Indirect Electrochemical Detection of Type-B Trichothecene

Anal. Chem. 1999, 71, 4075-4080

Indirect Electrochemical Detection of Type-B Trichothecene Mycotoxins Chen-Chan Hsueh, Yi Liu, and Michael S. Freund*,†

Department of Chemistry, Lehigh University, Bethlehem, Pennsylvania 18015

Trichothecene mycotoxins in animal feed and human food can cause fatalities in livestock and disease in humans. In addition, these toxins are suspected chemical warfare agents. Therefore, development of a simple and sensitive method for the screening of trichothecenes is important to prevent economic loss and health hazards. A simple and inexpensive method for the detection of type-B trichothecene mycotoxins has been developed in our laboratory. By hydrolyzing the toxin under basic conditions at 80 °C for 1 h it is possible to detect the toxin with simple electrochemical techniques. Deoxynivalenol (DON), commonly known as vomitoxin, was used as a representative compound for type-B trichothecenes in this detection scheme. The detection limit for DON using our procedures was determined to be 9.1 µM in solution, corresponding to 0.24 ppm in a 25-g grain sample if the final extraction volume is 2.2 mL. The linear dynamic detection range was determined to be from 0.32 ppm to greater than 32 ppm. In addition to standard solutions, this method was used on rice samples spiked with DON. It was demonstrated that there is no electrochemical interference from rice extract and that 1 ppm of DON in rice samples can be quantified. This method may be ideal for toxin screening in animal feeds or in runoff from sites that produce the compounds as chemical warfare agents. Since the active moiety in DON is common to virtually all type-B trichothecenes, our approach may be ideal for typespecific screening. Trichothecenes are metabolites produced by fungi such as Fusarium, Myrothecium, Stachybotrys, and others.1 The toxicity of trichothecenes to microorganisms, plants, animals, and humans has been known for decades. Their occurrences in agricultural products continue to be a worldwide problem.2-6 Toxicology † Current address: Beckman Institute, California Institute of Technology, Pasadena, CA 91125; (phone) 626-395-2420; (fax) 626-564-9672; (e-mail) [email protected]. (1) Riemann, H.; Bryan, F. L. Food-Borne Infections and Intoxications, 2nd ed.; Academic Press: New York, 1979. (2) Jelinek, C. F.; Pohland, A. E.; Wood, G. E. J. AOAC 1989, 72 (2), 223. (3) Park, J. J.; Smalley, E. B.; Chu, F. S. Appl. Environ. Microbiol. 1996, 62 (5), 1642-1648. (4) Kim, J.-C.; Kang, H.-J.; Lee, D. H.; Lee, Y.-W.; Yoshizawa, T. Appl. Environ. Microbiol. 1993, 59 (11), 3798-3802. (5) Abouzied, M. M.; Azcona, J. I.; Braselton, W. E.; Pestka, J. J. Appl. Environ. Microbiol. 1991, 57, 672-677. (6) Tanaka, T.; Hasegawa, A.; Yamamoto, S.; Lee, U.-S.; Sugiura, Y.; Ueno, Y. J. Agric. Food Chem. 1988, 36, 979-983.

