3'37“ 35 3; bi‘ i—‘i’."".t I 3.9;»; $123 a 99.4%.: ‘n-a‘v'ufifi ~fi‘t; :s ”“E‘Iei'iufl This is to certify that the dissertation entitled Quantitative characteristics of resistance to corn ear rot caused by Gibberella zeae (Schw.) Petch. and Immunochemistry of T-2 toxin 0 presented by Elie Hy Gendloff has been accepted towards fulfillment of the requirements for PhD degree in Plant Pathology flaw; Mast Major professor Lynn Patrick Hart Date 925' M85 0 MS U i: an Affirmative Action/Equal Opportunity Institution 0-12771 RETURNING MATERIALS: )V153I_} Place in book drop to LJBRARJES remove this checkout from “— your record. FINES will be charged if book is returned after the date stamped below. QUANTITATIVE CHARACTERISTICS OF RESISTAMJE TO CORN EAR ROT CAUSED BY GIBBERELLA ZEAE (SCHIL) FETCH. AND WSW 0F T-2 TOXIN BY Elie Hy Gendloff A DISSERTATICN Submitted to Michigan State University in partial fulfillment of the requirenents forthedegreeof WWW Department of Botany ard Plant Pathology 1985 ABSTRACT QJANTITATIVE CHARACTERISTICS OF RESISTAMZE TO CORN EAR RDT CAUSED BY GIEBERELLA ZEAE (scum FETCH. Immerng T-2 TOXIN BY Elie Hy Gerriloff Various generations of corn (_Zeg mag 12..) derived from four inbred lines were inoculated over a three year period with Gibberella zeae (Schw.) Petch. and rated for disease severity. Analyses of the inbred and F1 generations revealed significant genetic and environmental effects. Generation means analysis indicated that additivity has the predominant genetic effect. Seven corn inbreds were also inoculated with one of eleven isolates of either g. a or Fusariun gpgrotrichioides. _G_. E was more virulent than 3. mtrichioides. Disease reaction of the inbreds followed similar rankings when inoculated with any of the isolates. These data sugest that stable resistance to Gibberella ear rot could be bred into hybrid corn lines. Polylysine conjugates of three structurally unrelated mycotoxins were made via a mixed anhydride intermediate or an activated ester intermediate. Control conjugates, which included no mycotoxins, were also prepared. Two antisera elicited by mycotoxin—bovine serum albumin- mJJed anhydride ccnjugtes bourri to all four polylysine-mixed mhydride conjugates but bound only to the polylysine—activated ester conjugates vim homologous mycotaxin was used. Conjmatim of an unwanted imme- reective epitope mto polypeptides by the mixed amydride procedure vas hypothesised to account for these data. A ccnpetitive enzyme-linked inmmiosorbent assay (direct ELISA) was used to screen for T-2 toxin (T-2) in Fusarium guiding-infected corn. The assay detected T-2 at concentrations of 0.05 ng/ml in extracts of corn samples. In infected corn samples, direct ELISA and gas—liquid chromatography estimations of T—2 concentrations were similar. A polyclonal antibody was produced against T-2 by immunizing rabbits with a mixed anhydride conjugate of T-ZHS and bovine serun albumin (T-ZHS-BSA). The antibody was used to detect T-2 at 0.05 ng/ml by direct ELISA and 1 ng/ml by an indirect ELISA. Cross-reactivity of this antibody with trichothecenes other than T-2 was similar to previously described polyclonal antibodies. A monoclonal antibody against T-2 was produced using a T-ZHS-BSA that was conjumted using a carbodiimide reagent. Mice were succesful 1y immunized using an unconventional immunization protocol involving large antigen doses without adjuvant. The nonoclonal antibody was characterized by indirect ELISA. Sensitivity to T-2 was 10 ng/ml (0.5 pg/assay). The antibody cross-reacted less to I-IT-2 tran previously described T-2 mtibodies. I wcmld like to express my sincerest appreciation to Dr. L. P. Hart for his generous support ard guidance as my major professor durirg these studies and for his interest and ermiragemt in my work. I would also like to thank Dr. J. J. Pestka for his support and patience, Dr. E. C. Rossman for his generous help with my field studies, and Dr. R. P. Scheffer for technical assistance and critical reviews of this diaertaticn.‘ I would also like to thank my friends in the Departments of Botany and Plant Pathology arri Food Science ani Hunan Nutrition for their friendship, encouragement, and assistance during my graduate studies. Lastly, I would like to thank my wife, Theresa Conrad, for her loving patience, faith, and inspiration. ii TABLEOFOONTENTS LISTOFTABLES v LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . vii CENERALIN’IRDDUC‘I‘ION 1 LiteratureCited........................3 PARTI WOFRESISTANCETOGIBBERELIAZEAEINFIEIDCORN INTRODUCTIONS RESULTS“ DISCUSSION............................23 LITERATURECITED.........................26 PARTII MYCOTOXIN—PROTEIN CONJUGATES PREPARED BY THE MIXED ANHYDRIDE MEmoo: (mes-mm ANTIBODIES IN HETERoIoews ANTISERA ABSTRACTZQ mouse RESULTSANDDISCUSSION......................37 LITERA'IURECITED.........................45 PARTIII DETECTION OF T-2 TOXIN IN FUSARIUM SPCROI‘RICI-IIOIDES-INFECTH) (DEN BY ENZYME-LINKED IMMUIDSORBENT ASSAY iii ABSTRACT . INTRODUCTION MATERIALS AND METKJIB . RESULTS AND DISCUSSION LITERATURE CITED PART IV PRODUCTION OF POLYCIDNAL AND MONCX'JLONAL ANTIBODIES AGAINST T-2 TOXIN INTRODUCTION MATERIALS AND METIDDS . RESULTS . DISCUSSION LITERATURE CITED APPENDICES Appendix A: An ELISA for ochratoxin A . Appendix 8: Flow charts for direct and indirect ELISA . Appendix C: Data used to generate Table 3 of Part IV iv 48 49 50 54 60 63 64 66 71 77 80 83 87 90 LIST OF TABLES Table PART I 1. Analyses of variance for disease rating ofGibberel la zeae ear rot on various corn lines. Only inbred and F1 generations are included. See text for lines used. NS = not significant at the 596 level of probability. ** significant at the 196 level of probability .......... 2. Mean disease ratings1 for inbred and F1 generations of corn inoculated in the developing ear with GibberelLa zeae U5373. The lines are ranked from met resistant to most susceptible over all three years. See text for rating systen. Means followed by the same letter within a column are not significantly different at the 5% level by Duncan's multiple range test (17) ........... 3. Means and variances of disease ratings1 forvarious generations of corn crosses. In each year, three replications (rows) of 30-70 plants per row were inoculated in the ear with a toothpick infested with Gibberella zeae 05373 as described in the text. Plants were rated after the first frost ............... 4. Significant genetic effects determined by generation means analysis. Means and variances listed on Table 3 were analyzed for six genetic effects by the method of Hayman (8,9). These effects were mean (m), additive (a), dominant (d), and the three epistatic effects (a, ad, dd). Differences between the two years (y) and the interaction between years and genetic effects (ya, yd, etc.) were also analysed. The reciprocal crosses of B'ZSHt x A509 were analysed separately in 1983 but together in 1984; therefore no determination of year effects were made for these crosses ...... . ....... 5. Amlysis of variance for corn inbred x isolate effects in 1984. Seven inbreds were inoculated with each of eleven pathogen isolates. Three inbreds were susceptible (=5) and four resistant (=R) to pathogen isolate 05373. Seven isolates were g. zeae (=2) and Page 15 16 four E. aorotrichioides (=F). Each inbred-isolate combination was represented as one row in each of two blocks. NS=not significant at the 596 level. *=significant at the 596 level. "asignificant at the 1% level............ ................. 20 6. Average disease ratings of corn inbred x isolate in— oculations. a. Each combination ranked by disease rating. Means of inbreds or isolates followed by the same letter are not different by Duncan's multiple range test. b. Grouped by reaction on 05373 and fungal species. . . . ................. 21 PART II 1. Cross-Reactivities of mixed anhydride and activated ester coating conjugates in indirect ELISA. The coating conjugate was bound to microtiter immunoplates, then treated with mycotoxin antibody. Bound antibody was determined with antirabbit-peroxidase followed by peroxidase substrate. All values are A405. Each value represents the mean of three replications. Standard deviations were always less than 0.1A0. The PLL-AFBl assays were conducted separately from the others. N= preiimnune rabbit serum . . . . ................ 38 PART III 1. Specificity and sensitivity of T-2 antibody in ELISA for T-2 and other trichothecenes ............... 56 2. Ccnparison of ELISA and GLO methods for determination of total trichothecenes in g. sporotrichioides-infected cornsamples.a .......... ..............58 PARTIV 1. Sensitivity of T-2 reactive rabbit antibody to various trichothecenes in competitive indirect enzyme 2. Results of competitive indirect enzyme immunoassay testing of diluted (1/400) mass sera for T-2 antibody activity....... . . .......... 73 3. Sensitivity of T-2reactive monoclonal antibody to various trichothecenes in carpetitive indirect enzyme vi LIST OF FIGURES Figure PART II Mycotoxin-polypeptide conjugation procedures. A. Mixed anhydride. B. Activated ester ............... Competitive EIA for Ochratoxin A (circles) and T—2 toxin (triangles and x's). Each point represents the mean of three replications. The coating conjtgate was bOund to microtiter immunoplates, then treated with free mycotoxin plus mycotoxin antibody. Bourd antibody was determined with antirabbit-peroxidase fol lowed by peroxidase substrate. The assays enployed polylysine— mycotoxin conjugates which were conjugated by the activated ester method (circles and triargles) or the mixed anhydride method (x' s) ............ Isobutyl- compounds tested for reactivity to mixed anhydride—produced antibodies. A. Isobutylchlorofor- rate. B. Isobutanol. C. Isobutylhenisuccinate. PART III Structures of trichothecenes tested vii Page 32 4O 43 55 WWW The furgus Gibberella zeae (Schw.) Petch. (aseaoial state, Fusarium graminarum Schnabe) canes stalk rot, ear rot, ani seedling blight in corn (Leg Egg L.) in the United States (10,11) and other parts of the world (17,18). The most important aspect of the ear rot disease caused by this fungus is the production of mycotoxins, most commonly the trichothecene deoxynivalanol and the estrogenic lactate zearalanone (25). This dissertation reports research directed toward reducing the occurrence of these mycotoxins in the food and feed chain. Two approaches were taken. The first involved a genetic study of the components of resistance to ear mold in corn. The second involved imnrmochenical methods for detecting trichothecenes, using T-2 toxin as a model. There are only scattered reports in the literature concerning resistance to corn ear rot caused by various Fusarium species. Genetic and envirornental variation of resistance to g. zeae in segregating generaticns of com inoculated over several years has not been reported previously. Data on the variation in virulence of various (1; z_e_a__e_ isolates was also lackirg. Studies in these areas are reported in Part I. Many inmmochenical mthods have been developed for the detectim of mycotoxins, particularly the aflatoxins (4,6,8,9,12,13,14,20,21,23), ochratoxin A (1,2,19,24), and T-2 toxin (3,5,7,15,16,22). Immlogicel 1 detection can be sensitive, specific, and rapid. Parts II, III, and IV of this dissertation describe work involving various aspects of the immochanistry of mycotoxins, mrticularly T-2 toxin. Part II describes a cross-reaction that occurs among antisera produced by immunization with mycotoxin-protein conjugates produced by the same conjugation procedure. Part III describes immunological methods developed to detect T-2 toxin in corn infected by various isolates of giggling grotrichioides. New methods for the production ard characterization of polyclonal an! monoclonal antibodies nainst T-2 toxin are described in Part IV. An appendix is included which describes factors influencing the effectiveness of an enzyme immoassay for ochratoxin A and antibodies against ochratoxin A. The development of this assay was the result of studies described in Parts II and IV. A second appendix summarizes protocols used in the indirect and direct ELISA used in parts II, III, and IV. The third appendix gives data used to calculate the cross- reactivities of a monoclonal antibody with various trichotheca'ies. A pertinent literature review is given at the beginning of each section. LITERATURE CITED 1. Aalund, 0., K. Brunfeldt, B. Hald, P. Krogh, and Poulsen, K. 1975. A radioinlnunoassay for ochratoxin A: A preliminary investigation. Acta path. microbial. scarnd. Sect. C 83:390-392. 2. Chu, F. S., F. C. C. Chang, and R. D. Hinsdill. 1976. Production of antibody against ochratoxin A. Appl. Environ. Microbiol. 31:831- 835. 3. Chu, F. S., S. Grossman, R.-D. Wei, arnd C. J. Mirocha. 1979. Produc- tion of antibody against T—2 toxin. Appl. Environ. Microbiol. 37:104-108. 4. Chu, F. S. and I. Ueno. 1977. Production of antibody against afla- toxin Bl. Appl. Environ. Microbiol. 33:1125-1128. 5. Fan, T. S. L., G. S. Zhang, arnd F. S. Chu. 1984. An indirect enzyme linked iumunosorbent assay for T—2 toxin in biological fluids. J. Food Protection 47:964-967. 6. Fan, T. S. L., G. S. Zhang, and F. S. Chu. 1984. Production and characterization of arntibody against aflatoxin 01. Appl. Environ. Microbiol. 47:526-532. 7. Fontelo, P. A., J. Beheler, D. L. Bunner, arnd F. S. Chu. 1983. Detection of T-2 toxin by an improved radioimmunoassay. Appl. Environ. Microbiol. 45:640—643. 8. Gaur, P. K., 0. El-Nakib, and F. S. Chu. 1980. Comparison of anti- body production against aflatoxin Bi in goats and rabbits. Appl. Envirun. Microbial. 40:678-680. 9. Gaur, P. K., H. P. Lau, J. J. Pestka, and F. S. Chu. 1981. Produc- tion and characterization of Aflatoxin 82a antiserum. Appl. Enviran. Microbial. 41:478-482. 10Koeh1er, B. 1959. Corn ear rots in Illinois. Illinois Agric. Exp. Stn. Bull. 639. 87 p. 11.Kommedahl, T. and C. E. Windels. 