7:.“ ...J.. I 2. 2.9 MI . :3 ion; . I .I “lit-.4571. l. . . v. .4 I y .. 9 :- +9.. PI ...I..\._..r- £15....Qo. n. .. van... ‘Q at: 7. “HRH... . uh 0W: "it! no. VI I, II n W973 am Illllllllllllllllllllllllllllllilll 31293 00794 5219 This is to certify that the dissertation entitled SELF—REACTIVE POLYSPECIFIC IgA ANTIBODIES IN VOMITOXIN-INDUCED GLOMERULONEPHRITIS presented by Linda Rasooly has been accepted towards fulfillment of the requirements for Ph.D. degree inflififlbjfllflgy Major professor Date ll/l "/ ?- yz... MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771 _—___ l' “\ LIERARY Michigan State 1 University L w ‘— PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE ‘ i . . . l x v - A J O" Q‘— rfi . '—r=—= IL all If ' l u If ml MSU Is An Affirmative Action/Equal Opportunity Institution SELF-REACTIVE POLYSPECIFIC IgA ANTIBODIES IN VOMITOXIN- INDUCED GLOMERULONEPHRITIS BY Linda Rasooly A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1992 ABSTRACT SELF-REACTIVE POLYSPECIFIC IgA ANTIBODIES IN VOMITOXIN- INDUCED GLOMERULONEPHRITIS BY Linda Rasooly Vomitoxin is a secondary metabolite of Egsarinm and a major contaminant of cereal grains worldwide. In the mouse model, vomitoxin induces dysregulation. of IgA. that causes accumulation of serum IgA, IgA immune complexes, glomerular IgA deposition and glomerulonephritis. The purpose of this thesis was to (1) characterize the IgA and IgA-producing B cell population present following vomitoxin feeding in the mouse model and (2) assess the possible pathogenic potential of these IgA antibodies. Vomitoxin feeding suppressed total and antigen-specific serum IgG and IgM while increasing total and antigen-specific serum 19A for intestinal and self antigens. Immunofluorescent staining revealed that dietary vomitoxin increased the percentage of IgA+ germinal center B cells (PNA+) in the PP and increased.kappa¥“ IgA+ cells in the spleen. IgA-secreting hybridomas (122) were produced from the PP of exposed animals. iMonoclonal IgAs were autoreactive and polyspecific as well as predominantly trimeric when tested by PAGE and western blotting. Monoclonal IgAs appeared to be representative of the hyperelevated IgA following vomitoxin feeding because the specificity was similar to serum IgA, IgA eluted from kidneys and IgA-producing cells in the PP and spleen of vomitoxin-fed mice. Cytoplasmic staining for the CDS antigen showed. itsl presence in selected Ihybridomas, indicating possible descent from CBS precursors. Injection of monoclonal IgAs into control mice caused hematuria, suggesting they were pathogenic. Staining kidney sections with fluorescinated monoclonal IgAs showed bright staining of cells in the kidneys of vomitoxin-fed mice but not control mice. These cells were likely to be macrophages since the same monoclonal IgAs bound to peritoneal exudate cells which contain macrophages. Similar cells in the kidney from one of the mice injected with monoclonal IgA were noted upon staining with anti-IgA FITC. In conclusion, the findings in this thesis suggest that upon vomitoxin feeding, there were increased numbers of IgA+ CD5+ B cells that secreted autoreactive polyspecific IgA in the PP. The latter may play a role in the development of IgA immune complexes that accumulate in the kidney and cause kidney damage by activating phagocytic cells. To Daphne Rasooly, who gave me the best title of all iv ACKNOWLEDGMENTS I am primarily grateful to my major advisor Dr. James J. Pestka for his advice and guidance throughout my project. I am also indebted to Dr. John E. Linz, Dr. Gale Strasburg, Dr. Kathy Brooks and Dr. Walt Esselman who served on my guidance committee and helped me with invaluable suggestions and encouragement during this project. I am very grateful to Dr. Ronald Patterson who helped me with advice at a critical time and to Dr. Mohamed Abouzied.for his advice and practical help. Thanks are due to Craig Banotai who performed the cytoplasmic CD5 stain for the monoclonal IgA hybridomas. Finally, I am thankful for the friendship and support I had from all the people in the Pestka/Linz group who made everything much easier. TABLE OF CONTENTS EQQQ 1.0 IntrOductionOOOOOOOOOO00....0.0000000000000000.0.00.01 1.1 General characteristics of trichothecenes 1.1.1 History..................................1 Chemistry................................3 Trichothecene-mediated diseases..........9 Biochemical effects.....................11 Toxicity of trichothecenes. .......... ...12 Immunotoxicity..........................14 xin History.................................17 Toxin production in crops...............17 Regulation of vomitoxin in food.........19 Toxicity and metabolism.................23 Immunotoxicity..........................27 .1 Host resistance..................27 2 In vigg cellular effects.........27 3 In gitrg studies.................28 4 Effects on serum immunoglobulins.29 5 IgA glomerulonephritis...........30 1.3 IgA glomerulonephritis and autoimmunity........31 1.3.1 IgA glomerulonephritis..................31 1.3.2 Autoantibodies and glomerulonephritis...33 1.4. Rationale.....................................37 2.0 Materials and Methods...............................39 2.1 General experimental design....................39 2.1.1 Effect of dietary vomitoxin on TNP- specific serum immunoglobulin response to oral TNP-SRBC challenge..............39 2.1.2 Effect of vomitoxin on PP and spleen B cell populations and on IgA specific to intestinal and self antigens............40 2.1.3 IgA hybridoma production and characterization........................41 2.1.4 Relevance of monoclonal IgA in the mouse model...................................42 2.1.5 Pathogenesis of monoclonal IgA..........43 2.2 General procedures.............................44 2.2.1 Vomitoxin production....................44 Toxin extraction and purification.......45 TLC.....................................47 HPLC....................................48 Preparation of diet.....................48 Safety..................................49 vi a O O O O O hiNthiwtdh‘Htflh‘H eoeeerfeeeee 1.2 V Ht‘k‘Hh‘OIHF‘HrJF' unhuowrao<30h>co~ 1.2.5 1.2.5 1.2.5 1.2.5 1.2.5 NNNNN NNNNN 0.... GUI-5U” page 2.2.7 Animals........... ..... .................50 2.2.8 Cell preparation........................50 2.3 Immunological methods..........................51 2.3.1 Antigen preparation.......... ........ ...51 .3.2 Total and antigen specific immunoglobulin quantitation.............52 3 Antigen inhibition ELISA................55 4 IgA elution from kidneys................55 5 ELISPOT.................................56 6 Flow cytometry..........................57 7 Immunofluorescence......................59 8 Hybridoma production....................60 9 PAGE and Western blotting...............61 10 Reduction and alkylation of monoclonal IgA.........................61 2.3.11 Labeling monoclonal IgA with fluorescein............................62 2.4 Statistical analysis...........................62 3.0 Results.............................................63 3.1 Vomitoxin production...........................63 3.2 Effect of dietary vomitoxin on mouse weight and serum Igs..................................63 3.3 Effect of dietary vomitoxin on TNP-specific serum immunoglobulin response to oral TNP-SRBC challenge.............................68 4 Effect of dietary vomitoxin on phenotype of PP and spleen B cells..........................71 5 Effect of dietary vomitoxin on IgA specific to intestinal and self antigens...................79 6 IgA hybridoma production.......................88 7 Antigenic specificity of monoclonal IgAs.......91 8 9 1 NNNNNNNN Tissue reactivity of monoclonal IgAs..........107 Molecular size of monoclonal IgAs.............120 0 Presence of the CD5 molecule in monoclonal IgAs.........................................120 3.11 Pathogenicity of monoclonal IgA..............124 3.12 Relevance of monoclonal IgA to animal model..128 3.13 Specificity of IgA eluted from treatment mouse kidneys................................133 4.0 Discussion.........................................137 5.0 Further studies....................................152 6.0 List of references......................... ..... ...155 vii LIST OF TABLES BESS Table 1. Surveys for vomitoxin presence in food.... ..... 20 Table 2. Percentage of different cell populations in 86C3F1 mice fed 25ppm vomitoxin for 8 weeks....73 Table 3. Supernatant IgA in 86C3F1 mice fed 25ppm vomitoxin for 8 weeks..........................73 Table 4. Percentage of IgA+ cells in each cell cycle stage in mice fed 25ppm vomitoxin for 8 weeks..74 Table 5. Recovery of IgA secreting hybridomas from master wells following fusion..................90 Table 6. ELISA reactivity of monoclonal IgA supernatants (10,000 ng/ml) with antigen panel. Data represent the reactivity of each clone to each antigen as expressed by the 0.D................92 Table 7. ELISA reactivity of monoclonal IgAs with IgE and IgAOOOOOOOOOOOOOOOOOOOO00.00.000.000000000104 Table 8. Tissue reactivity of fluorescinated monoclonal IgA-...OCOOCOOOOOOOOOO...OCOOOOOOOOOOOOO00.0.0110 Table 9. Number of red blood cells in urine of mice injected with monoclonal IgAs.................125 Table 10. Reactivity of IgA eluted from treatment mouse idneYSOO...CIOOOOOOOOOOOOO00.0.00000000000000134 Table 11. Characteristics of antigens used on the screening paneIOOOOOOOOOOOOOOOOOOOOOOOOO0.0.0144 viii LIST OF FIGURES PS9! Pig. 1. Structure of some commonly identified trichothecenes (modified from Vidal, 1990).... ........... .5 Pig. 2. Classification of trichothecenes (from Bamburg' 1983)...OOOOOOOOOOOO...OOOOOOOCOOO0.0.0.0000000007 Pig. 3. Quantitation of vomitoxin amount by TLC and HPLC. A = standard curve following reading of standard vomitoxin amounts on a dual wavelength TLC scanner. B = standard curve of standard vomitoxin amounts by HPLC. AU = arbitrary units.....................................6S Pig. 4. Typical appearance of vomitoxin on HPLC. = standard 500 ng vomitoxin, B = purified sample of 500 ng vomitoxin. Arrows pointing to vomitoxin elution peak. AU = arbitrary units..............................66 Pig. 5. Effect of feeding 25ppm vomitoxin on B6C3F1 mouse weight. Data are means i SEM. * = significantly different (P<0.05) from matching control. ** = significantly different (P<0.01) from matching control..................................................67 Pig. 6. Effect of feeding 25ppm vomitoxin for 4 and 8 weeks on total serum immunoglobulins in B6C3F1 mice. Data are means i SEM (n=8) and are representative of 3 experiments. * = significantly different (P<0.05) from matching controls...................................70 Pig. 7. Effect of vomitoxin on Ig response in the B6C3F1 mouse to oral immunization with TNP-SRBC. Mice were fed 25ppm vomitoxin for 6 weeks and 18.75ppm vomitoxin for 2 more weeks prior to immunization and then gavaged for 4 consecutive days with 4x109 TNP-SRBC. Data are means 1- SEM (n=8). * = significantly different (P<0.05) from matching controls..........................72 ix Pig. 8. Correlation between serum IgA and spleen supernatant IgA levels in 8 weeks 25ppm vomitoxin fed and control B6C3F1 mice. Spleen cells were culture in RPMI-1640 media for 7 days withoug mitogens. Supernatants were collected and analyzed for total IgA. Each. data. point. is representative of one :mouse (n=16). Control r=0.447 (no significant correlation), treatment r=0.802 (correlation significant at P<0.01)..............75 Pig. 9. Correlation between serum IgA and PP supernatant IgA in B6C3F1 mice fed 25ppm vomitoxin or control diet for 8 weeks. PP cells were cultured in RPMI-1640 medium for 7 days and supernatants were collected and analyzed for total IgA concentration. Each data. point. is representative of one 'mouse (n=16). Control r=0.06, treatment r=0.44 (both not significant at p 2-8.8 3 8:8 88 82 .88 3883 2-3 8 8:8 58 F2 <8 82 .8 8 28: 8.8.8 8.8 8:8 883 82 82 .8 8 28: 8.8-88 ; 8.8 1.3. 8:8 82 838.8 Ede ”Ennis—9:8 55:89, accounted owefl 2583 we :88: 5:3 8388 35 .88» 35 .3850 8888 ._ 8:8... 23 On the other hand, in Canada and the USSR, the amount of vomitoxin is regulated. The Canadian tolerance limit is 2000ug/kg in uncleaned soft‘wheat and in the‘USSR.the official tolerance limit for vomitoxin is 500-1000ug/kg (Hietaniemi and Kumpulainen, 1991). 1.2.4 Toxicity and metabolism Vomitoxin is much less toxic than other trichothecenes such as T-2 toxin, HT-z toxin, diacetoxyscirpenol, nivalenol and fusarenon-X (Scott et al., 1980). In an experiment using LD,0 of brine shrimp larvae as a measure of toxicity, vomitoxin ‘was found to be 60-fold less toxic than.T-2 toxin and two fold less toxic than nivalenol (Scott et al., 1980). In mice, the vomitoxin L0” is 70 mg/kg i.p. (intraperitoneal) and 78 mg/kg p.o. (pg; gs) (Forsell et al., 1987). Vomitoxin is cytotoxic at 0.25-1 ug/ml in HE and HeLa cells (Ueno, 1983). This toxicity is probably not due to DNA damage since vomitoxin does not increase unscheduled DNA synthesis in rat hepatocytes (Bradlaw et al., 1985) nor does it cease to be cytotoxic in cells that are defective in DNA repair systems (Robbana-Barnat et al. , 1988) . The toxicity of vomitoxin may be due to protein synthesis inhibition that can occur at 2 ug/ml in rabbit reticulocytes (Ueno, 1983) . Human and mouse fibroblast cells lines were more sensitive in that 500-750 ng/ml vomitoxin caused.killing of one-half the cells (Abbas et al., 1984). Inhibition of lymphocyte cell growth was achieved at even lower concentrations - 115 ng/ml vomitoxin (Porcher et 24 al., 1987). Ingestion of large amounts of vomitoxin (LD,o or higher) results in classical acute symptoms of trichothecenes ranging from :necrosis of the intestinal tract, bone 'marrow' and lymphoid tissues as well as lesions in kidney and heart, to death (Forsell et al., 1987; Robbana-Barnat et al., 1987). Pathological changes include lesions and degeneration of the stomach and small intestine mucosa, enlargement and edema of mesenteric lymph nodes, vascular congestion and depletion of all lymphoid organs and liver (Cote et al., 1985; Robbana- Barnat et al., 1987). In farm animals, ingestion of vomitoxin can cause death when the concentration of toxin approaches the LB” level. At lower levels it is associated with feed refusal, reduced rate of growth, reproductive problems and reduced milk production (Forsyth et al., 1977; Friend et al., 1982; House, 1991; Khera et al., 1984; Morrissey and Vesonder, 1985; Trenholm et al., 1983; Trenholm et al., 1984). Blood characteristics such as hematocrit and hemoglobin are reduced as a result of vomitoxin feeding, probably due to reduced feed intake (Lun et al. , 1985). There is a sex-related sensitivity to vomitoxin in that females are less sensitive than males or castrated males (Cote et al., 1985; Greene et al., 1992). Farm animals differ in their sensitivity to vomitoxin presence in the feed. Swine are the most sensitive and will show feed refusal and decreased weight gain on diets containing as low as 0.3ppm vomitoxin (Trenholm et al., 1983; 25 Trenholm et al. , 1984) . Naturally contaminated feeds are more effective at inducing feed refusal than clean diet containing the same amount of purified vomitoxin, indicating possible additional factors in swine feed refusal (Forsyth et al., 1977) . Cattle are less susceptible to vomitoxin, probably due to the ability of rumen microorganisms to metabolize vomitoxin (King et al., 1984). Vomitoxin was detoxified (the toxic epoxide was removed) within 24 hours by rumen microorganisms at feed concentrations of up to 10ppm (King et al., 1984). However, feed refusal occurs at higher concentrations along with reduced weight gain. Decreased milk production in dairy cows feeding on more than 6ppm vomitoxin containing diets has been observed (Trenholm et al., 1984; Whitlow and Hagler, 1987). Poultry are also less susceptible than swine and can tolerate up to Sppm vomitoxin containing feed without adverse effects and can even gain more weight than poultry on a clean diet (Trenholm et al., 1984). Vomitoxin is rapidly metabolized in farm animals. In swine which received intravenous vomitoxin the half-life of vomitoxin was 2.08-3.65 hours (Coppock et al., 1985) . During that time vomitoxin was both secreted and reabsorbed from the serum into the renal tubules. Vomitoxin was not observed, following 12 and 24 hours after exposure, in the skeletal muscle or other tissues of swine (Coppock et al., 1985; Pollmann et al., 1985). The ability of vomitoxin to cause teratogenic effects differs depending on the experimental animal model used. Oral 26 exposure to vomitoxin by esophageal intubation at 2.5 mg/kg was found to have teratogenic effects on mouse embryos (Khera et al., 1982). Even though no maternal toxicity was observed at this toxin level, a specific lethal effect on the mouse embryo was found together with multiple teratogenic effects. In contrast to these observations, teratogenic effects of vomitoxin were not observed in rabbits even at levels that were toxic to the mothers (Khera et al., 1986). Association of vomitoxin with cancer, more specifically with esophageal cancer, has been shown in several studies. High levels of vomitoxin have been associated with human esophageal cancer in China and Africa (Luo et al., 1990 ; Marasas et al., 1979). In a study, where the natural occurrence of vomitoxin was tested, over 5 times higher levels of the toxin were found in wheat from Linxian (a high-risk area for human esophageal cancer in China) as compared to Shangqiu county (a low'risk area in China) (Luo et al., 1990). In Transkei, an area in Africa where the highest rate of esophageal cancer in Africa has been reported, the natural contamination of corn with vomitoxin was much higher than in low-incidence areas for esophageal cancer (Harasas et al., 1979). More supporting evidence for a carcinogenic potential of vomitoxin is its ability to transform mouse embryo cells in 21;;9 (Sheu et al., 1988). It should be noted, however, that the presence of fumonisin, a mycotoxin produced by Eyggrium mgnilifigzme, in contaminated cereals has also been recently implicated as a possible etiologic agent for cancer induction 27 subsequent to ingestion of moldy grains (Voss et al., 1990; Wang et al., 1991) 1.2.5 Immunotoxicity 1.2.5.1 Host Resistance Oral exposure to vomitoxin results in reduced time-to- death in mice challenged with L._._ monocngogeneg and reduced resistance to this pathogen as exhibited by the splenic clearance of Ligtgzig (Pestka et al., 1987; Tryphonas et al., 1986). Increased splenic Ligtgria counts following exposure to vomitoxin may be partly due to vomitoxin-induced feed refusal rather than directly as a result of modulation of the immune function of the host (Pestka and Bondy, 1990). 1.2.5.2 I; 1119 cellular effects Vomitoxin induces a dose dependent leukopenic effect in mice in that white blood cell, lymphocyte and monocyte numbers are decreased (Forsell et al., 1986; Robbana-Barnat et al., 1988) . Neutrophil levels in the blood are increased as a result of vomitoxin feeding (Forsell et al., 1986). As opposed to its suppressive effect, vomitoxin has a stimulatory effect on cells in the PP, especially with regards to IgA+ cells. 'This stimulatory effect was observed as an increase in the size and frequency of the germinal centers in the PP subsequent ot vomitoxin feeding (Pestka et al., 1990a). The more specific stimulatory effect on IgA+ cells and helper T cells was seen in the same experiment when the percentage of membrane IgA bearing cells and CD4+ T cells, as well as the ratio CD4+/CD8+, were all increased in PP and spleen of 28 vomitoxin-fed animals. Vomitoxin also increased the number of IgA-secreting cells in the PP of toxin-fed animals and decreased the number of IgG-secreting cells (Pestka et al., 1989; Pestka et al., 1990; Pestka et al., 1990a) . This information may suggest that while vomitoxin has an inhibitory effect on lymphocytes, it has a specific stimulatory effect on IgA producing cells, possibly mediated by increasing IgA specific helper T cells. 1.2.5.3 1g 21.-3:2 studies Following vomitoxin feeding, a significant increase in IgA secretion from both mitogen-stimulated and unstimulated PP lymphocytes was observed in culture (Pestka et al. , 1990) . This increase in the amount of IgA secretion was accompanied by an increase in the number of IgA secreting cells in both PP and spleen cultured with and without mitogens and taken from vomitoxin fed mice (Pestka et al., 1990). It is possible that following vomitoxin stimulation, PP B cells undergo terminal differentiation into IgA-secreting plasma cells. A large increase in the number of IgA secreting cells was found following only one day of in m culture suggesting a premature differentiation of IgA secreting cells in the PP (Bondy and Pestka, 1991). Also, the fact that IgA secretion and the number of IgA producing cells in PP and spleen cultures stimulated with lipopolysaccharide (LPS) is higher than 196 secretion, suggests an isotype specific effect possibly at the level of the PP. T cells may play a role in the vomitoxin-mediated terminal 29 differentiation of PP lymphocytes. The increase in CD4+ T cells in PP and spleen, as well as the increased CD4+/CDB+ ratio, implies an expansion of T cell help following vomitoxin ingestion (Pestka et al., 1990a). Further support for the involvement of T cells is the significant increase in IgA secretion when T cells from vomitoxin fed mice were cultured with control B cells (Bondy and Pestka, 1990). This T cell effect may be mediated by T lymphokines such as IL-4 and IL-5. In a study on IL-5 secretion by cultured T cells, an increase in IL-5 production by T cells cultured with vomitoxin was observed (Warner et al., 1992). 1.2.5.4 Effects on serum immunoglobulins Vomitoxin causes a dysregulation of serum immunoglobulin production by decreasing IgG and IgM and increasing IgA and IgE serum levels (Forsell et al., 1986; Pestka and.Dong, 1992; Tryphonas et al., 1986). Initially, the maximal effect of vomitoxin.on increased serum IgAuwas observed at a 10ppm level with serum IgA levels being somewhat less at the 25ppm level (Forsell et al., 1986). However, subsequent studies found 25ppm as the optimal level for IgA increase in the serum (Greene et al., 1992). Vomitoxin increases not only the total amount of serum IgA but also the ratio of polymeric IgA to monomeric IgA (Pestka et al., 1989) . In sera of animals exposed to 25ppm vomitoxin for 8 weeks the predominant IgA form was polymeric (pIgA) while control mouse serum contained mostly monomeric IgA. These data might also suggest a possible mucosal origin of the IgA produced since murine IgA 3O exists in a polymeric form in mucosal secretions whereas both monomeric and polymeric IgA are present in the serum. Vomitoxin also has an effect on reducing antigen-specific IgG and IgM and increasing antigen-specific IgA. In a study on the immunosuppressive effects of vomitoxin in mice, the amount of anti-SRBC (sheep red blood cells) antibody produced after an intraperitoneal injection of 10% SRBC was reduced following exposure to 10ppm vomitoxin diet (Robbana-Barnat et al., 1988). The number of plaque-forming cells against SRBC was also reduced following vomitoxin exposure (Pestka et al., 1987). Both studies tested for IgM antibodies. In another study, the IgA reactivity to intestinal antigens was tested. Vomitoxin feeding significantly increased IgA against casein (a component of .the diet) (Pestka et al., 1990). Cholera toxin (CT)-specific IgA was also increased in non CT-exposed mice but not in CT-exposed mice. Both casein and cholera toxin specific IgG were depressed in vomitoxin-fed mice. 1.2.5.5 IgA glomerulonephritis Beside dysregulation of IgA production, ingestion of vomitoxin causes glomerular IgA deposition (Pestka et al. , 1989). These symptoms are very similar to human IgA nephropathy ‘which is the most common glomerulonephritis worldwide and which has an unknown etiology. Similarity to human IgA glomerulonephritis is further supported by increased levels of serum IgA immune complex as well as hematuria found in ‘vomitoxin fed. mice (Dong et al., 1991). The only difference between ‘vomitoxin induced and. human. IgA 31 glomerulonephritis was found to be the absence of C3 complement component in the kidney mesangium of vomitoxin fed mice, which is usually present in human glomerulonephritis. These findings may suggest vomitoxin could be a possible etiological factor in IgA nephropathy. 