10.1021/ac981114k CCC: $18.00 Published on Web 08/11/1999

© 1999 American Chemical Society

effects of trichothecenes include enteritis, emesis, dermonecrosis, gastroenteritis, oral necrosis, and gastroenteric necrosis in livestock.7 Low-level affects in humans include nausea, vomiting, diarrhea, and immunosuppression.7 Due to the frequent occurrences of trichothecenes in crops, they continue to be a threat to human health as well as the economy. Trichothecenes have also been used as chemical warfare agents. Their use as chemical weapons has been reported in Southeast Asia and Afghanistan in the late 1970s (known as yellow rain).8-11 More recently, the Iraqi government has been suspected by the United Nations Special Commission (UNSCOM) of producing trichothecenes as a biological weapon.12 The trichothecenes of greatest concern as warfare agents include the T-2 toxin, 4-deoxynivalenol (DON), diacetylnivalenol, and nivalenol with LD50 (intraperitoneal, in mice) of 5.2, 70.0, 9.6 and 4.0 mg/kg, respectively.7,13 More than 80 different trichothecenes have been isolated and identified from natural sources and have been categorized into types according to their chemical structure, where the common element consists of a 12,13-epoxytrichothec-9-ene ring. Among the four types of trichothecenes (A-D), type-A and type-B (see Figure 1) are the most well studied due to their toxicity and frequent occurrences in agricultural products. A major portion of trichothecene contamination in agricultural products consists of T-2 toxin and scirpentriol (type-A) as well as DON and nivalenol (type-B) and their derivatives. However, the natural occurrence of T-2 toxin and scirpentriol is less frequent than that of DON and nivalenol.14 The concentrations of both type-A and -B trichothecenes produced by fungi are often correlated, and studies of both have shown that the concentrations of type-B trichothecenes are typically 2 times higher than type-A.3 Among these compounds, DON, commonly known as vomitoxin, is the most important and most well studied due to its frequent occurrence in North American grains. Due to the concern over contamination of grains by DON, the FDA has issued an advisory to federal and state officials recommending a (7) Smith, J. E.; Moss, M. O. Mycotoxins: Formation, Analysis and Significance; John Wiley & Sons: New York, 1985; Chapters 4 and 5. (8) Haig, A. M., Jr. Chemical Warfare in Southeast Asia and Afghanistan. Report to the congress from Secretary of State, special report No. 98, March 1982. (9) Shultz, G. P. Chemical Warfare in Southeast Asia and Afghanistan: An Update. Report from Secretary of State, special report No. 104, November 1982. (10) Ember, L. R. Chem. Eng. News 1984, (Jan 9), 8. (11) Wade, N. Science 1981, 214, 34. (12) Zilinskas, R. A. JAMA, J. Am. Med. Assoc. 1997, 278 (5), 418-424. (13) Tatsuno, T.; Saito, M.; Enomoto, M.; Tsunoda, H. Chem. Pharm. Bull. 1968, 16, 2519-2520. (14) Laure, R. D.; Agnew, M. P. J. Agric. Food Chem. 1991, 39, 502-507.

Analytical Chemistry, Vol. 71, No. 18, September 15, 1999 4075

Figure 1. Structures of some common type-A and -B trichothecenes.

level of concern for DON of 2.0 ppm in wheat entering the milling process, 1.0 ppm in finished wheat products for human consumption, and 4.0 ppm for animal feed ingredients.15 Several analytical methods have been developed to detect these toxins. The most commonly used methods include GC/ECD,14,16-19 TLC/MS,20 TLC/fluorescence,21-23 HPLC/UV,14 HPLC/MS,24,25 and GC/MS.26-28 These methods require significant amounts of time associated with labor-intensive cleanup, sophisticated instrumentation, and skilled operators. More recently, enzyme-linked immunosorbent assays (ELISA) have became more popular.29-32 (15) Wood, G. E. J. Anim. Sci. 1992, 70, 3941-3949. (16) Croteau, S. M.; Prelusky, D. B.; Trenholm, H. L. J. Agric. Food Chem. 1994, 42, 928-933. (17) Moller, T. E.; Gustavsson, H. F. J. AOAC Int. 1992, 75 (6), 1049. (18) Scott, P. M.; Kanhere, S. R.; Tarter, E. J. J. AOAC 1986, 69, 889-893. (19) Cohen, H.; Lapointe, M. J. AOAC 1984, 67, 1105-1107. (20) Brumley, W. C., Trucksess, M. W.; Adler, S. H.; Cohen, C. K.; White, K. D.; Sphon, J. A. J. Agric. Food Chem. 1985, 33, 326-330. (21) Eppley, R. M.; Trucksess, M. W.; Nesheim, S.; Thorpe, C. W.; Pohland, A. E. J. AOAC 1986, 69, 37-40. (22) Trucksess, M. W.; Flood, M. T.; Mossoba, M. M.; Page, S. W. J. Agric. Food Chem. 1987, 35, 445-448. (23) Shannon, G. M.; Peterson, R. E.; Shotwell, O. L. J. AOAC 1985, 68, 11261128. (24) Voyksner, R. D.; Hagler, W. M., Jr.; Swanson, S. P. J. Chromatogr. 1987, 394, 183-199. (25) Kostiainen, R.; Matsuura, K.; Nojima, K. J. Chromatogr. 1991, 538, 323330. (26) Mossoba, M. M.; Adams, S.; Roach, J. A. G.; Trucksess, M. W. J. AOAC Int. 1996, 79 (5), 1116. (27) Krishnamurthy, T.; Sarver, E. W. Anal. Chem. 1987, 59, 1272-1278. (28) D’agostino, P. A.; Provost, L. R.; Drover, D. R. J. Chromatogr. 1986, 367, 77-86. (29) Usleber, E.; Martlbauer, E.; Dietrich, R.; Terplan, G. J. Agric. Food Chem. 1991, 39, 2091-2095. (30) Casale, W. L.; Pestka, J. J.; Hart, L. P. J. Agric. Food Chem. 1988, 36, 663668. (31) Schmitt, K.; Martlbauer, E.; Usleber, E.; Gessler, R.; Lepschy, J.; Abramson, D. ACS. Symp. Ser. 1996, No. 621, 314-321, (Immunoassays for Residue Analysis).