1981. Root-, stalk-, and ear- infecting Fusarium species on corn in the USA. Pages 94-103 in: Fusarium: diseases, biology, and taxonomy. P. E. Nelson, T. A. Toussoun, and R. J. Cook, eds. The Pennsylvania State University Press, University Park. 12.Langone, J. J. and H. Van Vunakis. 1976. Aflatoxin Bl: Specific antibodies ad their use in radioinmunoassay. J. Nat. Cancer Inst. 56:591-595. 13.Lau, H. P., P. K. Gaur, ard F. S. Chu. 1981. Preparation and char- acterization of aflatoxin BZa—hemiglutarate and its use for the production of antibody against aflatoxin Bl. J. Food Safety 3:1-13. 14.Lawellin, D. W., D. W. Grant, and B. K. Joyce. 1977. Enzyme linked inmunosorbent analysis for aflatoxin Bi. Appl. Environ. Microbiol. 34:94-96. 15. Lee, S. and F. S. Chu. 1981. Radioimmunoassay of T-2 toxin in corn and wheat. J. Assoc. Off. Anal. Chem. 64:156-161. 16.Lee, S. arnd F. S. Chu. 1981. Radioimmunoassay of T-2 toxin in biological fluids. J. Assoc. Off. Anal. Chem. 64:684-688. 17. Mesterhazy, A. 1982. Resistance of corn to Fusarium ear rot and its relation to seedling resistance. Phytopath. Z. 103:218-231. 18.Miric, M. 1969. Fusarium ear rot of corn under irrigation condi- tions. Savrennena Poljoprivreda 17:489-495. 19.Morgan, M. R. A., R. McNerney, and H. W. S. Chan. 1983. Enzyme- linked immunosorbent assay of ochratoxin A in barley. J. Assoc. Off. Anal. Chan. 66:1481-1484. 20.Pestka, J. J. and F. S. Chu. 1984. Aflatoxin B1 dihydrodiol anti- body: Producion ad specificity. Appl. mvircm. Microbiol. 47:472— 477. 21.Pestka, J. J., P. K. Gaur, and F. S. Chu. 1980. Quantitation of aflatoxin 81 ad aflatoxin 81 antibody by an enzyme-linked imnuno— sorbent microassay. Appl. Environ. Microbiol. 40:1027-1031. 22.Pestka, J. J., S. Lee, H. P. Lau, and F. S. Chu. 1981. Enzyme- linked immunosorbent assay for T-2 toxin. J. Am. 011 Chem. Soc. 58:940A-944A. 23.Pestka, J. J, Y. Li, W. 0. Harder, and F. S. Chu. 1981. Comparison of radioilllmmoessay ad enzyme-linked immosorbant assay for de- termining aflatoxin M1 in milk. J. Assoc. Off. Anal. Chu. 64:294- 301. 24.Pestka, J. J., B. W. Steinert, and F. S. Chu. 1981. Enzyme-linked inlnlmosorbent assay for detecticxn of ochratcnxin A. Appl. Environ. Microbiol. 41:1472-1474. 25.Rodricks, J. V., C. W. Hesseltine, and M. A. Mehlman. 1977. Myco- toxins in hunan ad animal health. Pathotox Publishers, Inc., Park Forest South, Ill. PARTI WSOF'RESISTAICENGIWZEAEINFIELDCDRN Crosses involving various generations derived from two susceptible and two resistant inbred corn lines were irnoculated with Gibberel la zeae 05373 over a three year period. Analysis of variance of disease reaction in inbred and F1 generations revealed differences among lines arnd blocks, and a year x line interaction. Generation means analysis involving inbred, F1, F2, F3, backcross, and selfed backcross generations implicated adiitivity (lack of dominance) as the predominant genetic effect. A maternal influence was apparent in one set of reciprocal crosses. Severn inbred lines also were inoculated with seven g. Egg ad four Fusarinmn gmrotrichioiggg isolates in two blocks. g. _zeag was general ly'more virulent than F. sporotrichioides. Inbred x isolate interactions were observed. Disease reaction of these inbred lines followed similar rankings regardless of the pathogen isolate tested . INTROIXE'TION Gibberel la ear rot of corn, caused by Gibberella _z_e_ag (Schwabe) Petch (anamorph=Fusarium graminearum) is sometimes epidemic in the midwestern United States (20). The disease is a cause for concern, even when no significant yield loss occurs, because the causal furgus often produces deoxynivalanol, a cytotoxic trichothecene (21), and zearalenone, an estrogenic lactone (18). Other Fusarium species, such as g. moniliforme, F. nmilifonne var. s_ubglutinans, F_. culmorum (14), ard z. gmrotrichioides (=F. tricinctum) (8,14) cause similar ear rots, with or without the presence of mycotoxins. Sanewnrklasbeendoneindeteminingthenatureof resistance to ear rots caused by various Fusarium species, but variation in experimental methods and pathogen species tested have limited the amt of useful information obtained. Early studies involved E. moniliforme isolates. Boling and Grogan (3) used generation means analysis of a susceptible x resistant crms ad fourd differences between years in the significance of various genetic effects. There was also evidence of epistasis. Other workers found significant maternal effects using diallel analysis of resistance to g. maniliforme seedlirng blight (15). Variation in resistance to one or mny Fusariun species amng popular hybrids has also been found (2.6.16). Differences among isolates and isolate x hybrid interactions were apparent (2), but no large rarnk reversals among these interactions were evident. Host morphological caupgnents have also been implicated as factors influencing resistance (6,13). Recent work has focused on g. zeae, primarily because of concern about mycotoxins. Inbred and F1 analyses have been used to sort out significant genetic canponents. Cullen et al (5) found more resistance among Fl's than among the inbreds from which they were derived. Working along these lines, large differences in resistance were found ancng 58 inbreds (9). Diallel aalysis amng the 10 met resistant and susceptible inbreds revealed significant general containing ability but not specific combining ability effects. This work carries that analysis further, with a generation means analysis of some of those crosses. Environmental variation my inbred and F1 generations over three years also was analyzed. Lastly, isolates of g. gage; and g. _sporotrichioides were compared for virulence on seven inbreds. Generation n_neans analfiis. Generation means analysis is a method of determining significance of various genetic effects nning the means of several generations. Hayman (8,9) developed this method to its current state by guarding the digsnic epistatic theoretical models of Anderson and Kmpthorne (1). Hayman (8), using data previously generated with wheat, tanto, ad tobacco, also sinned that digenic epistatic effects, or the interactian of two nonallelic genes, can be significant. Gable (7) used this method with corn to determine that epistatic effects were inportant canpmants in com yield. Boling ad (3110931 (3) also used this method with com to shcw that resistance to Fusarimn miliforme ear rot involved additivity (a), domimnce (d), and a x d epistasis. 10 Anderson and Kempthorne (1) and Hayman (8,9) showed that the expectations of the mans of two inbred lines and their decadents can be listed (using the terminology of Ganble [7]) as follows: P1 =m+a -1/2d +aa-ai P2 =m-a -1/2d +aa+ad F1 = m + 1/2d F2 (=SF1) = m 91171 (=acl) = m + 1/2a + 1/4aa P2F1 (=BCZ) = m - 1/2a + 1/4aa F3 (=SF2) = m - 1/4d S(P1F1). = m + 1/2a - 3(9231) = m - 1/2a - + 1/4dd + 1/4dd + 1/4di + 1/16di 1/4d + 1/4aa -. 1/4ad + 1/16dd 1/4d + 1/4aa + 1/4ad + 1/16di. Note that the mean (m) is a statistical midpoint, defined as the F2 generation mean. The other generations are therefore defined in terms of the F2 generation. Estimates of these genetic effects can be derived from the Insane of the generations tested by solving the squatiorn listed above for each effect. For exanple, if means were obtained for the first six generations listed above, the varians effects are estimated as (7): Genetic effect 88“ (11 Means estimtg the effect F2 1”151-921" 1 ‘1/2P1'1/2P2+F1"F2+291F1+2P2F2 "F2+2P1F1+2P2F1 ‘1/2P1+1/2P2+P1F1'92F1 P1+P2+2F1+4F2“P1F1’4P2F1 The significance of each effect is tested by a 2-tailed t test, 11 where the variance for each effect is the variance of each generation summed as above. For example, the variance of aa would be 16(var.F2) + 4(var.P1F1) + 4(var.P2F1). WWW Inoculations were made as previously described (9), using the toothpick method modified from Young (22). Only the uppermost ear of each plant was inoculated. Disease ratings were taken after the first killing frost, as previously described (9). Husks were removed and disease was rated by amunt of mycelium visible on the ear. The ratings were: 0 = no disease present, 0.1 = a few kernels around the inoculation point infected, 1 = 10% or less of the ear infected, 2 a 11-25%, 3 = 26- 508, 4 = 51-75%, and 5 =- 75-100% of the ear infected. For the genetic studies, various crosses of the inbreds B79, B73Ht, A509, arnd Pa347 were inoculated in each of three years (1982- 1984). The first two inbreds (B79 and 8731-It) were consistently susceptible and the last two (A509 and Pa347) were consistently resistant in previous work (9). A randomized block design with three blocks was used. Each plot was a single row of one cross, and 30-70 plants in each row were inoculated with g. _zei, isolate we (Penn state # 05373,-g. graminearum R6576). Inbreds and all F1 combinations were inoculated in all three years. The only reciprocals included were those of A509 x B73Ht. In 1983 and 1984, F1 backcrosses (F1P1 and F1P2) and F2s were also included. In 1984, selfed F1 backcrosses (F1P181 and F9281) ad F38 were irnclnded when available. Significance of acuitive (a), dominance (d), and the three digenic epistatic (aa, ad, and dd) 12 13 effects were determined by the method of Hayman ( 8,9). This method is summarized in the Introduction of this section. Significance of interactions between each of the six genetic effects ad the two years (ym, ya, yd, yaa, yad, ydd) also was determined (19). Analysis of variance (19) of the inbred ad F1 generations was used to determine environmental effects. The analysis included the reciprocal generation described above. Orthogonal cantarisons (19) were scmetimes made. \ Seven inbreds were used to determine variability of the reaction to various isolates of g. & ad 3. smrotrichioides. Three (B79, B7311t, and Mount) were susceptible and four (A509, M874, Pa347, and ND100) were resistant to isolate 05373 in previous work (9). Each inbred was inoculated by the toothpick method with each of eleven pathogen isolates. A split plot design (19) was used, with each plot cuntaining one inbred inoculated with each isolate. Two blocks were used, and each block contained cme row of 20-40 plants per inbred-isolate canbimticrn. Seven isolates (05373, 05372, 05371, M3, 31, SA2, and WAl) were (_3. gag. The first four were from the culture collection of L. P. Hart; Si wasobtainedfranR. Stuckey, miversityofxentucky;adSA2andM1 were from D. Cullen, University of Wisconsin. The remaining four isolates (T-340, F27, W299, ad F38) were _F_. gerotrichioides (= g. tricintn- [8]) . T—340 was obtained from E. B. Smalley, University of Wisconsin; NRRL3299 and F38 were from C. J. Mirocha. University of Minnesota; ad F27 was from the culture collection of L. P. Hart. RESULTS Variation in g; zeae resistance. Analyses of variance of disease ratings of the inbred and F1 generations for individual years and for the three years combined are shown in Table 1. There were significant differences among these lines, implying genetic differences. The significant block effects in 1983 and 1984 suggested that location within a field also affects disease severity. Differences in overal 1 disease severity from year-to-year was not significant (Table 1d), although there were significant year x line interacticrns (Table 1d). The mean of the disease ratings used in the previous arnalysis (Table 1) is shown in Table 2. Although some year-to-year differences occured within sane lines, the rankings were similar from year to year. The disease severity of every F1 fel 1 between its two parents. There were no differences between the reciprocals of the cross B73Ht x A509 (Table 2); therefore no maternal effect was evident among the F1 generatian of this cross. Generation m analflis. Table 3 shcus the mean ad variance of thedisaaserating foreachgenerationandcrossussdinthegeneration means analysis. There was a large variance for some inbred and F1 generations, which are genetically hanogeneous determinations. Table 4 sumarizes the results of the generation means analysis. The major genetic effect in the susceptible x resistant crosses was the aiditive 14 15 Table 1. Analyses of variance for disease rating of Gibberel la zeae ear rot on various corn lines. Only inbred and F1 generations are included. See text for lines used. NS = not significant at the 53 level of probability. " = significant at the 1% level of probability. a. 1982 Source E Q F Si ficance Blocks 2 .568 1.3 NS Lines 10 5.133 11.9 ** Error 20 .432 Within plots 1550 .035 b. 1983 Source g1: Q 13: Si ficance Blocks 2 4.804 12.9’ " Lines 10 4.916 13.2 ** Error 19 .373 Within plots 673 . 104 c. 1984 Source D1 _ng _I'j Si ficarnce Blocks 2 4.337 7.8 ** Lines 10 3.094 5.5 ** Error 20 .559 Within plots 1961 .038 d. All three years Source 1F LB; 2 Si ficance Years 2 6.839 2.3 NS Blodts in Years 6 2.982 6.4 ** Lima 10 8.944 4.1 " Year x Line 20 2.199 4.7 " Error 59 . 467 16 Table 2. Mean disease ratings1 for inbred arnd F1 generations of corn inoculated in the developing ear with Gibberel la zeae 05373. The lines are ranked from most resistant to most susceptible over all three years. See text for rating system. Means followed by the sane letter within a column are not signif- icantly different at the 595 level by Duncan's nultiple rage test (17). Line All three years 1982 1983 1984 Pa347 .33a .53ab .23a .33a A509 x Pa347 .67ab .13a .40a 1.47ab A509 1.47abc .53ab 1.17ab 2.73 bc Pa347 x B73Ht 1.57abc .93ab 1.83 bod 2.03 bc B79 x Pa347 1.70abcd .80ab 2.37 od 1.94 bc B73Ht{F} x A509 1.90 bcd 1.70 bc 1.80 bod 2.17 bc A509(F) x B73Ht 2.13 bode 1.27abc 2.07 bod 3.00 cd A509 x B79 2.77 ode 1.57 be 3.70 ef 3.03 cd B731nlt 2.87 ode 2.17 c 2.77 de 3.23 cd B79 x B73Ht 3.17 de 2.33 c 4.47 f 2.70 bc B79 3.60 e 4.87 d 1.50 be 4.17 d ' 1Disease ratirgs— 0 = no disease present; 0.1 a few infected kernels around the inoculation point; 1 = 1096 or less of the ear infected; 2 = 11-25%; 3 = 26-50%; 4 = 51-75%; ad 5 = 75-10096 of the ear infected. 17 0v.N 00.N 0¢.N we.” mm.N 0N.N mm.N mn.H 00.0 hm.0 0N.m 00.H NN.H #0.0 0h.