1.3 IgA glomerulonephritis and autoimmunity 1.3.1 IgA glomerulonephritis IgA nephropathy (IgAN) , also known as Berger's disease, is the most common form of glomerulonephritis in the world with progressive potential and an unknown etiologic origin (Bene and Faure, 1987; D'Amico, 1987; Emancipator and Lamm, 1989). It appears to be more frequent in Asia than Australia, Europe and North America (Schena, 1990) . However, it has been observed that IgAN in Asia is commonly milder than in American and European patients who exhibit more severe renal lesions and have a higher rate of progression to renal insufficiency and end-stage kidney disease. This milder form of the disease probably results from intensive screening efforts for IgAN in Asia and earlier detection relative to absence of screening procedures in Europe and North America. IgAN is likely to begin with an aberration in the control of the mucosal immune response resulting in a large increase in serum IgA (especially polymeric IgA) and ends with IgA immune complexes that deposit in the kidney (Bene and Faure, 1987; D'Amico, 1987; Emancipator and Lamm, 1989; Jones et.al., 1990) . The unregulated production of IgA against dietary 32 antigens, pathogens and probably autoantigens (Emancipator and Lamm, 1989; Montinaro et al., 1991) leads to excessive IgA production and thus to glomerulonephritis. The IgA deposits are commonly associated with 196, 1g)! and C3 (complement component), the deposition of the latter increasing with time (Alamartine et al., 1990; Emancipator and Lamm, 1989). The kidney damage is manifested as hematuria and proteinuria with a 10-33% of the patients progressing to hemodialysis or renal transplantation (D'Amico, 1987; Emancipator and Lamm, 1989). There are also several abnormalities of the immune system such as an increase in T cells that have IgA Fc receptors following induction by IgA. Mechanisms which may be responsible for the observed increase in IgA in IgAN include: (1) defective clearing of polymeric IgA; (2) hyperresponsiveness of IgA producing B cells; (3) lack of IgA specific suppressor cells or augmented IgA specific helper functions; (4) induction of oral tolerance and (5) development of autoimmune reactivities. Susceptibility to IgAN has been linked by some investigators to sex, race and presence of a certain HHC component. Males are more likely to have IgAN than females especially in the second and third decade of their lives (Schena, 1990) . In addition African Americans have a lower incidence of IgAN (Emancipator and Lamm, 1989) . A genetic component of the MHC was also found to be relevant to susceptibility to IgAN. Since IgAN is commonly found in members of the same family, an investigation of the association between MHC alleles and susceptibility to IgA 33 nephropathy-revealed that presence of a DQw? allele at the HLA-DQB locus increases susceptibility of Caucasians to IgAN (Li et al., 1991). Many antigens have been postulated to be associated with the pathogenesis of the IgA generated in IgAN. For example, phosphorylcholine (PC) was found to be pathogenic in kidney deposits (Montinero et al. , 1991) . Here, injection of passively formed IgA/IgA-immune complex with pneumococcal C polysaccharide (PC containing antigen) caused glomerular injury in mice. Since C3 deposition in the glomerulus.was not correlated with the kidney damage, it is possible that activation of monocytes and the subsequent release of eicosanoids and oxygen radicals may cause kidney damage (Arima et al., 1991; Montinaro et al., 1991). This activation of monocytes can be mediated by deposition of immune complexes containing IgA with certain bacterial antigens (such as PC) in the kidney. The ability of activated macrophages to secrete cytokines such as interleukin 1 and interleukin 6 can further elicit kidney damage. These cytokines have the ability to induce mesangial proliferation and induction of HLA-DR antigens in cells resulting in mesangial damage (Arima et al. , 1991). 1.3.2 Autosntibodios and glomerulonephritis Putative pathogenic antibodies in various types of glomerulonephritis have been studied extensively. Autoantibodies have been detected against components of the mesangium as well as cellular components. IgG antibodies 34 against renal tubular epithelium were found in many cases of glomerulonephritis and in a study by Bruijn et al. (1990); a pathogenic role was suggested in naive mice following injection of the pathogenic IgG antibody. IgA antibodies against the laminin and collagen components of the renal tubule were found in IgAN patients (Shinkai et al., 1990). These antibodies were thought to cause mesangial IgA.deposits by direct binding between anti-laminin serum IgA and laminin in the kidney glomerulus. Other autoantibodies were also found, such as antibodies against neutrophil lysosomal enzymes whose pathogenic mechanism might be through activation of neutrophils in the kidney and tissue destruction (Kallenberg et al. , 1991) . In another study the connection between autoantibodies against renal tubule and activation of mononuclear cells was made. Naito and Sado (1989) found that the deposition of IgG anti-glomerular basement membrane antibody in the kidney resulted in the initiation of mononuclear cell mediated glomerulonephritis. Dysregulation of the idiotype-anti-idiotype network has also been raised as a possible mechanism in IgA glomerulonephritis. In studies by Gonzalez-Cabrero et al. (1989, 1991) cross-reactive idiotype bearing cells were found in most patients with IgA nephropathy as well as idiotypic antibodies in the kidneys of IgAN patients. These results indicated the possibility that the presence of an idiotypic antibody already bound to the kidney may cause deposition of anti-idiotypic antibody and increase the amount of IgA 35 deposited in the kidney facilitating tissue injury. While many investigators found specific autoantibodies in IgAN, polyclonal B cell stimulation has also been hypothesized to result in production of many antibodies, among them the self-reactive IgA antibodies. These self reactive IgA antibodies have been found to induce symptoms of autoimmunity. The definition of autoimmunity is ”a breakdown or failure of the mechanisms that are normally responsible for maintaining self tolerance" (Abbas et al., 1991). Such a breakdown can occur after a polyclonal stimulation of T or B lymphocytes, some of which are self-reactive clones that were not deleted during development.(Abbas et al., 1991). In the case of IgAN, polyclonal activation of IgA was observed together with a preferential production of polymeric IgA (Huso et al., 1991). The tissue damage by polymeric IgA produced following polyclonal stimulation is thought to result from binding to Fc receptors on mesangial cells. The binding of polymeric IgA to these Fc receptors on mesangial cells causes phagocytosis of the IgA and release of damaging metabolites from the mesangial cells (Muso et al., 1991). Another link between IgAN and autoimmunity is the postulated involvement of CBS (Ly 1) B cells in IgAN. CDS (Lyl) B cells are a primitive B cell lineage that produce polyspecific autoreactive antibodies and are sometimes mucosally-associated (Kasaian et al., 1991; Lalor, 1991) . CDS cells secrete polyreactive IgN against both self and foreign antigens such as proteins, DNA, cellular and bacterial 36 components (Kasaian et al., 1991). However, following immunization with bacterial cell wall antigens, CD5 B cells can also produce antibodies against phosphorylcholine (a bacterial antigen) and dextran (Lalor, 1991). The antibodies produced by CD5 B cells have been termed "natural antibodies" since they are normally present in the serum without previous immunization (Kasaian et al., 1991). The CD5 B cells are thought to play roles in autoimmunity due to their autoreactivity (Kasaian et al., 1991) and in protection against bacteria, serving as the first line of defense due to their ability to produce protective antibodies to bacteria (Lalor, 1991). There is a possible connection between CD5 B cells and IgA glomerulonephritis. CD5 B cells, when permanently stimulated by environmental and endogenous antigens in mucosal tissues, can clonally expand and start producing IgA (Jahn et al. , 1991). This polyspecific IgA might be responsible for the autoimmune-like disorders seen in IgA glomerulonephritis (Jahn et al., 1991). In addition, a study on B lymphocyte markers in IgAN found that B lymphocytes from IgAN patients had significantly higher populations of CD5 lymphocytes (Magyarlaki et al., 1990). These results are further supported by the presence of polyspecific IgA antibodies in serum from patients with IgAN, that can react against actin, tubulin, myosin, DNA, TNP and.albumin.(Matsiota et al., 1990). These antibodies are polyspecific and closely resemble the polyspecific natural antibodies produced from CD5 B cells as 37 a result of polyclonal stimulation (Matsiota et al., 1990). In another study by Guery et al. (1990) restricted sets of V genes were found to encode for the anti-laminin autoantibodies in autoimmune glomerulonephritis. This finding gives additional support. to ‘the jpossibility that CD5 B cells (producing natural antibodies encoded by a certain set of V genes) are polyclonally stimulated to produce polyreactive polymeric and autoreactive IgA antibodies. The ability of vomitoxin to cause IgAN resembles the capacity of other chemicals to cause autoimmune responses, and specifically glomerulonephritis. The most common experimental glomerulonephritis induced by chemicals is mercury-induced nephritis (Bigazzi, 1988; Guery et al., 1990; Pelletier et al., 1990). The administration of mercuric chloride to rats causes polyclonal B cell activation (Guery et al., 1990a) of restricted sets of V genes (Guery et al., 1990) possibly mediated.by anti-class II T cells (Guery et al., 1990a). This toxin mediated activation results in increased serum IgE (Pelletier et al., 1990), production of anti-laminin autoantibodies and finally glomerulonephritis. 1.4 Rationale Initially the similarity between oral tolerance and vomitoxin-induced suppression of total and antigen-specific IgG with concurrently increased total and antigen-specific IgA was observed. The possibility of oral tolerance induction by vomitoxin was studied involving the reaction of vomitoxin-fed 38 mice to oral gavage with a particulate antigen. In conjunction, B cell populations activated by vomitoxin feeding were analyzed. Cell surface markers and the cell cycle of IgA+ B cells were investigated in order to better characterize the B cell population activated by vomitoxin feeding. Since in previous studies an increase in IgA specific to two intestinal antigens (casein and cholera toxin) has been observed, the possibility that self reactive IgA was also elevated was tested. This information was significant in view of the abundant literature on the connection between IgA nephropathy and autoimmunity. Based on these initial studies I hypothesized that polyspecific autoreactive IgA is involved in IgAN in vomitoxin-fed mice. Specifically, the aims of this work were: (1) characterization of IgA and IgA-producing B cells stimulated by vomitoxin feeding in the mouse model; (2) assessment of the possible pathogenic potential of these IgA antibodies. 39 2.0 Materials and Hethods 2.1 General experimental design A series of experiments were undertaken to evaluate the effect of vomitoxin feeding on IgA antibodies and IgA producing B cells. Initially the effect of vomitoxin on the humoral response to oral gavage with a particulate antigen, TNP-SRBC was evaluated. Secondly, B cell populations in vomitoxin-fed mice were analyzed.by quantitating cell surface markers and cell cycle of IgA+ B cells. Finally the autoreactivity of the elevated IgA in vomitoxin fed mice was tested and possible pathogenic characteristics were assessed. 2.1.1. Bffect of dietary vomitoxin on TNP-specific serum immunoglobulin response to oral TNP-SRBC challenge. The purpose of this experiment was to test the effect of vomitoxin on antigen-specific response following oral antigen exposure. Due to an initial analytical problem, treatment BGC3F1 female mice (n=10) were fed 37.5ppm vomitoxin (instead of the optimal dose for increasing serum IgA - 25ppm) while control mice were fed clean control diet. One week after introduction of toxin food mice were bled and.preimmune serum was collected. At the same time a large decrease in the body weight of toxin fed mice was observed. I therefore decreased toxin level in the food by one half so that for the next 4 weeks 18.75ppm vomitoxin was fed to the treatment group. Control and treatment mice were gavaged with 2x106 TNP- SRBC/mouse on days 0, 1, 8 and 22 (counted after thejpreimmune blood was collected). One week after the last gavage (and 5 40 weeks since introduction of toxin food) mice were bled and sacrificed and TNP-specific 19 was quantitated (see below). h dayO -D day1—b day8 ——9 day22 -———> bled The failure of the oral immunization protocol in the initial experiment in increasing TNP-specific IgA was possibly attributed to insufficient length of vomitoxin feeding (5 weeks) combined with an ineffective immunization protocol. In a second experiment, treatment BGC3F1 female mice (n=8) were fed for 6 weeks with 25ppm vomitoxin, then 2 more weeks with 18.75ppm vomitoxin again due to excessive weight loss in the treatment group. Control mice (n=8) were fed clean control diet. After 8 weeks of feeding, mice were gavaged on 4 consecutive days with 4x109 TNP-SRBC/mouse. Five days later mice were bled and TNP-specific Igs were quantitated (see below) . + 4 day gavage ——-> bled 4 ----------------- 9’4 -------------------------- > 25ppm 18.75ppm 6 wk 3.3 wk 2.1.2 Effect of vomitoxin on PP and spleen B cell populations and on IgA specific to intestinal and self antigens. Experiments were undertaken in an attempt to quantitate the effect of vomitoxin feeding on serum Igs to naturally occurring intestinal antigens and self antigens in unimmunized 41 mice. B cell populations in PP and spleen were also studied to identify sub-populations of B cells that might be increased by vomitoxin feeding. Control and treatment BGC3F1 female mice (8 each) were fed for 8 weeks with 25ppm vomitoxin or clean control diet. Serum was collected at 4 and 8 weeks and analyzed for total 19 and phosphorylcholine (PC), inulin, HRBC (mouse red blood cells) and DNA-specific Ig. Following 8 weeks, PP, spleen and bone marrow cells were collected and stained for surface IgA, peanut agglutinin binding (PNA) , kappa light chain (K), both IgA+ and 10'"II (IgA+Km) and both IgA+ and peanut agglutinin binding (IgA+PNA*) . Cells were also cultured for 7 days and supernatants were assessed for total IgA. This experiment was repeated twice and in the second trial PP cells were also stained for the cell cycle of IgA+ cells. A year and a half later this cell cycle data on IgA-I- B cells was reassessed, following development of more information on the role of apoptosis on natural cell death in the immune system and its measurement by flow cytometry. Thus, the percentage of IgA+ apoptotic cells was calculated. 2.1.3 IgA hybridoma production and characterisation IgA-secreting hybridomas were produced to obtain more detailed information on the antigenic specificity and cross- reactivity of the hyperelevated IgA produced following vomitoxin feeding. In this experiment the animal model strain was changed from B6C3F1 mice to Balb/c mice for effective fusion with N81 cell line. Control and treatment Balb/c female mice were fed 25ppm vomitoxin or clean control diet for 42 8 weeks. Serum was collected and total and antigen specific IgA was measured. Two treatment mice with the highest serum IgA level were selected and sacrificed together with 2 control mice. Spleen and PP lymphocytes were isolated and fused with N81 myeloma cells. IgA secreting hybridomas were produced, scaled up and frozen for future characterization. Supernatants containing IgA were tested for reactivity against a panel of antigens for cross-reactivity. Supernatants were also fluorescinated to test for ability to directly bind tissue sections as a.measure of pathogenicity. The molecular size of the monoclonal IgAs was measured by Western blotting and by reduction alkylation of the IgA as an additional indication of pathogenicity, since polymeric IgA is thought to be more pathogenic. Fluorescent staining for presence of cytoplasmic CD5 molecule was performed to test if monoclonal IgAs descended from CD5+ precursors. 2.1.4 Relevance of monoclonal IgA in the mouse model Experiments were performed to ascertain that the IgA producing clones do, in fact, represent the population of lymphocytes being induced to secrete IgA by vomitoxin. Since previous studies concentrated on the B6C3F1 mouse model (a commonly used model in toxicological studies), it was important to find out whether Balb/c mice used for fusion exhibit the same effects following vomitoxin ingestion as B6C3F1 mice. Specifically, two experiments were set up to verify the autoreactivity of the monoclonal IgA in a Balb/c mouse model. In the first set, control and treatment Balb/c 43 female mice were fed 25ppm vomitoxin for 5 weeks and then 12.5ppm vomitoxin for another 3 weeks or clean control diet. In the second set treatment mice were fed 3 weeks 25ppm vomitoxin and then 5 more weeks 12.5ppm vomitoxin. In both trials the toxin amount was reduced due to rapid weight loss in the treatment mice group. After 8 weeks, serum was collected, mice were sacrificed, PP and spleen were removed for ELISPOt quantitation and kidneys were removed and stored at -80%L. Antigen specific spleen and PP cells were quantitated by ELISPOT against TNP-BSA, cardiolipin, sphingomyelin and PC. Serum was used to quantitate total IgA and antigen specific IgA as well as cross-reactivity to antigens. IgA antibodies were eluted from the kidneys of treatment mice and screened for ability to bind to several antigens in the same manner as the screening of monoclonal IgAs was performed. 2.1.5 Pathogenesis of monoclonal IgA In order to assess the possibility that the monoclonal IgA in itself is pathogenic two model monoclonal IgAs were injected into control Balb/c female mice, along with culture medium that was injected into 3 control mice. Mice were injected twice a week with 1mg IgA/mouse or with medium for 6 weeks. Urine was collected every 2 weeks and red blood cells were counted. Following 6 weeks of injections mice were sacrificed and the kidneys were sectioned and stained with anti-IgA FITC conjugate to measure IgA accumulation in the kidneys as additional measure of pathogenicity. 44 2.2 General procedures 2.2.1 Vomitoxin production Vomitoxin was produced in culture by the method of Witt et al. (1985). Potato dextrose agar plates were inoculated with stock soil cultures of Fusarium gramineargm (Qibbezella gene W8/U5373). Plates were sealed with.parafilm.and incubated in the dark at room temperature for 3-7 days until red colonies appeared. Plugs of 4mm size from the growing edge of the fungus were seeded into 100 ml autoclaved CMC medium (NHJflk, 1.0g; KHQPOH 1.0g; MgSO,.7H20, 0.5g; yeast extract, 1.09; carboxymethyl cellulose, 159; distilled. water' 1 liter - stirred and heated until CMC dissolves). Inoculated CMC flasks were agitated on a rotary shaker at 25°C for 3-5 days at 250 rpm. Samples were taken periodically to count spore concentration using a hemacytometer. After a concentration of 1x10°/flask or more was achieved, spores were filtered through 6 layers of sterile cheesecloth. Spores were inoculated onto autoclaved rice cultures (3509 enriched rice + 150ml distilled water, autoclaved at 121°C for 30 min.) in 2800ml Fernbach flasks stoppered with cotton plugs at 1x106 spores per flask. Inoculated rice cultures were incubated at 28°C in the dark for 13-18 days. At this stage considerable growth and red pigmentation appeared. Methanol (800ml) and water (600 ml) were added to each flask and the rice cultures were blended in a Waring blender with this solution. The homogenate was left overnight for complete extraction of the toxin into the solution. Blended material was filtered on a Buchner funnel 45 through a #4 Whatman filter paper (Whatman, Inc. , Clifton, NJ) and the resulting red filtrate was evaporated on a steam bath until the liquid was 10% of the original volume or less. The solution was saturated with sodium chloride and left overnight to precipitate the proteins. Precipitated proteins were filtered out by filtering on a Buchner funnel through #4 Whatman filter paper. 2.1.2 Toxin extraction and purification The culture filtrate was extracted 3 times with ethyl acetate (250ml ethyl acetate + 250ml culture filtrate) by gentle shaking (to prevent emulsion formation). Ethyl acetate, containing the toxin, was evaporated on a Buchi R110 rotary evaporator (Brinkmann Instruments, Inc., westbury, NY) at 45°C and the resultant brown concentrate was dried further in the hood until a viscous liquid was formed. At this point vomitoxin was dissolved in ethyl acetate and quantitated using TLC with a known vomitoxin control (purchased from Sigma). The toxin solution 'was dissolved in a small amount of methylene chloride and loaded onto a silica gel column to purify vomitoxin from other contaminants. A.Michael-Miller glass chromatographic column (37 mm i.d. x 300 mm) (Ace Glass, Inc., Vineland, NJ) was packed with silica gel (Adsorbosil, 200/450 mesh, Anspec) and fitted with safety shield, Teflon end fittings, connectors and tubing (2 mm i.d.). A pump (Model RP-SY-lCSC; Fluid Metering, Inc., Oyster Bay, NY) was first primed with methanol. Distilled water was pumped through the column, and the it was left to 46 equilibrate overnight. The column was then washed with methylene chloride and the toxin solution (also dissolved in methylene chloride) was loaded onto the top of the column. Extract movement through the column could be detected by following the movement of brown.pigment~ .As soon as the toxin started coming through the bottom part of the column, fractions were collected. Fractions were spotted onto a TLC plate together with a standard for 15-acetyl-deoxynivalenol and vomitoxin to test for the presence of these compounds. When 15-Acetyl-deoxynivalenol stopped coming through the column, the pump was washed with methanol and then with distilled water and the distilled water was pumped through the column. Fractions were not collected until the water passed through the column. Every third water fraction was tested for presence of vomitoxin on TLC and fractions containing vomitoxin‘were collected. Vomitoxin containing fractions were pooled together and further extracted. Water fractions containing vomitoxin were extracted with ethyl acetate (1:1 volume of each) 10 times“ ‘Water was tested for presence of vomitoxin.and.in.the case of remnant vomitoxin was further extracted. Ethyl acetate fractions containing vomitoxin were extracted on a rotary evaporator at 45°C until only a small volume remained. The solution was filtered through a #4 Whatman filter paper and sodium sulfate (to adsorb remaining water molecules) and further evaporated in a water bath with nitrogen or directly in the hood (in the case of a small volume). For crystallization, the solution was 47 transferred into a small beaker and seeded with vomitoxin crystals and kept refrigerated for several weeks until crystals formed. The pigment was washed off the crystals with ethyl acetate (and sometimes methanol) until the vomitoxin was white colored powder or crystals. At this time a small amount of vomitoxin was dissolved and quantitated by TLC and HPLC to ensure the purity and the amount of vomitoxin. Alternatively, the solution was quantitated before crystallization took place and used directly as a source of vomitoxin. 2.2.3 TLC (Thin Layer Chromatography) TLC chamber (inside dimension 27x7x26cm) was filled with a 1:3 v/v toluene ethyl acetate solution (25ml toluene and 75ml ethyl acetate) and equilibrated overnight. TLC plate (precoated 20x10cm silica gel G plate, Redi-Plates, Fisher Scientific Co. , Fair Lawn, NJ) was spotted with 0.25;:1, 0.51:1, lul, Zul, 5ul and 10:11 of a standard 1mg/ml solution of vomitoxin in methanol (purchased from Sigma) together with different amounts of sample. The TLC plate was put in the chamber until solvent rose almost to the top. The plate was taken out, dried at room temperature and sprayed with 15% aluminum chloride solution (159 aluminum chloride AlCl,.6I-120 + 85ml ethanol + 15ml distilled.water). Plate‘was dried at room temperature and put into 110°C for 5 minutes. Plate was quantitated by densitometry on a dual wavelength TLC scanner (Shimadzu, Japan) and the amount of vomitoxin was calculated according to the standard curve of the Sigma standard. 48 2.2.4 BPLC (High Pressure Liquid Chromatography) Vomitoxin samples were run at 20% methanol in water (previously filtered to prevent gas bubbles) (HPLC grade, Fisher Scientific Co.) on a Model 2300 HPLC pump and a V‘ variable wavelength absorbance detector (5-mm flow cell) (Isco, Inc., Lincoln, NE). The system had RP-18 Spheri-lo MPLC analytical (22 cm x 4.6 mm i.d.) and guard (3 cm x 4.6 mm i.d.) cartridges (Brownlee Labs, Inc., Santa Clara, CA). Column was flushed with methanol for 15 minutes and then 20% methanol was pumped through until the baseline stabilized. Detection was performed at 224nm. At a pump rate of lml/min the vomitoxin peak appeared after 4 minutes. At that time standard amounts of vomitoxin (purchased from Sigma) were injected into the column through a Valco C6U sample injector (Valco Instruments Co., Inc. Houston, TX) with a 10ul sample loop and the peak area was measured. Amount of vomitoxin in the sample was calculated by comparison with peak areas of standards. 2.2.5 Preparation of diet AIN-76A semi-purified diet (ICN Nutritional Biochemical, Cleveland, OH) was purchased commercially or prepared by mixing 2kg casein, 2.5kg sucrose, 5009 alphacel, 1.5kg cornstarch, 309 methionine, 209 bitartrate, 3509 mineral mix, 1009 vitamin mix and 5009 corn oil. Mixed diet was stored in plastic bags at 4°C and used within 2 months. Vomitoxin was dissolved in ethyl alcohol (200 proof) at 200mg vomitoxin.per 300ml ethyl alcohol. Vomitoxin.was mixed 49 in a large beaker with 2009 AIN-76A diet to make up a stock 1000ppm vomitoxin food. The alcohol was slowly evaporated while mixing carefully to ensure equal distribution of the toxin throughout the food. When the stock 1000ppm food was completely dry, it was used to prepare the desired concentration of toxic food by mixing with clean AIN-76A diet in a KitchenAid mixer (model KS-A, Hobart Mfg. Co., Troy, Ohio). To verify toxic diet concentration, a 59 food sample was extracted into 25ml acetonitrile-water (9:1) solution by shaking for 60 minutes. The extract was centrifuged at 450xg for 8 minutes and the supernatant containing the toxin was removed. Five ml of the supernatant was passed through a #215 charcoal column (Romer Labs, Washington, MD) by using a water pump and collected into a glass tube. The extract was evaporated on an analytical evaporator (Organomation Associates, South Berlin, MA) with nitrogen pressure and redissolved in 1 ml acetone-methanol (2:1) for TLC or 20% methanol-water for HPLC. TLC plates were run in toluene- acetone solution (1:1) as described above. HPLC was run in 20% methanol-water as described above. 2.2.6 Safety Face masks and vinyl gloves were used when handling the toxin. Concentrated toxin was handled in a fume hood. Contaminated glassware and equipment were detoxified by soaking in 10% bleach solution (Thompson & Wannemacher, 1984) . 