4076 Analytical Chemistry, Vol. 71, No. 18, September 15, 1999

Although ELISA may be ideal for detecting these toxins without extensive cleanup through the use of antibodies, a major drawback is that their high degree of specificity limits their use to individual toxins. As a result, ELISA does not work well for type-specific screening.33 A rapid, simple, and inexpensive method for type-specific trichothecene detection that does not require extensive cleanup and sophisticated laboratory equipment would be beneficial for routine screening of trichothecenes in food and animal feeds. Electrochemical methods can be ideal for such tasks since their selectivity often requires less intensive cleanup, in turn allowing rapid and inexpensive detection. The advantages of electrochemical detection are evident in the recent developments of chemical and biological sensors utilizing electrochemical detection schemes.34 However, due to the difficulty in reducing or oxidizing trichothecenes, electrochemical methods do not currently play a major role in their detection. Only a few studies have used electrochemical detectors (amperometric detection) combined with HPLC separation for detection of trichothecenes.35,36 The applied potentials used in those studies were either -1.4 or +1.0 V in basic solution. To the best of our knowledge, there are only two publications that have studied the electrochemistry of trichothecenes in detail.37,38 In those studies, it was demonstrated that type-A compounds are electrochemically inactive, while type-B and -C compounds could be reduced at potentials of ∼-1.4 V versus SCE on Hg electrodes. At such extreme potentials, however, electrochemical selectivity is compromised and extensive deoxygenation is required. In our laboratory, we have developed a novel detection scheme based on the electrochemical detection of the hydrolysis products of type-B trichothecenes. It has been reported in the literature that type-B compounds such as DON and acetyl-DON hydrolyze in basic solution upon heating.39,40 We have found some of the hydrolysis products to be easily oxidized at potentials near +0.6 V on glassy carbon electrodes and carbon fiber ultramicroelectrodes. In this paper, we discuss the electrochemical properties of these hydrolysis products and their use for the indirect detection of trichothecenes. Based on the reaction pathway reported in the literature, our work can be extended to virtually all type-B trichothecenes since they have the same functional groups in the R7 and R15 positions as DON. EXPERIMENTAL SECTION Chemicals and Electrodes. DON, T-2 toxin, verrucarin A, and verrucarol were provided by Professor Bruce Jarvis in Department of Chemistry & Biochemistry, University of Maryland at College Park. 2-(2-Hydroxyethoxy)phenol was purchased from (32) Park, J. J.; Chu, F. S. J. AOAC Int. 1996, 79 (2), 465-471. (33) Chu, F. S. Vet. Hum. Toxicol. 1990, 32 (Suppl.), 42-50. (34) Taylor, R. F.; Schultz, J. S. Handbook of Chemical and Biological Sensors; IOP Publishing Ltd.: Philadelphia, 1996. (35) Sylvia, V. L.; Phillips, T. D.; Clement, B. A.; Green, J. L.; Kubena, L. F.; Heidelbaugh, N. D. J. Chromatogr. 1986, 362, 79-85. (36) Childress, W. L.; Krull, I. S.; Selavka, C. M. J. Chromatogr. Sci. 1990, 28, 76. (37) Palmisano, F.; Bottalico, A.; Lerario P.; Zambonin, P. G. Analyst 1981, 106, 992-998. (38) Visconti, A.; Bottalico, A. Anal. Chim. Acta 1984, 159, 111-118. (39) Young, J. C.; Blackwell B. A.; ApSimon, J. W. Tetahedron Lett. 1986, 27 (9), 1019-1022. (40) Young, J. C. J. Agric. Food Chem. 1986, 34, 919-923.

Figure 2. Mechanism of DON hydrolysis in basic solution and its products.