~ 00.H 00.0 00.0 00.0 5.0 0H.H ev.0 00.0 am.0 mfi.0 HN.0 0H.m 0N.N 0mm 8.0 um> can: hemmmxmond .00unmm Hmmm nfimfimmm .cwuamm Hmmm nnmflhfim .Nm 00 oommonuxumn ah "Hmmm Am 0a UmmmOuo £000— Hh Nahum .009me ucmhwm 80000 "mm .UflumHH 988% #020.“ flamn .monszzvmuhm mum—m mogAhvam magmeuwh £09 EH92... 0% $00..” GH N .3082: .30 05 00 00200 n m 05 .xmhlnm n e .&00I0N u m .XDNIHH u N .Dwuomucw. .000 00.3 HO 08H .00 xom u a .Umuowucw usaoc 8302605.. 05 055.50 magma 33 m u 0.0 5808.3 0000ch o: u 0 300.030.“ 0000300 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 2.0 00.0 00.0 00.0 00.0 FM 00.0 HM> 3 0000000 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 0.0.0 00.0 00.0 00.0 00.0 00... 3.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 2.0 0mm 00.0 00> cam: mm.u ......0 cm... mm... mm; 00.N 0N.H H0.H 00.0 Madame: 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 3.0 00.0 00.0 00.0 00.0 3.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 2.0 00.0 00.0 00.0 8.0 00.0 3.0 0mm 00.0 nm> 5002 N¢.m «0.0 00.0 00.N mm.m 0h.m 0H.m 00.0 «0.0 Hh.N h0.N 0N.N HH.0 0H.m 00.N 00.0 00.N 00.0 mm.m 00.0 w«.N GH.N h0.N 00.0 00.0 No.0 #0.H DN.N n0.H Mmm can: hemmmxmnm monszzvmhm «monfiihvmfim 2.90002 fiN.N Hm.m 00.N ma.m 00.0 «n.0 Madam?— hvammumbm $0.0 00.N H mm. 00. 00. OH Hm. 0e. ('30) we. 00. HH 00.H Hm.N «0.0 HN.0 h0.¢ v0.n Acme amnmum Avmv awaken Acme mm v0 mm amen v0 mm aham am no uh v0 mm an #0 mm mm #0 mm Hm HMQV.M.GUO .umouu uthu 0:» umuum nopmu 0am! mucudm .quu may 5.. 030% mm .0000: omen 030.8053 503 cmumomca 0.03503 0 503 .30 any 5.. saunas... 0.003 30.0 and massed 00:00 no 3003 0:038:02 00.05 53> some 5 60000.00 E8 .00 9830,0806 950.09, .000 0.0.9.350 00803 no 00050an Em 0:00: .0 manna. 18 Table 4. Significant genetic effects determined by generation means analysis. Means ad variances listed on Table 3 were analyzed for six genetic effects by the method of Hayman (8,9). These effects were mean (an), additive (a), dominant (d), and the three epistatic effects (a, ad, dd). Differences between the two years (y) and the interaction between years and genetic effects (ya, yd, etc.) were also analysed. The reciprocal crosses of B730t x A509 were analysed separately in 1983 but together in 1984; tlnerefore nno determination of year effects were nade for these crosses. Cross _S_ignificant Effects1 B79 x Pa347 m a A509 x B79 m a B7311t(M) x A509 (83) m a d manna) x A509 (83) m a a ad B730t x A509 (84) m a ad B73Ht x Pa347 m a 879 x B73Ht m Y Ya A509 x Pa347 m 1Significant at 596 level using Student's t test. 19 effect. Nb genetic effects were indicated in the two crosses of inbreds with the same disease reaction.(879 x B73Ht- both susceptible, and A509 x Pa347- both resistant). There were interactions between years and genetic effects (Table 4) in only one of the seven crosses: the cross involving the two susceptible inbreds. Genetic differences were found between the two reciprocal crosses of B7SHt x A509. This indicates a possible nnaternnal effect in disease reaction to (_5. _zeae; in generationns later than the F1, since there were no differences between the F1 reciprocals of this cross (Table 2). An attenpt was made to fit a relatively simple Mendelian gene model using the ratings obtained for the various crosses involving the most susceptible (B79) ad most resistant (Pa347) inbreds. The ratings of the F1 generation of that cross, as a percentage of the total, were: 0 and 0.1 - 39.8%; 1 - 16.9%; 2 - 8.7%; 3 - 13.9%; 4 - 11.5%; 5 - 7.98. Since .all plants in the F1 generation are genetically’identical, this variability must be due to environmental factors. I concluded, therefore, that this large environmental influence nude it inpossible to fit this data to a simple gene model. _I_n£r;_ed gg isolate interactions. Partitioning of the variationn among inbred lines showed that the significant differences were between the resistant ad susceptible groups ad among the smoeptible linnes (Table 5). B79 was more susceptible than Mol7Ht and B73Ht when ratings were averaged over all isolates (Table 6), but tlnere were no differences anong the resistant inbreds (Table 5). Similarly, there were significant differences between isolates of g. gene; ad 2. smtrichioides (Table 5), with the former the more virulent (Table 6). There were also 20 Table 5. Analysis of variarnoe for corn inbred x isolate effects in 1984. Seven inbreds were inoculated with each of eleven pathogen isolates. Three inbreds were susceptible (=8) and four resistant (=R) to pathogen isolate 05373. Seven isolates were g. zeae (=2) and four E. sporotrichioides (=F). Each inbred- isolate oanbination was represented as one row in each of two blocks. NS=not significant at the 58 level. ‘=significant at the 5% level. "Insignificant at the 13 level. Source D__F_ Q E Si ficance Block 1 0.03 0.0 NS Inbred 6 19.80 38.6 " R vs S 1 82.77 162.3 ** Among R 3 1.11 2.2 NS Amng S 2 15.52 30.4 " Error a 6 0.51 Isolate 10 12 . 77 39 . 2 * * 2 vs F 1 93.66 283.8 " Amng Z 6 6.05 18.3 " Among F 3 0.03 0.1 NS Innbred x Isolate 60 0.79 2.4 ** (R vs. S)x(Z vs. F) 1 6.67 20.2 ** (R vs. S)x(annnng Z) 6 3.25 9.8 ** (R vs. S)x(anong F) 3 0.17 0.5 NS (2 vs. F)x(among S) 2 0.95 2.9 NS (ancng Z)x(ancrg S) 12 0.55 1.7 IS (snug F)x(amng S) 6 0.10 0.3 NS (2 vs. F)x(annong R) 3 1.22 3.7 * (anmg Z)x(anong R) 18 0.47 1.4 NS (anong F)x(anncng R) 9 0.02 0.1 NS Error b 70 0.34 Within plot 3804 0. 16 21 Table 6. Average disease ratings of corn inbred x isolate inoculations. a. Each oannbination ranked by disease rating. Means of inbreds or isolates followed by the sane letter are nnot different by Duncan's multiple rage test. b. Grouped by reaction on 05373 ad fungal species. a. Inbred W Res_____18tant B79 Moi7Ht B73Ht Ms74 ND100 A509 Pa347 Isolate than £5. zeae 05372 4.40 3.55 3.45 2.95 2.10 1.90 1.40 3.06s Si 4.20 3.75 3.65 2.45 2.30 1.75 1.00 2.79s 05373 4.75 4.05 4.55 1.25 0.60 0.55 0.25 2.42a SA2 4.50 2.60 3.05 1.65 1.45 1.75 1.45 2.41a VWAl 4.05 2.00 2.40 2.60 2.50 1.95 0.60 2.39s 05371 3.65 3.55 2.70 2.45 0.75 2.40 1.70 2.18a Ma 3.00 0.30 0.80 0.40 0 45 0.40 0.55 1.05b g. gmrotrichioides ’ F38 2.30 0.25 0.55 0.00 0.40 0.25 0.25 1.08b T-340 2.85 0.80 0.45 0.00 0.20 0.30 0.30 0.79b NRRL3299 2.50 0.85 . 0.35 0.10 0.35 0.20 0.30 0.78b F27 2.10 0.45 0.50 0.20 0.50 0.20 0.25 0.62b Mean 3.52a 2.07b 2.02b 1.33bc 1.07bo 1.04bc 0.71o 1.67 b. Reaction Susceptible Resistant man g. zeae 3.28 1.48 2.26a _F_'. gpgrotrichioides 1.16 0.24 0.63b Mean 2.51s 1.03b 1.67 22 differences in virulence among g. zeae isolates (M3 was significantly less virulent), but not among 2. §porotrichioides isolates. Although there were inbred x isolate interactions (Table 5), these did not gm gaior fl reversals; i.e. an inbred which was resistant to one isolate was never susceptible to annother (Table 6). Therefore, the inbred rankings determined with isolate 05373 were similar to other isolates. The inbred x isolate interaction involved differences among inbreds in the variation of their reactions to different isolates. For exanple, a large interaction was evident when comparing the difference in reaction of resistant linnes to susceptible lines irnoculated with g. zeae with the reaction of these lines inoculated with g. sporotrigigides (Table 5). An interaction was also evident when disease reaction of resistant lines were cannered to susceptible lines when inoculated with the g. gas isolates (Table 5). The only other significant inbred x isolate interaction occured when the virulence of the two species anonng the resistant inbreds was compared (Table 5). Among the resistant inbreds, therefore, the (_5. gage isolates were more virulent than the Z- gpgrotrichioides isolates (Table 6) . DISCUSSION These data general 1y support the contention that environmental influences have an innportant effect on the reaction of corn to g_._ z_§_a_e ear rot (13). The differences between blocks in 1983 and again in 1984 as well as the year x line interactionn among F1 ad inbred generations (Table 1) clearly implicate an environmental effect. The large error variances in Table 1 also indicate a substantial variation within rows. Other indications of an environmental influence are the large variances among inbred and F1 generations (Table 3) ad the inability to fit the cross B79xPa347 to a Mendelian model. In spite of the environmental influence, significant genetic differences were consistent from year to year among the inbred and F1 generations (Table 2). As previously nnoted, all F1s analysed were rated intermediate to the inbred parents in disease reaction. This lack of overdominance (since the hybrid disease ratings were not outside the rage definned by the inbreds) is contrary to the findings of Cullen et a1 (5), although they used different linnes. The generation means analysis indicated additivity was the predominant genetic effect, oocuring in five out of the eight analyses conducted (Table 4). Other effects (aside from the mean effect) were sometimes significant, but less often then was adiitivity. This implies that most segregating genes for resistance exhibited little daninanoe or 23 24 digenic epistasis (interactionn between two loci). One of the potential deficiencies or cautionns of generation means analysis is an inability to identify when opposing effects cancel each other (10-12). Therefore, dominant genes for resistance and susceptibility could occur within a particular cross ad would not be evident in the analysis due to their opposing effect. This possibility is minimized, however, when each inbred is analysed in more than one cross (12), as was done here. The mean, which is the statistical midpoint defined as the mean of the F2 generation, was significant in all cases. This means that the mean F1 rating is different frann zero. Trere were nno significant geetic effects between tre two crosses of inbreds with tie sane disease reactionn (B79 1: 873Ht and A509 x Pa347), suggesting few genetic differences between trese comparably rated inbred lines, sirnce the environmental differences were greater than the segregatinng geetic differences. The predanimnce of additivity in resistance to (_5. gage should make incorporation of resistance into agronanicel ly useful corn lines relatively easy, since genes for a susceptible or resistant disease reaction would not be masked by other dominant or epistatic alleles (12). Interesting in this analysis were the differences in significant genetic effects between the two reciprocals of the cross B73Ht x A509 (Table 4). As nnoted above, tl'ere were nno differences between the two Fl reciprocals (Table 2). Therefore tl'e geetic differences nust have been evident in later generations, even though the differences between various crosses involving m reciprocals were generally not great (Tables 2 and 3). The study of inbred x isolate effects (Table 5 and 6) clearly showed that F. sporotrichioides (=_F_. tricinctum) isolates were less virulent than those of g. 5:033. This has been reported previously (14). I-bwever, g. gmtrichioides was capable of causing significant disease an tie highly susceptible inbred B79 (Table 6). There were no major rank reversals in susceptibility of inbreds annong different Fusarium isolates tested (Table 6). Since trere were no major rank reversals aneng g. _zege; or _F. grotrichioides isolates, it is probable that these two species have similar methods of cauSing disease, with g. gage; simply teving more effective virulence gees. If these two species had different methods of causing disease, inbred rankings with Q. gag would likely be different than inbred rankinngs with F. gmrotrichioides, which did nnot happen. Since each inbred was ranked similarly with each isolate, resistance incorporated into an inbred is likely to be stable (4). In spite of tie environmental influence, tl'e large differences in resistance among inbreds and their crosses were readily evident. The large additivity component as well as the apparent "stable" reaction annng the lines tested indicate tret long lasting resistance to 9; E could be easily inncorporated into agronannicel ly useful lines. LITERA'IURE CITED 1. Anderson, V. L. and 0. Kempthorne. 1954. A model for the study of quantitative inheritance. Geetics 39. 883—898. 2. Atlin, G. N., P. M. Enerson, L. G. McGirr, and R. B. Hunter. 1983. Gibberella ear rot developnnnent and zearalenone and vanitoxin pro- duction as affected by maize genotype and G. zeae strain. Can. J. Plant Sci. 63:847-853. 3. Boling, M. B. and C. 0. Grogan. 1965. Gene action affectinng host resistance to Fusarium ear rot of maize. Crop Sci. 5:305-307. 4. Browninng, J. A., M. D. Simona, and E. Torres. 1977. Managing Host Gees: Epidenniologic and Geetic Concepts. Pages 191-212 in: Plant Disease. Volume 1. How disease is managed. J. G. Horsfall and B. B. Cowling, eds. Acaienic Press, New York. 5. Cullen, D., R. W. Caldwell, and E. B. Smalley. 1983. Susceptibility of maize to Gibberel la ear rot: Relationship of host genotype, pathogen virulence, and zearalenone contamination. Plant Dis. 67:89—91. 6. Enerson, P. M. and R. B. Hunter. 1980. Response of maize hybrids to artificially innoculated ear mold incited by Gibberel la zeae. Can. J. Plant Sci. 60:1463-1465. 7. Gamble, E. B. 1962. Gene effects in corn (Ze_a mag L.) I. Separation and relative importance of gene effects for yield. Can. J. Plant Sci. 42:339-348. 8. Gerlach, W. 1981. The present concept of Fusarium classification. Pages 413-426 in: Fusarium: diseases, biology, and tamy. P. E. Nelsen, T. A. Toussoun, and R. J. Cook, eds. The Pennsylvania State university Press, University Park. 9. Hart, L. P., B. Gendloff, and E. C. Rossman. 1984. Effect of corn genotypes on ear rot infection by Gibberel la zeae. Plant Dis. 68:296-298. 10. Hayman, B. I. 1958. The separation of epistatic from additive and daninance variation in generation neans. Heredity 12: 371-390. 11.Hayman, B. I. 1960. The separation of epistatic from additive and 26 27 dominance variation in generation meanns. II. Genetics 31:133-146. 12.1-Iallauer, A. R. and J. B. Miranda. 1981. Quantitative genetics in naize breeding. The Iowa State University Press, Ames. 468 pp. 13.xoehler, B. 1959. Corn ear rots in Illinois. Illinois Agric. Exp. Stn. Bull. 639. 