50 2.2.7 Animals B6C3Fl or Balb/c female mice (8-10 weeks old, Harlan/Sprague-Dawley, Indianapolis, IN) were :randomized, housed in protected-environment cages (Nalgene, Rochester, NY) and fed powdered AIN-76A semi-purified diet (ICN Nutritional Biochemical, Cleveland, OH). Distilled.water was provided ad up and feed was changed every 3-4 days. The mice were acclimatized to housing, feed, a 12-hr light/dark cycle and to a negative-pressure ventilation area for at least 7 days prior to initiating experiments. Mice were fed control diet or diet containing vomitoxin. At intervals, mice were anaesthetized with ether and bled at the retro-orbital plexus. Serum was analyzed by ELISA and mice were sacrificed at selected time points for cell preparation. If necessary, other tissues (such as kidneys, livers etc.) were removed at the same time and stored frozen at -80°C until further analysis. 2.2.8 Cell preparation PP were removed, teased apart in RPMI-l640 medium (Sigma), passed through a sterile 85-mesh stainless steel screen and resuspended in the same medium. Cells were centrifuged at 450xg for 10 minutes, resuspended in RPMI-1640 medium and counted. Spleens were removed, teased apart with tissue forceps in RPMI-1640 medium and left on ice for 10 minutes for tissue particles to settle. Supernatants were removed and centrifuged at 450xg for 10 minutes. Erythrocytes were lysed by resuspending in 0.83% ammonium chloride for 3 minutes at 51 room temperature. Cells were centrifuged, resuspended in RPMI-1640 medium and counted. 2.3 Immunological methods 2.3.1 Antigen preparation Phosphorylcholine bovine serum albumin conjugates (PC-BSA) and inulin-BSA, prepared as described by Gearhart and Cebra (1979) , were kindly provided by J. Cebra (University of Pennsylvania, Philadelphia, PA). Salmon sperm DNA, actin, sphingomyelin, collagen type I, mouse IgG, thyroglobulin, laminin and cardiolipin were purchased from Sigma. Salmon sperm DNA was extracted with phenol/chloroform (Ausubel et al., 1990). Mouse red blood cells (MRBC) were obtained by orbital bleeding of B6C3F1 female mice and bromelated as described by Klinman and Steinberg (1987) . Briefly, MRBC were centrifuged and diluted again to 5% in 0.1% bromelain in saline solution. MRBC were incubated at 37°C for 30 minutes followed by 3 washes in PBS by centrifugation at 450xg for 10 minutes and diluted to a concentration of 107 cells/m1. Bovine serum albumin (BSA) was trinitrophenylated (TNP- BSA) using picric acid (Good at al. 1980). One hundred (100) mg of BSA was mixed with Sml distilled water and 100 mg potassium carbonate and then 100 mg picrylsulfonic acid (purchased from Sigma) were added. The mixture was covered with aluminum foil and stirred at room temperature overnight. The conjugate was dialyzed in the dark at 4°C for 3 days against borate buffer (0.2M, pH 9.2) with several buffer 52 changes. Following dialysis the amount of BSA as well as the ratio between moles BSA and moles TNP were calculated by spectrophotometric reading at 278nm for protein and 348nm for TNP (extinction coefficient.= 15,400). It‘was calculated that the BSA was completely conjugated with 53 TNP molecules/one BSA molecule. Conjugation was also verified by ELISA using MOPC 315 anti-TNP antibodies. Sheep red blood cells were trinitrophenylated (TNP-SRBC) using picrylsulfonic acid (Sigma) as described by Rittenberg and Pratt (1969). Picrilsulfonic acid (Sigma) (1 ml) was diluted in 7 ml cacodylate buffer (0.28M, pH 6.9). Packed SRBC (1 ml) were added to this solution, mixed well and incubated at room temperature for 10 minutes. PBS was added to stop conjugation (3 volumes) and TNP-SRBC were centrifuged at 450xg for 10 minutes, washed 3 times and resuspended to a concentration of 1X109cells/ml. Conjugation of TNP to SRBC was verified by passive hemagglutination with MOPC 315 antibodies (anti-TNP IgA, Sigma) as described by Herbert (1978). 2.3.2 Total and antigen specific immunoglobulin quantitation Ig isotype concentrations and antigen-specific 19 were determined by enzyme linked immuno-sorbent assay (ELISA). For ELISA solid phase, Immunolon 2 (Dynatech) microtiter wells were incubated overnight at 4°C with 50ul of PC-BSA, inulin BSA, TNP-BSA, actin, sphingomyelin, mouse IgG, thyroglobulin, or anti-lg (heavy chain specific antibody) (Sigma Chemical Co.) at a concentration of long/ml in 0.5M carbonate buffer 53 (pH 9.6). Salmon sperm DNA (long/ml) was coated in 0.1M phosphate buffered saline (PBS) (pH 7.5) by air drying overnight at 40°C (Pestka and Chu 1984) in an isotemp oven (Fisher Scientific). Collagen was diluted to a concentration of 10u9/ml in 0.1M acetic acid and air dried overnight. Cardiolipin was diluted to a concentration of 50u9/ml in 70% alcohol and air dried overnight. Wells were washed on a MicrowashII automatic plate washer (Skatron, Norway) three times with PBS containing 0.2% Tween 20 (PBS-Tween) and then incubated with 0.3ml 1% (wt/vol) bovine serum albumin in PBS (BSA-PBS) for 30 minutes at 37°C to block nonspecific binding. In the case of MRBC, a monolayer of bromelated MRBC cells was produced in microtiter plates by pipeting 150 ul/well of 1x107 MRBC/ml suspension and centrifugation at 450xg for 5 minutes. PBS was gently aspirated and 200ul/well of 0.15% glutaraldehyde in PBS were added for 5 minutes in order to fix the cells onto the plastic well. Plates were washed twice more in PBS, air dried for 3 hours, blocked with 1% BSA in PBS, washed 3 more times in PBS and stored at 4°C until used. Serum samples were diluted in 1% BSA-PBS and added at 50ul/well to blocked and washed plates. Standard mouse reference serum (Bethyl Laboratories, Montgomery, TX) was added to anti-Ig wells that flanked the antigen coated wells for use in isotype reference curves. Plates were incubated 1 hour at room temperature or overnight at 4°C, washed and incubated for 1 h with anti-mouse Ig (heavy chain specific) alkaline phosphatase conjugate (Sigma Chemical Co.) diluted 54 1:500 in 1% BSA-PBS. Plates were washed six more times and then reacted with 100ul p-nitrophenyl phosphate (Sigma) (4mg/ml) in glycine buffer (pH 10.5) . Absorbance at 405nm was read on the EIA Minireader 2 (Dynatech, Alexandria, VA). Initially, to minimize day-to-day variability in absorbance readings encountered with ELISA, Igs bound to the antigens were assigned gravimetric equivalents (ng/ml) based on comparison of absorbance to values obtained in isotype reference curves run concurrently. Later, during screening of monoclonal IgA, reactivity to antigens was standardized by a reference 100 ng/ml MOPC 315 IgA precoated onto wells. Monoclonal IgAs were diluted in 0.05% Tween in PBS at 10ug/ml. Antigen coated wells were blocked by 30 minute incubation with 0.3 ml/well of 1% BSA-PBS (w/v) at room temperature. Plates were washed and incubated for 1 hour at room temperature with 50111 of monoclonal IgA/well. Plates were washed again and incubated with 50ul/well of anti-mouse IgA alkaline phosphatase conjugate diluted 1:500-1:2000 (depending on the titer of the conjugate) in 1% BSA-PBS for 1 hour at room temperature. Plates were washed and reacted with 100ul/well p-nitrophenyl phosphatase for 1 hour. Positive reactivity to any antigen was obtained when the 0.D. at 405 nm of the tested supernatant was higher than the average blank + 2 fold the standard deviation of the blank. Blank was uncoated wells that were blocked with 1% BSA-PBS, and reacted with the monoclonal IgAs and enzyme conjugate. 0.D. of each monoclonal IgA to each antigen was normalized by dividing with 55 the 0.D. of MOPC 315 IgA which ranged between approximately 0.5 and 1.5. This method standardized the ELISA to different antigens in such a way that a comparison of the 0.D. could be made as a measure of ability of antibody to bind to that antigen. 2.3.3 Antigen inhibition ELISA Antigen coating and blocking were preformed in the same way as described in the ELISA section. During the blocking stage SOul of the antibody tested was premixed with 50ul of the inhibiting antigen. Following the blocking step the premixed solution was added at 100u1 per well and the ELISA proceeded as previously described. Results were read as the % of maximal absorbance that was obtained when the antibody is premixed with an equal solution of PBS. 2.3.4 IgA elution from kidneys IgA was eluted from kidneys by a modification of the procedure by Matsiota et al. (1990). Kidneys from four (4) treatment mice were sectioned onto glass slides. The slides for each mouse were put back to back into a Coplin jar and immersed in glycine buffer with a pH 2.8 for 1 hour while shaking at 80 rpm. The buffer was collected and sections were washed again in glycine buffer (pH 2.8) for another hour and then twice more in glycine buffer with pH 2.2. Buffer was pooled from the four washes and centrifuged at 450 x g for 10 minutes to remove pieces of tissue. The pH of the eluate was adjusted to 7.8 and it was precipitated with an equal volume of ammonium sulfate (pH 7.8). The eluate was centrifuged at 56 10,000 rpm for 20 min. and the precipitated protein was resuspended in PBS with 0.02% azide. Eluent was washed 3 more times by centrifugation on Centriprep 10 concentrators (Amicon, W.R. Grace & Co., Conn. Beverly, MA). The eluent was finally resuspended in 1% BSA-PBS with 0.02% azide and the IgA concentration was measured by ELISA as previously described. IgA concentration in all eluates was adjusted to 10,000 ng/ml and they were screened against 7 antigens. 2.3.5 ELISPOT Immunolon 4 wells (Dynatech, Chantilly, Virginia) were coated with coating buffer containing long/ml PC-BSA or sphingomyelin, 50pg/ml cardiolipin or 100ug/ml TNP-BSA. Plates were incubated overnight at 4°C followed by 5 washes in PBS-Tween (Harlow and Lane, 1988) . They were blocked with 300ul/well 1% BSA-PBS and incubated at room temperature for 30 minutes. Plates were washed 6 times in PBS-Tween and 100ul/well of different concentrations of cell suspension were added. Plates were then incubated at 37°C in a C02 incubator, washed 6 times with PBS-Tween and 6 more times with distilled water (to lyse the cells). Anti-IgA alkaline phosphatase (Sigma) was added at 50ul/well (1:200 dilution in 1%BSA-PBS) and plates were incubated overnight at 4°C. Substrate consisted of 6.7 ml BCIP (5-bromo-4-chloro-3-indolyl phosphate) (0.15 g in 100 ml AMP buffer: 9.62 ml 1M 2-amino-2- methyl-l-propanol + 15 mg magnesium chloride + 10 ul Triton X- 405 + 0.01 9 sodium azide + 1 liter distilled water), 100 pl nitro blue tetrazolium (10mg/ml in 70% dimethylformamide) , 1.2 57 ml distilled water and 3 ml boiled 3% agarose (Sigma fA6013) which were mixed together and kept in a water bath at 45%: until ready to use. The plates were washed 6 times with PBS— Tween and 100ul/well tempered substrate were added. Plates were wrapped in aluminum foil and incubated at 37°C for 2 hours. Following the incubation plates were viewed with an inverted microscope with a piece of Kim-wipes on top (to diffuse the light) and positive spots of antibody producing cells were counted. 2.3.6 Flow cytometry Cell suspensions of 1X106 cells were washed by centrifugation at 450xg for 3 minutes and placed into Eppendorf tubes. Cells were incubated on ice for 20 minutes with 1% BSA-PBS azide (0.02% azide w/v) to block non-specific binding sites. Cells were centrifuged at 450xg for 3 minutes and washed twice in 1%BSA-PBS azide. The first antibody diluted in 1%BSA-PBS azide was added at 100ul/tube (anti kappa-biotin 1:500 dilution, PNA (peanut agglutinin)-biotin 1:33.3 dilution, rat IgG-biotin 1:1000 dilution) and tubes were incubated for 20 minutes on ice. Tubes were centrifuged and the cells were washed 3 times in 100ul PBS azide. The secondary incubation solution diluted in 100ul/tube 1%BSA-PBS azide was added (PE-avidin 1:500 dilution, anti-IgA FITC 1:25 dilution, rat IgG-FITC 1:1000 dilution) and tubes were incubated on ice for another 45-60 minutes. Cells were washed 3 times in 100ul/tube PBS azide. Cells were resuspended in 1ml 1% formaldehyde in PBS and stored at 4°C in the dark until 58 sorted. For cell cycle staining cells were resuspended in 0.2ml PBS + 0.2ml FCS (fetal calf serum), fixed with 1.2ml 70% ice cold alcohol (dropwise while mixing) and held on ice overnight. Following alcohol fixation protein precipitate formed which was washed with 2% FCS-PBS and cells were resuspended in 1ml PI stain (3.76ml PBS + 0.2ml propidium iodide (1mg/ml) + 0.04ml RNAse A (5mg/m1)) and incubated for 1 hour at room temperature. Samples were kept cold and in darkness until analysis. For cytoplasmic CD5 staining the procedure of Schmid et al. (1991) was used, Cells were washed in PBS and resuspended at 1x105/0.875ml cold PBS. Cold 2% paraformaldehyde was added (0.125m1) and the cells were vortexed immediately. Samples were incubated at 4°C for 1 hour. Cells were centrifuged at 250xg for 5 minutes and supernatant was removed. Pellet was resuspended in 1 ml of 0.2% Tween 20 in PBS and incubated at 37°C for 15 minutes. One mililiter (1 ml) of 2% fetal calf serum (FCS) in PBS was added. Cells were spun down again and resuspended in minimal amount of 2% FCS PBS. A rat anti-mouse Fc gamma II receptor antibody solution was added at 5 ul of 1:5 antibody in order to prevent non-specific binding through the Fc gamma receptor. The samples were incubated on ice for 5 minutes. Half ml (0.5 ml) of 0.2% Tween 20 in PBS were added and the cells were centrifuged. The supernatant was removed and FITC labeled.anti-CD5 antibody was added (10 pl of 200 ug/ml). Cells were incubated on ice for 20 minutes and washed 3 times in 0.2% Tween 20 in PBS. Cells were finally 59 resuspended in 1% paraformaldehyde and stored at 4°C and in the dark until analyzed. Cell cycle analysis was performed on FACScan cell sorter (Becton Dickinson) and analyzed by CellFIT cell-cycle analysis program version 2.0 at Sparrow hospital (Lansing, MI). All other analyses were performed using an Ortho Diagnostics Cytofluorograph 50-H Fluorescence-Activated Cell Sorter (FACS)/2150 computer system in MSU Giltner Hall. 2.3.7 Immunofluorescence Immediately following sacrifice of mice, kidneys, liver, spleen and tails were frozen in liquid nitrogen onto a solid support (cork) and stored at -80°C. Tissues were sectioned on a cryostat and sections were stained with diluted f luorescinated antibody in PBS. Peritoneal exudate cells were obtained by washing the peritoneal cavity of Balb/c female mice with PBS, collecting the exudate and incubating the cell suspension on poly-L-lysine coated slides for 10 min. Each section was thawed out and circled with PAP pen (Daido Sangyo Co. , Japan). The sections were blocked with 1%BSA-PBS for 15 minutes, the BSA-PBS was shaken off and then 501:1 of the fluorescinated antibody was added. Tissue sections were incubated with the fluorescinated antibody for 45 minutfi3 1“ a dark and moist chamber. Following staining slides were sh washed briefly 3 times by immersion into a Coplin jar of fire with PBS and then further washed by 3 washes in a Cop]. in ‘3 a: o“ ‘5 clean PBS for 10 minutes at a time at a speed of 601'?” 1.19““; shaker. Excess PBS was shaken off and a drop of FA 1“ 60 fluid (pH 9, Difco laboratories, Detroit, MI) was placed on top of each section. Sections were then covered with a cover slide and either looked at immediately or frozen at -20°C until further analysis. 2.3.8 Hybridoma production PP lymphocytes were fused with N81 cells according to the protocol of Galfre and Milstein (1983) with the modifications by Abouzied et al. (1990). Mouse PP and spleen cells (2.2x107) from two control and vomitoxin fed mice were fused with NS-l myeloma cells (2x107) with PEG as described by Galfre and Milstein (1983). Following fusion the cells were suspended in 20% fetal bovine serum (PBS) in Dulbecco's modified medium (Sigma) with 1% NCTC medium (Sigma), 10 mM sodium pyruvate and pen/ strep solution (Sigma) and distributed into 96 well tissue culture plates. The plates were then incubated at 37°C with humidity and 7-8% CO2 in air. Every 3 days the wells were fed with 20% FBS-DMEM containing hypoxanthine, aminopterin and thymidine (HAT medium). One week after fusion the fusion efficiency was checked. After 2 weeks wells were fed with 20% FBS-DMEM containing hypoxanthine and thymidine (HT medimm) and screening for IgA production began. Supernatants were collected from wells exhibiting growth and tested by ELISA for production of IgA antibodies. All wells that contained cells producing IgA were transferred into 24 well tissue culture plates and then into 50 and 250 ml tissue culture flasks. Hybridomas were cloned by limiting dilution in 20% macrophage conditioned medium, scaled up to 61 large flasks and then supernatants and cells were frozen for further analysis. 2.3.9 PAGE and Western blotting. Monoclonal IgA supernatants were precipitated 3 times with 50% ammonium sulfate, resuspended in PBS at a concentration of 1 mg/ml IgA and 25ul was loaded onto a precast Mini-PROTEAN II native 4-17% gradient gel (Bio-Rad Laboratories, Richmond, CA) . PAGE and Western blotting were performed using Mini- PROTEAN II electrophoresis cell , Mini Trans-Blot electrophoretic transfer cell and power supply model 200/2.0 (Bio-Rad Laboratories) (Coligan et al. 1992). PAGE was done at 100v for 4 hours and the gel was blotted onto PVDF transfer membranes (Biotechnology Systems, Boston, MA) for 1 hour at 100v in 0.01% SDS in the running buffer. Membranes were blocked with 5% BSA and then detected with 1:1000 anti IgA alkaline phosphatase (Sigma) in 1% BSA. 2.3.10 Reduction and alkylation of monoclonal IgA. Monoclonal IgA supernatants were reduced and alkylated as previously described (Pestka et al. 1989). IgA supernatants were precipitated with ammonium sulfate (12.5ul saturated ammonium sulfate added to 251:1 sample in PBS), incubated overnight at 4°C and centrifuged at 14OOXg for 30 minutes. Supernatant was removed and the protein was reduced by adding 201:1 DTT (dithiothreitol) (10mg/ml in 0.275M Tris-HCl buffer). The tubes were flushed with nitrogen and incubated with nitrogen for 1h at 37°C. Alkylation was performed by addition of Sul iodoacetamide (22mg/ml iodoacetic acid in 0.275M Tris- 62 HCl buffer) flushing again with nitrogen and 15 minutes incubation at 4W3. Staining solution was added and samples were loaded onto gel. Matching non reduced alkylation samples were precipitated with ammonium sulfate and resuspended in 25;:1 PBS. 2.3.11 Labeling monoclonal IgA.with fluorescein. Monoclonal antibody supernatants were brought to 1 mg/ml IgA concentration and then labeled with FITC as described (Coligan et al. 1992) . Supernatants were first dialyzed against labeling buffer (0.05M boric acid, 0.2M NaCl, pH 9.2) for 3 days at 4°C with several buffer changes. Optical density at 280nm was read and protein concentration calculated (protein mg/ml = A280 x 0.74 x dilution factor). For each 1 mg of protein 4ul of 5mg/ml FITC in DMSO ‘were added. Conjugation to FITC proceeded at room.temperature for 2 hours while gently shaking the solution. Unbound FITC was removed by extensive dialysis against 0.1M Tris buffer (pH 7.4) containing 0.1% vol/vol NaN3 and 0.2M NaCl for 3 days with several buffer changes. FITC/protein ratio was determined by reading optical density at 280nm (for protein) and 492nm (for FITC). A ratio FITC/protein of 5-6 was considered optimal. Fluorescinated antibodies were stored with 0.02% NaN3 in PBS at 4°C in the dark. 2.4 Statistical analysis Differences between vomitoxin-treated and control groups were analyzed by Student's t test using SigmaPlot as well as correlation coefficients (Jandel Sci., Corte Madera, CA). 63 3.0 RESULTS 3.1 Vomitoxin production Vomitoxin was produced on rice cultures with an approximate yield of 2 grams toxin per 5 kg. rice (400ppm) . This was similar to yields reported by Witt et al. (1985) . Following extraction and purification, the toxin was quantitated by HPLC and TLC (Fig. 3) . During quantitation by HPLC under the conditions described, the vomitoxin peak appeared approximately 3.7 minutes after injection onto the column (Fig. 4) . According to HPLC the vomitoxin concentration in the crude solution was 12.6 mg/ml. By TLC quantitation, using the dual length TLC scanner, the vomitoxin concentration in the crude solution was 13.9 mg/ml. Thus, both methods gave very similar results on vomitoxin amount in the preparation. Following chromatography, vomitoxin fraction did not appear to contain contaminants based on the TLC or HPLC. Since this fraction was relatively pure and.also because of loss of large amounts of toxin during crystallization, it was used directly for feeding studies. 3.2 Effect of dietary vomitoxin on mouse weight and serum Igs The most commonly known effect of vomitoxin feeding is feed refusal. When B6C3F1 female mice were fed a diet of 25ppm vomitoxin they exhibited feed refusal. Their body weight did not increase with time. After 4 weeks of feeding the difference between treatment and control mice was significant and this trend increased toward the termination of the experiment at 8 weeks (Fig. 5). These results were 64 Fig. 3. Quantitation of vomitoxin amount by TLC and HPLC. .A = standard curve following reading of standard vomitoxin amounts on a dual wavelength TLC scanner. B = standard curve of standard vomitoxin amounts by HPLC. AU = arbitrary units. 10.. AU x 1000 o" 65 120- r 80-- AU x 10,000 40 '1?- E 3 Vomitoxin (ug) (”i .b-J- 200 460 600 360 1dbo Vomitoxin (n9) CH- Pig. 3 66 AU A 7 h”) 1.47 1.25 5' .E E K :39? “ AU 1; 3 0 E 1.45 "'5 1.27 E 2.52 E :3): *— f Fig. 4. Typical appearance of vomitoxin on HPLC. A - standard 500 ng vomitoxin, B - purified sample of 500 ng vomitoxin. Arrows pointing to vomitoxin elution peak. AU - arbitrary units. 67 35 l T I I I 0 control 0 treatment 30 ~ §> - ”a? E O L. CD \«I 25 — /////// e z . .. .9 M: Q) 3 20 — - 15 l 1 l l l O 2 4 6 8 Weeks feeding Fig. 5. Effect of feeding 25ppm vomitoxin on B6C3F1 mouse weight. Data are means :1: SEM. (P<0.05) from matching control. (P<0.01) from matching control. * a significantly different ** = significantly different 68 similar to those found by Forsell et al. (1986). Analysis of serum immunoglobulins at 4 and 8 weeks exhibited serum IgA dysregulation that was noted in previous studies. There was a significant increase in total serum IgA coupled to a decrease in total IgG and IgM both after 4 and 8 weeks vomitoxin feeding although 196 levels recovered after 8 weeks (Fig. 6). Total serum IgA was increased 2 and 13 fold after 4 and 8 weeks (Fig. 6) of feeding 25ppm vomitoxin, respectively. 3.3 Effect of dietary vomitoxin on TNP-specific serum immunoglobulin response to oral TNP-SRBC challenge Further efforts to understand vomitoxin dysregulation of serum IgA were directed at studying the effect of vomitoxin feeding on oral immunization with a particulate antigen. TNP- SRBC was chosen since it is a model antigen commonly used in immunological studies to gauge the immune response. Another reason was the ease of using TNP in detecting TNP-specific responses by ELISA. TNP is easily conjugated to proteins which allows rapid preparation of the oral immunogen TNP-SRBC as well as the TNP-BSA conjugate used to coat plates. Oral administration of TNP-SRBC was used to investigate whether vomitoxin had a specific adjuvant effect in enhancing serum IgA response to an exogenous antigen bolus. Repeated gavaging of control mice with TNP-SRBC caused a large increase in TNP- specific IgG with a slight IgM response, whereas an IgA response to TNP was not detectable (data not shown). When 69 Pig. 6. Effect of feeding 25ppm vomitoxin for 4 and 8 weeks on total serum immunoglobulins in B6C3F1 mice. Data are means 1 SEM (n=8) and are representative of 3 experiments. * = significantly different (P<0.05) from matching controls. 48 2.2.4 HPLC (High Pressure Liquid Chromatography) Vomitoxin samples were run at 20% methanol in water (previously filtered to prevent gas bubbles) (HPLC grade, Fisher Scientific Co.) on a Model 2300 HPLC pump and a V‘ variable wavelength absorbance detector (S-mm flow cell) (Isco, Inc., Lincoln, NE). The system had RP-18 Spheri-lo MPLC analytical (22 cm x 4.6 mm i.d.) and guard (3 cm x 4.6 mm i.d.) cartridges (Brownlee Labs, Inc., Santa Clara, CA). Column was flushed with methanol for 15 minutes and then 20% methanol was pumped through until the baseline stabilized. Detection was performed at 224nm. At a pump rate of 1ml/min the vomitoxin peak appeared after 4 minutes. At that time standard amounts of vomitoxin (purchased from Sigma) were injected into the column through a Valco C6U sample injector (Valco Instruments Co., Inc. Houston, TX) with a 10ul sample loop and the peak area was measured. Amount of vomitoxin in the sample was calculated by comparison with peak areas of standards. 2.2.5 Preparation of diet AIN-76A semi-purified diet (ICN Nutritional Biochemical, Cleveland, OH) was purchased commercially or prepared by mixing 2kg casein, 2.5kg sucrose, 5009 alphacel, 1.5kg cornstarch, 309 methionine, 209 bitartrate, 3509 mineral mix, 1009 vitamin mix and 5009 corn oil. Mixed diet was stored in plastic bags at 4°C and used within 2 months. Vomitoxin was dissolved in ethyl alcohol (200 proof) at 200mg vomitoxin per 300ml ethyl alcohol. Vomitoxin was mixed 49 in a large beaker with 2009 AIN-76A diet to make up a stock 1000ppm vomitoxin food. The alcohol was slowly evaporated while mixing carefully to ensure equal distribution of the toxin throughout the food. When the stock 1000ppm food was completely dry, it was used to prepare the desired concentration of toxic food by mixing with clean AIN-76A diet in a KitchenAid mixer (model KS-A, Hobart Mfg. Co., Troy, Ohio). To verify toxic diet concentration, a 59 food sample was extracted into 25ml acetonitrile-water (9:1) solution by shaking for 60 minutes. The extract was centrifuged at 450xg for 8 minutes and the supernatant containing the toxin was removed, Five ml of the supernatant was passed through.a #215 charcoal column (Romer Labs, Washington, MD) by using a water pump and collected into a glass tube. The extract was evaporated on an analytical evaporator (Organomation Associates, South Berlin, MA) with nitrogen pressure and redissolved in 1 ml acetone-methanol (2:1) for TLC or 20% methanol-water for HPLC. TLC plates were run in toluene- acetone solution (1:1) as described above. HPLC was run in 20% methanol-water as described above. 