A Convenient Method for Assessing Mycotoxin Production in Cultures of Aspergilli and Penicilliat

642

D. ABRAMSON1' and R. M. CLEAR2

lAgriculture and Agri-Food Canada, Winnipeg Research Center, Winnipeg, Manitoba R3T 2M9; Downloaded from http://meridian.allenpress.com/jfp/article-pdf/59/6/642/1660166/0362-028x-59_6_642.pdf by guest on 28 September 2021 and 2Grain Research Laboratory, Canadian Grain Commission, Winnipeg, Manitoba, Canada

(MS#95-171:Received17July 1995/Accepted27October1995)

ABSTRACT time-consuming step for concentration of the extract. The agar-sampling methods, while more rapid, involve a subjec- Production of aflatoxins, sterigmatocystin, citrinin, and ochra- tive decision as to which sector(s) of the culture to sample, toxin A by Aspergillus and Penicillium species on agar in 50-ml and skillful manipulation of agar plugs onto the TLC flasks after 21 days was detected and confirmed by thin-layer surface. Care is needed while dripping solvent onto the plug chromatography. Toxins were extracted with 2.5 ml of acidified in attempting to produce a compact spot at the TLC origin. methylene chloride for 20 to 60 min. Mycotoxin concentrations Since all the extract is used, confirmatory analysis of the were estimated by liquid chromatography.Addition of formic acid to the methylene chloride extraction solvent effected an increase in same agar sample is not possible. Furthermore, there is the recovery of ochratoxin A from Czapek-yeast (CY) agar from 20% hazard of exposure to toxic vapors during the solvent to 91%. Recoveries of added mycotoxins were generally higher extraction of the plug. In both methods the recoveries of from CY agar than from yeast extract-sucrose (YES) agar and toxins from the media are unknown, and weak toxin ranged from 11% (citrinin from YES agar) to 91% (ochratoxin A producers might be construed as nonproducers. from CY agar). The method described is convenient, rapid, and The proposed method was developed using four com- reproducible, since it samples the entire culture surface. Human mon Aspergillus and Penicillium mycotoxins. It offers a exposure to volatile solvents is also minimized. time-saving alternative to the broth method and an extraction and TLC application which are more precise, more Key words: Mycotoxin, Aspergillus, Penicillium, production reproducible, and more convenient than the agar-sampling method. This procedure also reduces the possibility of Although many analytical methods have been devel- exposure of laboratory workers to toxic vapors. oped to detect mycotoxins in agricultural products (4, 13), it is not feasible to assay all commodities for all mycotoxins MATERIALS AND METHODS (9). To reduce this problem to manageable dimensions, it is expedient to determine the fungi characteristically present in Recovery studies agricultural products from specific locales, and the mycotox- For recovery studies of mycotoxins from Czapek-yeast (CY) and yeast extract-sucrose (YES) agars, 5.0 ml of sterile melted agar ins that can be expected (6). (Difco Laboratories, Detroit, MI) was added to 50-ml Erlenmeyer To do this effectively, many fungal isolates need to be flasks to give a nutrient thickness of approx 4 mm. Before the agar screened for mycotoxins and other secondary metabolites. solidified, I00 ~I of ethanol, containing 50 to 60 ~g of mycotoxin Various methods have been proposed. Fungi have been (aflatoxin B[ and ochratoxin A from Aldrich Chemical Co., cultured using semimicro broth techniques (11, 12) and by Milwaukee, WI; sterigmatocystin and citrinin from Sigma Chemi- sampling cultures from agar plates (5). The metabolites have cal Co., St. Louis, MO), was aseptically added to the flasks to give been analyzed by thin-layer chromatography (TLC) and 10 to 12 ~g of mycotoxin per ml of nutrient; agar in control flasks high-performance liquid chromatography (HPLC) (7). Many contained no mycotoxins. Flasks were plugged with cotton and left Aspergillus and Penicillium species and chemotypes have at 22°C for 21 days. been screened in this manner. After addition of 1.0 ml of methylene chloride or methylene In the examination of Aspergillus and Penicillium chloride/formic acid 25:I (using formic acid 97.7%, Sigma Chemi- cal Co., St. Louis, MO), the flasks were capped with aluminum foil species for their mycotoxin production, difficulties often to minimize evaporation. After 20 min, 40 min, or 60 min at 22°C, a arise using the above techniques. The semimicro broth 0.2 ml aliquot of extraction solvent was mixed with 1.0 ml of method involves a lengthy extraction with hot solvent and a methanol in a quartz cuvette, and the absorbance of trial samples against control samples was measured. The following wavelengths * Authorforcorrespondence.Tel:204-984-5536;Fax:204-983-4604. were used: ochratoxin A, 333 nm; aflatoxinB1'362 nm; sterigmato- t Contributionnumber1636oftheWinnipegResearchCenter,andnumber cystin, 329 nm; citrinin, 332 nm. The recoveries are shown in 740oftheGrainResearchLaboratory. Table 1 and Table 2. ASSESSING MYCOTOXIN PRODUCTION IN CULTURES 643