87 p. 14.1(ommedahl, T. and C. B. Winndels. 1981. Root-, stalk-, and ear- infecting Fusarium species on corn in the USA. Pages 94-103 in: Fusarium: diseases, biology, and taxonomy. P. E. Nelson, T. A. Toussoun, and R. J. Cook, eds. The Pennsylvania State University Press, University Park. 15.Lunsford, J. N., M. C. Futrell, and G. E. Scott. 1975. Maternal influence on response of corn to Fusarium moniliforme. Phytopatho logy 65: 223—225. 16Mesterhazy, A. 1982. Resistance of corn to Fusarium ear rot and its relation to seedling resistance. Phytopath. Z. 103:218-231. 17.Miric, M. 1969. Fusarium ear rot of corn under irrigation condi- tions. Savrennena Pol Joprivreda 17:489-495. 18. Mirocha. C. J., J. Harrison, A. A. Nichols, and M. McClintock. 1968. Detectien of a fungal estrogen (F-2) in hay associated with infer- tility in dairy cattle. Appl. Microbiol. 16:797-798. 19.Steel, R. G. D. and J. H. Torre. 1980. Principles and procedures of statistics. McGraw-Hill, New York. 633 pp. 20. Tuite, J., G. Shaner, G. Rambo, J. Foster, and R. W. Caldwell. 1974. The Gibberel la ear rot epidemics of corn in Indiana in 1965 and 1972. Cereal Science Today 19:238-241. 21.Vesonder, R. F., A. Ceiger, A. H. Jensen, W. K. Rohwedder, and D. Weisleder. 1976. Co-identity of the refusal and emetic principle from Fusariun-infected corn. Appl. Environ. Microbiol. 31:280-285. 22.Young, H. C. 1943. The toothpick method of inoculating corn for ear and stalk rots. Phytopathology 33:16. PARTIEI MYCOTOXIN-PROTEIN CONJUGATES PREPARE! BY THE MIXEDANHYDRIDE lfiflflDD:(JKEEFRENGIHEEANTDIIEES DNIEHERCUJIIEBANTHflEUK 28 Polylysine conjugates of three structurally unrelated mycotoxins were made by a mixed anhydride intermediate or an activated ester intermediate. Central conjugates, in which nno mycotoxin was involved, were also prepared by each method. Two antisera, made by T-2-toxin- and aflatoxin Bl-bovine serum albumin-mined anhydride conjngates, bound to all mycotoxin- and the one control-polylysine-mixed anhydride conjugates, but bound only to the polylysine-activated ester conjugates produced using tie same (homologue) mycotoxin trat was used to produce the antibody. Binding of antisera to their homologous polylysine conjugates was always inhibited by free hapten when activated ester- mycotoxin conjngates where used to coat immunoplates, but not when mixed anhydride-mycotoxin conjugates were used. The origin of this cross- reactivity was hypotlesised to be due to tie connjngatien of an unwanted immune-reactive epitope onto polypeptides by the mixed anhydride procedure, but this could nnot be proven. 29 One of the goals of my immunochemical studies was to develop methods for tie production of monoclonal antibodies to T-2 toxin (T-2) and other mycotoxins. Mycotoxins are low molecular weight compounds (haptens) which do not by themselves elicit an antibody response when injected into an animal. In order to elicit this response, they require conjugation to a larger immunogenic molecule, such as bovine serum albumin (BSA). Some of the antibodies elicited against this conjugate wil 1 be against the mycotoxin. To isolate hybridomas which produced antibodies to these toxins, a rapid, sensitive, and reliable assay for detection of these antibodies was required. Tierefore, I investigated various published procedures to conjugate mycotoxins to poly-L-lysine (polylysine) for the development of an indirect enzyme immunoassay (ELISA) for mycotoxin-specific antibody detection. In the indirect ELISA tl'e mycotoxin is innnobilized on a solid phme, usually by conjugating the mycotoxin to a compound like polylysine, which readily sticks to plastic solid m (immoplates). As part of these investigations, I found that binding occured between polylysine-mycotoxin conjugates and antisera elicited by different (heterologous) protein-mycotoxin conjugates when both were prepared by a popular conjugation method involving a mixed anhydride intermediate. Tnese cross reactionns did nnot occur when canjngates were prepared via an activated ester intermediate. 30 31 The mixed anhydride procedure, original ly developed for peptide preparation (19,20), has been extensively used in immunochemistry for conjugation of a variety of carboxyl-containing haptens to proteins, either for antigen preparationn (5,10) or for enzyme labeling for enzyme immunoassay (3,11,15). This procedure has proved useful for preparing conjngates with a high ratio of hapten to protein (13). The reaction proceeds in two steps (Figure 1A). First, the hapten containing a carboxyl group is conjugated to an alkylchlorocarbonate, usually isobutylchloroformate (=isobuty1chlorocarbonate), in the presence of a tri-n—alkylanine, usually triethylamine or tributylamine, at low temperature and anhydrous conditions. This results in the formation of a mixed anhydride. The mixed anhydride is then added to a polypeptide solution where the hapten reacts with free amino groups, usually lysine side chains, linking tie hapten to tie polypeptide via an amide bond (4). The activated ester method (12) is an alternative method for conjugating carboxyl-containing haptens. This method is illustrated in Figure 13. The hapten is first conjugated ("activated") to N- hydroxysuccinimide with dicyclotexylcarbodiimide. This activated ester is then aided to the polypeptide ad a mycotoxin-polypeptide anide bond is formed which is equivalent to that formed by the mixed anhydride reaction (Figure 1A ad 1B) . 32 (mycotoxin) -g-OH + C-g-C-O-g-Cl (Triethylanine) Hapten Isobutylchloroformate 9 9 S3 (mycotoxin) -c’:-o—c-o—c—c—c + HzN-Polypeptich Mixed Anhydride 9 ( mycotoxin) -C—N-Polypept ide Conjngate B a}: 9 g + (mycotoxin) -C-OH N-Hydroxysuccinimide Hapten + (01clohexyl)-N=C=IN-(Cyclotexyl) Dicyclohexylcarbodiimide (D00) 9 0 /C —C (mycotoxin) -C-O-N\q _l + HzN—Polypept Me 6 Activated Ester (mycotoxin) -N-Polypeptide Conjngate Figure 1. Mycotoxin-polypeptide conjugation procedures. A. Mixed anhydride. B. Activated ester. MATERIAISAND mos Materials. All inorganic chemicals and organic solvents were reagent grade or better. BSA (fatty acid free and fraction V), polyoxyethylenesorbitan monolaurate (Tween 20), 2,2'-azino-di(3- ethylbenzthiazoline sulfonic acid) (ABTS), polylysine (M.W. 22,000), ochratoxin A (ochratoxin; crytallized frann benzee), ovalbunin (crude), dicyclohexylcarbodiimide, and N-hydrounysuccinimide were obtained from Sigma Chemical Co., St. Louis, MO; Freund's adjuvants from Difco, Detroit, MI; isobutylchloroformate and triethylamine from Aldrich Chemical 00., Milwaukee, WI; goat antirabbit IgG conjugated to horseradish peroxidase (antirabbit-peroxidase) fran Cooper Biomedical, Malvern, PA; immunoassay microtiter plates (immunnoplates) from Nunc Internned, Roskilde, Denmark; filter paper fran Whatnnann, Inc., Clifton, NJ; T-2 toxin (T-2) from MycoLab Co., Chesterfield, MO; Adsorbosil (200/425 mesh) from the Anspec 00., Ann Arbor, MI; Silica gel-G thin layer chrunatografiny (TLC) plates (Redi-plates) from Fisher Scientific 00., Pittsburgh, PA; antisera reactive against ochratoxin was provided by F. S. Chu, University of Wisconsin; polylysine conjugated to aflatoxin B1-oxime (polylysine-aflatoxin) by both tl'e activated ester and mixed anhydride methods were provided by B. P. Ram, as was both mixed anhydride- and activated ester-produced antisera reactive to aflatoxin B1. Isobutylhenisuccinate was synthesized and conjugated to 33 34 polylysine by w. L. Casale. W o_f_ conjngates. A carboxyl group was introduced to T—2 by adding a hennisuccinnate group to carbon C-3 (T2HS) (21). Conjugatiens of T-ZHS to polylysine or BSA by tne mixed anhydride method were carried out by the method of Lau, et a1 (13). Briefly, 10 mg of dried mycotoxin or eerboxyl-containing derivative was dissolved in 5 ml tetrahydrofuran and cooled to -5 C. To this, 5 ul triethylamine and 5 ul isobutylchloroformate were added. After 20 min at -5 C, tl'e mixture was slowly added to a stirring solution of 25 mg polypeptide (BSA or polylysine) in 15ml water plus 7.5 ml pyridine at 4 C. This mixture was stirred for 30 min at 4 C, then overnight at room tennperature. Dialysis for 3 days againnst frequent changes (four liters per day) of distilled water followed. Conjugations by tl'e activated ester nnnethod were by a modificationn of the method of Kitagaa, et a1. (12). Mycotoxin or arboxyl-containing derivative (1 mg) was mixed with an equimolar amount of both dicyclohexylcarbodiimide and N-hydroxysuccinimide in 0.1 ml dry tetrahydrofuran, then stirred 30- 60 min at room temperature. Precipitate was filtered using Whatman #1 paper, ad then washed with 2- 3 ml tetrahydrofuran. The tetrahydrofuran was evaporated ad the residue dissolved in 0.2 ml dimethylformamide. Dropwise addition of the dissolved residue to 5 ng polypeptide (BSA or polylysine) dissolved in 0.5 ml of 0.13 M sodium bicarbonate followed. This mixture was slowly stirred30minthendialyzed for3daysagainnst four litersperdayof 0.1 M sodium bicarbonate. Control conjugates of polylysine prepared without mycotoxin were 35 made by both the mixed anhydride and activated ester methods to investigate tle effects of these procedures on tle polypeptide when used in the indirect ELISA described below. Inlmmization protocols. TZI-IS-BSA conjugated by tie mixed anhydride method was used as an immunogen. Initial immunization was by a modification of the multiple site method of Vaitukaitis et a1 (17). Here, 0.5 mg conjugate in 0.5 ml of 0.9% saline was emulsified with 1.5ml Freund's complete adjuvant. The preparation was injected intradernnnal ly into 30-40 sites on the shaved back of a New Zealand white doe rabbit. Subsequent injections were nade at six week intervals using Freund's incomplete adjuvant emulsified _in the same ratios and concentrations as described above, but at one-half tie volune. rabbits were bled through the marginal ear vein and sera purified by three annmonium sulfate (35% saturated) precipitations (8). Indirect ELISA. This method is sumarized as a flow chart in Appendix B, under indirect ELISA. Polylysine-mycotoxin conjugate (200ul ) , diluted to 5 ug/ml in 50 mM carbonate-bicarbonate buffer, pH 9.6, was placed in each well of 96-well immunoplates. In some cases, polylysine or a polylysine control conjugate at tte same concentration was substituted. The plates were incubated overnnight at 4 C. Trey were then wasted two tines with sodiunn phosphate buffered saline (PBS- 0.1 M; pH 7.5) cantaining 0.05% Tween 20 as previously described (7), except a twelve-channel aspirator was used. After washing, 200 ul PBS conntaining 1% (wt/vol) ovalbumin (PBS-ovalbumin) was added to each well after insoluble matter in the preparation was removed by a low speed centrifugation. After incubation for 30 min at 37 C, the plates were 36 washed twice as above. Next, 25 ul PBS-ovalbumin was added to each well, followed by 50 ul purified antisera which had been adjusted to 100 ug/ml by tl'e spectrophotanetric method of Burn ad Chantler (9). Washing (four times) followed a 1 h incubation at 37 C. In some cases, a cannpetitive procedure was used by adiing 50 ul unnconjugated mycotoxin at various dilutions prior to antibody addition. Next, 50 ul antirabbit- peroxidase (100 ul in tl'e cometitive procedure), diluted 1/2000 in PBS + 1% BSA (wt/vol) + 0.1% Tween 20 was added, followed by a 30 min incubation at 370. After washing eight times, bounnd peroxidase was assayed by incubating 100 ul ABTS-H202 substrate in each well for 5-10 min; the reaction was terminated with 100ul stopping solution (14). Absorbance at 405 nm was determined on an EIA reader EL307 (Bio-Tech, Inc., Burlington, VT). There were three replications of each treatment. RESULTS AND DISCUSSION of the anti-T-2 sera (produced using a mixed anhydride-conjugated imnncgen) when an activated ester-T2HS-polylysine conjugate, but not when a mixed anhydride-TZHS-polylysine conjugate, was bound to the inmunnoplates (Figure 2), even thongh the T-2 antisera strongly binds to tre mined amydride—Ter-polylysine conjugate (Table 1). Tne regression analysis equation was y = .52 - 0.0688 log(x), where y = A405 absortance readings and x = concentratien of free T-2, in ng/ml. The r2 was 0.94. A similar phencnnenon occured with aflatoxin, where free aflatoxin strongly innhibited binding of both mixed anhydride— and activated ester—produced aflatoxin antisera to the activated ester-polylysine-aflatoxin conjugate bound to the inmunoplates. Free aflatoxin also strongly inhibited binding of the activated ester-aflatoxin antisera to the mixed anhydride-polylysine—aflatoxin bound to the inmunoplates. However, free aflatoxin mly weakly inhibited binding of the mixed anhydride-af latoxin antisera to mined annhydride—pOlylysine—aflatoxin bound to imnneplates (B. P. Ram, personal connmnunication). This suggested the presence of annotlner epitope an the mixed anhydride-polylysine conjugates that only mined anhydride-produced antisera can bind. . Free ochratoxin innhibited binding of the ochratoxin antisera to tie activated ester-ochratoxin-polylysine conjugate (Figure 2). The 37 38 Table 1. Cross-reactivities of mined anhydride (MA) and activated ester (AE) coating conjugates in indirect ELISA. The coating conjugate was bound to microtiter implates, then treated with mycotoxin antibody. Bound antibody was determined with antirabbit-peroxidase followed by peroxidase substrate. All values are A405. Each value represents the mean of three replications. Standard deviations were always less than 0.1AU. The polylysine-aflatoxin assays were conducted separately tron the others. N= preimune rabbit serun. Polylysine Myootoxin Method med 39 make ooatirg comm coating conmgte antim Mined anlydjride Activated ester T-2 T-2 (MA) 1.54 0.59 Aflatoxin (M) 1.36 0.09 Aflatoxin (AB) 0.30 0.07 N 0.03 0.02 Aflatoxin T-2 (MA) 0.93 0.15 Aflatoxin (MA) 0.93 0.56 Aflatoxin (AZ) 0.40 0.85 N 0.01 0.01 Ochratoxin T-2 (M) 1.56 0.19 Aflatoxin (M) 1.40 0.07 Aflatoxin (AB) 