2.2.6 Safety Face masks and vinyl gloves were used when handling the toxin. Concentrated toxin was handled in a fume hood. Contaminated glassware and equipment were detoxified by soaking in 10% bleach solution (Thompson & Wannemacher, 1984) . 50 2.2.7 Animals B6C3F‘ or Balb/c female mice (8-10 weeks old, Harlan/Sprague-Dawley, Indianapolis, IN) were randomized, housed in protected-environment cages (Nalgene, Rochester, NY) and fed powdered AIN-76A semi-purified diet (ICN Nutritional Biochemical, Cleveland, OH). Distilled water was provided 39 11b and feed was changed every 3-4 days. The mice were acclimatized to housing, feed, a 12-hr light/dark cycle and to a negative-pressure ventilation area for at least 7 days prior to initiating experiments. Mice were fed control diet or diet containing vomitoxin. At intervals, mice were anaesthetized with ether and bled at the retro-orbital plexus. Serum was analyzed by ELISA and mice were sacrificed at selected time points for cell preparation. If necessary, other tissues (such as kidneys, livers etc.) were removed at the same time and stored frozen at -80°C until further analysis. 2.2.8 Cell preparation PP were removed, teased apart in RPMI-1640 medium (Sigma), passed through a sterile SS-mesh stainless steel screen and resuspended in the same medium. Cells were centrifuged at 450xg for 10 minutes, resuspended in RPMI-1640 medium and counted. Spleens were removed, teased apart with tissue forceps in RPMI-1640 medium and left on ice for 10 minutes for tissue particles to settle. Supernatants were removed and centrifuged at 450xg for 10 minutes. Erythrocytes were lysed by resuspending in 0.83% ammonium chloride for 3 minutes at 51 room temperature. Cells were centrifuged, resuspended in RPMI-1640 medium and counted. 2.3 Immunological methods 2.3.1 Antigen preparation Phosphorylcholine bovine serum albumin conjugates (PC-BSA) and inulin-BSA, prepared as described by Gearhart and Cebra (1979) , were kindly provided by J. Cebra (University of Pennsylvania, Philadelphia, PA). Salmon sperm DNA, actin, sphingomyelin, collagen type I, mouse IgG, thyroglobulin, laminin and cardiolipin were purchased from Sigma. Salmon sperm DNA was extracted with phenol [chloroform (Ausubel et al., 1990). Mouse red blood cells (MRBC) were obtained by orbital bleeding of B6C3F1 female mice and bromelated as described by Klinman and Steinberg ( 1987) . Briefly, MRBC were centrifuged and diluted again to 5% in 0.1% bromelain in saline solution. MRBC were incubated at 37°C for 30 minutes followed by 3 washes in PBS by centrifugation at 450xg for 10 minutes and diluted to a concentration of 107 cells/ml. Bovine serum albumin (BSA) was trinitrophenylated (TNP- BSA) using picric acid (Good et al. 1980) . One hundred (100) m9 of BSA was mixed with Sml distilled water and 100 mg potassium carbonate and then 100 mg picrylsulfonic acid (purchased from Sigma) were added. The mixture was covered with aluminum foil and stirred at room temperature overnight. The conjugate was dialyzed in the dark at 4°C for 3 days against borate buffer (0.2M, pH 9.2) with several buffer 52 changes. Following dialysis the amount of BSA as well as the ratio between moles BSA and moles TNP were calculated by spectrophotometric reading at 278nm for protein and 348nm for TNP (extinction coefficient.= 15,400). It was calculated.that the BSA was completely conjugated with 53 TNP molecules/one BSA molecule. Conjugation was also verified by ELISA using MOPC 315 anti-TNP antibodies. Sheep red blood cells were trinitrophenylated (TNP-SRBC) using picrylsulfonic acid (Sigma) as described by Rittenberg and Pratt (1969). Picrilsulfonic acid (Sigma) (1 ml) was diluted in 7 ml cacodylate buffer (0.28M, pH 6.9). Packed SRBC (1 ml) were added to this solution, mixed well and incubated at room temperature for 10 minutes. PBS was added to stop conjugation (3 volumes) and TNP-SRBC were centrifuged at 450xg for 10 minutes, washed 3 times and resuspended to a concentration of 1X10’cells/ml. Conjugation of TNP to SRBC was verified by passive hemagglutination with MOPC 315 antibodies (anti-TNP IgA, Sigma) as described by Herbert (1978). 2.3.2 Total and antigen specific immunoglobulin quantitation Ig isotype concentrations and antigen-specific 19 were determined by enzyme linked.immuno-sorbent assay (ELISA). For ELISA solid phase, Immunolon 2 (Dynatech) microtiter wells were incubated overnight at 4°C with 50111 of PC-BSA, inulin BSA, TNP-BSA, actin, sphingomyelin, mouse IgG, thyroglobulin, or anti-lg (heavy chain specific antibody) (Sigma Chemical Co.) at a concentration of long/ml in 0.5M carbonate buffer 53 (pH 9.6). Salmon sperm DNA (10ug/ml) was coated in 0.1M phosphate buffered saline (PBS) (pH 7.5) by air drying overnight at 40°C (Pestka and Chu 1984) in an isotemp oven (Fisher Scientific). Collagen was diluted to a concentration of long/ml in 0.1M acetic acid and air dried overnight. Cardiolipin was diluted to a concentration of 50ug/ml in 70% alcohol and air dried overnight. Wells were washed on a MicrowashII automatic plate washer (Skatron, Norway) three times with PBS containing 0.2% Tween 20 (PBS-Tween) and then incubated with 0.3ml 1% (wt/vol) bovine serum albumin in PBS (BSA-PBS) for 30 minutes at 37°C to block nonspecific binding. In the case of MRBC, a monolayer of bromelated MRBC cells was produced in microtiter plates by pipeting 150 ul/well of 1x107 MRBC/ml suspension and centrifugation at 450xg for 5 minutes. PBS was gently aspirated and 200ul/well of 0.15% glutaraldehyde in PBS were added for 5 minutes in order to fix the cells onto the plastic well. Plates were washed twice more in PBS, air dried for 3 hours, blocked with 1% BSA in PBS, washed 3 more times in PBS and stored at 4°C until used. Serum samples were diluted in 1% BSA-PBS and added at 50ul/well to blocked and washed plates. Standard mouse reference serum (Bethyl Laboratories, Montgomery, TX) was added to anti-Ig wells that flanked the antigen coated wells for use in isotype reference curves. Plates were incubated 1 hour at room temperature or overnight at 4°C, washed and incubated for 1 h with anti-mouse Ig (heavy chain specific) alkaline phosphatase conjugate (Sigma Chemical Co.) diluted 54 1:500 in 1% BSA-PBS. Plates were washed six more times and then reacted with 100111 p-nitrophenyl phosphate (Sigma) (4mg/ml) in glycine buffer (pH 10.5) . Absorbance at 405nm was read on the EIA Minireader 2 (Dynatech, Alexandria, VA). Initially, to minimize day-to-day variability in absorbance readings encountered with ELISA, Igs bound to the antigens were assigned gravimetric equivalents (ng/ml) based on comparison of absorbance to values obtained in isotype reference curves run concurrently. Later, during screening of monoclonal IgA, reactivity to antigens was standardized by a reference 100 ng/ml MOPC 315 IgA precoated onto wells. Monoclonal IgAs were diluted in 0.05% Tween in PBS at long/ml. Antigen coated wells were blocked by 30 minute incubation with 0.3 ml/well of 1% BSA-PBS (w/v) at room temperature. Plates were washed and incubated for 1 hour at room temperature with 50111 of monoclonal IgA/well. Plates were washed again and incubated with 50ul/well of anti-mouse IgA alkaline phosphatase conjugate diluted 1:500-1:2000 (depending on the titer of the conjugate) in 1% BSA-PBS for 1 hour at room temperature. Plates were washed and reacted with 100ul/well p-nitrophenyl phosphatase for 1 hour. Positive reactivity to any antigen was obtained when the 0.D. at 405 nm of the tested supernatant was higher than the average blank + 2 fold the standard deviation of the blank. Blank was uncoated wells that were blocked with 1% BSA-PBS, and reacted with the monoclonal IgAs and enzyme conjugate. 0.D. of each monoclonal IgA to each antigen was normalized by dividing with 4.4 55 the 0.D. of MOPC 315 IgA which ranged between approximately 0.5 and 1.5. This method standardized the ELISA to different antigens in such a way that a comparison of the 0.D. could be made as a measure of ability of antibody to bind to that antigen. 2.3.3 Antigen inhibition ELISA Antigen coating and blocking were preformed in the same way as described in the ELISA section. During the blocking stage 50ul of the antibody tested was premixed with SOul of the inhibiting antigen. Following the blocking step the premixed solution was added at 100ul per well and the ELISA proceeded as previously described. Results were read as the % of maximal absorbance that was obtained when the antibody is premixed with an equal solution of PBS. 2.3.4 IgA elution from kidneys IgA was eluted from kidneys by a modification of the procedure by Matsiota et al. (1990). Kidneys from four (4) treatment mice were sectioned onto glass slides. The slides for each mouse were put back to back into a Coplin jar and immersed in glycine buffer with a pH 2.8 for 1 hour while shaking at 80 rpm. The buffer was collected and sections were washed again in glycine buffer (pH 2.8) for another hour and then twice more in glycine buffer with pH 2.2. Buffer was pooled from the four washes and centrifuged at 450 x g for 10 minutes to remove pieces of tissue. The pH of the eluate was adjusted to 7.8 and it was precipitated with an equal volume of ammonium sulfate (pH 7.8). The eluate was centrifuged at 56 10,000 rpm for 20 min. and the precipitated protein was resuspended in PBS with 0.02% azide. Eluent was washed 3 more times by centrifugation on Centriprep 10 concentrators (Amicon, W.R. Grace & Co., Conn. Beverly, MA). The eluent was finally resuspended in 1% BSA-PBS with 0.02% azide and the IgA concentration was measured by ELISA as previously described. IgA concentration in all eluates was adjusted to 10,000 ng/ml and they were screened against 7 antigens. 2.3.5 ELISPOT Immunolon 4 wells (Dynatech, Chantilly, Virginia) were coated with coating buffer containing long/ml PC-BSA or sphingomyelin, 50ug/ml cardiolipin or 100ug/ml TNP-BSA. Plates were incubated overnight at 4°C followed by 5 washes in PBS-Tween (Harlow and Lane, 1988) . They were blocked with 300ul/well 1% BSA-PBS and incubated at room temperature for 30 minutes. Plates were washed 6 times in PBS-Tween and 100ul/well of different concentrations of cell suspension were added. Plates were then incubated at 37°C in a C02 incubator, washed 6 times with PBS-Tween and 6 more times with distilled water (to lyse the cells). Anti-IgA alkaline phosphatase (Sigma) was added at 50ul/well (1:200 dilution in 1%BSA-PBS) and plates were incubated overnight at 4°C. Substrate consisted of 6.7 ml BCIP (5-bromo-4-chloro-3-indolyl phosphate) (0.15 g in 100 ml AMP buffer: 9.62 ml 1M 2-amino-2- methyl-l-propanol + 15 mg magnesium chloride + 10 ul Triton X- 405 + 0.01 g sodium azide + 1 liter distilled water), 100 pl nitro blue tetrazolium (10mg/ml in 70% dimethylformamide) , 1.2 57 ml distilled water and 3 ml boiled 3% agarose (Sigma #A6013) which were mixed together and kept in a water bath at 45%: until ready to use. The plates were washed 6 times with PBS- Tween and 100ul/well tempered substrate were added. Plates were wrapped in aluminum foil and incubated at 37°C for 2 hours. Following the incubation plateswere viewed with an inverted microscope with a piece of Kim-wipes on top (to diffuse the light) and positive spots of antibody producing cells were counted. 2.3.6 Flow cytometry Cell suspensions of 1X106 cells were washed by centrifugation at 450xg for 3 minutes and placed into Eppendorf tubes. Cells were incubated on ice for 20 minutes with 1% BSA-PBS azide (0.02% azide w/v) to block non-specific binding sites. Cells were centrifuged at 450xg for 3 minutes and washed twice in 1%BSA-PBS azide. The first antibody diluted in 1%BSA-PBS azide was added at 100ul/tube (anti kappa-biotin 1:500 dilution, PNA (peanut agglutinin)-biotin 1:33.3 dilution, rat IgG-biotin 1:1000 dilution) and tubes were incubated for 20 minutes on ice. Tubes were centrifuged and the cells were washed 3 times in 100ul PBS azide. The secondary incubation solution diluted in 100ul/tube 1%BSA-PBS azide was added (PE-avidin 1:500 dilution, anti-IgA FITC 1:25 dilution, rat IgG-FITC 1:1000 dilution) and tubes were incubated on ice for another 45-60 minutes. Cells were washed 3 times in 100ul/tube PBS azide. Cells were resuspended in 1ml 1% formaldehyde in PBS and stored at 4°C in the dark until 58 sorted. For cell cycle staining cells were resuspended in 0.2ml PBS + 0.2ml FCS (fetal calf serum), fixed with 1.2ml 70% ice cold alcohol (dropwise while mixing) and held on ice overnight. Following alcohol fixation protein precipitate formed which was washed with 2% FCS-PBS and cells were resuspended in lml PI stain (3.76ml PBS + 0.2ml propidium iodide (1mg/ml) + 0.04ml RNAse A (5mg/ml)) and incubated for 1 hour at room temperature. Samples were kept cold and in darkness until analysis. For cytoplasmic CD5 staining the procedure of Schmid et al. (1991) was used, Cells were washed in PBS and resuspended at 1x105/0.875ml cold PBS. Cold 2% paraformaldehyde was added (0.125ml) and the cells were vortexed immediately. Samples were incubated at 4°C for 1 hour. Cells were centrifuged at 250xg for 5 minutes and supernatant was removed. Pellet was resuspended in 1 ml of 0.2% Tween 20 in PBS and incubated at 37°C for 15 minutes. One mililiter (1 ml) of 2% fetal calf serum (FCS) in PBS was added. Cells were spun down again and resuspended in minimal amount of 2% FCS PBS. .A:rat anti-mouse Fc gamma II receptor antibody solution was added at 5 pl of 1:5 antibody in order to prevent non-specific binding through the Fc gamma receptor. The samples were incubated on ice for 5 minutes. Half ml (0.5 ml) of 0.2% Tween 20 in PBS were added and the cells were centrifuged. The supernatant was removed.and FITC labeled anti-CD5 antibody was added (10 ul of 200 ug/ml). Cells were incubated on ice for 20 minutes and washed 3 times in 0.2% Tween 20 in PBS. Cells were finally 59 resuspended in 1% paraformaldehyde and stored at 4°C and in the dark until analyzed. Cell cycle analysis was performed on FACScan cell sorter (Becton Dickinson) and analyzed by CellFIT cell-cycle analysis program version 2.0 at Sparrow hospital (Lansing, MI). All other analyses were performed using an Ortho Diagnostics Cytofluorograph 50-H Fluorescence-Activated Cell Sorter (FACS)/2150 computer system in MSU Giltner Hall. 2.3.7 Immunofluorescence Immediately following sacrifice of mice, kidneys, liver, spleen and tails were frozen in liquid nitrogen onto a solid support (cork) and stored at -80°C. Tissues were sectioned on a cryostat and sections were stained with diluted fluorescinated antibody in PBS. Peritoneal exudate cells were obtained by washing the peritoneal cavity of Balb/c female mice with PBS, collecting the exudate and incubating the cell suspension on poly-L-lysine coated slides for 10 min. Each section was thawed out and circled with PAP pen (Daido Sangyo Co., Japan). The sections were blocked with 1%BSA-PBS for 15 minutes, the BSA-PBS was shaken off and then 50ul of the fluorescinated antibody was added. Tissue sections were incubated with the fluorescinated antibody for 45 minutes in a dark and moist chamber. Following staining slides were washed briefly 3 times by immersion into a Coplin jar of fresh PBS and then further washed by 3 washes in a Coplin jar with clean PBS for 10 minutes at a time at a speed of 60rpm on a shaker. Excess PBS was shaken off and a drop of FA mounting 60 fluid (pH 9, Difco laboratories, Detroit, MI) was placed on top of each section. Sections were then covered with a cover slide and either looked at immediately or frozen at -20°C until further analysis. 2.3.8 Hybridoma production PP lymphocytes were fused with NSl cells according to the protocol of Galfre and.Milstein (1983) with the modifications by .Abouzied. et al. (1990). iMouse PP and spleen. cells (2.2x107) from two control and vomitoxin fed mice were fused with NS-l myeloma cells (2x107) with PEG as described by Galfre and Milstein (1983). Following fusion the cells were suspended in 20% fetal bovine serum (PBS) in Dulbecco's modified medium (Sigma) with 1% NCTC medium (Sigma), 10 mM sodium pyruvate and pen/ strep solution (Sigma) and distributed into 96 well tissue culture plates. The plates were then incubated at 37°C with humidity and 7-8% CO2 in air. Every 3 days the wells were fed with 20% FBS-DMEM containing hypoxanthine, aminopterin and thymidine (HAT medium). One week after fusion the fusion efficiency was checked. .After 2 weeks wells were fed with 20% FBS-DMEM containing hypoxanthine and thymidine (HT medium) and screening for IgA production began. Supernatants were collected from wells exhibiting growth and tested by ELISA for production of IgA antibodies. All wells that contained cells producing IgA.were transferred into 24 well tissue culture plates and then into 50 and 250 ml tissue culture flasks. Hybridomas were cloned by limiting dilution in 20% macrophage conditioned medium, scaled up to 61 large flasks and then supernatants and cells were frozen for further analysis. 2.3.9 PAGE and Western blotting. Monoclonal IgA supernatants were precipitated 3 times with 50% ammonium sulfate, resuspended in PBS at a concentration of 1 mg/ml IgA and 2511l was loaded onto a precast Mini-PROTEAN II native 4-17% gradient gel (Bio-Rad Laboratories, Richmond, CA) . PAGE and Western blotting were performed using Mini- PROTEAN II electrophoresis cell , Mini Trans-Blot electrophoretic transfer cell and power supply model 200/2.0 (Bio-Rad Laboratories) (Coligan et al. 1992). PAGE was done at 100v for 4 hours and the gel was blotted onto PVDF transfer membranes (Biotechnology Systems, Boston, MA) for 1 hour at 100v in 0.01% SDS in the running buffer. Membranes were blocked with 5% BSA and then detected with 1:1000 anti IgA alkaline phosphatase (Sigma) in 1% BSA. 2.3.10 Reduction and alkylation of monoclonal IgA. Monoclonal IgA supernatants were reduced and alkylated as previously described (Pestka et al. 1989). IgA supernatants were precipitated with ammonium sulfate (12.5111 saturated ammonium sulfate added to 25111 sample in PBS), incubated overnight at 4°C and centrifuged at 1400Xg for 30 minutes. Supernatant was removed and the protein was reduced by adding 201:1 DTT (dithiothreitol) (10mg/ml in 0.275M Tris-HCl buffer). The tubes were flushed with nitrogen and incubated with nitrogen for 1h at 37°C. Alkylation was performed by addition of 5111 iodoacetamide (22mg/ml iodoacetic acid in 0.275M Tris- 62 HCl buffer) flushing again with nitrogen and 15 minutes incubation at 4W3. Staining solution was added and samples were loaded onto gel. Matching non reduced alkylation samples were precipitated with ammonium sulfate and resuspended in 2511.1 PBS. 2.3.11 Labeling monoclonal IgA.with fluorescein. Monoclonal antibody supernatants were brought to 1 mg/ml IgA concentration and then labeled with FITC as described (Coligan et al. 1992). Supernatants were first dialyzed against labeling buffer (0.05M boric acid, 0.2M NaCl, pH 9.2) for 3 days at 4°C with several buffer changes. Optical density at 280nm was read and protein concentration calculated (protein mg/ml = A280 x 0.74 x dilution factor). For each 1 mg of protein 4ul of 5mg/ml FITC in DMSO were added. Conjugation to FITC proceeded at room temperature for 2 hours while gently shaking the solution. Unbound FITC was removed by extensive dialysis against 0.1M Tris buffer (pH 7.4) containing 0.1% vol/vol NaN3 and 0.2M NaCl for 3 days with several buffer changes. FITC/protein ratio was determined by reading optical density at 280nm (for protein) and 492nm (for FITC). A ratio FITC/protein of 5-6 was considered optimal. Fluorescinated antibodies were stored with 0.02% NaN3 in PBS at 4°C in the dark. 2.4 Statistical analysis Differences between vomitoxin-treated and control groups were analyzed by Student's t test using SigmaPlot as well as correlation coefficients (Jandel Sci., Corte Madera, CA). 63 3.0 RESULTS 3.1 Vomitoxin production Vomitoxin was produced on rice cultures with an approximate yield of 2 grams toxin per 5 kg rice (400ppm). This was similar to yields reported by Witt et al. (1985) . Following extraction and purification, the toxin was quantitated by HPLC and TLC (Fig. 3) . During quantitation by HPLC under the conditions described, the vomitoxin peak appeared approximately 3.7 minutes after injection onto the column (Fig. 4) . According to HPLC the vomitoxin concentration in the crude solution was 12.6 mg/ml. By TLC quantitation, using the dual length TLC scanner, the vomitoxin concentration in the crude solution was 13.9 mg/ml. Thus, both methods gave very similar results on vomitoxin amount in the preparation. Following chromatography, vomitoxin fraction did not appear to contain contaminants based on the TLC or HPLC. Since this fraction was relatively'pure and also because of loss of large amounts of toxin during crystallization, it was used directly for feeding studies. 3.2 Effect of dietary vomitoxin on.mouse weight and serum Igs The most commonly known effect of vomitoxin feeding is feed refusal. When B6C3F1 female mice were fed a diet of 25ppm vomitoxin they exhibited feed refusal. Their body weight did not increase with time. After 4 weeks of feeding the difference between treatment and control mice was significant and this trend increased toward the termination of the experiment at 8 weeks (Fig. 5). These results were 64 Fig. 3. Quantitation of vomitoxin amount by TLC and HPLC. .A = standard curve following reading of standard vomitoxin amounts on a dual wavelength TLC scanner. B = standard curve of standard vomitoxin amounts by HPLC. AU = arbitrary units. AU x 1000 AU )1 10,000 65 A 1 2 T O 1 0 .J- / / 8 u / O 5 " O 4 u / O 2 .. o .1 1 #1 1 2 3 5 Vomitoxin (09) e O 120 ‘r / O 80 .1. / O 40 v / /O 0 1 + 4 : 4. 0 200 400 600 800 1000 Vomitoxin (n9) Pig. 3 66 AU A 7 3 1.47 1 26 5' .5 AU 3; a 0 g 1.45 ”'3 1.27 g 2.52 E fine H f Fig. 4. Typical appearance of vomitoxin on HPLC. A - standard 500 ng vomitoxin, B - purified sample of 500 ng vomitoxin. Arrows pointing to vomitoxin elution peak. AU - arbitrary units. 66 AU A T} E; 1.47 1.26 :3 .E E f, 33.72 *— AU t a E 1.45 ' 1.27 a: 2.52 E 73.72 F f Fig. 4. Typical appearance of vomitoxin on HPLC. A - standard 500 ng vomitoxin, B - purified sample of 500 ng vomitoxin. Arrows pointing to vomitoxin elution peak. AU - arbitrary units. 67 35 T T I 1 I 0 control 0 treatment 30 — %’ d ”a? E O L CD \v’ 25 — /////// — E .. .. .9 .. a) E 20 — - 15 l 1 l l l O 2 4 6 8 Weeks feeding Fig. 5. Effect of feeding 25ppm vomitoxin on B6C3F1 mouse weight. Data are means 1 SEM. (P<0.05) from matching control. (P<0.01) from matching control. * a significantly different ** = significantly different 45 through a #4 Whatman filter paper (Whatman, Inc. , Clifton, NJ) and the resulting red filtrate was evaporated on a steam bath until the liquid was 10% of the original volume or less. The solution was saturated with sodium chloride and left overnight to precipitate the proteins. Precipitated proteins were filtered out by filtering on a Buchner funnel through #4 Whatman filter paper. 2.1.2 Toxin extraction and purification The culture filtrate was extracted 3 times with ethyl acetate (250ml ethyl acetate + 250ml culture filtrate) by gentle shaking (to prevent emulsion formation). Ethyl acetate, containing the toxin, was evaporated on a Buchi R110 rotary evaporator (Brinkmann Instruments, Inc., Westbury, NY) at 45°C and the resultant brown concentrate was dried further in the hood until a viscous liquid was formed. At this point vomitoxin was dissolved in ethyl acetate and quantitated using TLC with a known vomitoxin control (purchased from Sigma). The. toxin solution ‘was dissolved in. a small amount of methylene chloride and loaded onto a silica gel column to purify vomitoxin from other contaminants. A.Michael-Miller glass chromatographic column (37 mm i.d. x 300 mm) (Ace Glass, Inc., Vineland, NJ) was packed with silica gel (Adsorbosil, 200/450 mesh, Anspec) and fitted with safety shield, Teflon end fittings, connectors and tubing (2 mm i.d.). A pump (Model RP-SY-lCSC; Fluid Metering, Inc., Oyster Bay, NY) was first primed with methanol. Distilled water was pumped through the column, and the it was left to 46 equilibrate overnight. The column was then washed with methylene chloride and the toxin solution (also dissolved in methylene chloride) was loaded onto the top of the column. Extract movement through the column could be detected by following the movement of brown,pigment~ .As soon as the toxin started coming through the bottom, part of the column, fractions were collected. Fractions were spotted onto a TLC plate together with a standard for 15-acetyl-deoxynivalenol and vomitoxin to test for the presence of these compounds. When 15-Acetyl-deoxynivalenol stopped coming' through the column, the pump was washed with methanol and then with distilled water and the distilled water was pumped through the column. Fractions were not collected until the water passed through the column. Every third water fraction was tested for presence of vomitoxin on TLC and fractions containing vomitoxin were collected.