Qualitative examination for mycotoxins TABLE 1. Recovery of ochratoxin A from Czapek-yeast agar after YES agar with 0.5% magnesium sulfate (8) and CY agar were extraction with methylene chloride-formic acid 25: 1 or methylene steam sterilized, and 5.0 ml was added to 50-ml Erlenmeyer flasks. chloride alone. Ochratoxin A (10 I-Lg/ml)was added to the liquid The flasks were capped and set aside to cool. Spore suspensions agar following steam sterilization; flasks were assayed after 3 from fungi cultured on potato dextrose agar plates were made and weeks at 22°C diluted with sterile 0.85% NaCl containing 0.1 % peptone (Difco) Solvent Time (min) Recovery (mean %) SE n to give 103 spores per ml. Three flasks per fungal isolate were each inoculated with 0.5 ml, plugged with cotton, and loosely capped Methylene chloride 20 16.1 2.4 4 with foil. 40 19.6 3.7 4 After 21 days in the dark at 22OC, the flasks were opened and 60 19.5 2.0 4 2.5 ml of methylene chloride/formic acid 25: I were added. Each Methylene chloride/ 20 81.5 6.8 4 flask was tightly recapped with foil and, after 60 min, a 10-/-11 formic acid 25: 1 40 89.0 3.3 4 disposable glass capillary pipette was inserted through the foil and 60 90.8 2.5 4 filled. The pipette contents were applied to the origin of the TLC