0.14 0.07 N 0.02 0.02 Control conjugate T-2 (MA) 1.58 0.21 (no mycotoxin) Aflatoxin (MA) 1.37 0.21 Aflatoxin (AB) 0.19 0.11 N 0.04 0.03 Polylysine T-2 (MA) 0.14 (no conjugation) Aflatoxin (M) 0.08 Aflatoxin (HS) 0.10 N 0.06 39 regression analysis equation was y = 0.278 - 0.021 log (x), where y = A405, and x = concentration of free ochratoxin, in ng/ml. The r2 was 0.73. Among all competitive indirect ELISAs described, structurally unrelated mycotoxins had no inhibitory effect on the binding of a specific antisera to its respective polylysine activated ester- ccnjugated mycotoxin. Since the activated ester ccnjtgation nethcd was useful in producing effective and specific polylysine coating conjugates with three unrelated mycotoxins (ochratoxin, aflatoxin, and T-2) (16) it appears to be an effective general method for conjugations of this type. grgs reactivities g; flious ccnjnngate—antiserg cunbinations. All mixed anhydride-produced polylysine conjugates (including the control conjugates, which did not have mycotoxin attached) bound both the T-2— and aflatoxin-mined anhydride-produced antisera (Table 1). In contrast, the activated ester-produced polylysine ccnjtgates bound only antisera reactive to the homologous mycotoxin, regardless of the conjugation procedure used in producing the antisera. As noted above, the extensive antibody binding found in the mined anhydride-polylysine/mined anhydride antisera combinations was inhibited poorly or not at all when free homologous mycotcntin was added, in contrast to the strong canpetitive inhibitiarn found with activated ester-polylysine/homclogous antisera combinations (Figure 2). All these data were consistent with the hypothesis that mined anhydride-specific cross-reactions resulted fran an immoreactive epitope being introduced mto polypeptides airing the mixed anhydride procedure. Thus, antisera produced from a mixed anhydride-conjugated mycotoxin would cantain antibodies tl'nt bound to this epitope on mined anhydride-conjugated polylysine conjugates, and 40 Figure 2. Canpetitive indirect ELISA for Ochratoxin A (circles) and T-2 toxin (triangles and x's). Each point represents the man of three replications. The coating conjugate was bound to microtiter immunoplates, then treated with free mycotoxin plus mycotoxin antibody. Bound antibody was determined with antirabbit peroxidase followed by peroxidase substrate. The assays employed polylysine-mycotoxin conjugates which were conjugated by the activated ester method (circles and triangles) or the mined anhydride method (x's). 41 Percent maximum cbsorbance ‘00 men I 033.8%: > I ...!» ...oxm: I HIN How»: o; _ no ab Bro 30.2. 38% 2750.853: ooaomiwoeozlzo\3_ 42 free homologous mycotoxin would not be expected to compete for antibodies specific to this epitope. rmvig. I hypothesized that the cross-reactivities described above were due to the connjugation of isobutyl fornnate, the "other" side of the mined anhydride (Figure 1A), to the polypeptide when the mind anhydride was added to tlne polypeptide. This would give R'OOON-Polypeptide rather than RCON-Polypeptide, where R' is (CH3)ZCHCH2 and R is the desired hapten. If this 12' epitope was immunoreactive, then antibodies raised against that would be available for reaction to the sane epitope on the polylysine-mycotoxin coating conjugate when produced by the mixed anhydride method. This side reaction does occur when the mined anhydride reaction is used in dipeptide synthesis (1). This hypothesis was tested by adding either free isobutylchlorofornate (Figure 3A) or free isobutannol (Figure 38) as a napten in the coupetitive indirect ELISA as described earlier. Neither cannpound at 10 ug/ml inhibited birding of any mined anhydride-produced antibodies to any mixed anhydride-produced polylysine conjugates. However, isobutylchlorofonnnate nay nnct have been an effective innhibitor because of possible nonspecific covalent binding with the coating cunjngate, anntibcdy, or ovalbumin blocking protein. To further test the above hypothesis, W. L. Casale conjugated isobutylhemisuccinate to polylysine by the activated ester method. Isobutylhenisuccinate (Figure 30) was selected because it has the sane isobutylformate misty as isobutylchloroformte as well as a arbcnnyl group for ccnjngation via the activated ester method. If isobutyl foruate was present in mined anhydride-produced canjugates than antibodies made 43 A c c—é—co—g-m B c c-b-c—onn c c o c—é-c-o—é-c—c—g-onn Figure 3. Isobutyl- connpounds tested for reactivity to mined anhydride- produced antibodies. A. Isobutylchloroforuate. B. Isobutarnol. C. Isobutylhsnnisuccinate. 44 by a mixed anhydride procedure should react to this isobutylhemisuccinate-polylysine. However, none of the antisera elicited against mixed anhydride-mycotoxin-BSA reacted with isobutylhenisuccinate—polylysine, nnor did free isobutylhenisuccinate at long/ml inhibit binding of either of the mixed anhydride—produced antibodies to the polylysine-mined anhydride-control conjugate. There is, therefore, no evidence for an isobutylformate moiety present on mind anhydride connjngates. Other explanations for the unusual cross- reactivities described here, such as a polypeptide alteratian peculiar to the mined anhydride procedure, should be enmined. To my knowledge, this is the first report of an apparent epitope associated with a hapten-polypeptide cmjngation procedure. Due to the cross-reactivity described herein, I reconnend that the mined anhydride procedure not be used for both immunogen preparation and antibody detection. The mined anhydride procedure ras been succesfully used in each aspect of a system, however, both in indirect ELISA for aflatoxin, as indicated here, and in direct ELISA (18). The problems encountered with the mined anhydride method would nnot arise in imnncassays where this ccnjngatim procedure is used only for immogen preparaticrn, such as radioinmnunoassay. However, in light of thensultsreportedhere, I reccmdtl‘atwherethesannecmjugatian procedure is used in imunogen preparation and antibody detection, proper controls ensuring detection of a similar phenomenon should be anployed. LITERAm CITED 1. Bodanszky, M. and J. C. Tolle. 1977. Side reactions in peptide synthesis V. A reexaninaticn of the mined anhydride method. Int. J. Peptide Protein Res. 10:380-384. 2. Chu, F. S. and I. Ueno. 1977. Production of antibody against aflatoxin Bl. Appl. Environn. Microbial. 33:1125—1128. 3. Comoglio, S. and F. Celada. 1976. An immuno-enzymatic assay of cortisol using E. coli b—galactosidase as label. J. Immol. Meth. 10:161-170. 4. Erlanger, B. F. 1973. Principles and methods for the preparation of drug protein conjugates for immunological studies. Pharmacol. Rev. 25:271-280. 5. Erlanger, B. F., S. M. Beiser, F. Borek, F. Edel and S. Lieberman. 1967., The preparation of steroid-protein conjugates to elicit antihormonal antibodies. Meth. Immunol. Immunochem. 1:144-150. 6. Fan, T. S. L., G. S. Zhang, and F. S. Chu. 1984. An indirect enzyme linked immosorbent assay for T-2 toxin in biological fluids. J. Food Protectionn 47:964—967. 7. Gendloff, E. H., J. J. Pestka, S. P. Swanson and L. P. Hart. 1984. Detection of T-2 toxin in Fusarium mtrichioides-infected corn by enzyme-linked imunosorbent assay. Appl. Environ. Microbiol. 47: 1161-1163 . 8. Hebert, G. A., P. L. Pelham and B. Pittman. 1973. Determination of the optimal mine sulfate cancentratian for the fractionation of rabbit, sleep, horse, and goat antisera. Appl. Microbiol. 25:26- 36. 9. Hurn, B. A. L. and S. M. Chantler. 1980. Production of reagent antibodies. Meth. Enzyml. 70:104-142. 10.Jaffe, B. M., J. W. Smith, W. T. Newton and C. W. Parker. 1971. Radioineunosssay for prostagla'dine. Science 171:494-496. 11.Joyce, B. G., G. F. Read and D. R. Fahmy. 1977. A specific enzymeimmunoassay for progesterone in human plasma. Steroids 29:761-770. 45 46 12.Kitagawa, T., T. Shinozano, T. Aikawa, T. Yoshida, and H. Nishi- mura. 1981. Preparation and characterization of hetero-bifunctianal cross-linking reagents for protein modifications. Chen. Pharm. Bull. 29:1130-1135. 13.Lau, H. P., P. K. Gaur and F. S. Chu. 1981. Preparation and char- acterization of aflatoxin Bza-hemiglutarate and its use for the production of antibody againnst aflatoxin Bl. J. Food Safety 3:1-13. 14.Pestka, J. J., P. K. Gaur and F. S. Chu. 1980. Quantitation of aflatoxin B1 and aflatoxin B1 antibody by enzyme-linked immuno- sorbent microessay. Appl. Environ. Microbiol. 40:1027-1031. 15.Rajkowski, K. M. and N. Cittanova. 1981. The efficiency of dif- ferent coupling procedures for the linlege of oestriol-iGa-glucur- cnide, oestrone-a-glucuronide and pregnanediol-Sa—glucurcnide to four different enzymes. J. steroid Biochen. 14:861-866. 16.Rodricks, J. V., C. W. Hesseltine and M. A. Mehlman. 1977. Myco- toxinns in human and animal health. Pathotcnn Publishers, Inc., Park Forest South, Ill. 17.Vaitukaitis, J., J. B. . Robbins, E. Neischlag and G. T. Ross. 1971. A method for producing specific antisera with smal 1 doses of im- munogen. J. Clin. Endocr. 33:988-991. 18.Van Weeman, B. K. and A. H. W. M. Schuurs. 1972. Immunoassay using hapten-enzyme conjugates. FEBS letters 24:77-81. 19.Vaughan, J. R. Jr. and R. L. Osato. 1951. Preparation of peptides using mixed carboxylic acid anhydrides. J. Am. Chem. Soc. 73:5553- 5555. 20.Vaughan, J. R. Jr. and R. L. Osato. 1952. The preparation of peptides using mixed carbonic-carboxylic acid anhydrides. J. Am. Chen. Soc. 74:676-678. 21.Wei, R., F. M. Strong, E. B. Smalley and H. K. Schnoes. 1971. MCI]. intercenversian of T-2 and inn-2 tannins and related can- pcmds. Biochen. Biophys. Res. Conmnun. 45:396-401. PART III DEMION OF T-2 TOXIN IN FUSARIUM SPORO‘IRICHIOIDES-Im COIN BY ENZYME-LINKED W ASSAY 47 A coupetitive enzyme—linked imnnosorbent assay was used to screen for T-2 toxin in Fusarium sporotrichioides-infected corn. The assay detected T-2 toxin in diluted methanol extracts of corn samples at concenntraticnns of 0.05 ng/ml. In infected corn samples, enzyme—linked immunosorbent assay and gas-liquid chromatography estimatianns of T—2 toxin concentrations were similar. 48 INTROIIICI'ION Analytical methods developed to detect T-2 toxin (T-2) include biological assays, thin-layer chromatography, high-pressure liquid chranatography, gas-liquid chronnatography (GLC) , and gas chronnatography— mass spectroscopy (6). These methods lack sensitivity and/or specificity, or are laborious and require expensive equipnent, and are therefore inadequate as rapid screening assays. The developnent of a radioinnnmunoassay (2) and an enzyme linked inmmosorbent assay (direct ELISA) (8) for T-2 has shown that immunological assays may be useful alternatives for detection of this toxin. In order to furtlner establish the usefulness of immunological methods for detectionn of T-2 toxin, a comparison was maie of an ELISA screening method for T-2 after a simple extraction of corn infected with T-2-producing strains of Fusarium smrotrichioides with a GLC analysis of the same corn after an extensive cleanup. 49 MAWANDMETHODS Antisera reactive against T-2 was provided by J. J. Pestka, Michigan State University, and was purified by precipitation with ammonium sulfate (4) by means of a 3596 saturated anmnonium sulfate mixture. T-2 (MycoLab Co., Chesterfield, Mo.) was converted to T-2- hemisuccinate (T-2HS) by the method of Chu et al. (2). T-2HS was connjugated to horseradish peroxichse by means of a modification of the method~ of Pestka et al. (8). In this variation, 150 ug of T-2HS dissolved in 1.0 ml of ethanol-3.0 ml of water was mined with 150 ng of 1-ethy1-3-(din'ethylenincpropyl) carbodiimide (Signs) and 3.0 ng of type VI homeradish peroxidase (Signa) in 1.0 ml of 25% aqueous ethanol. The mixture was stirred for 30 min at room tenperature, and then an additional 150 mg of carbodiimide was added. This mixture was stirred for 20hat 4Candthenndialyzedagainst threechangesof 0.01 Msodiun phosphate buffer (pH 7.5) for 3 days. The T-2HS-horseradish peroxidase conjugate (0.3 mg/ml) was stored frozen in one ml aliquots. As needed, the aliquots were thawed and distributed into single-use aliquots (30 ul), and these were frozen. The 1:20 ratio of T-2HS:horseradish per- oxidase used in preparation of the conjugate resulted in less nonspecific binding of the conjugate in the ELISA than did the 3:10 ratio used by Pestka et al (8). Antisera titration and T-2 quantitation by direct ELISA were nodified from the method of Pestle et a1 (8). Appendix B sunnnarizes this 50 51 direct ELISA protocol in a flow diagram. Falcon 3070 polystyrene microtissue culture plates (Becton Dickinson and Co., Oxnard, Calif.) were prepared by air drying 50 ul of 0.2% (vol/vol) fraction V (Signa) BSA solution (0.2 mg/ml of water) in each well. The wells were then reacted with 50 ul of 0.2% (vol/vol) glutaraldehyde in 0.1 M phosphate- buffered saline (PBS) (pl-l 7.5) for 30 min, washed extennsively with distilled water, air dried, and stored under desiccation. Purified antisera were diluted fifty-fold in PBS, and 50 ul aliquots were air dried in each well under a warm air current (ca. 40 C) and then stored under desiccation. When used, the antisera coated plates were first washed three times by filling each well with 0.2 ml of 0.05% (vol/vol) _ Tween 20 in PBS (PBS—Tween 20) and aspirating the contents with a single-well aspirator under a vacuun of 580 mm I-g. Nanepecific binding was decreased by incubating each wel l for 1 h at 37 C with 0.2 ml of 1% (wt/vol) BSA in PBS, follaved by two more washes with PBS-Theen 20. Antisera were titrated by diluting the T-2HS-horseradish percnxichse conjugate 1:300 with 5% (wt/vol) BSA and 0.1% (vol/vol) Men 20 in PBS (PBS-BSA-Tween 20) and aching 50 ul aliquots to wells previously treated with serial dilutians of antisera or preinlnnne sera. Incorporation of Tween 20 with the cmjngate allaed less nonspecific canjngate binding than did the method of Pestka et al (8). In the competitive direct ELISA, this step was performed by sinnnltaneously incubating 25 ul of T-2 