‘Vomitoxin containing fractions were pooled together and further extracted. Water fractions containing vomitoxin were extracted with ethyl acetate (1:1 volume of each) 10 times. Water was tested for presence of vomitoxin.and.in.the case of remnant.vomitoxin was further extracted. Ethyl acetate fractions containing vomitoxin were extracted on a rotary evaporator at 45°C until only a small volume remained. The solution was filtered through a #4 Whatman filter paper and sodium sulfate (to adsorb remaining water molecules) and further evaporated in a water bath with nitrogen or directly in the hood (in the case of a small volume). For crystallization, the solution was 47 transferred into a small beaker and seeded with vomitoxin crystals and kept refrigerated for several weeks until crystals formed. The pigment was washed off the crystals with ethyl acetate (and sometimes methanol) until the vomitoxin was white colored powder or crystals. .At this time a small amount of vomitoxin was dissolved and quantitated by TLC and HPLC to ensure the purity and the amount of vomitoxin. Alternatively, the solution was quantitated before crystallization took place and used directly as a source of vomitoxin. 2.2.3 TLC (Thin Layer Chromatography) TLC chamber (inside dimension 27x7x26cm) was filled with a 1:3 v/v toluene ethyl acetate solution (25ml toluene and 75ml ethyl acetate) and equilibrated overnight. TLC plate (precoated 20x10cm silica gel G plate, Redi-Plates, Fisher Scientific1Co., Fair Lawn, NJ) was spotted with 0.25ul, 0.5ul, 1111, 2111, 5111 and 10111 of a standard 1mg/ml solution of vomitoxin in methanol (purchased from Sigma) together with different amounts of sample. The TLC plate was put in the chamber until solvent rose almost to the top. The plate was taken out, dried at room temperature and sprayed with 15% aluminum chloride solution (159 aluminum chloride AlCl3.6H20 + 85ml ethanol + 15ml distilled.water). Plate‘was dried at room temperature and put into 110°C for 5 minutes. Plate was quantitated by densitometry on a dual wavelength TLC scanner (Shimadzu, Japan) and the amount of vomitoxin was calculated according to the standard curve of the Sigma standard. 48 2.2.4 HPLC (High Pressure Liquid Chromatography) Vomitoxin samples were run at 20% methanol in water (previously filtered to prevent gas bubbles) (HPLC grade, Fisher Scientific Co.) on a Model 2300 HPLC pump and a V‘ variable wavelength absorbance detector (5-mm flow cell) (Isco, Inc., Lincoln, NE). The system had RP-18 Spheri-lo MPLC analytical (22 cm x 4.6 mm i.d.) and guard (3 cm x 4.6 mm i.d.) cartridges (Brownlee Labs, Inc., Santa Clara, CA). Column was flushed with methanol for 15 minutes and then 20% methanol was pumped through until the baseline stabilized. Detection was performed at 224nm. At a pump rate of lml/min the vomitoxin peak appeared after 4 minutes. At that time standard amounts of vomitoxin (purchased from Sigma) were injected into the column through a Valco C6U sample injector (Valco Instruments Co., Inc. Houston, TX) with a 10ul sample loop and the peak area was measured. Amount of vomitoxin in the sample was calculated by comparison with peak areas of standards. 2.2.5 Preparation of diet AIN-76A semi-purified diet (ICN Nutritional Biochemical, Cleveland, OH) was purchased commercially or prepared by mixing 2kg casein, 2.5kg sucrose, 5009 alphacel, 1.5kg cornstarch, 309 methionine, 209 bitartrate, 3509 mineral mix, 1009 vitamin mix and 5009 corn oil. Mixed diet was stored in plastic bags at 4°C and used within 2 months. Vomitoxin was dissolved in ethyl alcohol (200 proof) at 200mg vomitoxin per 300ml ethyl alcohol. Vomitoxin.was mixed 49 in a large beaker with 2009 AIN-76A diet to make up a stock 1000ppm vomitoxin food. The alcohol was slowly evaporated while mixing carefully to ensure equal distribution of the toxin throughout the food. When the stock 1000ppm food was completely dry, it was used to prepare the desired concentration of toxic food by mixing with clean AIN-76A diet in a KitchenAid mixer (model K5—A, Hobart Mfg. Co., Troy, Ohio). To verify toxic diet concentration, a 59 food sample was extracted into 25ml acetonitrile-water (9:1) solution by shaking for 60 minutes. The extract was centrifuged at 450xg for 8 minutes and the supernatant containing the toxin was removed. Five ml of the supernatant was passed through a #215 charcoal column (Romer Labs, Washington, MD) by using a water pump and collected into a glass tube. The extract was evaporated on an analytical evaporator (Organomation Associates, South Berlin, MA) with nitrogen pressure and redissolved in 1 ml acetone-methanol (2:1) for TLC or 20% methanol-water for HPLC. TLC plates were run in toluene- acetone solution ( 1:1) as described above. HPLC was run in 20% methanol-water as described above. 2.2.6 Safety Face masks and vinyl gloves were used when handling the toxin. Concentrated toxin was handled in a fume hood. Contaminated glassware and equipment were detoxified by soaking in 10% bleach solution (Thompson & Wannemacher, 1984) . 50 2.2.7 Animals B6C3F1 or' Balb/c female :mice (8-10 ‘weeks old, Harlan/Sprague-Dawley, Indianapolis, IN) were randomized, housed in protected-environment cages (Nalgene, Rochester, NY) and fed powdered AIN-76A semi-purified diet (ICN Nutritional Biochemical, Cleveland, OH). Distilled.water was provided :9 up and feed was changed every 3-4 days. The mice were acclimatized to housing, feed, a 12-hr light/dark cycle and to a negative-pressure ventilation area for at least 7 days prior to initiating experiments. Mice were fed control diet or diet containing vomitoxin. At intervals, mice were anaesthetized with ether and bled at the retro-orbital plexus. Serum was analyzed by ELISA and mice were sacrificed at selected time points for cell preparation. If necessary, other tissues (such as kidneys, livers etc.) were removed at the same time and stored frozen at -80°C until further analysis. 2.2.8 Cell preparation PP were removed, teased apart in RPMI-1640 medium (Sigma) , passed through a sterile 85-mesh stainless steel screen and resuspended in the same medium. Cells were centrifuged at 450xg for 10 minutes, resuspended in RPMI-1640 medium and counted. Spleens were removed, teased apart with tissue forceps in RPMI-1640 medium and left on ice for 10 minutes for tissue particles to settle. Supernatants were removed and centrifuged at 450xg for 10 minutes. Erythrocytes were lysed by resuspending in 0.83% ammonium chloride for 3 minutes at 51 room temperature. Cells were centrifuged, resuspended in RPMI-1640 medium and counted. 2.3 Immunological methods 2.3.1 Antigen preparation Phosphorylcholine bovine serum albumin conjugates (PC-BSA) and inulin-BSA, prepared as described by Gearhart and Cebra (1979) , were kindly provided by J. Cebra (University of Pennsylvania, Philadelphia, PA). Salmon sperm DNA, actin, sphingomyelin, collagen type I, mouse IgG, thyroglobulin, laminin and cardiolipin were purchased from Sigma. Salmon sperm DNA was extracted with phenol/chloroform (Ausubel et al., 1990). Mouse red blood cells (MRBC) were obtained by orbital bleeding of B6C3F1 female mice and bromelated as described by Klinman and Steinberg ( 1987) . Briefly, MRBC were centrifuged and diluted again to 5% in 0.1% bromelain in saline solution. MRBC were incubated at 37°C for 30 minutes followed by 3 washes in PBS by centrifugation at 450xg for 10 minutes and diluted to a concentration of 107 cells/ml. Bovine serum albumin (BSA) was trinitrophenylated (TNP- BSA) using picric acid (Good at al. 1980) . One hundred (100) mg of BSA was mixed with Sml distilled water and 100 mg potassium carbonate and then 100 mg picrylsulfonic acid (purchased from Sigma) were added. The mixture was covered with aluminum foil and stirred at room temperature overnight. The conjugate was dialyzed in the dark at 4°C for 3 days against borate buffer (0.2M, pH 9.2) with several buffer 52 changes. Following dialysis the amount of BSA as well as the ratio between moles BSA and moles TNP were calculated by spectrophotometric reading at 278nm for protein and 348nm for TNP (extinction coefficient = 15,400). It was calculated that the BSA was completely conjugated with 53 TNP molecules/one BSA molecule. Conjugation was also verified by ELISA using MOPC 315 anti-TNP antibodies. Sheep red blood cells were trinitrophenylated (TNP-SRBC) using picrylsulfonic acid (Sigma) as described by Rittenberg and Pratt (1969). Picrilsulfonic acid (Sigma) (1 ml) was diluted in 7 ml cacodylate buffer (0.28M, pH 6.9). Packed SRBC (1 ml) were added to this solution, mixed well and incubated at room temperature for 10 minutes. PBS was added to stop conjugation (3 volumes) and TNP-SRBC were centrifuged at 450xg for 10 minutes, washed 3 times and resuspended to a concentration of 1X109cells/ml. Conjugation of TNP to SRBC was verified by passive hemagglutination with MOPC 315 antibodies (anti-TNP IgA, Sigma) as described by Herbert (1978). 2.3.2 Total and antigen specific immunoglobulin quantitation Ig isotype concentrations and antigen-specific 19 were determined by enzyme linked immuno-sorbent assay (ELISA). For ELISA solid phase, Immunolon 2 (Dynatech) microtiter wells were incubated overnight at 4°C with 50111 of PC-BSA, inulin BSA, TNP-BSA, actin, sphingomyelin, mouse IgG, thyroglobulin, or anti-Ig (heavy chain specific antibody) (sigma Chemical Co.) at a concentration of long/m1 in 0.5M carbonate buffer 53 (pH 9.6). Salmon sperm DNA (long/ml) was coated in 0.1M phosphate buffered saline (PBS) (pH 7.5) by air drying overnight at 40°C (Pestka and Chu 1984) in an isotemp oven (Fisher Scientific). Collagen was diluted to a concentration of 10119/ml in 0.1M acetic acid and air dried overnight. Cardiolipin was diluted to a concentration of 50ug/ml in 70% alcohol and air dried overnight. Wells were washed on a MicrowashII automatic plate washer (Skatron, Norway) three times with PBS containing 0.2% Tween 20 (PBS-Tween) and then incubated with 0.3ml 1% (wt/vol) bovine serum albumin in PBS (BSA-PBS) for 30 minutes at 37°C to block nonspecific binding. In the case of MRBC, a monolayer of bromelated MRBC cells was produced in microtiter plates by pipeting 150 111/well of 1x107 MRBC/ml suspension and centrifugation at 450xg for 5 minutes. PBS was gently aspirated and 200111/well of 0.15% glutaraldehyde in PBS were added for 5 minutes in order to fix the cells onto the plastic well. Plates were washed twice more in PBS, air dried for 3 hours, blocked with 1% BSA in PBS, washed 3 more times in PBS and stored at 4°C until used. Serum samples were diluted in 1% BSA-PBS and added at 50111/well to blocked and washed plates. Standard mouse reference serum (Bethyl Laboratories, Montgomery, TX) was added to anti-Ig wells that flanked the antigen coated wells for use in isotype reference curves. Plates were incubated 1 hour at room temperature or overnight at 4°C, washed and incubated for 1 h with anti-mouse Ig (heavy chain specific) alkaline phosphatase conjugate (Sigma Chemical Co.) diluted 54 1:500 in 1% BSA-PBS. Plates were washed six more times and then reacted with 100111 p-nitrophenyl phosphate (Sigma) (4mg/ml) in glycine buffer (pH 10.5) . Absorbance at 405nm was read on the EIA Minireader 2 (Dynatech, Alexandria, VA). Initially, to minimize day-to-day variability in absorbance readings encountered with ELISA, Igs bound to the antigens were assigned gravimetric equivalents (ng/ml) based on comparison of absorbance to values obtained in isotype reference curves run concurrently. Later, during screening of monoclonal IgA, reactivity to antigens was standardized by a reference 100 ng/ml MOPC 315 IgA precoated onto wells. Monoclonal IgAs were diluted in 0.05% Tween in PBS at 10119/ml. Antigen coated wells were blocked by 30 minute incubation with 0.3 ml/well of 1% BSA-PBS (w/v) at room temperature. Plates were washed and incubated for 1 hour at room temperature with 50111 of monoclonal IgA/well. Plates were washed again and incubated with 50ul/well of anti-mouse IgA alkaline phosphatase conjugate diluted 1:500-1:2000 (depending on the titer of the conjugate) in 1% BSA-PBS for 1 hour at room temperature. Plates were washed and reacted with 100111/well p-nitrophenyl phosphatase for 1 hour. Positive reactivity to any antigen was obtained when the 0.D. at 405 nm of the tested supernatant was higher than the average blank + 2 fold the standard deviation of the blank. Blank was uncoated wells that were blocked with 1% BSA-PBS, and reacted with the monoclonal IgAs and enzyme conjugate. 0.D. of each monoclonal IgA to each antigen was normalized by dividing with 55 the 0.D. of MOPC 315 IgA which ranged between approximately 0.5 and 1.5. This method standardized the ELISA to different antigens in such a way that a comparison of the 0.D. could be made as a measure of ability of antibody to bind to that antigen. 2.3.3 Antigen inhibition ELISA Antigen coating and blocking were preformed in the same way as described in the ELISA section. During the blocking stage 50ul of the antibody tested was premixed with 50u1 of the inhibiting antigen. Following the blocking step the premixed solution was added at 100ul per well and the ELISA proceeded as previously described. Results were read as the % of maximal absorbance that was obtained when the antibody is premixed with an equal solution of PBS. 2.3.4 IgA elution from kidneys IgA was eluted from kidneys by a modification of the procedure by Matsiota et al. (1990). Kidneys from four (4) treatment mice were sectioned onto glass slides. The slides for each mouse were put back to back into a Coplin jar and immersed in glycine buffer with a pH 2.8 for 1 hour while shaking at 80 rpm. The buffer was collected and sections were washed again in glycine buffer (pH 2.8) for another hour and then twice more in glycine buffer with pH 2.2. Buffer was pooled from the four washes and centrifuged at 450 x g for 10 minutes to remove pieces of tissue. The pH of the eluate was adjusted to 7.8 and it was precipitated with an equal volume of ammonium sulfate (pH 7.8). The eluate was centrifuged at 56 10,000 rpm for 20 min. and the precipitated protein was resuspended in PBS with 0.02% azide. Eluent was washed 3 more times by centrifugation on Centriprep 10 concentrators (Amicon, W.R. Grace 8 Co., Conn. Beverly, MA). The eluent was finally resuspended in 1% BSA-PBS with 0.02% azide and the IgA concentration was measured by ELISA as previously described. IgA concentration in all eluates was adjusted to 10,000 ng/ml and they were screened against 7 antigens. 2.3.5 ELISPOT Immunolon 4 wells (Dynatech, Chantilly, Virginia) were coated with coating buffer containing 10119/ml PC-BSA or sphingomyelin, 50ug/ml cardiolipin or 100ug/ml TNP-BSA. Plates were incubated overnight at 4°C followed by 5 washes in PBS-Tween (Harlow and Lane, 1988) . They were blocked with 300111/well 1% BSA-PBS and incubated at room temperature for 30 minutes. Plates were washed 6 times in PBS-Tween and 100111/well of different concentrations of cell suspension were added. Plates were then incubated at 37°C in a C02 incubator, washed 6 times with PBS-Tween and 6 more times with distilled water (to lyse the cells). Anti-IgA alkaline phosphatase (Sigma) was added at 50ul/well (1:200 dilution in 1%BSA-PBS) and plates were incubated overnight at 4°C. Substrate consisted of 6.7 m1 BCIP (5-bromo-4-chloro-3-indolyl phosphate) (0.15 g in 100 ml AMP buffer: 9.62 ml 1M 2-amino-2- methyl-l-propanol + 15 mg magnesium chloride + 10 111 Triton X- 405 + 0.01 g sodium azide + 1 liter distilled water), 100 ul nitro blue tetrazolium (10mg/ml in 70% dimethylformamide) , 1.2 57 ml distilled water and 3 ml boiled 3% agarose (Sigma #A6013) which were mixed together and kept in a water bath at 45%: until ready to use. The plates were washed 6 times with PBS- Tween and 100ul/well tempered substrate were added. Plates were wrapped in aluminum foil and incubated at 37°C for 2 hours. Following the incubation plates were viewed with an inverted microscope with a piece of Kim-wipes on top (to diffuse the light) and positive spots of antibody producing cells were counted. 2.3.6 Flow cytometry Cell suspensions of 1X106 cells were washed by centrifugation at 450xg for 3 minutes and placed into Eppendorf tubes. Cells were incubated on ice for 20 minutes with 1% BSA-PBS azide (0.02% azide w/v) to block non-specific binding sites. Cells were centrifuged at 450xg for 3 minutes and washed twice in 1%BSA-PBS azide. The first antibody diluted in 1%BSA-PBS azide was added at 100111/tube (anti kappa-biotin 1:500 dilution, PNA (peanut agglutinin)-biotin 1:33.3 dilution, rat IgG-biotin 1:1000 dilution) and tubes were incubated for 20 minutes on ice. Tubes were centrifuged and the cells were washed 3 times in 100u1 PBS azide. The secondary incubation solution diluted in 100ul/tube 1%BSA-PBS azide was added (PE-avidin 1:500 dilution, anti-IgA FITC 1:25 dilution, rat IgG-FITC 1:1000 dilution) and tubes were incubated on ice for another 45-60 minutes. Cells were washed 3 times in 100ul/tube PBS azide. Cells were resuspended in lml 1% formaldehyde in PBS and stored at 4°C in the dark until 58 sorted. For cell cycle staining cells were resuspended in 0.2m1 PBS + 0.2ml FCS (fetal calf serum), fixed with 1.2ml 70% ice cold alcohol (dropwise while mixing) and held on ice overnight. Following alcohol fixation protein precipitate formed which was washed with 2% FCS-PBS and cells were resuspended in lml PI stain (3.76ml PBS + 0.2ml propidium iodide (1mg/ml) + 0.04m1 RNAse A (5mg/ml)) and incubated for 1 hour at room temperature. Samples were kept cold and in darkness until analysis. For cytoplasmic CD5 staining the procedure of Schmid et al. (1991) was used" Cells were washed in PBS and resuspended at 1x105/0.875ml cold PBS. Cold 2% paraformaldehyde was added (0.125ml) and the cells were vortexed immediately. Samples were incubated at 4°C for 1 hour. Cells were centrifuged at 250xg for 5 minutes and supernatant was removed. Pellet was resuspended in 1 ml of 0.2% Tween 20 in PBS and incubated at 37°C for 15 minutes. One mililiter (1 ml) of 2% fetal calf serum (FCS) in PBS was added. Cells were spun down again and resuspended in minimal amount of 2% FCS PBS. A rat anti-mouse Fc gamma II receptor antibody solution was added at 5 pl of 1:5 antibody in order to prevent non-specific binding through the Fc gamma receptor. The samples were incubated on ice for 5 minutes. Half ml (0.5 ml) of 0.2% Tween 20 in PBS were added and the cells were centrifuged. The supernatant was removed.and.FITC labeled anti-CD5 antibody was added (10 pl of 200 pg/ml). Cells were incubated on ice for 20 minutes and washed 3 times in 0.2% Tween 20 in PBS. Cells were finally 59 resuspended in 1% paraformaldehyde and stored at 4°C and in the dark until analyzed. Cell cycle analysis was performed on FACScan cell sorter (Becton Dickinson) and analyzed by CellFIT cell-cycle analysis program version 2.0 at Sparrow hospital (Lansing, MI). All other analyses were performed using an Ortho Diagnostics Cytofluorograph 50-H Fluorescence-Activated Cell Sorter (FACS)/2150 computer system in MSU Giltner Hall. 2.3.7 Immunofluorescence Immediately following sacrifice of mice, kidneys, liver, spleen and tails were frozen in liquid nitrogen onto a solid support (cork) and stored at -80%L. Tissues were sectioned on a cryostat and sections were stained with diluted fluorescinated antibody in PBS. Peritoneal exudate cells were obtained by washing the peritoneal cavity of Balb/c female mice with PBS, collecting the exudate and incubating the cell suspension on poly-L-lysine coated slides for 10 min. Each section was thawed out and circled with PAP pen (Daido Sangyo Co., Japan). The sections were blocked with 1%BSA-PBS for 15 minutes, the BSA-PBS was shaken off and then 50pl of the fluorescinated antibody was added. Tissue sections were incubated with the fluorescinated antibody for 45 minutes in a dark and moist chamber. Following staining slides were washed briefly 3 times by immersion into a Coplin jar of fresh PBS and then further washed by 3 washes in a Coplin jar with clean PBS for 10 minutes at a time at a speed of 60rpm on a shaker. Excess PBS was shaken off and a drop of FA mounting 60 fluid (pH 9, Difco laboratories, Detroit, MI) was placed on top of each section. Sections were then covered with a cover slide and either looked at immediately or frozen at -20°C until further analysis. 2.3.8 Eybridoma production PP lymphocytes were fused with NSl cells according to the protocol of Galfre and.Milstein (1983) with the modifications by Abouzied et al. (1990). Mouse PP and spleen cells (2.2xldU from two control and vomitoxin fed mice were fused with NS-l myeloma cells (2x107) with PEG as described by Galfre and Milstein (1983). Following fusion the cells were suspended in 20% fetal bovine serum (FBS) in Dulbecco's modified medium (Sigma) with 1% NCTC medium (Sigma), 10 mM sodium pyruvate and pen/ strep solution (Sigma) and distributed into 96 well tissue culture plates. The plates were then incubated at 37°C with humidity and 7-8% C02 in air. Every 3 days the wells were fed with 20% FBS-DMEM containing hypoxanthine, aminopterin and thymidine (HAT medium). One week after fusion the fusion efficiency was checked. After 2 weeks wells were fed with 20% FBS-DMEM containing hypoxanthine and thymidine (HT medium) and screening for IgA production began. Supernatants were collected from wells exhibiting growth and tested by ELISA for production of IgA antibodies. All wells that contained cells producing IgA.were transferred into 24 well tissue culture plates and.then into 50 and 250 m1 tissue culture flasks. Hybridomas were cloned by limiting dilution in 20% macrophage conditioned medium, scaled up to 61 large flasks and then supernatants and cells were frozen for further analysis. 2.3.9 PAGE and Western blotting. Monoclonal IgA supernatants were precipitated 3 times with 50% ammonium sulfate, resuspended in PBS at a concentration of 1 mg/ml IgA and 25pl was loaded onto a precast Mini-PROTEAN II native 4-17% gradient gel (Bio-Rad Laboratories, Richmond, CA) . PAGE and Western blotting were performed using Mini- PROTEAN II electrophoresis cell , Mini Trans-Blot electrophoretic transfer cell and power supply model 200/2.0 (Bio-Rad Laboratories) (Coligan et al. 1992). PAGE was done at 100v for 4 hours and the gel was blotted onto PVDF transfer membranes (Biotechnology Systems, Boston, MA) for 1 hour at 100v in 0.01% SDS in the running buffer. Membranes were blocked with 5% BSA and then detected with 1:1000 anti IgA alkaline phosphatase (Sigma) in 1% BSA. 2.3.10 Reduction and alkylation of monoclonal IgA. Monoclonal IgA supernatants were reduced and alkylated as previously described (Pestka et al. 1989). IgA supernatants were precipitated with ammonium sulfate (12.5p1 saturated ammonium sulfate added to 25pl sample in PBS), incubated overnight at 4°C and centrifuged at 1400Xg for 30 minutes. Supernatant was removed and the protein was reduced by adding 20pl DTT (dithiothreitol) (10mg/ml in 0.275M Tris-HCl buffer). The tubes were flushed with nitrogen and incubated with nitrogen for 1h at 37°C. Alkylation was performed by addition of 5pl iodoacetamide (22mg/ml iodoacetic acid in 0.275M Tris- 62 HCl buffer) flushing again with nitrogen and 15 minutes incubation at 4%:. Staining solution was added and samples were loaded onto gel. Matching non reduced alkylation samples were precipitated with ammonium sulfate and resuspended in 25p1 pas. 2.3.11 Labeling monoclonal IgA with fluorescein. Monoclonal antibody supernatants were brought to 1 mg/ml IgA concentration and then labeled with FITC as described (Coligan et al. 1992) . Supernatants were first dialyzed against labeling buffer (0.05M boric acid, 0.2M NaCl, pH 9.2) for 3 days at 4°C with several buffer changes. Optical density at 280nm was read and protein concentration calculated (protein mg/ml = A280 x 0.74 x dilution factor). For each 1 mg of protein 4pl of 5mg/ml FITC in DMSO ‘were added. Conjugation to FITC proceeded at room temperature for 2 hours while gently shaking the solution. Unbound FITC was removed by extensive dialysis against 0.1M Tris buffer (pH 7.4) containing 0.1% vol/vol NaN3 and 0.2M NaCl for 3 days with several buffer changes. FITC/protein ratio was determined by reading optical density at 280nm (for protein) and 492nm (for FITC). A ratio FITC/protein of 5-6 was considered optimal. Fluorescinated antibodies were stored with 0.02% NaN3 in PBS at 4°C in the dark. 2.4 Statistical analysis Differences between vomitoxin-treated and control groups were analyzed by Student's t test using SigmaPlot as well as correlation coefficients (Jandel Sci., Corte Madera, CA). 63 3.0 RESULTS 3.1 Vomitoxin production Vomitoxin was produced on rice cultures with an approximate yield of 2 grams toxin per 5 kg rice (400ppm) . This was similar to yields reported by Witt et al. (1985) . Following extraction and purification, the toxin was quantitated by HPLC and TLC (Fig. 3). During quantitation by HPLC under the conditions described, the vomitoxin peak appeared approximately 3.7 minutes after injection onto the column (Fig. 4) . According to HPLC the vomitoxin concentration in the crude solution was 12.6 mg/ml. By TLC quantitation, using the dual length TLC scanner, the vomitoxin concentration in the crude solution was 13.9 mg/ml. Thus, both methods gave very similar results on vomitoxin amount in the preparation. Following chromatography, vomitoxin fraction did not appear to contain contaminants based on the TLC or HPLC. Since this fraction was relatively pure and also because of loss of large amounts of toxin during crystallization, it was used directly for feeding studies. 3.2 Effect of dietary vomitoxin on mouse weight and serum Igs The most commonly known effect of vomitoxin feeding is feed refusal. When B6C3F1 female mice were fed a diet of 25ppm vomitoxin they exhibited feed refusal. Their body weight did not increase with time. After 4 weeks of feeding the difference between treatment and control mice was significant and this trend increased toward the termination of the experiment at 8 weeks (Fig. 5). These results were 64 Fig. 3. Quantitation of vomitoxin amount by TLC and HPLC. .A = standard curve following reading of standard vomitoxin amounts on a dual wavelength TLC scanner. B = standard curve of standard vomitoxin amounts by HPLC. AU = arbitrary units. AU x 1000 AU )1 10,000 121 104 120-- 80~ 404 r V 65 l l l I I U 2 3 4 Vomitoxin (L19) 4 _L I I 200 400 600 800 1dbo Vomitoxin (n9) Pig. 3 66 AU A 77 8 1.47 1.26 5' .E E fr 33.72 +— AU 1 a E 1.45 ' 1.27 g 2.52 E 73.72 F f Fig. 4. Typical appearance of vomitoxin on HPLC. A - standard 500 ng vomitoxin, B - purified sample of 500 ng vomitoxin. Arrows pointing to vomitoxin elution peak. AU - arbitrary units. 