to use as a thiol-selective detector for capillary electrophoresis, which will provide not only multicomponent

5760

Analytical Chemistry, Vol. 72, No. 23, December 1, 2000

CONCLUSIONS A modified electrode sensor system has been described and well characterized for the detection of thiols. Indirect detection of cysteine, homocysteine, N-acetylcysteine, glutathione, and penicillamine at micromolar to nanomolar levels is achieved by the incorporation of the coenzyme, PQQ, into a PPy conducting polymer film. PQQ acts as a catalyst by initiating the chemical oxidation of thiols in solution. Thiols are then quantitatively detected by monitoring the amperometric current resulting from the direct regeneration of the PQQ catalyst by the oxidation of PQQH2 at the GC electrode substrate. A key advantage of the PQQ/PPy electrode is the ability to enhance sensitivity and selectivity through adjustments in the pH of the analysis solution. Although other factors such as thiol size and redox potential influence the response, more basic conditions tend to increase the sensitivity of the sensor due to the formation of the more easily oxidized RS- form. Furthermore, as the pH is increased, the oxidation potential for the PQQ redox reaction shifts to more negative potentials. Such tuning of the redox potential allows the detection potential to be varied to minimize the response from electroactive substances commonly found in biological samples. A further advantage is the simplicity of the sensor design that utilizes an entrapped enzyme cofactor as the biocatalyst. PQQ directly undergoes electron transfer with the electrode without the need for a mediator redox couple that is found in many enzyme biosensor designs. ACKNOWLEDGMENT Financial support from the deArce Memorial Foundation at The University of Toledo and The Ohio Board of Regents Investment Fund for the development of a microanalytical laboratory is gratefully acknowledged.

Received for review June 21, 2000. Accepted September 25, 2000. AC000716C (36) Inoue, T.; Kirchhoff, J. R., manuscript in preparation.