standard or extracts of infected corn diluted in 10% methanol in PBS with 25 ul of the T—2HS-horseradish peroxiwe canjngate diluted 1:150 in PBS-BSA—Tween 20. In both assays, incubation at 37 C for 1 h followed. The plates were then washed six times with PBS-Tween 20 as 52 described above. Bound horseradish peroxidase per well was assayed (7), and absorbence at 410 nm was determined with a Microelisa Mini Reader MR590 (Dynnatech Laboratories, Inc., Alexandria, VA). Since other trichothecenes might be present in samples of _F. §porotrichioides-infected corn, various standards were tested by canpetitive direct ELISA to determine cross reactivity in the T-2 ELISA. The trichothecenes tested were HT-2 toxin, verrucarcl , diacetoxyscirpenol, and roridin A, all from Sigma Chenical Co., deaxynnivalanol fran MycoLab, and acetyl T-2, nneosolaniol, T—2 triol, and T-2 tetraol, prepared as previously described (1,11). The structures of these trichothecenes are shown in Figure 1. The effectiveness of direct ELISA in detecting T—2 in E. smrotrichioides-infected corn was tested. Ears of corn (inbred B79) were inoculated at the milk stage of developnent by inserting toothpicks infested with a single strain of F. mtrichioides through the husks into the center of the ear (3). E. sporctrichicides strains used were NRRL 3299, F27, T-2, F38, and T-340. Strains T-2 and T-340 were obtained from E. B. Smalley and R. W. Caldwell (Department of Plant Pathology, University of Wisconsin); strains NRRL 3299 and F38 were obtained from C. J. Mirocha (Department of Plant Pathology, University of Minnesota); and strain F27 was from tlne culture collection of L. P. Hart. For determinatians by direct ELISA, 25 g of infected kernels were entracted by blending for 5 min with 250 ml of methannol-water (60:40). The solids were renoved by filtration through no. 4 filter. paper (Rnatnan, Inc., Cliftan, N.J.), and the extracts were diluted to various degrees (up to 4 x 106 times) with 10% netlanol in PBS. 53 The detection of T—2 in corn by direct ELISA was compared with detection by GLC for nine different samples of corn inoculated in the field with various isolates of E. smrotrichioides. Visibly molded kernels were removed from infected ears and dried at 65°C for 2 days, fol lowed by dry chopping for 45 s in a blender sat at low speed. For sample 1 (Table 2), the entire cob, which was extensively molded, was dry chopped for 90 5. Each sample was divided into two subsanples, and direct ELISA determination of T-2 concentration was made on one subsample after dilution of the methanol-water extract as described above. The other subsample was subjected to GLC with electron capture determination of toxin concentration by S. P. Swanson, University of Illinois, by the methOd of Scott et al (9). Thin-layer chranatcgraghy by the method of Takitani et al. (10) was used by S. P. Swanson to confirm the presence of the various trichothecenes tested. RESULTS AND DISCUSSIW The reactivities of the trichothecenes in the direct ELISA relative to T—2, as indimted by the concentration required for 50% inhibition of conjugate binding, is summarized in Table 1. Both T-2HS and T-2 inhibited conjngate binding at 50 pg/ml. Only trichothecenes containing the isovaleroxy moity at the R1 positionn (Figure 1) inhibited conjugate binding at a concentration conparable to that of T—2. As an illustration of this point, nneosolaniol, similar in structure except for this misty, inhibited conjugate birding at onnly 0.1% of the level of T-2. However, as indicated by the low affinity of T-2 trial for the T-2 antibody (Table 1), not all annalogs with the isovaleroxy moiety are cross reactive. The minor reaction of trichothecenes without the C-8 isovaleroxy moiety would be insignificant in assays employing this direct ELISA method. The macrocyclic trichothecene roridin A was the only trichothecene tested that did nnot inhibit canjugate binding at 500 ug/ml. Similar cross-reactivities of various trichothecenes for T-2 antibody have beenn found in T-2 radioinmnunoassay (2) (Table 1), suggesting that no significant differences in specificity occur when radiolabeled or horseradish peroxidase-labeled T-2 is used as an immunoassay ligand. Concentration determinations of T-2 in 60% methanol—extracted, noninfected corn by direct ELISA revealed that the extract could be diluted minimally (sixfold) with PBS without causing significant 54 55 ococazcnzcooa Figure 1. Structures of trichotlnecennes tested. Nane R1 R2 R3 R4 122 (xxonfnucnanz a (xxnng (IXCHa an T-2HS ocoaazcmcnna)2 Hococnnaococnn3 Acetyl n—2 oaocnzau(aH3)2 H ococn3 ocoana ocean3 nnez cnaamgnucuanz H (IIEHa an as you T-2* acnocnnn2cnnn(c:113)2 H ococnna ococnna on s'an nn-2* OCOCH200H(CH3)2 H OCCCHQ an OH T-Zixdol «xxcnfnucnanz H <11 an an Molaniol OH H 000C113 OOOCHa 0H T52 tetraol OH H OH OH OH Verrucarol H H 0H OH H Diacetoxyscerpenol H H 000C113 00061-13 0H Deoxynivalanol =0 0H OH H OH Roridin A H H 'a 14 C 'diester H 'ThesecamoundsarestndiedinSectianIV 56 Table 1. Specificity and sensitivity of T-2 antibody in ELISA for T-2 and other trichothecenes % Cross Trichothecene reactivity relative to T-2a 'n-2 100 (WP T-2HS 100 (ND) Acetyl T-2 100 (ND) HT-2 3.4 (17.5) Nsosolaniol 0.1 (0.2) T-2 trial 0.1 (2.1) T-2 tetracl <0.1 (0.07) Deoxynivalanol . 100,000 a Nanograms of T-2 required for 50% inhibition/nanograms of trichothecene required for 50% inhibition X 100. b Nanograms of trichothecene per milliliter required for first significant inhibition (_13 = 0.05 by Student's t test on four repl icatians per ELISA) of peroxidase conjugate binding. C Numbers in parentheses indicate percentage of cross-reactivity relative to T-2 determined for T-2 radioimmunoassay by Chu et al. (2) . ND,Notdeterminned. 57 interference of peroxidase conjngate bindin the nethanol was evaporated from the extra extract to 10% (vol/vol), interference witl occured. Therefore, preparations for dire not include methanol evaporation. Prel gporotgichigifides F38-infected and T-2-spil indicated that T—2 could be detected by din by methanol extraction with recoveries of spiked samples had equivalent concentra The results obtained when direct ELI thesamesanplesareshamin'l'able 2. A. direct ELISA versus GLC for T-2 concentr: 0.01 (df = 1,7;F = 13.95). The regression e where X = T-2 concentration as determir. coefficient (3) was 0.816. For sample 3 (Table 2), T-2 was not d 10.9 ppm was detected by GLC. This dispa sanpling problens. Further sanples fran the s for reanalysis. It is also possible th= substances that interfere with the inter during direct ELISA, or cause a peak at th heptafluorobutyrate during GLC analysis. This is the first report of T-2 dete immunological methane. The results mt could be used as a seniquantitative tool f spgrotrichioides-infected corn. Acetyl '. 58 Table 2. Couparison of ELISA and GLC metlnods for determination of total trichothecenes in F. smrctrichioides—infected corn sannples.a E-.§29£2£ri- 9995 Sample chioides ELISA-mean T-2 Neoso— strain (3 determinations)b T-2 nrr—z triol laniol 1 NRRL 3299 20.0 (15, 20, 25) 9.5 10.0 trd 2 F27 23.3 (15, 25, 30) 11.4 1.6 tr 4.2 3 F27 <0.05 (all <0.05) 10.9 0.3 tr 4 T-2 0.03 (0.1, 0.1, 0.05) 1.4 1.3 5 n-2 0.5 (0.3, 0.6, 0.6) 2.3 0.4 6 T-2 25.0 (20,.25, 30) 19.9 21.2 7 n-340 33.3 (15, 25, 60) 47.7 42.4 3.5 2.8 a T-340 1.2 (1.0, 1.0, 1.6) 2.0 4.0 9 T-340 0.3 (0.4, 0.4, 0.15) 1.0 0.9 aAll values are in parts per million. ELISA results are of total trichothecenes in the sanple with affinity for the T-2 antibody. bEach determination is the average of four replications of infected corn extract per ELISA plate canpared with replications in the sane plate of nnoninfected corn extracts diluted to the same degree and spiked with various ancnmts of T-2. cLevels unncorrected for spike recoveries (85% for T-2, 76% for HT-2) . GLC analysis was performed by S. P. Swanson by the method of Scott, et al. (9). 5 tr, Trace (<0.5 ppm) 59 (5), also could be detected by direct ELISA (Table 1). The high sensitivity exhibited is nuch greater than is required to detect T-2 at toxic concentrations (6), and the extraction procedures are simpler tran those used in other assays (6,8). This assay could therefore be very useful for routine screening of large numbers of corn samples for T-2 in the time required to analyze a much smaller number by conventional chemical means (6). LITERATURE CITED 1. Chi, M. S., T. S. Robison, C. J. Mirocha, K. R. Reddy. 1978. Acute toxicity of 12,13-epoxytrichothecenes in one—day-old broiler chicks. Appl. Environ. Microbiol. 35:636-640. Chu, F. S., S. Grossman, R.-D. Wei, and C. J. Mirocha. 1979. Productionn of antibody against T-2 toxin. Appl. Environ. Microbiol. 37:104-108. [0 3. Hart, L. P. W. E. Braselton, Jr., and T. C. Stebbens. 1982. Production of zearalenone and deoxynivalannol in corn. sweet cornn. Plannt Dis. 66: 1133-1135. 4. Hebert, G. A., P. L. Pelham, and B. Pittman. 1973. Determination of tlne optimal anmnmium sulfate concentration for the fractionation of rabbit, sheep, horse, and goat antisera. Appl. Microbiol. 25:26-36. 5. Kotsonis, F. N., R. A. Ellison, and E. B. Smal ley. 1975. Isolation of acetyl T-2 toxin from Fusarium poae. Appl. Microbiol. 30:493- 495. 6.Pathre, S. V., and C. J. Mirocha. 1977. Assay methods for trichothecenes and review of their natural occurance, p. 229-253. I_n_nJ. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman (ed.), Mycotoxins in human and aninal health. Pathctcnx Publishers, Inc., Park Forest South, Ill. Pestka, J. J., P. K. Gaur, and F. S. Chu. 1980. Quantitation of aflatoxin B1 and aflatoxin B1 antibody by enzyme-linked innmosorbent micrroassay. Appl. Envirann. Microbial. 40:1027-1031. :4 8. Pestka, J. J., S. L. Lee, H. P. Lau, and F. S. Chu. 1981. Enzyme- linked immunosorbent assay for T-2 toxin. J. Am. 011 Chem. Soc. 58:940A-944A. 9. Scott, P. M., P. Lau, and S. R. Kanhere. 1981. Gas chromatography with electrcrn capture and mass spectnunetric detection of deoxyni- valannol in wheat and other grains. J. Assoc. Off. Anal. Chen. 64:1364-1371. 10.Takitani, S., Y. Asabe, T. Kato, M. Suzuki, and Y. Ueno. 1979. Spectrodensitanetric. determination of trichothecene mycotoxins with 4-(p-nitrobenmyl)pyridine on silica gel thin-layer chrcnnato- 60 61 grans. J. Chranatogr. 172:335—342. 11.Wei, R., F. M. Strong, E. B. Smalley, and H. K. Schnoes. 1971. Chenical conversion of T-2 and HT-2 tannins and related compounds. Biochem. Biophys. Res. Cammnn. 45:396-401. PART IV PREDICTION OF POLYCIDIQL AND W ANTIBGDIES AGAINST T-2 TOXIN 62 A polyclonal antibody was produced against T—2 tannin by imnnizing a rabbit with T-2HS connjugated to bovine serum almmin (T-ZHS-BSA) by a mixed anhydride intermediate. The rabbit did not produce a titer detectable by direct ELISA until after one year of repeated immunization, and two other rabbits did nnot produce a titer after more . than one year. A titer was detected by indirect ELISA in sera from the earliest bleedings, however. The arntibody was used to detect T-2 toxin at 50 pg/ml by direct ELISA and 1 ng/ml by indirect ELISA. Cross- reactivity with other trichothecenes, determined by indirect ELISA, was similar to previously described polyclonal antibodies. Cross-reactivity against the two T-2 metabolites 3'0H T-2 and 3'0H HT-2 was 2% and 1%, respectively, that of T-2 toxin. A monoclonal antibody gainst T-2 toxin was produced using a T-2HS-BSA conjugated by a water-soluble carbodiimide method. Mice were succesfully immunized using a uncanventienal inmmnnizatian protocol consisting of large arntigen doses injected subcutaneously without adjuvant. This antibody was characterized by indirect ELISA. Sesitivity to T-2 toxin was 10ng/ml (0.5pg/assay). The antibody cross-reacted less to HT-2 toxin than previously described T-2 antibodies. Tl'ere were streg cross-reaction to 3'0H T-2 and 3'0H HT-Z. 63 W10)! There are various difficulties in the met cannonly used Chenical methods for detecting trichothecenes (16). Therefore, immunological nethods have been developed for the detection of T-2 (1,4,6,8,13,14,18) and other trichothecenes (2,3). My project, started before some of those reports appeared, had the goal of furthering the developnent of T-2 immunochemistry. Specifically, I wanted to produce polyclonal and monoclonal antibodies to T-2 and canpare them to each other and to other T-2 antibodies described in the literature, with respect to both sensitivity and to specificity to T—2 and other trichothecenes. I hoped to ascertain the feasability of producing a nunber of different antibody preparations, each with different cross-reactivities to various trichothecenes. If this was possible, then samples contaminated with trichothecenes could be analyzed for the various trichothecenes by simply conflicting a series of imglogical tests with the sanple and determining which antibodies reacted stragly and which weakly with the sample. By loaning the cross-reactivities of the antibodies with various trichothecenes, the trichothecene profile of the sample could be determined. I wanted to know if monoclonal antibodies had greater or less variability than polyclonal antibodies with regard to cross- reactivity to various trichothecenes. Large variation in cross- reactivity of monoclonal antibodies to haptens has been found in investigations with certain steroids (5), so this approach seemed 64 65 reasonable. Accomplishment of this goal involved investigations into conjugation procedures, ELISA optimization, and immization protocols, as detailed in in other sections of this dissertation. MATERIALSANDBETI'HDS Materials. All solvents and inorganic chemicals were reagent grade or better. Sources of materials used in mixed anhydride, water soluble carbodiimide, and activated ester conjugations of polypeptides to T-ZHS are given in Part II of this dissertation. The trichothecenes 3'OH T-2 and 3'0H PIT-2 were obtained fran Steven Swanson, University of Illinois. Sources of other trichothecenes used are given in the Part III of this dissertation. Ovalbumin (fraction