67 35 j T l 1 I 0 control 0 treatment 4 _ A U) E O L. CD \«I 25 — /////// — z ,. .. g M Q) g 20 — - 15 l l l l l O 2 4 6 8 Weeks feeding Pig. 5. Effect of feeding 25ppm vomitoxin on B6C3F1 mouse weight. Data are means 1 SEM. (P<0.05) from matching control. (P<0.01) from matching control. * - significantly different ** = significantly different 68 similar to those found by Forsell et al. (1986). Analysis of serum immunoglobulins at 4 and 8 weeks exhibited serum IgA dysregulation that was noted in previous studies. There was a significant increase in total serum IgA coupled to a decrease in total IgG and IgM both after 4 and 8 weeks vomitoxin feeding although IgG levels recovered after 8 weeks (Fig. 6). Total serum IgA was increased 2 and 13 fold after 4 and 8 weeks (Fig. 6) of feeding 25ppm vomitoxin, respectively. 3.3 Effect of dietary vomitoxin on TNP-specific serum immunoglobulin response to oral TNP-SRBC challenge Further efforts to understand vomitoxin dysregulation of serum IgA were directed at studying the effect of vomitoxin feeding on oral immunization with a particulate antigen. TNP- SRBC was chosen since it is a model antigen commonly used in immunological studies to gauge the immune response. Another reason was the ease of using TNP in detecting TNP-specific responses by ELISA. TNP is easily conjugated to proteins which allows rapid preparation of the oral immunogen TNP-SRBC as well as the TNP-BSA conjugate used to coat plates. Oral administration of TNP-SRBC was used to investigate whether vomitoxin had a specific adjuvant effect in enhancing serum IgA response to an exogenous antigen bolus. Repeated gavaging of control mice with TNP-SRBC caused a large increase in TNP- specific IgG with a slight IgM response, whereas an IgA response to TNP was not detectable (data not shown). When 69 Fig. 6. Effect of feeding 25ppm vomitoxin for 4 and 8 weeks on total serum immunoglobulins in B6C3F1 mice. Data are means 1 SEM (n=8) and are representative of 3 experiments. * = significantly different (P<0.05) from matching controls. serum l9 (ug/ml) 1 1 A1 E1 \ 01 31 serum lg 70 JOOOT 4 wk vomitoxin 25004 I: control -treatment 2000-— 1500-» * 1000.2 * ‘Isoo-w H t o e e Fl... IgA 190 W t 80007 8 week vomitoxin 60004.. [Zlcontrol 4000-- treatment ZOOOT ooooJ- 8000-~ \ 6000“ § 4000-- § 2000* § [I] Q ["1 7 PL: 19A 196 lgM Fig. 6 71 mice were fed 25ppm vomitoxin for 8 weeks prior and during TNP- SRBC challenge, the TNP-specific IgG response was almost completely abrogated (Fig. 7). Toxin exposure did not enhance the TNP-specific IgA or IgM response. 3.4 Effect of dietary vomitoxin on phenotype of PP and spleen B cells Failing to detect differences in IgA reactivity to TNP between control and treatment in deliberately immunized female B6C3F1 mice, attentionnwas directed at the effect of vomitoxin on B cells. In an effort to characterize the B cell population that is activated as a result of vomitoxin feeding, cell surface markers and cell cycle analysis were undertaken. Vomitoxin feeding tended to increase the percentage of peanut agglutinin (PNA) binding IgA+ cells and the percentage of kappa (light chain) high IgA+ cells in the PP and spleen (Table 2). Ability to bind PNA is indicative of germinal center B cells, while density of kappa light chain (K) on the cell surface increases during the maturation and differentiation of B cells (Fyfe et al., 1987). In this respect, an increase in IgA+/PNA+ cells in PP as a result of vomitoxin feeding might indicate an activation process of germinal center IgA+ B cells in the PP of vomitoxin-fed mice. The increase in IgA+I("i'h B cells in the spleen suggests an increased population of fully differentiated IgA+ B cells that secrete IgA into the systemic compartment. This notion is reinforced by the significant 72 2000-— g E __ \ 1 500 cc" "‘ 1:] control 3’ treatment I; 1000 4 '3 1. D. (D a. SOD-- z '— F1 :1: MA IgG lgM Pig. 7. Effect of vomitoxin on 19 response in the B6C3F1 mouse to oral immunization with TNP-SRBC. Mice were fed 25ppm vomitoxin for 6 weeks and 18.75ppm vomitoxin for 2 more weeks prior to immunization and then gavaged for 4 consecutive days with 4x109 TNP-SRBC. Data are means :1: SEM (n=8) . * - significantly different (P<0.05) from matching controls. 73 Table 2. Percentages of different cell populations in B6C3F1 mice fed 25ppm vomitoxin for 8 weeks‘ 9mg Trial 319A 316*“ %PNA[IgA_ gxmzlgg PP 1 control 4.6 39.4 2.9 14.7 treatment 5.8* 36.5 4.9 16.1 2 control 2.0 40.2 7.1 30.6 treatment 3.0 32.8* 12.0* 39.2 Spleen 1 control 4.0 39.0 1.3 5.2 treatment 3.3 30.3* 1.3 5.4 2 control 3.6 30.9 5.3 7.3 treatment 3.3 22.3 7.3 11.1* Bone 1 control 2.6 20.4 4.2 6.4 Marrow treatment 3.3* 26.6 1.5* 8.7 2 control 2.7 9.4 8.9 11.8 treatment 2.6 10.6 6.7 10.8 ‘ Data were collected on two trials. Data are average of 8 mice. * = significantly different (P<0.05) from matching controls. Table 3. Supernatant IgA in B6C3F1 mice fed 25ppm vomitoxin for 8 weeksfi green _ti__T 'al Treatment Wei C tur m PP 1 control 413.3 1 62.5 treatment 3900.2 1 2079.2 2 control 1396.1 1 496.8 treatment 99319.0 1 76331.1 Spleen 1 control 94.5 i 11.8 treatment 396.9 f 66.7* 2 control 250.7 1 27.9 treatment 1092.4 1 154.2* Bone 1 control 504.3 1 132.6 Marrow treatment 994.2 f 467.9 2 control 1126.6 1 550.4 treatment 3371.3 1 2552.5 ' Data are average of 8 mice 1 SEM. * = different (P<0.05) from matching controls. significantly 74 correlation between serum IgA levels with spleen supernatant IgA levels (Fig. 8) and a similar trend between serum IgA and PP supernatant IgA (Fig. 9). No significant correlation was found between serum IgA and bone marrow supernatant IgA. The amount of IgA secretion (Table 3) as well as the percentage of IgA? cells in the G2+M cell cycle in the spleen (Table 4) were also increased by vomitoxin feeding. Table 4. Percentages of IgA+ cells in each cell cycle stage in mice fed 25ppm vomitoxin for 8 weeksh sateen Erasmus LLSIGO AL 352.211 Spleen control 90.1 5.5 4.4 treatment 90.1 3.6 6.2* PP control 55.4 19.5 25.2 treatment 59.7 14.0 26.3 ‘ Data are means of 8 mice. * = significantly different (P<0.05) from matching controls. These data support the possibility that fully differentiated IgA+ cells in the spleen are increased as a result of vomitoxin feeding. The effect of vomitoxin treatment on apoptosis (programmed cell death) in the IgA+ population was subsequently assessed by reanalyzing cell cycle data based on a protocol developed by W. Telford in Dr. P. Fraker's laboratory (MSU department of Biochemistry). Treatment mice appeared to show a reduced apoptotic peak compared to controls (Fig. 10, 11) . In the first trial there appeared to be an inverse correlation between serum IgA levels and apoptosis percentage in the 75 0 control r=0.45 30 + . treatment r=0.8 . 25 —- E 20 4 DD 5 E4 15 4 E a m 10 ~— 5 __ 0 r t’ i i i 0.0 0.5 1.0 1.5 2.0 Spleen supernatant IgA (ug/ml) Fig. 8. Correlation between serum IgA and spleen supernatant IgA levels in 8 weeks 25ppm vomitoxin fed and control B6C3F1 mice. Spleen cells were culture in RPMI-1640 media for'7 days without mitogens. Supernatants were collected and analyzed for total IgA” Each data point is representative of one mouse (n=16). Control r=0.447 (no significant correlation), treatment r=0.802 (correlation significant at P<0.01). 76 O control . treatment 0 30 —~ 25 —- 0 Serum IgA (mg/m1) 0 10 20 30 40 PP supernatant IgA (ug/ml) Fig. 9. Correlation between serum IgA and PP supernatant IgA in B6C3F1 mice fed 25ppm vomitoxin or control diet for 8 weeks. PP cells were cultured in RPMI-1640 medium for 7 days and supernatants were collected and analyzed for total IgA concentration. Each data point is representative of one mouse (n=16). Control r=0.06, treatment r=0.44 (both not significant at P<0.05). 77 35 —— O 30 «— 0 control r=0.66 0 treatment r=0.899 25 —- E >0 20 -~ :5. 35., 15 ~— E g 10 —— U) 5 _._ 0 1 (EEK? 1 1 iilgg>4 {D 1 l 1" l r l l l I l O 510152025303540455055 % apoptotic cells Fig. 10. Correlation of serum IgA level and IgA+ apoptotic cell percent in PP of B6C3F1 mice fed control or 25ppm vomitoxin diet for 8 weeks. Each data point is representative of one mouse. r=0.899 is statistically significant at P<0.05. 78 26 I l ‘ a .2 j . O -I U‘ . 3 : il b. q II s I I “ I .d u “ I! I, C - 3 _. I. a .1 u - d «Ill‘i . {gll I ‘ II I II .1 .HJI; ' “II" UNIIII'jIIII! a as see see 1999 BFL2- R 21 j 3 0 .L‘ j . a -I m s '5 I L . a - a -I c . a C O U — I I q} ‘Iliil‘lWII‘JH Ii! .1 IIII‘ILII I'I III.I,I" a 499 BBB BBB 19! FLQ-R Fig. 11. Cell cycle analysis of IgA+ cells from control and treatment mice. A a control mouse, B s 8 week 25 ppm vomitoxin—fed mouse. Arrow is pointing to the apoptotic peak area . 79 vomitoxin-fed mice, whereas no such correlation occurred in the control group (Fig. 10). However, upon later repetition of the experiment in Balb/c mice, no correlation was found between serum IgA levels and.apoptosis inmeither the vomitoxin fed or the control group (data not presented). 3.5 Effect of dietary vomitoxin on IgA specific to intestinal and self antigens Ig reactivity to phosphorylcholine (PC) and inulin in control and treatment mice were compared to assess the effects of vomitoxin on immune response against naturally occurring intestinal bacterial antigens (Gearhart and Cebra, 1979) . PC- specific IgA was significantly increased at both 4 and 8 weeks (Fig. 12). This increase was coupled with a significant decrease in specific IgMZat.8*weeks and an initial decrease in specific IgG at.4 weeks that.recovered after 8 weeks. Inulin- specific IgA was significantly increased after 8 weeks (Fig. 13) whereas there were significant decreases in specific IgM at 4 and 8 weeks. DNA and MRBC were used as model antigens to assess the effect of vomitoxin feeding on autoantibody production. While MRBC-specific IgA increased 4-fold (Fig. 14) DNA-specific IgA increased 15-fold in sera from treated mice compared to controls after' 8 *weeks of ‘vomitoxin feeding (Fig. 15). Increased DNA-specific IgA was coupled with decreased specific IgG and IgM'after'4 weeks (Fig. 15) and.a recovery of specific 196 at 8 weeks. There was also a decrease in MRBC-specific IgM after 4 and 8 weeks of toxin exposure (Fig. 14). 80 Fig. 12. Effect of vomitoxin exposure on PC-specific Igs in unimmunized BSC3F1 mice. Data are means 1 SEM (n=8) and are representative of 3 experiments. * = significantly different (P<0.05) from matching controls. PC specific lg (ng/ml) PC specific lg (ng/ml) 1400~ 1200- 1000 - 800 - 600 - 400 - 200 - I I l I 81 [:1 control treatment 4 wk vomitoxin l 1400- 1200- 1000- 800- 600- 400~ 200- I I 1 I IgA 7 IgG lgM * 8 wk vomitoxin [:1 control treatment I IgA IgG Fig. 12 lgM 82 Fig. 13. Effect of vomitoxin exposure on inulin-specific Igs in unimmunized 86C3F1 mice. Data are means i SEM (n=8) and are representative of 3 experiments. * = significantly different from matching controls. lnulin specific lg (ng/ml) lnulin specific lg (ng/ml) 1600- 1400- 1200‘ 1000- 800- 600- 400‘ 200- m 83 [:1 control treatment II. 4 wk vomitoxin I 1400- 1200- 1000- 800- 600- 400- 200- I IgA Fl lgG CI control treatment Fla lgM 8 wk vomitoxin l lgAV lgG Fig. 13 lgM 84 Fig. 14. Effect of vomitoxin exposure on.MRBC-specific Igs in unimmunized 86C3F1 mice. Data are means 1 SEM (n=8) and are representative of 3 experiments. * = significantly different (P<0.05) from matching controls. 85 2000 T 4 wk vomitoxin ‘E E 5 .. Cjcontrol 9 1500 -treatment .9 g 1000 -- (I) 33 1 a: 2 500-~ 0 Fl ms 9A lgG lgM 8 wk vomitoxin [3 control treatment N N (A 0 01 O O O O O O O i l i mm? It IgA lgG lgM Fig. 14 86 Fig. 15. Effect of vomitoxin exposure on DNA-specific Igs in unimmunized BGC3F1 mice. Data are means i SEM (n=8) and are representative of 3 experiments. * = significantly different (P<0.05) from matching controls. DNA specific lg (ng/ml) DNA specific lg (ng/ml) 87 500 -- 4 wk vomitoxin 400 -+ [3 control ‘i’ treatment .300 -~ I: 200 -- * 100 .. § 0 IgA lgG lgM 8 . . 1 500 ‘r wk vomItoxm [:1 control * treatment 1000 -- 500 ~- 0 ‘ m _ a , Q IgA lgG lgM Fig. 15 88 3.6 IgA hybridoma production Following the finding of large increases in IgA specific to self antigens as well as intestinal antigens after dietary vomitoxin exposure, the self reactivity of vomitoxin-induced IgA was further assessed. In addition, the possibility that this IgA reacted with multiple antigens was evaluated. To further characterize the autoreactive IgA population, IgA- secreting hybridomas were produced from spleen and PP of control and vomitoxin-fed Balb/c mice. The mouse strain was changed from the standard 86C3F1 to Balb/c for enabling fusion with N81 myeloma cell line (derived from Balb/c mice) during hybridoma production. Prior to hybridoma production the effect of increased serum IgA subsequent to vomitoxin feeding was verified in Balb/c mice (Fig. 16). Similar results in Balb/c mice have been found by D. Greene in our laboratory (personal communication). Mice were sacrificed after 8 weeks of 25ppm vomitoxin or clean control diet and spleen and PP lymphocytes were fused with NS1 myeloma cell line. The fusion efficiency, defined as the percent of wells that showed growth out of the total wells seeded, was calculated: spleen control 90%, spleen treatment 97%, PP control 24%, PP treatment 37%. All IgA secreting hybridomas were transferred from the original 96-well plate into 24-well plates and scaled up. During this transfer step, some IgA secreting cells were lost in transfer, possibly due to their sensitivity to dilution with a larger volume of medium. In addition, IgA secreting hybridomas were most likely overgrown 89 *III \1 —i C3 control -treatment CD i (J1 4 1 ..s J I HI! //////// O Fig. 16. Effect of vomitoxin feeding on serum IgA in Balb/c mice. Mice were fed 25ppm vomitoxin or clean control diet for 8 weeks. Data are means i SEM (n=6). Data are representative of 2 experiments. ** = significantly different (P<0.01) from matching control. 90 by other hybridomas (such as IgG and IgM secreting hybridomas) which are more commonly found in the spleen. Table 5. Recovery of IgA secreting hybridomas from master wells following fusion. IgA producing lost in ___!§ll§_____ ££QD§£§I QXEEQIQED 12:; PP control 13 6 4 3 PP treatment 65 14 8 43 spleen control 63 15 47 1 spleen treatment 67 14 49 4 As a result of the massive loss of IgA secreting hybridomas from control tissues and spleen of treatment mice, analysis of IgA reactivity was focused on clones isolated from PP of treatment mice. PP were used for hybridoma production since they are likely sites for action of vomitoxin. In previous studies PP exhibited expanded germinal center development and increased IgA-producing cells following vomitoxin feeding as compared to other organs (Bondy and Pestka, 1991; Pestka et al., 1990). To ensure true clonality of the hybridomas, the number of precursor cells/well was calculated according to the Poisson distribution of probability. This is: u =-lnFo where u=number of cells/number of wells and Fo=% of culture that did not grow. According to this calculation the number of precursor cells/well in hybridomas produced from PP of treatment mice was 0.45, indicating that hybridomas were 91 monoclonal after fusion (i.e. only one clone per well or less). To further ensure clonality, hybridomas were cloned again at 0.5 cells/well and 3 clones were isolated from each hybridoma. The result was 122 IgA producing hybridoma clones. 3.7 Antigenic specificity of monoclonal IgAs The IgA containing culture supernatants produced by the hybridomas were screened against a panel of autoantigens, dietary antigens (casein), intestinal bacterial antigens (PC and inulin) and TNP-BSA. Results based an OD with blank subtracted are presented in Table 6. A large number of clones were specific toITNP and thyroglobulin (Fig. 17). A.very high percentage of the clones were cross-reactive in that they'were able to react to more than one antigen. The highest avidity (as measured by OD) was found for PC and TNP (Fig. 18). The avidity of the individual monoclonal IgAs was lower than that of MOPC 315 IgA but higher than reference serum IgA at the same concentration (Table 6). Detection of reactivity of monoclonal IgAs with IgA and IgE presented a problem sice both antibodies were reactive with the commercially available IgA-specific alkaline phosphatase canjugate. The reason for this reactivity in the case of MOPC 315 is both isotypic and idiotypic specificity since the IgA-specific antibodies were raised against the MOPC 315. In the case of the IgE, the reactivity with the IgA- specific antibodies is probably due to idiotypic interaction with the DNP binding site which is shared by both the MOPC 315 92 8.8.8.. I 13.69.233.388 I E. 59.228 I and .838 u 58 sage—ups... I are. gaseous? n sin ”8... 33.32% . .5 8.6 «8.6 as... as... 3.3 _ s. s. s. a. a. 3.3 _ s. 8 Fee on was 5.2. a. 83 23 so... on... 5.2. _ 2 $3 «93 23 - a. s. 83 8:. a. 5.: _ a. 88.6 as 83 s. s. a; on... 8.8 E... 8.3 _ s. 8.8 23 n8... 2 s. 33 83 s. 8.6 80.3 — __ s. 2. «8... 3.8.6 - s. B s. “8.8 a. 8-3 — __ s. B as... 2 - E a. as... s. a. «93 E __ .2. a. 48.6 s. B s. s. s. s. a. 8.3 «3.6 a. 83 a. s. s. .a a. s. 2.3 — _ E s. 86.6 :88 - a. s. s. 2 82. 3.3 = a. s. as... 8.6 - a. s. B 2 s. 8."... s. s. 488 E; - s. s. 2 a. a. SAL .3 2 R2. SS - .8 £3 8... as c as... 8.3.. a. a. s. E... - s. s. 8 a. s. 8..-... a. 2. 39¢ an... - s. s. s. a. a. 8.3.. £an .58 on. 83% £2: 58.5 o.— -'*-'*"":‘r'n% u 063006thon —l 0 ..e 01 01 Fig. 28. Polyacrylamide gel electrophoresis of monoclonal IgA supernatants. Supernatants of several clones were precipitated with 50% ammonium sulfate and run on a native 4- 17% gradient gel at 100v for 4 hours together with MOPC 315. Following the electrophoresis the gel was stained with Coomassie blue stain for protein identification. 123 29RA 25RA 22RA 12RA 9’» erg ”,00alyl Fig. 29. Western blot of reduced and alkylated monoclonal IgA. Monoclonal IgA supernatants were precipitated with 50% ammonium sulfate 3 times. The supernatants were reduced by incubation with dithiotheitol, alkylated with iodoacetamide and run next to the non-reduced and alkylated supernatants at 100v for 4 hours. The gel was blotted onto PVDF membrane at 100v for 1 hour with 0.01% SDS in running buffer and detected with anti-IgA alkaline phospatase. U) C“ 124 IgA glomerulonephritis, the possibility that the clones were derived from CD5 precursors was evaluated. Monoclonal IgAs (3 representative clones) were stained for presence of cytoplasmic CD5 molecule. All showed positive staining for cytoplasmic CD5 molecule (Fig. 30) suggesting that the hybridomas descended from CD5+ precursors. 3.11 Pathogenicity of monoclonal IgA As a preliminary test for the ‘possibility that the monoclonal IgA in itself can be pathogenic (possibly by binding to self or circulating dietary antigens and forming large immune complexes) 2 representative monoclonal IgAs were injected into 2 control Balb/c mice. As a control, an equal volume of dialyzed, ammonium sulfate precipitated culture media was injected into 3 control mice. Glomerular injury was measured every two weeks by urinanalysis as a measure of the number of red blood cells in the urine. The results are summarized in Table 9. Following 6 weeks of injection mice were sacrificed and the kidneys were sectioned and stained with anti-IgA FITC antibody to test for accumulation of glomerular IgA. There was no difference in glomerular IgA accumulation between media injected and IgA injected mice. However in 2 of 4 sections of kidneys from.:mouse injected. with 3-1-D2 IgA. there ‘were brightly staining cells (Fig. 31) . The preliminary data suggests that polyspecific IgA can be pathogenic in itself and can cause hematuria in the mouse model. 125 Table 9. Number of red blood cells in urine of mice injected ‘with monoclonal IgAsh Red blood cellszfield ;Mate;ial injected Weekz Week4 week; weekfi 1 6 IMedia (3mice) 2 4 3-1-D2 (1 mouse) 2 2325 >300 350 7-1-E4 (1 mouse) 2 2125 267 54 'Mice were injected twice a week with the monoclonal IgA or 'with an equal amount of media. 126 B x x u I 1 1 ‘ .4 E? r ’ ’0‘e‘ 9‘. . 3' 3 x ’ . . conflo ’ ’. cfi."f ‘ ’ o. '- 'f'": " ‘ 3 -1»f’:-.o‘%.h-r,~’~’¢=’c;>,/'. 4’: ° ...4 0.0.0.0?- "g'l'l '6‘. $.‘i : : . (a '0 .. ' I! ' . 1":: ' J; —:~‘:‘- “ I”. ‘5‘” ’ " ,4" :_.._a.:. I'. e e . "- a .- ‘ ’ ‘ [IITIIII1TIIUIITITTj] “‘1 P000 1 FL)! C D 'UUII'IUIIIVVUIIIIT helm .Peetlm Fig. 30. Cytoplasmic staining of monoclonal IgAs wigh anti- CD5 fluorescinated antibody. A= clone 47-2-B6, B - clone 47- 2-85, C = clone 47-2-D4, D = control, stain of clone 47-2-B6 stained with isotype matched control antibody. 127 Fig. 31. Reactivity of kidney from mouse injected with IgA from clone 3-1-D2 with anti-IgA FITC. Arrows are pointing at brightly staining cells in the kidney. Bar size is 1mm. 128 3.12 Relevance of monoclonal IgA to animal model The relevance of the monoclonal IgAs produced to the mouse model had. to be demonstrated for two reasons: (1) the possibility that the clones isolated may not represent the true population in the mouse and ( 2) the different strain used - Balb/c, which might be different in the IgA specificity from previously used B6C3F1 mice. For these reasons, serum IgA and IgA secreting cells in PP and spleen of vomitoxin-fed and control Balb/c mice were quantitated. After 8 weeks of vomitoxin feeding there was a significant increase in the serum IgA specific to PC, TNP and cardiolipin (Fig. 32A). There was a large increase in sphingomyelin-specific IgA but it was not statistically significant due to the large variation between mice. A trend toward increased IgA secreting cells specific ‘to tall four’ antigens was also observed in both PP and spleen (Fig. 32B and C). This information suggests that the antigenic specificity of the IgA in vomitoxin-fed Balb/c mice has a similar pattern to the monoclonal IgAs produced from PP of vomitoxin-fed mice. Polyspecificity of serum IgA was verified by performing inhibition studies in. the same :manner as was done for monoclonal IgAs. Serum IgA reactivity to phosphorylcholine was inhibited by casein and collagen, reactivity to sphingomyelin was inhibited by TNP and casein, reactivity to casein was inhibited by TNP and casein and reactivity to cadiolipin was inhibited by casein and inulin (Fig. 33). These results show that serum IgA is polyreactive in that 129 Fig. 32. Effect of vomitoxin feeding on antigen-specific IgA response of Balb/c mice. Mice were fed 25ppm vomitoxin for 5 weeks and 12.5ppm vomitoxin for 2 more weeks or clean control diet. Serum was tested for antigen-specific IgA by ELISA and spleen and PP cells were tested for antigen-specific IgA producing cells by ELISPOT. A = antigen-specific serum IgA diluted 1:6 (serum pooled from treatment mice); B = antigen- specific IgA producing cells in PP; C = antigen-specific IgA producing cells in spleen. Results are mean 1 SEM (n=5). ** = significantly different (P<0.01) from matching control. Antigen—specific serum lgA (00) PP EUSPOT cells Antigen-specific l Antigen—specific lgg spleen EUSPOT spots/10 cells spots/10 4.000 - 3.000 ‘- 2.000 . I I 130 O. c. D control A treatment .0 n ..E . 300 4 250 4 N O O 150» mol )- D I r PC TNP Cardiolipin Sphlngomyelin Eli fl TNP Cardiolipin Sphingomyelin PC TNP Cardiolipin Sphingornyelin Fig. 32 131 Fig. 33. Inhibition of reactivity of serum IgA against several antigens. Serum was pooled from 5 treatment mice, 10001 each, and diluted 1:15. D looug/ml an 50009/ml 132 O. 8 §8838° (090W tomb“ it) Damn Wounds 2 8898‘” 100 (”mono W” x) 60mm mmiknudooua Fig. 33 2 § 8 8 3 8 ° (”Mimixows) fiwmndumo mildew mm- casein 3 8 f é é «.3 (nwpmwx) WWW 133 reactivity to one antigen can be inhibited by another. These results are another support for the validity of the monoclonal IgA data in the mouse model. 3.13 Specificity of IgA eluted from treatment mouse kidneys Following 8 weeks of vomitoxin feeding there was considerable accumulation of IgA in the glomerulus of treatment mouse kidneys (Fig. 34) . The pathogenic IgA was eluted from kidney sections of treatment mice and its specificity was tested against a panel of antigens. The results are summarized in Table 10. These data give additional support to the monoclonal IgAs relevance to the mouse model since both the monoclonal IgAs and the pathogenic IgA eluted from kidneys were higly reactive to TNP as well as self, dietary and bacterial antigens (Table 7, Fig. 17, 18). 134 Table 10. Reactivity of IgA eluted from treatment mouse kidneys‘ ‘ e OD cardiolipin 1.88 i 0.45 TNP >4.0 casein 0.49 r 0.06 sphingomyelin 0.7 i 0.06 PC 1.5 i 0.19 inulin 0.98 i 0.2 DNA 1.3 i 0.13 'IgA was eluted by four acid washes. The eluted IgA concentration was adjusted to 10 ug/ml and the IgA was screened for reactivity to antigens by ELISA with an overnight incubation at 4°C. Data are means i SEM (n=4). 