Aldrich. Al2O3 powder (neutral, 150 mesh) was purchased from Aldrich, and decolorizing carbon power (Norit-A, alkaline) was obtained from Fisher. All compounds were used as received without any further purification. A glassy carbon disk electrode was used as the working electrode. A silver wire was used as a quasi-reference electrode. Pt foil was used as a counter electrode. Working electrodes were polished with 3-µm diamond paste and then colloidal silica. After polishing, electrodes were ultrasonicated for at least 3 min in distilled/deionized water. Instrumentation and Sample Preparation. A BAS 100A potentiostat was used for all electrochemical measurements. Cyclic voltammetry was performed from +200 to +800 mV with scan rates ranging from 5 to 500 mV/s. Waveform parameters for square-wave voltammetry (SWV) included the following: step potential of 4 mV, square-wave amplitude of 25 mV, and squarewave frequency of 15 Hz. After each measurement, the surface concentration of analytes was allowed to reestablish initial conditions by stirring the solution. Sample preparation (hydrolysis) was accomplished in the following steps. A 3-mL aqueous solution of DON in 0.1 M NaOH was heated at 80 °C for 1 h. To prevent evaporation, the solution was kept in a capped vial while being heated. After the solution returned to room temperature, it was then acidified with 2 mL of 1 M H2SO4. Extraction of DON from Rice Samples. A commercial sample of rice was ground into fine powder at 12 200 rpm in a mill for 1.5 min. A 25-g portion of rice powder was then spiked with a 12.5 µg/mL DON/methanol stock solution. The spiked sample was allowed to dry before addition of 100 mL of an 85:15 acetonitrile/water extraction solution. The sample was then vigorously stirred for 1 h. The extract was filtered through a cleanup column41 (1.5 cm diameter) packed with a mixture (41) Romer, T. R. J. AOAC 1986, 69 (4), 699-703.

containing 5.0 g of Al2O3 powder and 0.15 g of carbon powder. After filtration, 70 mL of the filtrate was evaporated to dryness with a rotor evaporator at 50 °C. The residue was redissolved in 1.5 mL of 0.1 M NaOH and heated at 80 °C for 1 h. After being cooled to room temperature, the extract was acidified with 0.7 mL of 1.0 M H2SO4. The final volume of the extract was 2.2 mL. RESULTS AND DISCUSSION Electrochemical Properties of the Hydrolysis Products of DON. It has been reported that DON and acetyl-DON in basic solutions at temperature of 75 °C degrade to at least four products, norDON-A, norDON-B, norDON-C, and lactone. The mechanism of forming norDON-A, -B, and -C has been proposed by Young’s group and is shown in Figure 2.39,40 We have found some of the hydrolysis products to be electroactive. On the basis of the structural information, norDON-B and norDON-C were suspected to be responsible for the electroactivity due to the catechol-like functional group on norDON-B and norDON-C. A model compound, 2-(2-hydroxyethoxy)phenol, whose structure is similar to the suspected electroactive portion of norDON-B and norDON-C was used to verify this hypothesis. Figure 3 shows the cyclic voltammograms of 2-(2-hydroxyethoxy)phenol and a hydrolyzed DON sample. For purposes of comparison, the response of 2-(2hydroxyethoxy)phenol has been normalized to the same peak current obtained for the hydrolyzed DON sample. The redox wave of 2-(2-hydroxyethoxy)phenol was found to be almost identical to that of the hydrolysis products of DON, suggesting that the hydrolysis products and 2-(2-hydroxyethoxy)phenol share the same electroactive moiety. The anodic peak potential of the hydrolyzed sample occurs at ∼600 mV at pH 0.5. The absence of a cathodic peak in the return scan suggests that the oxidation products of norDON-B and norDON-C are rapidly converted to another, electrochemically inactive, species. We propose that the redox mechanism involves 2 e-, 1 H+ transfer followed by an Analytical Chemistry, Vol. 71, No. 18, September 15, 1999

4077

Figure 3. Cyclic voltammograms of 1.2 mM 2-(2-hydroxyethoxy)phenol (dashed line) and hydrolysis products of 1.2 mM DON (solid line). The voltammogram of 2-(2-hydroxyethoxy)phenol was scaled down for the comparison. Scan rate was 10 mV/s. The electrolyte was 0.37 M H2SO4/0.03 M Na2SO4. The reference electrode was a Ag quasi-reference electrode (AgQRE).