VII), aminopterin, B—azaguanine, polyethylene glycol MW 1450, sodium pyruvate, insulin, 8- mercaptoethanol, oxaloacetate, pristane, thymidine, and hypoxanthene were obtained from Sigma Chenical Co., St. Iouis, MO; goat antimouse IgG conjtgated to horseradish peroxidme (antimuse—peroxidase) fran Cooper Bianedicel, Malvern, PA; and various cell culture reagents frcm GIBCO Laboratories, Grand Island, NY. Preparation _o_f_ conflugates. Preparation of mixed anhydride and activated ester conjugates are described in Part II of this dissertation. T-ZHS conjugation to ovalbmnin (fraction VII) via the activated ester method was also described in Part II of this dissertation. Conjugation of T-ZHS to bovine serum albumin via water soluble carbodiimide was by the nethod of Chu et al (1). m immization mtoool. This protocol is described in Part II of this dissertation. Three rabbits were injected repeatedly for more 66 67 than one year. They were bled weekly and the titer was checked by the direct ELISA method described in Part III of this dissertation as well as in Appendix B. The titer was also checked by the canpetitive indirect ELISA described in Part II of this dissertation. Mouse immuniggion protocols. Two immunization protocols were compared. Both used the T-2HS-BSA carbodiimide conjugate described above. Three BALB/c mice were injected by each protocol. In the first group, 100 ug T-ZHS-BSA conjugate in 0.1 ml saline was emulsified with 0.1 ml Freund's complete adjuvant and injected into the peritoneal cavity of each mouse. Booster injections were made at one month intervals as above except that Freund's inequplete adjuvant was used. In the second group, 1 ng T-ZHS-BSA conjtgate in 0.5 ml saline was injected subcutaneously into the shoulder. These were repeated at two week intervals, except that 0.5 ng cmjugate was used. Both groups were bled througi tl'e tail vein 79 days after the initial injection. The red cells were pelleted and the serum was diluted and tested for T-2 antibody activity as described below. Cystitive indirect ELI—SA. A cometitive indirect ELISA procedure was always used in screening for mouse antibody activity against T-2. This procedure is detailed in a flow diagram in Appendix 3 under "Indirect ELISA". T-2HS-polylysine or T-2HS-ovalbumin conjugate (200 ul), diluted to 2.5 \g/ml in 50 mM carbonate-bicarbonate buffer, pH 9.6, was placed in each well of 96-well immunoplates. The polylysine conjugate was used in initial hybridoma screening and the ovalbumin conjugate was used later. The plates were incubated overnight at 4 C. They were then wasted with sodium phosdaate battered saline (PBS- 0.1 M; 68 pH 7.5) containing 0.05% Tween 20 as previously described (8), except that a 12 hole aspirator was used. After washing, 200 ul PBS containing 1% (wt/vol) ovalbumin (PBS-ovalbumin) was added to each well after insoluble matter in the preparation was removed by a low speed centrifugation. After incubation for 30 min at 370, the plates were wasted as described above. Next, when undiluted hybridona supernatant fractions were being screened, 40 ul of eitler 10% methanol in PBS (PBS- MEOH) or T-2 toxin (10 ug/ml) in PBS—MEOH were put in each well, followed by 40 ul of hybridoma supernatant. Supernatants from each colony tested were put in four wells, two with free T-2 and two without. Since often less than 200 ul of supernatant could be sampled from the hybridoma colonies, 40 ul was used here instead of 50 ul so that each colony could be tested. in four wells. When preparations other than undiluted hybridoua supernatants were assayed, this step consisted of adding 25 ul PBS-ovalbunin to each well, followed by 50 ul trichothecene in PBS-MEOH, then 50 ul antibody preparation diluted in PBS, as described in Part II. The preparations were incubated for 1 h at 37 C, then were washed. Next, antimouse-peroxidase (80 ul when screening hybridoua supernatants and 100 ul at other times), diluted 1/500 in PBS + 1% fracticn V BSA (wt/vol) + 0.1% Tween 20 was added; followed by a 30 min incubation at 370. After washing, bound peroxidase was assayed by incubating 100ul ABTS-Hzoz substrate in each well for 5-30 min; the reaction was terminated with 100ul stopping solution (17). Absorbance at 405m was determined on an EIA reader EL30'I (Bio-Tech, Inc., Burlington, VT). {lybridoma preparation. Dulbecco's modified eagle median with 20% 69 fetal bovine serum, 50 units penicillin/ml, 50 ug streptanycin/ml, 10% NCTC (Gibco), 5 uM oxaloacetate, 5 uM sodium pyruvate, 75.5 mg insulin/L, and 50 uM B—mercaptoethanol (20% FBS) was routinely used as cell culture medium. The mouse with the highest titer was injected intraperitoneally after five weeks rest with 300 ug T-2HS-BSA in 300 ul saline. The fusion protocol described by 01 and Herzenberg (15) and the spleen cell preparation nethod of Kennett (11) were followed. Tie spleen cell to myeloma ratio in the fusion was 7:1. Three hundred wells were seeded at 7 x 105 total cells per well. Supernatants were screened when tie colonies were at least half-confluent (ie. when at least one-half of the bottom of the microtiter plate well was covered with cells). Colonies positive for anti-T-2 antibody were expanded in 20% FBS, to which hypoxanthine and thymidine (HT) had been added. Half of this nedium had previously beenn conditiaed by myelona growth in log mase for 2 days (half strength conditioned media). Cloning was performed by limiting dilution in eitl'er half strength conditioned median or 20% FBS + HT which contained 30% myeloma conditioned and 20% macrophage conditioned (19) media. Ascites fluid was collected fran pristane primed mice which were injected intraperitaeally with 107 hybridoma cells in 0.5ml 20% FBS twelve days earlier. Masclonal antib_ogy _cl'_na_racterization. The supernatant fraction from an anti-T-2 antibody producing colmy that was clued at 1 cell per well in the limiting dilution step was precipitated three times in a 50% ammonium sulfate solution. The precipitated fraction was diluted sixteen-fold over tie original supernatant volume and this was used in tl‘e indirect ELISA described above. This fraction was also used with the 7O subclass determination kit of Boehringer Mannteim (Indianapolis, IN) to determine tte subclasses of the heavy ad light chains of the antibody. RESULTS Production and characterization 9_f polyclonal antibody reactive gating; 2:3. Only one of three rabbits repeatedly immized for greater than one year showed a titer by direct ELISA. The earliest bleedings from all rabbits, made seven weeks after the initial injection, had a titer of at least 1/500 in the indirect ELISA, however. Sensitivity to T—2 toxin in direct ELISA (performed exactly as in Part III) was the same as for a previously characterized antibody (8), 0.05 ng/ml. The sensitivity of this antibody in the indirect ELISA to various trichothecenes is shown in Table 1. The structures of these trichothecenes are shown in Figure 1 of Part III. This antibody was used at a dilution of 1/1000. The sensitivity to T-2 in the indirect ELISA was 1 ng/ml, 1/20 that of the same antibody in the direct ELISA. The relative cross-reactivity to T-2 of tle otl'er trichotlecenes tested was similar to previously characterized T-2 reactive rabbit antibodies (1,8). _MO_us_e immunization protocolg. Table 2 summarizes the results obtained when a 1/400 dilution of the meme sera was tested by indirect ELISA. Tie mice injected by protocol A, a cannon immimtim protocol, shaved nno activity against T-2, either in anount of antibody binding or in specific inhibition of antibody binding by free T-2. In contrast, strong specific binding occured in mice 1 and 3, since free T-2 at 1 71 72 Table 1. Sensitivity of T-2 reactive rabbit antibody to various tricho- tlecaes in carpetitive indirect enzyme immunoassaya Trichctlecene minimal inhibitionb 5095 inhibitionc 'r—2 1.0 3.2 n—zns 1.0 2.7 HT—2 1.0 13.5 3'0}: 'r-2 so 300 3'OH m-z 100 350 n—z triol 1000 1150 Decanynivalenol 10000 38000 'aAll values are in ng/ml of trichothecene in 10% methanol/PBS. See text for assay protocol. bng/ml trichothecene required for first significant inhibition of binding of antibody to the T-ZHS-ovalbumin solid piece. c:ng/ml trichotlecene required to inhibit binding of antibody by 50% to tie T-2HS—ovalhlmin solid finase. Calculated by regression analysis. 73 Table 2. Results of connpetitive indirect enzyme imnlnnoassay testing of diluted (1/400) mouse sera for T-2 antibody activity. A405 Protocola Mouse No free T—2b free T-2c No specific coat‘d A 1 .17 .17 .17 2 .15 .19 .14 3 .12 .12 .14 B 1 .47 .21 .23 2 .61 .70 .60 3 .44 .23 .19 aProtocols: A = 100119 T-2HS-BSA in 0.1ml saline + 0.1ml adjuvant intra- peritoneally at 1 month intervals. B = 0.5 to 1.0ng T-ZHS-BSA in 0.51111 ‘ saline subcutaeously in the shoulder at two week intervals. b10% methanol in PBS (PBS—MEOH) added with equal volumes of diluted serum ~ Clug/ml T-2 in PBS-MED}! added with equal volumes of diluted serum. d’l'tme wel ls were coated with a conjugate prenared similar to tie T-ZHS- PLL connjtgate, except no T-ZHS was used. 74 ng/ml inhibited binding to levels occuring with uncoated wells. Mouse 2 with this protocol showed nonspecific binding of antibody, since free T- 2 could not inhibit antibody binding ad wel ls coated with a conjugate that was not made with T-2 had a similar amount of binding (Table 2). Mouse 3 under protocol B was used in tie fusion; at a dilution of U100, serum from this mouse showed stronger T-2 antibody activity than did mouse 1. fiybridana preparation. Tre fusionn efficiency (nunber of wells with clones/number of wells seeded) was greater tran 95% (286/300). Of these, nine (3.1%) sinned strong specific binding to tl'e T-2HS—PLL solid plnase. Other colonies showed an equivalent amount of binding but this binding was not inhibited in the wells where free T-2 was present. The nine colonies showing specific binding were cloned at both one and five cells per well in half strength conditioned media. None of these clonings yielded a colony with stable T-2 antibody activity and eight of these either lost activity before they could be frozen or were no longer active when they were thawed. One colony did retain activity upon freezing and thawing and this was successfully cloned (one positive clone) at one cell per well in 20% FBS + HT with 30% myeloua conditioned and 20% macroflnme caditioned media. This positive clone was stable, and all three mice injected with this clone yielded ascites fluid active at a dilution of 1/5000 after ammonium sulfate precipitation. When cloned a second time at one cell per well, more tram 90% of the colonies which grew showed specific binding. Monoclonal antim characterization. Tie antibody described above was an 1961 with a kappa light chain. Table 3 shows tre results of tests 75 Table 3. Sensitivity of T-2 reactive monoclonal antibody to various trichottecenes in coupetitive indirect enzyme innlnnunnoassaya Trichotlece'e minimal inhibitionb sons inhibitionc T-2 0.01 0.023 T-2HS 0.01 0.072 Acetyl T-2 0.01 0.094 3'OH T—2 0.005 0.018 3'OH HT-2 0.05 0.19 HT-2 0.5 1.0 Neosolaniol 1.0 1.6 T-2 triol 5.0 14.9 T-2 tetracl 100. 447. verrucarcl 500. >500. Deoxynivalenol >1000. >1000. Rcridin.A >5000. >5000. 8All values are in \g/ml of trichothecene in 10% methanol/PBS.See text for assay protocol. bug/ml trichothecene required for first significant inhibition of binding of antibody to the T-2HS—ovalhlmin solid pm. Chg/ml trichotl'ecee required to inhibit binding of antibody by 50% to tl'e T-ZHS-ovalbumin solid phase. Calculated by regression analysis. 76 of various trichotlecenes for antibody reactivity. The structures of the trichothecenes tested for cross-reactivity with this antibody are shown in Figure 1 of Part III. The data used to determine tl'e cross-reactivity of tie most cross-reactive trichothecenes listed in Table 5 is sham in Appendix C. This monoclonal antibody tad lees reactivity to HT-z toxin relative to T-2 (2%) than did other monoclonal or polyclonal T-2 antibodies (1,8,10). A strong cross reaction was found with the 3'OH netabolites of T-2 ad HT-2 as well as with acetyl T-2 (Table 3). These netabolites showed much greater cross-reactivity to this antibody than did the polyclonal antibody described above (Table 1). The other trichothecenes tested (other than HT-2) had similar percentage cross reactivities relative to T-2 as did other T-2 rabbit or monoclonal antibodies (1,8,10). Sensitivity to T-2, at 0.01ug/ml, was 0.5ng/assay, which is more sesitive than the previously described T-2 monoclonal antibody/ELISA system (10), but less sensitive than syste-s using rabbit antisera (1,4,6,8,18), inclnding the one described above (Table 1). DISCUSSION Tie rabbit antibody described above produced a titer detectable by indirect ELISA quite easily. This does nnot agree with the assertion by Hunter et al (10) that T-2 is a relatively ineffective hapten. A titer detectable by direct ELISA could only be produced with great difficulty, hmever. This demonstrates tie importance of tie assay. The direct assay needed a much higher titered antisera, and this was onnly obtained after one year. Since the antibody could be used at 1/ 1000 in the indirect ELISA but only stared a minimal titer at 1/50 in tie direct assay, this is also evidence that the indirect ELISA requires less antibody than does the direct ELISA, in agreement with Fan et a1 (4). However, the direct ELISA had a 20—fold greater sensitivity to T-2 than did the indirect ELISA. The mouse immunization protocol used in monoclonal antibody production, in which large antigen doses were given without adjuvant, was clearly superior to the more traditional immunization protocol (Table 2). Two of the three mice immunized with the former protocol ahead titers at the dilutien tested, whereas none of tie mice immunized with tlne latter protocol were positive. The nonspecific binding shown by serum of wise B2 (Table 2) illustrates that antibody binding does not necessarily mean a specific reaction has taken place. As stated previously, when tie l'nybridana supernatant fractions were being tested, 77 78 antibody binding would sanetimes occur that could nnot be inhibited with free T-2 toxin. Therefore, I considered routine employment of the connpetitive procedure crucial in identifying false positives. I experienced none of the fusion problems found by Hunter et al (10). I had hdgh fusion efficiency and a reasonable number of positive colonies. Perhaps the immunization protocol we used overcame the possible inmmlnotoxicity experienced in treir work. The najor problem I had was in retaining a stable colony through cloning. Althongh cloning efficiencies approached 100% for both myeloma ad hybridana lines when half strength conditioned medium was used, no positive colony was obtained until macroprage conditioned medium (19) was used. Presunebly, (the hybridoma needed growth factors present in the macrophage conditioned median and not present in the myelona conditioned medium. The monoclonal antibody has cross-reactivities similar to T-2 antibodies previously characterized (1,8,10). except for its relatively weak cross-reactivity to HT-2 and relatively strong cross-reactivities to the two 3'OH metabolites tested (Table 3). As explained in the introduction, these differing specificities could be useful in assaying for T-2 in biological systems. If samples were assayed with two antibodies of different specificities, the relative amounts of each trichothecene may be determined. This also lends credence to the possibility that a series of monoclonals can be obtained, each with different specificities for many trichothecenes.