135 rig. 34. Accumulation of glomerular IgA in kidneys of A = control mouse, B = vomitoxin-fed mouse. Kidney sections of control and treatment mice were stained with anti-IgA FITC conjugate. Data representative of 6 control and 5 treatment mice. Bar size is 10 pH. 136 rig. 34 137 4.0 Discussion The were several major findings in this thesis. First, vomitoxin feeding suppressed total and antigen-specific serum IgG and IgM while increasing total and antigen-specific serum IgA to intestinal and self antigens. Second, dietary vomitoxin increased the percentage of germinal center 8 cells committed to IgA secretion in PP and increased kappa+ IgA+ (fully differentiated) B cells in the spleen. Third, analysis of IgA secreting hybridomas produced from PP of vomitoxin-fed mice revealed that monoclonal IgAs were both autoreactive and, in some cases, polyspecific. Reactivity to one antigen could be inhibited by another. Fourth, specificity of monoclonal IgAs was very similar to immunoglobulins produced by CD5+ B cells which are a cell lineage possibly connected to autoimmune diseases. Representative hybridomas contained cytoplasmic CD5 molecules suggesting their descent from CDS+ ’precursors. This is unique because CD5 derived cells have been isolated from peritoneal cavity and lamina propria in other studies but not from the PP. Fifth, IgAs bound to cells present in the kidneys of vomitoxin-fed mice but not of control mice. The possibility that these cells are macrophages is strengthened by the ability of IgAs to bind peritoneal exudate cells. Sixth, the monoclonal IgAs are trimeric, a characteristic that might make them less easily cleared and more pathogenic. The monoclonal IgAs were pathogenic upon injection to control mice in that they caused hematuria after 4 weeks of injections. Finally, it appeared 138 that monoclonal IgAs were representative of the hyperelevated IgA.following vomitoxin feeding for the following reasons: (a) Increased serum IgA and IgA secreting cells in PP and spleen were specific to TNP, PC, cardiolipin and sphingomyelin after vomitoxin feeding; (b) Specificity of serum IgA to antigens could.be inhibited by another antigen (casein) in an analogous manner to inhibition of monoclonal IgAs and (c) IgA eluted from kidneys of vomitoxin-fed mice exhibited the same antigenic specificity as monoclonal IgAs. Down-regulation of total and antigen specific serum IgG and IgM during vomitoxin feeding mimics the phenomenon of "oral tolerance" that has been described following high dose oral antigen exposure (Emancipator and Lamm, 1989). However, it differs from the classical definition in that the elevated specific IgA is detectable in the systemic compartment rather than the mucosal compartment (McGhee et al. , 1989) . The mechanism for vomitoxin-induced down-regulation of IgG and IgM isotypes is unclear but may involve previously described migration of T regulatory cells from the mucosal to the systemic immune compartments (Mestecky and McGhee, 1987) . Alternatively, the increase in serum IgA coupled with a decrease in IgG and IgM might involve a class-switching mechanism in activated lymphocytes. The decrease in serum IgG and IgM might stem from a technical problem in detecting these Igs. Monoclonal IgAs have been demonstrated to be able to bind IgG and IgE (Tables 6, 7), therefore it is possible that a large increase in serum IgA, with similar reactivity to the 139 monoclonal IgA, will cause formation of IgA-IgG and possibly IgA-lgM immune complexes that will decrease the amount of IgG and IgM estimated by the ELISA technique. In a study by Sakai et al. (1989) a large increase in the IgA-specific helper T cell population was found in IgAN patients. This increased population was considered to account for increased serum IgA by specific conversion of IgM- producing B cells into IgA-producing B cells. In this thesis research, the suppression in total and antigen-specific IgG at 4 weeks and the subsequent recovery seen at 8 weeks (Figs. 6, 12, 13, 14, 15) might indicate the progressive switching from 196 to IgA at 4 weeks and later from IgM to IgA at 8 weeks. This suppression of specific IgG and 1914 production could have a negative influence on the ability of a host to mount an effective immune response against pathogens or clear antigens from the systemic compartment. The inability of dietary vomitoxin to enhance the specific serum IgA response to orally administered TNP-SRBC as described here (Fig. 7) , or to the potent mucosal antigen, CT, as described previously (Pestka et al., 1990b) suggests that ingested vomitoxin does not act as an adjuvant for inducing specific IgA.when an exogenous antigen is orally administered as a bolus; however, suppression of the specific 196 response is evident. In contrast, the serum IgA reactive with chronic and/or endogenous antigens such as PC, inulin, DNA, MRBC, cardiolipin, sphingomyelin, TNP or casein is increased when animals are fed vomitoxin (Figs. 12, 13, 14, 15, 32). This 140 may suggest that there is a temporal requirement for continuous antigen exposure during vomitoxin-induced elevation of specific IgA. Alternatively, vomitoxin may stimulate polyclonal differentiation and secretion among antigen- activated IgA committed B cells at mucosal sites such as the PP that is subsequently expressed at the systemic level. The potential involvement of T cells and their cytokines in this latter effect is supportable by previous observations that ConA-stimulated.T cells from vomitoxin-fed.mice stimulate IgA secretion in B-cells of control mice (Bondy and Pestka, 1991) and that the T helper population is expanded in toxin-exposed mice (Pestka et al., 1990a). This T cell population may have a role in abrogating specific IgG and IgM. Either of the above two models would result in a polyclonal increase in IgA- secreting cells. The stimulation of TNP-specific IgA in unimmunized vomitoxin-fed Balb/c mice as compared to very low levels of TNP-specific IgA in BGC3F1 mice may be due to the strain difference as ‘well as 'the jpossibility' that attempts at immunization actually cause suppression of the specific IgA. Strain differences might also account for the low titer of TNP-specific IgA in BGC3F1 mice when compared to Balb/c mice since it is known that antigenic reactivity is dependent on the genetic ability to respond to the antigen. On the other hand it is possible that deliberate immunization with TNP caused generation of suppressor cells that decreased the TNP- specific reactivity. The last idea is supported by the 141 finding that immunization with cholera toxin (CT) did not increase CT-specific IgA in CT-immunized vomitoxin-fed 36C3F1 mice as compared to control mice (Pestka et al., 1990b). However, without prior immunization the titer to CT was increased in vomitoxin-fed BGCBFl mice. The increase in IgA specific for endogenous bacterial or self antigens as described here may be a factor in immunopathologic effects such as elevated IgA immune complexes and mesangial IgA deposition that have been observed in animals fed vomitoxin (Dong et al., 1991). For example, elevated levels of anti-PC IgA in animals fed vomitoxin might play a role in causing glomerular injury when the animals become exposed to bacterial products. Recently Montinaro et al. (1991) determined, in a mouse model, that mesangial bound anti-PC IgA could capture, in sing, circulating PC-containing antigens including pneumococcal C polysaccharide and this contributed greatly to the extent of glomerular damage. Additional support for this idea is the pathogenicity of the polyspecific autoreactive monoclonal IgAs. Upon injection of these IgAs into a control mouse, hematuria developed after 4 weeks (Table 9) suggesting a pathogenic potential in antibodies that can react with endogenous bacterial and self antigens. The ability of dietary vomitoxin to increase the percent of germinal center 8 cells (PNA*) committed to IgA secretion in the PP together with increased kappa“ (fully differentiated) IgA“ B cells in the spleen (Table 2) suggests 142 a mechanism of expansion of PP B cells that are committed to IgA production. These germinal B cells have been recognized as cells that undergo antigen presentation, activation, proliferation, maturation and isotype switching (Fyfe et al. , 1987). After oral vomitoxin exposure, these germinal center B cells may migrate into the spleen where they complete differentiation and start secreting IgA into the systemic compartment. This mechanism of vomitoxin activation is further supported by the significant correlation between serum IgA levels with spleen IgA secretion (Fig. 8), by the increased percentage of IgA+ B cells found in the 62+}! stage of the cell cycle in spleen from treatment mice (Table 4) and by the increase in the size of the PP germinal centers seen in a previous study (Pestka et al., 1990a). Vomitoxin-mediated expansion event of IgA+ B cells in the PP may involve direct activation of B cells by vomitoxin and/or activation through a vomitoxin effect on T cells in the PP. Vomitoxin, as other trichothecenes, is a potent protein synthesis inhibitor. The effect of vomitoxin on apoptosis (programmed cell death) was studied because it is known that protein synthesis inhibitors will inhibit the process of apoptosis due to the requirement for d_e_ mg protein synthesis (Telford et al., 1991). In this study dietary vomitoxin apparently reduced apoptotic IgA+ B cells in the PP of BGC3F1 female mice (but not in Balb/c female mice) as measured by flow cytometry (Fig. 10). This phenomenon was inversely correlated with the increase in serum IgA. Ongoing studies of 143 vomitoxin effect on apoptosis in cultured cells currently suggest that high levels of vomitoxin (500-1000 ng/ml) can inhibit apoptosis in T cells (Ding, personal communication). During hybridoma production, PP lymphocytes were used for fusion with a myeloma cell line. In the fusion process large antibody forming plasma cells fuse with myeloma cells (Kuus- Reichel et al., 1991). As a result, hybridomas produced are likely to represent B cells differentiated toward the plasma cell population. The 122 IgA secreting hybridomas that were isolated from the PP of vomitoxin fed mice in this work, were highly cross-reactive and self-reactive (Table 6, Figs. 17, 18, 20, 21). The specificity of the hybridomas isolated in this work was also similar to the specificity of antibodies produced by CD5+ (Ly1) B cells and named "natural antibodies" which are typically reactive with diverse antigens such as DNA, PC, MRBC, TNP, BSA, dextran, T cells, myosin, actin and ovalbumin (Klinman, 1992; Klinman and.Holmes, 1990; Lalor and Morahan, 1990; Zoller and Achtnich, 1991). Furthermore, CD5 B cells are a population of B cells that are thought to be involved in autoimmune disease (Kasaian et al., 1991). Antibodies secreted by CD5 B cells are mainly of the lgM isotype but they can switch into IgA production (Kroese et al., 1989). The possibility that the monoclonal IgAs were derived from CD5 B cells is further strengthened by the verification of the cytoplasmic C05 molecule in representative clones (Fig. 30). Table 11 lists the panel of antigens used to screen the 144 Table 11. Characteristics of antigens used on the screening panel. Midget: Grammatical: IgG The most common serum Ig. Can activate complement efficiently. IgA Ig commonly found in mucosal secretions. IgE Ig associated with allergic reactions. Levels of IgE in IgAN patients are elevated. DNA Genetic material. Has a sugar phosphate backbone (Stryer, 1988). Sphingo- The most abundant sphingolipid (Bohinski, 1973) . myelin Found in large quantities in cellular membranes of nerve and brain tissue. Participates as an insulator for nerve fibers. Has membrane phospholipid.that contains.aihydrophobic fatty acid unit and a hydrophylic phosphorylcholine unit (Stryer, 1988). Thyro- A protein of the thyroid gland that is composed of globulin two peptide linked hormones - triiodotyrosine and thyroxine (Bohinski, 1973). During thyroid stimulation thyroglobulin undergoes an enzymatic cleavage by proteolytic enzymes with the resultant secretion of the two hormones into the blood. PC Phosphorylcholine - a phosphoglyceride. A membrane component of bacteria (Stryer, 1988). Precursor of phosphatidyl choline - a membrane component in mammals. Has hydrophilic properties due to presence of positive and negative charges. Inulin A polysaccharide present in bacterial cell walls (Budavari et al., 1989). Casein Protein found in milk. A major component of the AIN-76 diet. TNP A small synthetic hapten with a phenol ring (Budavari et al., 1989). Collagen A protein commonly found in the connective tissues of vertebrates (Bohinski, 1973). Unusually rich in glycines (Stryer, 1988) . A fibrous protein that forms insoluble fibers in the form of a triple- stranded helical rod. Represents a major element in skin, bone, tendon, cartilage, and blood vessels. Cardio- A. phosphoglyceride (Bohinski, 1973). A. major lipin compound lipid occuring most abundantly in membranes. 145 hybridomas and their characteristics. One common motif among some of these antigens is the presence of phosphate groups in sphingomyelin, PC, cardiolipin and DNA. Another common feature is the presence of aromatic rings in TNP, DNA and. all proteins (due to their content of aromatic amino acids). This ability to bind to aromatic haptens might be one explanation of the polyspecific nature of these antibodies since in a study by Michaelson et al. (1992) reactivity with an aromatic hapten was concluded to enable TNP-specific hybridoma antibodies to react to other aromatic amino acids. In that study five different TNP—reactive hybridomas (all producing IgG) showed unexpected interaction with Superose HPLC gel filtration resins" More detailed study demonstrated that the interaction between the antibodies and the resins occurred through the Fab portion of the IgGs and that it was due to an aromatic interaction between the antibodies and the resins. This interaction is assumed to enable TNP and DNP-specific antibodies to react with other aromatic haptens. The apparent high avidity for TNP-BSA among the hybridomas isolated is intriguing. The reactivity of the monoclonal IgAs was, however, lower than the reactivity of MOPC 315 IgA. MOPC 315 IgA is an antibody with a high-affinity combining site that reacts with TNP ligands. It has an intrinsic association constant of 1.6x107 M" (Eisen et al., 1968; Jarvis and Voss, 1983) and in our study was found to cross-react with several other panel antigens. The TNP reactivity of monoclonal IgAs may be partly due to the high molar haptenation of BSA. The 146 molar ratio of TNP to BSA was 53:1 which can account for high avidity, especially with polymeric IgA antibodies. Increased avidity to high molar haptenation with TNP has also been observed with IgM secreted by CD5 B cells. Specifically, in a study by Lalor and Morahan (1990) all IgM antibodies from CD5 B cells were able to bind to TNPwBSA but not to TNP3BSA. This would be characteristic of low affinity antibodies with increased avidity due to repetitious hapten. The ability of TNP-reactive antibodies to cross react with self antigens has been previously reported. In a study by Zoller and Achtnich (1991), IgM-producing TNP-specific hybridomas were shown to be able to bind several self antigens. This is very similar to the hybridomas produced in this study in that most of the hybridomas were able to bind TNP as well as other antigens (Table 6). Studies on human IgA nephropathy also demonstrated TNP-specific IgA antibodies in human patients that were reactive to other antigens (Matsiota et al., 1990; Zoller and Achtnich, 1991) although no explanation on the reason for this cross reactivity was offered. As previously' mentioned, the results presented here suggest that in vomitoxin-fed mice there is a polyclonal stimulation of IgA+ cells in the germinal centers of the PP. These cells may be arising from C05 precursors because of the similarity of the antibody specificity typically produced by this B cell subset and due to the presence of the C05 molecule in IgA-producing hybridomas (Fig. 30). Thus, polyclonal 147 stimulation resulted in elevated production of serum IgA which is both autoreactive and multispecific. Large amounts of autoreactive IgA in the serum could form immune complexes with self antigens, including serum immunoglobulins such as IgG and IgE, and deposit in the kidney causing mesangial IgA accumulation. Another possible mechanism of pathogenicity to kidney tissue that was tested in this work was direct binding between IgA and the kidney tissue. In previous studies, specificity of IgA to laminin and other mesangial proteins was thought to be a contributing factor to IgA deposition in the kidney (Shinkai et al., 1990). However, in this study there was no observable binding between any of the monoclonal IgAs and the glomerular tissue (where IgA accumulation was shown to occur) . On the other hand, the hybridomas were able to bind to blood vessels in kidneys and spleen as well as to hair and hair follicles in the tail (Figs. 22, 25) . This binding is probably due to the ability of the hybridomas to bind to connective tissue proteins such as collagen, as well as to membrane components such as PC, sphingomyelin and cardiolipin. Even though pathogenesis of the vomitoxin-induced IgA.may be mediated through self-reactivity it is probably not mediated by direct binding to kidney mesangium- Rather, self reactive IgAs may form extremely large IgA immune complexes that cannot be filtered by the kidney and will cause disease by immune complex deposition. In vomitoxin-induced glomerulonephritis there is no 148 increase in complement C3 component deposition in the kidney of vomitoxin fed mice (Dong et.al., 1991), suggesting that the damage to kidney tissue is not mediated by complement deposition. An alternative possibility for kidney damage is through binding to Fc receptors for IgA on phagocyte cells in the kidney. This will result in phagocyte activation, release of damaging metabolites and ultimately kidney damage. An interesting observation in this work was the binding of several IgAs to cells in kidneys from vomitoxin fed mice, but not in kidneys from.control mice (Fig. 24). This was true for kidneys from 3 treatment mice but none of the control mouse kidneys. In addition these IgAs were able to bind to large cells in the peritoneal exudate (Fig. 23). These results are suggestive of binding to phagocytes that are present in the kidney as ‘well as in. the, peritoneal exudate. It was previously reported that phagocytic cells are able to bind IgA and IgA immune complex in IgAN patients (Daha et al., 1989). This binding triggers a respiratory burst in the phagocytes resulting in release of damaging metabolites such as 8,0,. Thus, deposition of large IgA immune complexes in the kidney may cause binding to and activation of resident phagocytes. This event will lead to release of damaging metabolites and damage of kidney tissue. In a study by Montinaro et a1. (1991) the involvement of phagocytes in kidney damage in IgA nephropathy was suggested. The ability of monoclonal IgAs to bind to 1913 (Table 7) also suggests the possibility of binding to mast cells through the IgE molecules they have bound on the 149 IgE Fc receptors” The activation of mast cells could occur by crosslinking of these IgE molecules. Hybridomas isolated in this work from PP of vomitoxin fed mice produced mainly trimeric IgA as determined by Coomassie Blue staining (Fig. 28) although small amounts of different polymeric sizes as well as monomeric IgA were also produced (Fig. 27). This ability to produce a mixture of antibody sizes is not unique in that it is common to other IgA hybridomas and myelomas such as MOPC 315. However, the predominant production of trimeric IgA is notable since in a previous study it was found that polymeric IgA was predominantly found in sera from vomitoxin fed mice, where it was assumed to be of dimeric form based on Western blots (Pestka et al., 1989). The finding' that. predominantly trimeric IgA is produced is important since larger molecular size IgA is considered more pathogenic in that it is more difficult to filter through the glomerulus and will more readily deposit in the kidney. In addition, while small IgA immune complexes can.be cleared into the bile by the polymeric IgA receptor on epithelial cells and hepatocytes, this receptor is unable to clear large IgA immune complexes (Mr>.1,000,000) (Daha et al., 1989). Thus further investigation of the molecular size of hyperelevated IgA is warranted. The production of hybridomas from PP of vomitoxin fed mice allowed a detailed analysis of the specificity, cross- reactivity and size of the IgA produced subsequent to 150 vomitoxin feeding. However, there remained the question of the validity of these results in an animal model. There was a possibility that the clones isolated were a population that does not truly represent the population of IgA secreting cells that is induced in the mouse. For this reason, IgA in the Balb/c mouse model was tested for similarities with monoclonal IgA. The results showed that the specificity of the IgA produced following vomitoxin exposure was very similar to that produced by the hybridomas. There was increased serum IgA level against PC, TNP, cardiolipin and sphingomyelin in vomitoxin-fed mice, as well as a trend for increased IgA secreting cells specific for the same antigens in both PP and spleen (Fig. 32). Serum IgA specificity to antigens could be inhibited by casein (Fig. 33) suggesting polyspecific reactivity of serum IgA analogous to monoclonal IgAs. Finally, the pathogenic IgA eluted from kidneys of vomitoxin- fed mice exhibited antigenic reactivity similar to monoclonal IgAs (Table 10). In conclusion, the findings in this thesis suggest a scenario whereby vomitoxin induces hyperelevation of serum IgA+ by altering expansion and/or differentiation of CD5+ B cells in the PP. These B cells produce polyspecific autoreactive and possibly polymeric IgA that apparently accumulates in the systemic compartment. This accumulation may be a result of increased IgA production as well as decreased catabolism and increased half-life of serum IgA. Due to their polyspecific reactivity, these IgA antibodies 151 have the capacity to form large immune complexes with both self-antigens (such as serum Igs and cellular components) as well as bacterial and dietary antigens that penetrate into the systemic compartment. IgA.immune complexes accumulate in the serum and are trapped in the kidney because of their large size. Resident phagocytic cells may be activated either through binding to the antigen in the immune complexes or through binding to the Fc portion of the IgA through IgA Fc receptors on phagocytic cells. Following activation the phagocytic cells undergo a respiratory burst and release damaging metabolites and radicals which cause non-specific injury to kidney cells. 152 5.0 Further studies The appearance of CD5-derived B cells in the PP of vomitoxin-fed mice is unusual since CD5 B cells are mostly found in the peritoneal cavity and lamina propria. It would be valuable to obtain information on the presence of cytoplasmic C05 in IgA+ cells in PP and spleen of vomitoxin- fed mice as compared to control animals. Another interesting subject is the pathogenic potential of monoclonal IgA antibodies. It seems important to find out whether the size is a major factor in pathogenesis. This could be achieved by digestion of the J chain that links the monomeric IgAs into trimers prior to injection into control mice. Inability of monomeric IgA to cause hematuria upon injection would suggest the importance of size in pathogenesis, possibly in combination with the antigenic specificity. Avidity to several antigens, most notably TNP- BSA, is expected to decrease as measured by ELISA following digestion of the J chain. An attempt at comparing the size of serum IgA to monoclonal IgA could be important. In this work, resolution of molecular‘weight.of serum IgA was not successful due to the different charges of serum IgA. This caused serum IgA to appear as a smear on the'Western blot. Further attempts could be made at neutralizing the charges of serum IgA in order to measure the molecular size or at trying to measure the molecular size by HPLC. Alternatively, a measurement of the J chain amount in both serum IgA and supernatant IgA from PP 153 and spleen of vomitoxin-fed mice will give an indication of increased polymeric IgA. In this case IgM, which also has J chain, is not expected to cause misinterpretation of the results since 1914 levels decrease in vomitoxin-fed mice, while an increase in polymeric IgA is the expected result. The activation of B cells in vomitoxin-fed mice through T cells is of major importance. Studies on the effect of vomitoxin on the T cells are ongoing and will shed light on the mechanism of IgA hyperelevation during vomitoxin feeding. Elevation of serum IgE in vomitoxin-fed mice was previously reported (Pestka and Dong, 1992) . It would be interesting to stain for IgE in the kidneys of vomitoxin-fed mice. Glomerular IgE accumulation is likely to be observed due to both the increased levels of serum IgE as well as the high affinity of monoclonal IgA to IgE which might indicate deposition of IgA-IgE immune complexes in the kidney. The presence of IgE in the kidney will have implications in both phagocyte activation and possibly in allergic reactions resulting in kidney damage. Finally, one major aspect of vomitoxin feeding has not been studied, namely, the penetration of vomitoxin from the intestinal tract into the PP. Studies of gavaging with radioactively labeled vomitoxin and measurement of radioactivity in different tissues and organs in the mouse should be done at several time intervals after gavage. These studies will supply important information on the concentrations of vomitoxin in the different tissues, most 154 importantly the PP, as well as give an estimate of the time it takes for vomitoxin to penetrate from the intestine into the internal organs. This information is especially relevant in ‘view of the ongoing use of different vomitoxin concentrations in tissue culture to study its effect on T and B cells. LIST OF REFERENCE 6.0 LIST 0? REFERENCES Abbas H.K., Shier W.T. and Mirocha C.J. (1984) Sensitivity of cultured human and mouse fibroblasts to trichothecenes. l; A§§Q§1 Q££1,Anflll,9h§ml 57: 607-610- Abbas H. R., Mirocha C. J., Pawlosky R. J. and Pusch D. J. (1985) Effect of cleaning, milling and baking on deoxynivalenol in wheat- 89.211 Enxirszn... uisrgnLll 50. 482-486- Abbas K.A., Lichtman A.H. and Pober J.S. (1991) Cellular and Molecular Immunology (ed. M.J. Wonsiewicz), W.B. Saunders Company, Philadelphia, PA. Abouzied M.M. , Asghar A. , Pearson A.H. , Gray J.I. , Miller E.R. and Pestka J.J. (1990) Monoclonal anibody-based enzyme-linked immunosorbent assay for C19-de1ta-16-steroids in sera of boar piQS- Journal at 5911211331 and £9.92 smug-.2 38. 