Figure 5. Square-wave voltammograms of different concentrations (1.2 mM, 0.24 mM, 60 µM, and 12 µM) of DON and the blank solution. The electrolyte was 0.37 M H2SO4/0.03 M Na2SO4. The reference electrode was a AgQRE.

Figure 6. Calibration curve of DON (peak current versus concentration).

Figure 4. Proposed mechanism for the electrochemical reactions of norDON-B and norDON-C.

irreversible chemical reaction with water (Figure 4) in a manner similar to that reported for tocopherol.42 A subsequent ringopening reaction is possible resulting in quinone species. However, no redox peaks are observed at +0.2 V, which is the expected for a quinone species. Therefore we conclude that the ring-opening reaction does not occur or occurs slowly on the time scale of the cyclic voltammetric experiments. The structural differences between norDON-B and norDON-C are too minor to exhibit any measurable differences in their electrochemical properties as indicated by the single peak in Figure 3. As a result, it is only possible to determine the conversion efficiency from DON to all electroactive products. By comparing the peak current of 1.2 mM hydrolyzed DON to the peak current of 1.2 mM 2-(2-hydroxyethoxy)phenol under identical conditions, the conversion efficiency was estimated to be 18%. The redox reaction of the electroactive products produced diffusioncontrolled responses with no evidence of adsorption as indicated by a slope of 0.5 from a log-log plot of peak current versus scan rate (not shown). (42) Dryhurst, G.; Kadish, K. M.; Scheller, F.; Renneberg, R. Biological Electrochemistry; Academic Press: New York, 1982; Vol. I, p 72.

4078 Analytical Chemistry, Vol. 71, No. 18, September 15, 1999

The redox reaction of the hydrolyzed DON proved to be very well behaved. For example, the scan-to-scan reproducibility of the responses shows that the oxidation products do not foul the electrode surface, which is typically a concern for electrochemical methods. After scanning 35 times, the final cyclic voltammogram was almost identical to the first with only a slight decrease in peak height. Following an initial decrease in peak current of 2.1% in the first four scans, the peak current decreased by only 1.4% over a subsequent 30 scans. Thus it was concluded that surface fouling is minimal under our solution conditions. Quantification of DON through Its Hydrolysis Products. Cyclic voltammetry is a powerful technique for the characterization of redox reactions. However, it is less sensitive for analytical applications due to its relatively large background. For calibration and quantification in this work, SWV was used. Figure 5 shows the SWVs of various concentrations of DON in solution. Although glassy carbon electrodes were used for all the data presented in this work, we have obtained similar results with carbon fiber ultramicroelectrodes. Figure 6 shows a calibration curve for DON using our procedure. The number of measurements at each concentration was 12. The sensitivity for this method was determined to be 4.2 µA/mM. The limit of detection for DON, based on a 99% confidence interval, was determined to be 9.1 µM in solution. On the basis of these results, if a 25-g sample was used and the total final extraction volume was 2.2 mL, the limit of detection would

Figure 7. (A) Square wave voltammograms of 0, 1, 3, 4, and 5 ppm of DON in 25 g of rice samples. The final extraction volume was 2.2 mL. (B) Background-subtracted square-wave voltammograms of 1, 3, 4, and 5 ppm of DON in rice samples. Subtraction was performed by subtracting an estimated linear baseline beneath the peaks.