‘The strong cross- reactivity of this unnoclonal antibody to tie 3'OH netabolites of T-2 and HT-2 opens the possibility that this antibody may’be used in.assays of T-2 toxicosis, since tress netabolites are diagnostic ad present in 79 significant amounts wren T—2 tonnicosis occurs (20—22). Tie three monoclonal antibodies against T—2 described thus far (two by Hunter et al [10] and one here) all show more variability to HT-Z than do the polyclonal systems (1,8). This indicates a greater variability in cross-reactivities among monoclonal antibodies than among polyclonal antibodies. Therefore, a number of monoclonal antibodies against T-2 would probably give a greater range in cross-reactivities than tre sane number of different rabbit antisera. LITERATURE CITED Chu, F. S., S. Grossman, R. D. Wei, and C. J. Mirocha. 1979. Praductiann of antibody against T—2 toxin. Appl. Ennviron. Microbial. 37:104-108. . Chu, F. S., M. Y. C. Liang, and G. S. Zhang. 1984. Production and characterization of antibody against diacetoxyscirpenal. Appl. Environ. Microbiol. 48:777-780. Chu, F. S, G. S. Zhang, M. D. Williams, and B. B. Jarvis. 1984. Production and characterization of antibody against deoxyverrucarol. Appl. Environ. Microbial. 48: 781-784. . Fan, T. S. L., G. S. Zhang, and F. S. Chu. 1984. An indirect enzyme linked inmunasarbent assay for T-2 tannin in biological fluids. J. Food Pratectiann 47:964-967. . Fantl, V. E., D. Y. Wang, and R. E. Knyba. 1982. The production of high affinity monoclonal antibodies to progesterone. J. Steroid Biochem. 17:125—130. Fontelo, P. A., J. Beheler, D. L. Bunner, and F. S. Chu. 1983. Detection of T-2 toxin by an improved radioimmunoassay. Appl. Environ. Microbial. 45:640-643. . Forsell, J. H., J. R. Kately, T. Yoshizawa, and J. J. Pestka. 1985. Inhibitiann of mitagen-induced blastagenesis in human lynphacytes by T-2 tannin ad its netabolites. Appl. Environ. Microbiol. 49:1523- 1526. . Gendloff, E. H., J. J. Pestka, S. P. Swanson, and L. P. Hart. 1984. Detection of T-2 tannin in hearium sporotrichiaides-infected corn by enzyme-linked imunosorbent assay. Appl. Environ. Microbial. 47: 1161-1163 . . Gendloff, E. H., B. P. Ram, W. L. Casale, J. H. Tai, J. J. Pestka, and L. P. Hart. (Submitted). Hapten-protein conjngates prepred by the mixed anhydride method: cross reactive antibodies in leterala- gous antisera. J. Imnunnal. Methads_- . 10.Hunter, K. W. Jr., A. A. Brimfield, M. Miller, E. D. Finkelman, and F. S. Chu. 1985. Preparation and characterization of monoclonal antibodies to the trichothecene mycotoxin T-2. Appl. Environ. 80 81 Microbial. 49: 168—172. 11. Kennett, R. H. 1980. Fusion by centrifugation of cells suspended in polyethylene glycol, p. 365-367. y; R. H. Kennett, T. J. McKearn, and K. B. Bechtol (ed.), Monoclonal antibodies. Plenum Press, New York. 12. Kitagawa, T., T. Shimozano, T. Aikawa, T. Yoshida, and H. Nishimura. 1981. Preparation and characterization of hetero-bifunctianal cross-l inking reagents for protein modifications. Chem. Pharm. Bull. 29:1130-1135. 13.Lee, S. and F. S. Chu. 1981. Radioimmunoassay of T—2 toxin in corn and wheat. J. Assoc. Off. Anal. Chem. 64:156-161. 14.Lee, S. and F. S. Chu. 1981. Radioimmunoassay of T-2 toxin in biological fluids. J. Assoc. Off. Anal. Chan. 64:684—688. 15.01, V. T. and L. A. Herzenberg. 1980. Immunoglobulin-producing hybrid cell lines. p. 351-372. I_r_1 B. B. Mishell, and S. M. Shiigi (ed.), Selected methods in cellular inmmology. W. H. Freeman, San Francisco. 16Pathre, S. V. and C. J. Mirocha. 1977. Assay methods for trichothe- cenes and review of their natural occurence, p. 229—253. 19 J. V. Rodricks, C. W. Hesseltine, and M. A. Mehlman (ed.), Mycotoxins in human and animal health. Pathotox Publishers, Inc., Park Forest South, Ill. 17.Pestka, J. J., P. K. Gaur, and F. S. Chu. 1980. Quantitation of aflatoxin B1 and aflatoxin 81 antibody by an enzyme-linked inanim— sorbent microassay. Appl. Environ. Microbiol. 40:1027-1031. 18. Pestka, J. J., S. Lee, H. P. Lau, and F. S. Chu. 1981. Enzyme-linked ' immunosorbent assay for T-2 toxin. J. Am. Oil Chem. Soc. 58:940A- 944A. 19.8ugasawara, R. J., B. E. Cohaon, and A. E. Karu. 1985. The influ- ence of mine necroptage-ca'riitimed radium on cloning efficiency, antibody synthesis, and growth rate of hybridomas. J. Immunol. Methods (In press). ‘ 20.Yashizawa, T., C. J. Mirocha, J. C. Behrens, and S. P. Swanson. 1981. Metabolic fate of T-2 toxin in a lactating cow. Fd. Cosmet. Toxicol. 19:31-39. 21.Yoshizawa, T., T. Sakamoto, Y. Ayano, and C. J. Mirocha. 1982. 3'- hydroxy T-2 and 3'-hydroxy FIT-2 toxins: new metabolites of T-2 toxin, a trichothecene mycotoxin, in animals. Agric. Biol. Chem. 46: 2613-2615 . 22.Yoshizawa, T., T. Sakamoto, and K. Okamota. 1984. In vitro form- 82 ation of 3'-hydroxy T—2 and 3'-hydroxy HT—2 toxins from T-2 toxin by liver hanogenates from mice and nmkeys. Appl. Environ. Micro- bial. 47:130-134. APPENDIXA AN ELISA FOR OCHRATOXIN A 83 84 Ochratoxin A is a furgal metabolite first isolated from Aspergillus ochraceus (9) arri subsequently found in various agricultural camodities (2). This nephrotoxin is the most toxic of the ochratoxins and is also produced by other gggggillus as well as certain Penicilli_u__m species (2,5). As with the trictnthecenes arri aflatoxins, immmochemical methods have been developed to detect ochratoxin A more rapidly than by conventional Chenical means (1,6,7,8). As part of investigations detailed in Part II of this dissertation. an indirect ELISA was developed for ochratoxin A. Most of the details of this assay are discussed there, and the reader is referred there for materials, methods, and results of the assay when conducted as described. I was interested in developing that assay further so trat it might be used as a screen for hybridana-produced antibodies against ochratoxin A, much as the equivalent assay was used with T-2 in Part III of this dissertation. As described, the assay could detect ochratoxin A at 50 ng/ml, since this was the toxin concentration giving the first significant inhibition of antibody bindim (Figure 2, part II). I have found tl'at various mdifications to that assay affect this sensitivity. These mdificaticms will rm be described. Since ochratoxin A is a rather polar molecule (5), I performed the assaybydilutingthe free ochratoxinA inPBS, rather thanPBSwith 10% methanol. This was effective, reducirg the toxin conceMration necessary to inhibit antibody binding to 10 ng/ml. I next eliminated _the 25ul ovalbumin put in with the antibody and free toxin, since Chu (3,4) had found that ochratoxin A binds to bovine serum albumin (BSA). If the 85 toxin also binds to ovalbumin, this may prevent it from specifical 1y binding to the antibody. When the ovalbumin was eliminated from this step, antibody binding was inhibited by a smaller concentration of ochratoxin A, 5 ng/ml. Since hybridana supernatants normally contain 20* fetal bovine serum, there would be a considerable amount of BSA. Based on the above results, the BSA in the medium might interfere with an indirect screening proceedure for antibodies against ochratoxin if performed as described in Part III of this dissertation. The possibility of interference by media was tested by diluting the rabbit anti- ochrataxin A antibody in 20% ms media. This increased the concentration of free toxin necessary to inhibit birding of the antibody to 500 rig/m1. In conclusion, this assay is more effective if methanol and albumins are eliminated from the step where free toxin is added with antibody. If this is not possible, then greater concentrations of ochratoxin A must be used to get competitive inhibition. I recommend, therefore, that free ochratoxin A be diluted in PBS to a concentration of 10 ug/ml when using this assay to screen for anti-ochratoxin A antibodies in hybridana supernatants. LITERATURE CITED 1. Aalund, O., K. Brunfeldt, B. Hald, P. Krogh, and Poulsen, K. 1975. A radioimunoassay for ochratoxin A: A preliminary investigation. Acta path. microbial. scand. Sect. C 83:390—392. 2. Applegate, K. L. and J. R. Chipley. 1973. Ochratoxins. Adv. Appl. Microbial. 16297-109. 3. Chu, F. S. 1971. Interaction of ochratoxin A with bovine serum albumin. Archs. Biochem. Biomys. 147:359-366. 4. Chu, F. S. 1974. A comparative study of the interaction of ochratoxins with bovine serum albmin. Biochen. Pmrmacol. 23:1105— 1113. 5. Chu, F. S. 1974. Studies on Ochratoxins. CRC Crit. Rev. Toxicol. 2:499-524. 6. Chu, F. S., F. C. C. Chang, and R. D. Hinsdill. 1976. Production of antibody against ochratoxin A. Appl. Environ. Microbiol. 31:831- 835. 7. Morgan, M. R. A., R. McNerney, and H. W. S. Chan. 1983. Enzyme-linked immnosarbent assay of ochratoxin A in barley. J. Assoc. Off. Anal. Chem. 66:1481-1484. 8. Pestka, J. J., B. W. Steinert, and F. S. Chu. 1981. Enzyme-linked immosorbent assay for detection of ochratoxin A. Appl . Environ. Microbiol. 41:1472-1474. 9. Van der Merwe, K. J., P. S. Steyn. L. Fourie, D. B. Scott, and J. J. Theron. 1965. Ochratoxin A, a toxic metabolite produced by Asperwgillus achraceus Wilh. Nature 205:1112-1113. 86 APPENDIXB MOHARTSFORDIRHITANDINDIRECTMSA 87 88 All quantities are per well of microtiter plate. Dilutions for reagents that vary fran batch to batch (eg. antisera, enzyme conjugates) will not be given. a Direct ELISA—see Part III for reagent sources and conjtgation protocols 50u.1 Zing/ml BSA in H20 air dry 50ul 0.2% (vol/vol) glutaraldehyde in PBS incubate 30min, room tanperature; wash in H20; air dry air dry 50u1 diluted, purified antisera ’ wash 3x with PBS/0.05% Mean 20 200.11 1% BSA in PBS 1 incubate 1h 37°C; wash 2x 25ul sanple or T-2 standard in 1096 methanol/PBS + 25ul Té-ZHS-l'nrseradish peroxidase in 596 BSA/PBS/OJx Tween 20 l incubate 1h 37°C; wash 6x 100u1 ABTS/HZOZ solution incubate 3min, room tanperature 100ul HF/EDTA stopping solution Realms 89 Indirect ELISA-see Parts II and III for reagent sources and conjugation protocols . 200ul T-2HS-poly1ysine or T-2HS-ovalbunin in 50mM carbonate- bicarborate, pH 9.6 incubate 4°C overnight; wash2x PBS/0.0596 Ween 20 200ul 196 ovalbumin in PBS 1 incubate 37°C 3min; wash 2x 25u.1 1% ovalbimin in PBS 4- 50u.1 purified antisera, diluted in PBS 4- 5011.1 toxin in 10% mettanol/PBS 1 incubate 37°C 1h; hash 4x 100ul goat antirabbit (or mouse) diluted in 1% BSA/PBS/O.le 'meen 20 incubate 37°C 30min; 8381': Ex looul ABTS/HZOZ solution 1 incutnte 3min, room tanperature looul HF/EDTA stopping solution Beams APPENDIXC DATA USED TO GENERATE TABLE 5 OF PART III 90 91 See text or Apperriix B for assay protocol. Includes only trichothecenes with 50% inhibition at less than 10 ug/ml. All values are A405 Trichathecene- high; 12? L 19 Avergg (_3 replications) T-2 .815 0051 0.001 .828 (0748' 0907' 0830) 0.005 .740 (.721, .756, .745) 0.01 .531 (.556, .695, .543) .263 (.281, .275, .232) .159 (.155, .156, .165) .066 (.047, .099, .052) Foo T-2HS .451 .000 .416 (.304, .470, .474) .370 (.325, .442, .344) .365 (.323, .407, .365) .268 (.218, .234, .351) .131 (.119, .129, .145) .059 (.064, .054, .060) .052 (.044, .063, .048) P‘C)C)C>C>C>C) C>O|h5C> C>FIC> a288 m (are (n .383 (.384, .347, .419) .341 (.314, .344, .366) .184 (.187, .187, .177) .126 (.112, .144, .121) .045 (.039, .050, .047) .040 (.037, .045, .039) Acetyl 'r-2 .451 .000 h¢C>C>C>C>C> CDOIhtCDCDEg Uih‘ 3'0H T-2 .708 .039 .635 (.632, .650, .622) .502 (.527, .455, .523) .436 (.418, .408, .484) .163 (.136, .166, .188) .082 (.079, .080, .086) .024 (.008, .032, .031) .012 (.011, .024, .002) Oil-l . . FPPP°PP °°°“8888 °°“8888 3'0H HT-2 .442 .000 .432 (.384, .433, .479) .459 (.431, .414, .531) .422 (.420, .444, .403) .363 (.356, .344, .389) .286 (.287, .264, .306) .132 (.124, .134, .137) .101 (.092, .101, .111) .078 (.054, .099, .081) Gibb Olth>C>C>CJC>C> \O N Trichathecene; mg; l__ L 19 Am _(3 mutations) A405 HT-z 710 047 0.1 .774 (.721, .754, .846) 0.5 .520 (.502, .524, .534) 1.0 .454 (.440, .439, .484) 5.0 .153 (.151. .148, .161) 10.0 .087 (.093, .080, .089) 50.0 .052 (.048, .057, .052) Neosalaniol .710 .047 0.5 .607 (.592, .642, .587) 1.0 .480 (.499, .475, .465) 5.0 .179 (.169, .193, .176) 10.0 .105 (.094, .112, .109) 50.0 .054 (.049, .051, .061) 100.0 .046 (.042, .051, .046) ahigh= average absorbance value when no free trichothecene is included (0 inhibition value); law= average absorbance value when plate is coated with bicarbonate coating buffer only (100% inhibition value). bCoi'icentration of free trichothecene, ug/ml.