plates. Silica plates (Whatman Inc., Clifton, NJ) were developed Downloaded from http://meridian.allenpress.com/jfp/article-pdf/59/6/642/1660166/0362-028x-59_6_642.pdf by guest on 28 September 2021 with the following solvents: for ochratoxin A, ethyl acetate/ chloroform/formic acid 60:40: 1; for aflatoxins, chloroform/acetone Large differences in recovery of mycotoxins from CY 9: 1; for sterigmatocystin, benzene/acetic acid 9: 1; for citrinin, ethyl and YES agars are apparent in Table 2. Generally, recoveries acetate/acetone/water 5:5:2. CIS reverse-phase plates (E. Merck, followed the series ochratoxin A> aflatoxin Bl > sterigmato- Darmstadt, Germany) were simultaneously run for confirmation of cystin > citrinin, and ranged from 90.8% (ochratoxin A, CY identity, using methanol-acidified water mixtures of 7:3, 7:3, 9: I agar, 60 min) to 9.5% (citrinin, YES agar, 20 min). Because and 6:4, respectively (2). The mycotoxins were visualized by their of relatively low recoveries of sterigmatocystin and citrinin, natural fluorescence at 365 nm and 254 nm; sterigmatocystin was the final amount taken for TLC or HPLC analysis when detected by additional spraying with 10% AICI 3 in ethanol, heating these toxins are suspected should be increased four- or at 110°C, and examining under 365 nm light. five-fold to ensure adequate sensitivity. Low recoveries of citrinin from most substrates are common, and may be due Measurement of mycotoxin production to reaction with cysteine (10). Except for sterigmatocystin, One milliliter of the methylene chloride/formic acid 25: 1 recoveries of mycotoxins from CY agar were generally mixture was taken from each flask and evaporated in a 1.5-ml better than from YES agar. This may be related to the polypropylene microcentrifuge tube. Residues were redissolved in 1.0 ml of acetonitrile and clarified by centrifugation at 13,000 X g for 1 min. Aflatoxins BI and G1 were converted to the B2a and G2a TABLE 2. Recovery of aflatoxin B1 (AFBl), sterigmatocystin (ST), derivatives (3) and sterigmatocystin was determined after acetyla- ochratoxin A (OA) and citrinin (CT) from Czapek-yeast (CY) and tion (1). yeast extract-sucrose (YES) agars with methylene chloride-formic Mycotoxins and their derivatives were determined using a acid 25:1. Mycotoxins (10-12 I-Lg/mleach) were added to the liquid Hewlett-Packard 1084 liquid chromatograph and a Schoeffel agar following steam sterilization; flasks were assayed after 3 FS-970 fluorescence detector. Aliquots of 20 IJl were injected onto weeks at 22°C a 250-mm-long by 4.6-mm-inner-diameter column of 10 flIl1 Lichrospher RP-8 (E. Merck) and run at 45°C with mixtures of Mycotoxin Agar Time (min) Recovery (mean %) SE n methanol and acidified water (containing 0.2% phosphoric acid). Sterigmatocystin was assayed after acetylation using 50% metha- AFBI CY 20 60.9 1.6 4 nol for elution and 256 nm fluorescence irradiation. Aflatoxins, 40 70.3 2.3 4 citrinin, and ochratoxin A were assayed using 35%, 50%, and 60% 60 78.4 0.8 4 methanol for elution, and 365, 333 and 333 nm fluorescence YES 20 29.9 0.7 4 irradiation, respectively. 40 41.7 2.2 4 60 42.8 2.8 4 ST CY 20 16.5 1.1 4 RESULTS AND DISCUSSION 40 19.2 0.2 4 60 19.2 0.6 4 Methylene chloride was initially chosen because it YES 20 24.7 2.3 4 23.8 1.3 4 extracted aflatoxin B, efficiently and was volatile. The 40 60 24.3 1.6 4 presence of formic acid in the extraction solvent proved OA CY 20 81.5 6.8 4 necessary for good recovery of acidic mycotoxins such as 40 89 3.3 4 ochratoxin A. The absence of formic acid resulted in low 60 90.8 2.5 4 recoveries of ochratoxin A (Table 1). The pH of liquid CY YES 20 25.9 4.8 4 agar was approx 7.3, and the agar retained most of the added 40 41.2 2.0 4 ochratoxin A (10 fig/ml agar) unless the extraction solvent 60 42.1 3.3 4 was acidified. The presence of formic acid in the methylene CT CY 20 10.1 0.5 4 chloride- formic acid 25: 1 mixture provides enough acidity 40 16 2.0 4 to counter the buffering action of the agar. Formic acid is 60 16.5 2.5 4 volatile enough to evaporate quickly when the solvent YES 20 9.5 0.6 4 0.6 4 mixture is spotted onto TLC plates or dried under nitrogen 40 10.9 60 11.1 1.8 4 for storage. 644 ABRAMSON AND CLEAR sucrose content of the medium; compared to YES agar with The method described above has several advantages in 15% sucrose, CY agar, with only 3% sucrose, may present a evaluating mycotoxin production in fungal cultures. First, less hydrophilic layer more permeable to lipophilic solvents. the entire culture surface is sampled, minimizing effects of In all cases, most of the extraction occurred in the first 20 differential toxin production in sectors of the culture having min, with incremental improvements at longer times. For ex- different color, morphological characteristics, or age. Sec- ample, with aflatoxin B, and CY agar, 61% recovery occurred ond, there is very little sample manipulation, and no special after 20 min, 70% after 40 min, and 78% after 60 min. training is necessary for simple operations such as applying In trials with well-characterized toxigenic cultures, the extracts to TLC plates with disposable capillary pipettes. expected mycotoxins were detected and confirmed by TLC Third, the sample volume can be controlled and applied to a and HPLC (Table 3). YES agar with 0.5% magnesium very small area for good TLC performance and well- sulfate was used for the studies involving A. jlavus and A. separated compact spots; this is especially important for CI8 parasiticus; this medium is known to enhance aflatoxin TLC. Fourth, exposure to evaporating solvents is mini- production (8). In Table 3, the SE values for AFB, from mized; only a few microliters are allowed to evaporate