correspond to 0.24 ppm by weight in the sample. The linear dynamic range was determined to be from 12 µM to greater than 1200 µM, corresponding to a range of 0.32 ppm to greater than 32 ppm. The level of concern for DON, issued by the FDA, is of 2.0 ppm DON for wheat entering the milling process, 1.0 ppm in finished wheat products for human consumption, and 4.0 ppm for animal feed ingredients. Our results suggest that the detection range of our method is suitable for detecting trichothecenes in typical agricultural samples. On the basis of literature reports, the time required to obtain a maximum yield of the electroactive products under conditions of 0.1 M NaOH and heating at 75 °C is 1 h. Ideally, a shorter hydrolysis time will make this method more appealing. Since the hydrolysis rate is greatly dependent on the temperature as well as the pH, the hydrolysis time may be shortened if a higher temperature and/or pH is used. Quantification of DON in Spiked Rice Samples. Rice samples were used to demonstrate the utility of this method for the detection of DON in complex samples. Figure 7A shows the voltammograms of DON obtained from filtered rice extracts. The concentrations of DON in the original rice samples were 0, 1, 3, 4, and 5 ppm. The control sample (0 ppm) shows no redox peak on top of the sloping background. The spiked samples show similar sloping background currents with a visible oxidation peak at ∼700 mV, ∼100 mV more positive than that observed in standard solutions. The sloping background and shifted peak potential are likely due to matrix effects in the rice extract but do not adversely influence the analysis using this method. Following background subtraction, by extrapolating a linear background under the peaks, symmetrical responses were obtained (Figure 7B). Figure 8 shows the calibration curve determined for DON in the filtered rice extracts. The sensitivity derived from the calibration curve was 0.18 µA/ppm for a 25-g sample with a 2.2mL final extraction volume. The average peak current for 1 ppm DON was determined to be 0.102 µA (n ) 6) at a peak potential of 736 mV. The standard deviation of the background current for the blank rice extract at 736 mV was 0.009 µA (n ) 5). This results in a S/Nblank for 1 ppm DON in the rice extract of 11, which exceeds the level of quantitation (S/N ) 10). It can therefore be

Figure 8. Calibration curve of DON in rice samples. The sensitivity is 0.134 µA/ppm and the correlation coefficient is 0.995.

concluded that this method is useful for quantifying DON concentrations in rice samples to less than 1 ppm.43 Type-Specific Detection with the Indirect Electrochemical Method. In this work, DON was used as a representative compound for the type-B trichothecenes. Young’s group has reported that acetyl-DON and diacetyl-DON rapidly convert to DON, and triacetyl-DON hydrolyzes to isoDON in less than 30 s in basic solutions at a temperature of 75 °C. Consequently, acetyl derivatives of DON and DON have the same hydrolysis products. Close examination of the hydrolysis reaction pathway suggests that the formation of the electroactive portion of the products only involves the R7 and R15 functional groups in DON. On the basis of this reaction pathway, most of common type-B toxins such as DON, nivalenol, triacetyl-DON, diacetyl-DON, acetyl-DON, and fusarenon-X should produce similar electroactive products since at some point during their hydrolysis they all have the same R7 and R15 functional groups. As a result, our method can be used for type-B-specific screening. We have also explored the possibility that other trichothecenes may hydrolyze and form electroactive products in a manner similar to type-B. However, we did not find similar behavior with type-A (T-2, and varrcarol) or type-C (verrucarin-A) trichothecenes. Thus, this method appears to be specific for type-B trichothecenes only. CONCLUSIONS In summary, a novel indirect electrochemical detection scheme for type-B trichothecenes based on their hydrolysis products in basic solutions has been developed. The hydrolysis products, with catechol-like functional group, are responsible for the electrochemical activity. Due to the low oxidation potentials of these hydrolysis products and the intrinsic selectivity of electrochemical methods, type-B trichothecenes may be detected without extensive cleanup or separation. This may lead to the development of electrochemical sensors for type-B trichothecene detection for agricultural products and runoff from sites that produce the compounds as chemical warfare agents. The potential utility of this approach for the analysis of real-world samples was demonstrated using rice samples spiked with DON. Despite the fact that data contained a sloping background and that peak potential were (43) ACS Guidelines for Data Acquisition and Data Quality Evaluation in Environmental Chemistry. Anal. Chem. 1980, 52, 2242-2249.

Analytical Chemistry, Vol. 71, No. 18, September 15, 1999

4079

slightly shifted, our results show that rice extracts do not introduce significant electrochemical interference and that DON can be quantified down to the 1 ppm level. Although not an exhaustive validation of this method, our results suggest that this approach is an excellent means for the electrochemical detection of DON in real-world samples. Further, our results suggest that a colorimetric test can be devised using redox indicators with sufficient oxidizing power to

test in addition to studies designed to optimize the rate of hydrolysis. ACKNOWLEDGMENT We are grateful to Professor Bruce Jarvis in the Department of Chemistry & Biochemistry, University of Maryland at College Park, for providing the trichothecene samples, and to Lehigh University for financial support.

react with the hydrolysis products of the type-B trichothecenes.

Received for review October 6, 1998. Accepted June 18, 1999.

Currently work is underway in our laboratory to devise such a

AC981114K

4080 Analytical Chemistry, Vol. 71, No. 18, September 15, 1999