NRRL-3357, NRRL-465 and NRRL-2999 are atypically during TLC sample application. Fifth, the method can be Downloaded from http://meridian.allenpress.com/jfp/article-pdf/59/6/642/1660166/0362-028x-59_6_642.pdf by guest on 28 September 2021 high. This may be due to variable extraction of the analyte scaled up easily using larger flasks, for cases where concen- due to occlusion of the agar by surface mycelium. With tration of trace amounts of mycotoxins is required. Lastly, exception of AFB2 from NRRL-3494 (20.2%), the rest of the aliquots of extract can be removed from the cultures, dried, data shows SE values :S 11.6% of the mean, with most below and saved for preparative HPLC, or for further study by 8%. In the small number of A. versicolor isolates studied, immunochemical assay or mass spectrometry. YES agar resulted in higher sterigmatocystin production than CY agar. CY agar was used rather than YES agar in ACKNOWLEDGMENTS studies on citrinin and ochratoxin production due to higher recoveries (Table 2). We thank Diane Smith and Susan Patrick for technical assistance. Fungal cultures were provided by Dr. S. W. Peterson, United States Department of Agriculture, Peoria, IL; Dr. John Mills, Agriculture and TABLE 3. Mycotoxin production by some Aspergillus and Penicil- Agri-Food Canada, Winnipeg, MB; and Dr. Ulf Thrane, Danish Technical lium species on yeast extract-sucrose (YES) and Czapek yeast (CY) University, Lyngby, Denmark. agar media after 3 weeks. For incubation and extraction details, see text. AFB1, AFB2, AFG1, AFG2: ajiatoxins Bj, B2' GJ, G2; ST: REFERENCES sterigmatocystin; CT: citrinin; OA: ochratoxin A. Duplicate jiasks were assayed by HPLC, except where only a single replicate was I. Abramson, D., and T. Thorsteinson. 1989. Analysis of sterigmatocys- analyzed (*) tin in barley by acetylation and high-pressure liquid chromatography. J. Assoc. Off. Ana!. Chern. 72:342-344. Mycotoxins 2. Abramson, D., T. Thorsteinson, and D. Forest. 1989. Chromatography produced by isolates of mycotoxins on precoated reverse-phase thin-layer plates. Arch. Environ. Contam. Toxico!. 18:327-330. TLC- HPLC assay, 3. Beebe, R. M. 1978. Reverse phase high pressure liquid chromato- positive, mean graphic determination of afiatoxins in foods. J. Assoc. Off. Ana!. Fungus Isolate Agar confirmed flg/5 ml agar SE Chern. 61:1347-1352. 4. Cole, R. J. (ed.). 1986. Modern methods in the analysis and structural A. jiavus NRRL-3357 YES AFBI 21 14.8 elucidation of mycotoxins. Academic Press, Orlando, FL. AFB2 1.5 0.4 5. Filtenborg, 0., and J. C. Frisvad. 1980. A simple screening method for NRRL-3251 YES AFBI 105 1.8 toxigenic molds in pure cultures. Lebens. Wiss. Techno!. 13: 128-130. AFB2 2 0 6. Frisvad, J. C. 1986. Taxonomic approaches to mycotoxin identifica- tion, p. 415--457. In R. J. Cole (ed.), Modern methods in the analysis NRRL-3494 YES AFBI 1174 81.3 and structural elucidation of mycotoxins. Academic Press, Orlando, FL. AFB2 49 9.9 7. Frisvad, J. c., O. Filtenborg, and U. Thrane. 1989. Analysis and A. parasiticus NRRL-465 YES AFBI 13.5 9.5 screening for mycotoxins and other secondary metabolites in fungal AFB2 1 0 cultures by thin-layer chromatography and high-performance liquid AFGl 164 8.1 chromatography. Arch. Environ. Contam. Toxico!. 18:331-335. AFG2 3 0 8. Frisvad, J. C., O. Filtenborg, U. Thrane, and P. V. Nielsen. 1992. Collaborative study on media for detecting and enumerating toxigenic NRRL-2999 YES AFBI 49 34.6 Penicillium and Aspergillus species, p. 275-284. In R. A. Samson, AFB2 3 0 A. D. Hocking, J. I. Pitt, and A. D. King (ed.), Modern methods in AFGl 168 4.2 food mycology. Elsevier, Amsterdam. AFG2 6 0 9. Krogh, P. (ed.). 1987. Mycotoxins in food. Academic Press, New York. A. versicolor NRRL-3499 YES ST 2830 * 10. Scott, P. M. 1977. Penicillium mycotoxins, p. 283-356. In T. D. NRRL-3499 CY ST 7 * Wyllie and L. G. Morehouse (ed.), Mycotoxic fungi, mycotoxins, mycotoxicoses, an encyclopedic handbook. Marcel Dekker Inc., New NRRL-5219 YES ST 567 * York. NRRL-5219 CY ST 50 * 11. Scott, P. M., J. W. Lawrence, and W. van Wa1beek. 1970. Detection of WRC-A385 YES ST 989 * mycotoxins by thin-layer chromatography: application to screening of WRC-A385 CY ST 19 * fungal extracts. App!. Microbio!. 20:839-842. A. ochraceus NRRL-3174 CY OA 5.5 0.4 12. van Walbeek, W., P. M. Scott, and F. S. Thatcher. 1968. Mycotoxins p. citrinum NRRL-1841 CY CT 1561 42.1 from food-borne fungi. Can. J. Microbio!. 14: 131-137. NRRL-1844 CY CT 12 1.4 13. Wilson, D. M., and D. Abramson. 1992. Mycotoxins, p. 341-391. In D. B. Sauer (ed.), Storage of cereal grains and their products, 4th ed. NRRL-5907 CY CT 2132 80.6 American Association of Cereal Chemists Inc., St. Paul, MN.