331-335. Abouzied M.M., Azcona J.I., Braselton W.E. and Pestka J.J. (1991) Immunochemical assessment of mycotoxins in 1989 grain foods: evidence for deoxynivalenol (vomitoxin) contamination. A2211 Enxl Miszghl 57: 572-577- Alamartine E., Sabatier J.C. and Berthoux F.C. (1990) Comparison of pathological lesions on repeated renal biopsies in 73 patients with primary IgA glomerulonephritis: value of quantitative scoring and approach to final prognosis. 911nm; HEPDIQlQQX 34: 45'51- Arima S., Nakayama M., Naito M., Sato T. and Takahashi K. (1991) Significance of mononuclear phagocytes in IgA nephropathy- Eisner International 39. 684-692. Ausubel F.A., Brent R., Kingston R.E., Moore 0.D., Seidman J.G., Smith.J.A. and.Struhl.K. (eds.) (1990) Current protocols in. ‘molecular’ biology. Greene Publishing and. Wiley- Interscience, New York. Bamburg J.R. (1983) Biological and biochemical actions of trichothecene myotoxins, in ”Progress in molecular and subcellular biology" (F.E. Hahn) vol 8 (pp. 41-110) . Springer- Verlag, Berlin. 155 156 Bamburg J .R. and Strong F.M. (1971) 12,13-Epoxitrichothecenes, in "Microbial toxins" (S. Kadis, A. Ciegler and S.J. Ajl) vol. VII (pp. 207-292) Academic Press, London, New York. Bene M.C. and Faure G. (1987) IgA nephropathy. Sm; min... Mambo]... 9. 387-394. Bigazzi P.E. (1988) Autoimmunity induced by chemicals. L. : fling Iozlgol. 26, 125-156. Bohinski R.C. (1973) Modern concepts in biochemistry. Allyn and Bacon, Inc., Boston. Bondy G.S. and Pestka J.J. (1991) Dietary exposure to the trichothecene vomitoxin (deoxynivalenol) stimulates terminal differentiation of peyer' s patch B cells to IgA secreting plasma cells. mnglggy ggg Apgligg Phamgcglogy 108, 520- 530. Bradlaw J.A. , Swentzel R.C. and Alterman E. (1985) Evaluation of purified 4-deoxynivalenol (vomitoxin) for unscheduled DNA synthesis in the primary rat hepatocyte-DNA repair essay. BL M Toxig. 23, 1063-1067. Brian P. W. and McGowan J. C. (1946) Biologically active metabolic products of the mold mm W S. Pope. Nature (Lond. ) 157, 334. Bruijn J.A., van Leer E.H.G., Baelde H.J.J., Corver W.E., Hogendoorn P.C.W. , Jan Fleuren G. (1990) Characterization and in vivo transfer of nephritogenic autoantibodies directed against dipeptidyl peptidase IV and laminin in experimental lupus nephritis. Lam 1.029.513; 63, 350-359. Budavari S., O'Neil M.J., Smith A. and Heckelman P.E. (1989) The Meck Index. An encyclopedia of chemicals, drugs and biologicals. Eleventh edition. Published by Merck and Co., Inc. Rahway, N.J. USA. Coligan J.E. , Kruisbeek A.H. , Margulies D.H. , Shevach E.M. and Strober W. (1992) Current Protocols in Immunology. Creene Publishing and Wiley-Interscience, New York. Cooray R. and Lindahl-Kiessling K. (1987) Effect of T-2 toxin on the spontaneous antibody-secreting cells and other non- lymphoid cells in the murine spleen. Eggs: m W 25, 25-29. Coppock R.W. , Swanson S.P. , Gelberg H.B. , Koritz G.D. , Hoffman W.E. , Buck W.B. and Vesonder R.F. (1985) Preliminary study of the pharmacokinetics and toxicopathy of deoxynivalenol (vomitoxin) in swine. m L yet, figs; 46, 169-174. 157 Corrier D.E. (1991) Mycotoxins: mechanisms of immunosuppression. yet. lmmunol, Immunoggthol, 30, 73-87. Cote L.M., Reynolds J.D., Vesonder R.F., Buck W.B., Swanson S.P., Coffey R.T. and Brown D.C. (1984) Survey of vomitoxin- contaminated feed grains in midwestern United States and associated health problems in swine. 5L. MEL; Vet, m 198, 184. ' Cote L.M., Beasley V.R., Bratich P.M., Swanson S.P., Shivaprasad H.L. and Buck W.B. (1985) Sex-related reduced weight gains in growing swine fed diets containing deoxynivalenol. lenrnnl 2f Aninnl Seienee 61. 942-950- Daha M.R., Gorter A., Rits M., van Es L.A. and Hiemstra P.S. (1989) Interaction of immunoglobulin A with complement and phagocytic cells. Biochemistry of the acute allergic reactions: fifth international symposium, pages 247-261. D'Amico G. (1987) The commonest glomerulonephritis in the world: IgA nephropathy- Querierlx Jennnel 2f Medieine 245. 709-727. Dong W., Sell J.E. and Pestka J.J. (1991) Quantitative assessment of mesangial immunoglobulin A (IgA) accumulation, elevated circulating IgA immune complexes, and hematuria during vomitoxin-induced IgA nephropathy. m ML 192112,. 17, 197-207. Eisen H.N., Simms E.S. and Potter M. (1968) Mouse myelome proteins with antihapten antibody activity. The protein produced by plasma cell tumor MOPC-315. mm 7, 4126- 4134. El-Banna A.A. , Lau P.-Y. and Scott P.M. (1983) Fate of mycotoxins during processing of foodstuffs II - deoxynivalenol (vomitoxin) during making of egyptian bread. L m Breteetien 46. 484-486. Emancipator S.N. and Lamm M.E. (1989) Biology of disease. IgA nephropathy: pathogenesis of the most common form of glomerulonephritis. lam 1mm 60, 168-182. Forsell J.H., Witt M.F., Tai J.-H., Jensen R. and Pestka J.J. (1986) Effects of 8-week exposure of the B6C3F1 mouse to dietary deoxynivalenol (vomitoxin) and zearalenone. E... m nglgg_24, 213-219. Forsell J.H., Jensen R., Tai J.-H., Witt M., Lin W.S. and Pestka J.J. (1987) Comparison of acute toxicities of deoxynivalenol (vomitoxin) and 15-acetyldeoxynivalenol in the B6C3F1 mouse. £13. M 119312;. 25, 155-162. 158 Forsyth D.M., Yoshizawa T., morooka N. and Tuite J. (1977) Emetic and refusal activity of deoxynivalenol to swine. Agglg Enylrggg Mlggobiol. 34, 547-552. Friend D.W., Trenholm H.L., Elliot J.I., Thompson B.K. and Hartin K.E. (1982) Effect of feeding vomitoxin-contaminated wheat to pigs. gag, Jg Aging Sci, 62, 1211-1222. Fyfe G. , Cebra-Thomas J.A. , Mustain E. , Davie J.M. , Alley C.D. and Nahm M.H. (1987) Subpopulations of B lymphocytes in germinal centers. 13g Journal g; lmmgnglggy 139, 2187-2194. Galfre G. and Milstein C. (1983) Preparation of minoclonal antibodies: Strategies and.procedures. uggngggflgnzymglg 1983, 3-460 Gearhart P.J. and Cebra J.J. (1979) Differentiated B lymphocytes potential to express particular antibody variable and constant regions depends on site of lymphoid tissue and antigen load. J; Exp; Med, 149, 216-227. Gonzalez-Cabrero J., Egido J., Mampaso F., Rivas M.C. and Hernando L. (1989) Characterization of circulating idiotypes containing immune complexes and their presence in the glomerualr mesangium in patients with IgA nephropathy. Cling Exni Immune]... 76. 204-209- Gonzalez-Cabrero J., de Nicolas R., Hernando L. and Egido J. (1991) Expression and synthesis of cross-reactive idiotypic antibodies by peripheral blood lymphocytes and their suppression. by’ anti-idiotypes in jpatients ‘with IgA nephropathy. Ihg Journal gfi lmmgnglggy 147, 4162-4166. Good A.H., Wofsy L., Henry L. and Kimora J. (1980) In: Selected Methods in Cellular Immunology, eds. B.B. Mishell and S.M. Shiigi (Freeman, San Francisco, CA) pp. 345-347. Greene D.H., Bondy G.S., Azcona-Olivera J.I. and Pestka J.J. (1992) The role of strain and gender in susceptibility to vomitoxin-induced IgA.dysregulation. Meeting of the Institute of Food Technologists, June 20-25 1992, New Orleans, LA. Guery J.C., Druet E., Glotz D., Hirsch F., Mandet C., De Heer E. and Druet P. (1990) Specificity and cross-reactive idiotypes of anti-glomerular'basement.membrane.autoantibodies iangCl sub(2)-induced autoimmune.glomerulonephritis. Egzg_lg Innuneli 20: 93-100- Guery J.C., Tournade H., Pelletier L., Druet E. and Druet P. (1990a) Rat anti-glomerular basement membrane antibodies in toxin-induced autoimmunity and in chronic. graft-vs.-host reaction share recurrent idiotypes. Egzg_qgwlmmgnglg 20, 101- 105. 159 Hagler W.M., Tyczkowska K. and Hamilton P.B. (1984) Simultaneous occurrence of deoxynivalenol, zearalenone, and aflatoxin in 1982 scabby wheat from the midwestern United States. M m Microbiol, 47, 151-154. Harlow E. and Lane D. (1988) Antibodies. A laboratory Manual. Cold Spring Harbor Laboratory. Hashimoto Y. , Kawamura M. , Ichikawa K. , Suzuki T. , Sumida T. , Yoshida S., Matsuura E., Ikehara S. and Koike T. (1992) Anticardiolipin antibodies in NZW x BXSB F1 mice. A model of antiphospholipid syndrome- The Jenrnal.ef Innnnelegiete 149. 1063-1068. Herbert W.J. (1978) Passive haemagglutination with special reference to the tanned cell technique. In: W 91 W lmunology. Edited by D.M. Weir, pp. 20.1-20.2. Hietaniemi V. and Kumpulainen J. (1991) Contents of W toxins in Finnish and imported grains and feeds. figgd Addiiixee and Qenieninenie 8. 171-182. House C. (1991) Vomitoxin reported in several states. Eeedetnffe May 27. 1991 pp- 5 8 46- Jahn S., Schwab J., Hansen A., Heider H., Schroeder C., Lukowsky A. , Achtman M. , Matthes H. , Kiessig S.T. , Volk H.D. , Krueger D.H. and Von Baehr R. (1991) Human hybridomas derived from CD5+ B lymphocytes of patients with chronic lymphocytic leukemia (B-CLL) produce multi-specific natural IgM (kappa) antibodies. £111}... £394 Mggl 83, 413-417. Jarvis M.R. and Voss E.W. (1983) Determination of dissociation constants and ligand specificity of detergent solubilized surface membrane immunoglobulin A from MOPC-315. uglggglg; M193! 20: 125-135- Jelinek C.F., Pohland A.E. and Wood G.E. (1989) Worldwide occurence of mycotoxins in foods and feeds - an update. 1,, 5.53.994. Slit... MAL. M 73: 223-230. Joffe A.z. (1971) Alimentary toxic aleukia, in ”Microbial toxins” (S. Kadis, A. Ciegler and S.J. Ajl) vol. VII (pp. 139- 189). Academic Press, London, New York. Jones C.L., Powell H.R., Kincaid-Smith P. and Roberton D.M. (1990) Polymeric IgA and immune complex concentrations in IgA- related renal disease. Kidney lntemtigngl 38, 323-331. Kasaian M.T., Ikematsu H. and Casali P. (1991) Minireview. 005+ B lymphocytes- Wings 2: the seeieix fer Exnerinenial Bielegx and Hedieine 197. 226-240. 160 Kallenberg C.G.M, Cohen-Tervaert J.W., van der WOude F.J., Goldschmeding R. , von dem Borne A.E.G.K. and Weening J .J . (1991) Autoimmunity to lysosomal enzymes: New clues to vasculitis and glomerulonephritis? Immgnglg ngay 12, 61-64. Khera K.S., Whalen C., Angers G., Vesonder R.F. and Kuiper- Goodman T. (1982) Embryotoxicity of 4-deoxynivalenol (vomitoxin) in mice. Bull, Envirgmg, ggnt_a_m_,_ M 29, 487- 491. Khera K.S., Arnold D.L., Whalen C., Angers G. and Scott P.M. (1984) Vomitoxin (4-deoxynivalenol): effects on reproduction of mice and rats. Texiealm and Applied W 74. 345- 356. Khera K.S., Whalen C. and Angers G. (1986) A teratology study on vomitoxin (4-deoxynivalenol) in rabbits. Egg gngng,2gxigg 24, 421-424. King R.R., McQueen.R.E., Levesque D. and Greenhalgh R. (1984) Transformation of deoxynivalenol (vomitoxin) by rumen microorganisms. fig Aggigg Egod gngmg 32, 1181-1183. Klinman D.M. (1990) Polyclonal B cell activation in Lupus- prone mice precedes and predicts the development of autoimmune disease. gain; Ingest; 86, 1249-1254. Klinman D.M. (1992) Similarities in B cell repertoire development between autoimmune and aging normal mice. The lemma]. ef Innunelmx 148. 1353-1358. Klinman D.M. and Holmes K.L. (1990) Differences in the repertoire expressed by peritoneal and splenic Ly-1 (CD5)+ B cells. The learnal af Imam 144. 4520-4525- Klinman D.M. and Stenberg A.D. (1987) Novel ELISA and ELISA- spot assays used to quantitate B cells and serum antibodies specific for T cell and bromelated mouse red blood cell autoantigens- Jeanna]. ef Magical manna: 102. 157-164- Kroese F.G.M., Butcher E.C., Stall A.M., Lalor P.A., Adams S. and Herzenberg L.A. (1989) Many of the IgA producing plasma cells in murine gut are derived from self-replenishing precursors in the peritoneal cavity. Int; lmmgnglg 1, 75-84. Kumar V., Bartholomew W., Shieh M.-T., Pierce C. and Kaul N. (1991) Standardization of ELISA for the detection of anti- cardiolipin antibodies - effect of non-specific IgG binding. Innanaieaieai Innetigaiiena 20. 583-593. Runs-Reichel K., Beebe A. and Knott C. (1991) Isolation of hybridoma forming cells from immune spleens by unit gravity sedimentation. Hybrldgma 10, 529-538. 161 Lafarge-Frayssinet C., Lespinats G., Lafont P., Louisillier F., iMousset S., Rosenstein Y. and C. Frayssinet (1979) Immunosuppressive effects of Eggggigm extracts and trichothecenes: blastogenic response of mmrine splenic and thymic cells to mitogens (40439). Proceedings of the society for experimental biology and medicine 160, 302-311. Lalor P.A. (1991) An evolutionarily-conserved role for murine Ly-l B cells in protection against bacterial infections. Anteimmaniix 10. 71-76. Lalor P.A. and Morahan G. (1990) The peritoneal Ly-1 (CD5) B cell repertoire is unique among murine B cell repertoires. Bar... 1.. Immune]... 20. 485-492. Lee U.-S., Jang H.S., Tanaka T., Hasegawa A., Oh Y.-J. and Ueno Y. ( 1985) The coexistance of the m mycotoxins nivalenol, deoxynivalenol and zearalenone in Korean cereals harvested in 1983. ngd A1115; 9.931292]. 2, 185-192. Lee U.-S., Jang H.S., Tanaka T., Oh Y.-J., Cho C.-M. and Ueno Y. (1987) Effect of milling on decontamination of m mycotoxins nivalenol, deoxynivalenol and zearalenone in Korean wheat. J_,_ Ag:_19_._ Fm Chem. 35, 126-129. Li P.K.-T., Burns A.P., So A.K.L., Pusey C.D., Feehally J. and Rees A.J. (1991) The DQw7 allele at the HLA-DQB locus is associated with susceptibility to IgA nephropathy in caucasians. 8151119.! W 39, 961-965. Lun A.K., Young L.G. and Lumsden J.H. (1985) The effect of vomitoxin and feed intake on the performance and blood characteristics of young pigs. lggmgl gt Animal figlgnge 61, 1178-1185. Luo Y. , Yoshizawa T. and Katayama T. (1990) Comparative study on the natural occurrence of W mycotoxins (Trichothecenes and Zearalenone) in corn and wheat from high- and low-risk areas for human esophageal cancer in China. Aggl_,_ BBL. limb... 5‘: 3723-3725- Magyarlaki T. , Davin J .C., Szabados E., Kocsis B. and Nagy J. (1990) Peripheral B-lymphocyte markers and function in IgA nephropathy. eliniea]. Nenmrelm 33. 123-129. Marasas W.F.O., van Rensburg S.J. and Mirocha C.J. (1979) Incidence of m species and the mycotoxins, deoxynivalenol and zearalenone, in corn produced in esophageal cancer areas in Transkei. L Ag11_<_:_,_ Eggg Mg 27, 1108-1112. Matsiota P., Dosquet P., Louzir H., Druet E. , Druet P. and Avrameas S. (1990) IgA polyspecific autoantibodies in IgA nephropathy. 9.111]... axe. Ml... 79: 351-356- 162 Matsuoka Y. and Kubota K. (1981) Studies on mechanisms of diarrhea indued by fusarenone-X, a trichothecene mycotoxin from REALM species. onicol. Agglg PM 57, 293-301. Matsuoka Y. and Kubota K. (1985) Studies on mechanisms of diarrhea induced by fusarenone-X, a trichothecene mycotoxin from w species; the effeccts of fusarenone-X and various cathartics on the digestion and the absorption in the mouse intestine. Yaggkg Zagshi 105, 77-82. McGhee J. R., Mestecky J., Elson C. O. and Kiyono H. (1989) Regulation of IgA synthesis and immune response by T cells and interleukins. Jenrnal ef elinieal Immnneleex 9. 175-199. Mestecky J. and McGhee J.R. (1987) Immunoglobulin A (IgA): molecular and cellular interactions involved in IgA biosynthesis and immune response. m in W 40, 153-245. Michaelsen T.E., Lofsgaard.M.F., Aase A. and Heyman B. (1992) Unexpected interaction of some anti-TNP hybridoma antibodies with Superose HPLC gel filtration resins. Jggzngl g: Immuneleeieal Heineee 146. 9-16. Montinaro V., Esparza R., Cavallo T. and Rifai A. (1991) Antigen as mediator of glomerular injury in experimental IgA nephropathy. Laneraeerx.Inxeanieaaien 64. 503-519. Morrissey R . E . and Vesonder R. F . ( 1985) Effect of deoxynivalenol (vomitoxin) on fertility, pregnancy, and postnatal development of Sprague-Dawley rats . Aggl_._ Emu m 49, 1062-1066 . Muso E., Yoshida H., Takeuchi E., Shimada T., Yashiro M., Sugiyama T. and Kawai C. (1991) 9.1.1.1]... 5252:. M 84, 459- 465. Naito I. and Sado Y. (1989) Early changes in rat experimental autoimmune glomerulonephritis induced with the nephritogenic antigen from bovine renal basement membrane. L 911m mm mm... 33: 187-193 . Pelletier L. , Rossert J. , Pasquier R. , Vial M.C. and Druet P. (1990) Role of CD8 super (+) T cells in mercury-induced autoimmunity or immunosuppression in the rat. m L MEL. 31: 55-74 . Pestka J .J . and Bondy G.S. (1990) Alteration of immune function following dietary mycotoxin exposure. Cam 1., Enxeieli Eharmaeeli 68. 1009-1016. Pestka J.J. and Chu F.S. (1984) Aflatoxin B1 dihydrodiol antibody: production and specificity . ApglLed and 163 Enxirenmeneal Eierenieleex 47. 472-477. Pestka J.J. and Dong W. (1992) Serum IgE hyperelevation in B6C3F1 mice following pulsed dietary exposure to the trichothecene ‘vomitoxin. FASEB, .April 1992, .Anaheim, California. Pestka J.J. and Forsell J.H. (1988) Inhibition of human lymphocyte transformation by the macrocyclic trichothecenes roridin A and verrucarin A. ngigglggy Lgttggg 41, 215-222. Pestka.J.J., Tai.J.-H., Witt M.E., Dixon D.E. and Forsell J.H. (1987) Suppression of immune response in the B6C3F1 mouse after dietary exposure to the fusarium. mycotoxins deoxynivalenol (vomitoxin) and zearalenone. Egg gngmg,1gxlgg 25, 297-304. Pestka J.J. , Moorman M.A. and Warner R.L. (1989) Dysregulation of IgA production and IgA nephropathy induced by the trichothecene vomitoxin. Egg gngmg Igzgglg;L 27, 361-368. Pestka J.J., Dong W., Warner R.L., Rasooly L. and Bondy G.S. (1990) Effect of dietary administration of the trichothecene vomitoxin (deoxynivalenol) on IgA and IgG secretion by Peyer's patch and splenic lymphocytes. Egg Qngmg,1gxigg 28, 693-699. Pestka J.J., Dong W., Warner R.L., Rasooly L., Bondy G.S. and Brooks K.H. (1990a) Elevated membrane IgA+ and CD4+ (T helper) populations in murine peyer's patch and splenic lymphocytes during dietary administration of the trichothecene vomitoxin (deoxynivalenol). Egg ghgmg ngigg 28, 409-420. Pestka J .J . , Moorman M.A. and Warner R.L. (1990b) Altered serum immunoglobulin response to model intestinal antigens during dietary exposure to vomitoxin (deoxynivalenol). Texieeleex Letters 50. 75-84. Pollmann D.S., Koch B.A., Seitz L.M., Mohr H.E. and Kennedy G.A. (1985) Deoxynivalenol-contaminated wheat in swine diets. Jenrnal.ef Animal Seienee 60. 239-247. Porcher J.H., Lafarge-Frayssinet C. and Frayssinet C. (1987) Determination of cytotoxic trichothecenes in corn by cell culture toxicity assay. J; Aggggg Qfifig Anglg gnemg 70, 844- 849. Rittenberg H.B. and Pratt K.L. (1969) Antitrinitrophenyl (TNP) plaque essay. Primary response of Balb/c mice to soluble and particulate immunogen. 21.9.2]. Sggl m 5.19... M; 132, 575-581. Ramakrishna Y., Bhat R.V. and Vasanthi S. (1990) Natural occurence of mycotoxins in staple foods in India. 2; Agzigl Eggg,gh§mg 38, 1857-1859. 164 Richardson K.E. , Hagler W.M. , Campbell C.L. and Hamilton P.B. (1985) Production of zearalenone, T-2 toxin and deoxynivalenol by Eusangm spp. isolated from plant materials grown in North Carolina. Mygggathologig 90, 155-160. Robbana-Barnat S., Loridon-Rosa B., Cohen H., Lafarge- Frayssinet C., Neish G.A. and Frayssinet C. (1987) Protein synthesis inhibition and cardiac lesions associated with deoxynivalenol ingestion in mice. EM Adm gag Centaminanee 4. 49-55- Robbana-Barnat S. , Lafarge-Frayssinet C. , Cohen H. , Neish G.A. and Frayssinet C. (1988) Immunosuppressive properties of deoxynivalenol. ngigglggy 48, 155-166. Rosen J.D., Cohen H., Mirocha C.J. and Schiefer H.B. (1985) Yellow rain and the bee faeces theory. ugtgrg 313, 271. Sakai H., Miyazaki M., Endoh.M. and Nomoto Y. (1989) Increase in IgA-specific switch T cells in patients with IgA nephrOPathy- Cline en... Inmnnel... 78. 378-382. Schena F.P. (1990) A retrospective analysis of the natural history of primary IgA nephropathy worldwide. The Amgnlgan leernal ef Menieine 89. 209-215. Schiefer H.B. (1988) The differential diagnosis: Are there natural explanations for what is called "yellow rain" and its alleged effects? W M 2, 51-62. Schmid I. , Uittenbogaart C.H. and Giorgi J .V. (1991) A gentle fixation and permeabilization method for combined cell surface and intracellular staining with improved precision in DNA quantification. gytgmgtzy 12, 279-285. Scott P.M., Harwig J. and Blanchfield B.J. (1980) Screening Fusarium strains isolated from overwintered Canadian grains for trichothecenes. flyggggtnglgglg 72, 175-180. Scott P.M., Lau P.-Y. and Kanhere S.R. (1981) Gas chromatography with electron capture and mass spectrometric detection of deoxynivalenol in wheat and other grains. L m eff... Anal... Chem. 64. 1364-1370. Scott P.M., Nelson K., Kanhere S.R., Karpinski S.P., Hayward S. , Neish G.A. and Teich A.H. (1984) Decline in deoxynivalenol (vomitoxin) concentrations in 1983 Ontario winter wheat before harvest. ABEL. £11m... WILL]. 48: 884-885- Sheu C.W., Moreland F.M., Lee J.K. and Dunkel V.C. (1988) Morphological transformation of Balb/ 3T3 mouse embryo cells in m by vomitoxin. Ed; Chm M12... 26, 243-245. 146 molar ratio of TNP to BSA was 53:1 which can account for high avidity, especially with polymeric IgA antibodies. Increased avidity to high molar haptenation with TNP has also been observed with IgM secreted by CD5 B cells. Specifically, in a study by Lalor and Morahan (1990) all IgM antibodies from CD5 B cells were able to bind to TNPlgBSA but not to TNP3BSA. This would be characteristic of low affinity antibodies with increased avidity due to repetitious hapten. The ability of TNP-reactive antibodies to cross react with self antigens has been previously reported. In a study by Zoller and Achtnich (1991), IgM-producing TNP-specific hybridomas were shown to be able to bind several self antigens. This is very similar to the hybridomas produced in this study in that most of the hybridomas were able to bind TNP as well as other antigens (Table 6). Studies on human IgA nephropathy also demonstrated TNP-specific IgA antibodies in human patients that were reactive to other antigens (Matsiota et al., 1990; Zoller and Achtnich, 1991) although no explanation on the reason for this cross reactivity was offered. As previously mentioned, the results presented here suggest that in vomitoxin-fed mice there is a polyclonal stimulation of IgA+ cells in the germinal centers of the PP. These cells may be arising from CD5 precursors because of the similarity of the antibody specificity typically produced by this B cell subset and due to the presence of the CD5 molecule in IgA-producing hybridomas (Fig. 30). Thus, polyclonal 147 stimulation resulted in elevated production of serum IgA which is both autoreactive and multispecific. Large amounts of autoreactive IgA in the serum could form immune complexes with self antigens, including serum immunoglobulins such as IgG and IgE, and deposit in the kidney causing mesangial IgA accumulation. Another possible mechanism of pathogenicity to kidney tissue that was tested.in this work was direct binding between IgA and the kidney tissue. In previous studies, specificity of IgA to laminin and other mesangial proteins was thought to be a contributing factor to IgA deposition in the kidney (Shinkai et al., 1990). However, in this study there was no observable binding between any of the monoclonal IgAs and the glomerular tissue (where IgA accumulation was shown to occur). On the other hand, the hybridomas were able to bind to blood vessels in kidneys and spleen as well as to hair and hair follicles in the tail (Figs. 22, 25). This binding is probably due to the ability of the hybridomas to bind to connective tissue proteins such as collagen, as well as to membrane components such as PC, sphingomyelin and cardiolipin. Even though.pathogenesis of the vomitoxin-induced IgA.may be mediated through self-reactivity it is probably not mediated by direct binding to kidney mesangium. Rather, self reactive IgAs may form extremely large IgA immune complexes that cannot be filtered by the kidney and will cause disease by immune complex deposition. In vomitoxin-induced glomerulonephritis there is no 148 increase in complement C3 component deposition in the kidney of vomitoxin fed.mice (Dong et al., 1991), suggesting that the damage to kidney tissue is not mediated by complement deposition. An alternative possibility for kidney damage is through binding to Fc receptors for IgA on phagocyte cells in the kidney. This will result in phagocyte activation, release of damaging metabolites and ultimately kidney damage. An interesting observation in this work was the binding of several IgAs to cells in kidneys from vomitoxin fed mice, but not in kidneys from control mice (Fig. 24). This was true for kidneys from 3 treatment mice but none of the control mouse kidneys. In addition these IgAs were able to bind to large cells in the peritoneal exudate (Fig. 23). These results are suggestive of binding to phagocytes that are present in the kidney as well as in. the peritoneal exudate. It ‘was previously reported that phagocytic cells are able to bind IgA and IgA immune complex in IgAN patients (Daha et al., 1989). This binding triggers a respiratory burst in the phagocytes resulting in release of damaging metabolites such as H202. Thus, deposition of large IgA immune complexes in the kidney may cause binding to and activation of resident phagocytes. This event will lead to release of damaging metabolites and damage of kidney tissue. In a study by Montinaro et al. (1991) the involvement of phagocytes in kidney damage in IgA nephropathy was suggested. The ability of monoclonal IgAs to bind to IgE (Table 7) also suggests the possibility of binding to mast cells through the IgE molecules they have bound on the 149 IgE Fc receptors” The activation of mast cells could occur by crosslinking of these IgE molecules. Hybridomas isolated in this work from PP of vomitoxin fed mice produced mainly trimeric IgA as determined by Coomassie Blue staining (Fig. 28) although small amounts of different polymeric sizes as well as monomeric IgA were also produced (Fig. 27). This ability to produce a mixture of antibody sizes is not unique in that it is common to other IgA hybridomas and myelomas such as MOPC 315. However, the predominant production of trimeric IgA is notable since in a previous study it was found that polymeric IgA was predominantly found in sera from vomitoxin fed mice, where it was assumed to be of dimeric form based on Western blots (Pestka et al., 1989). The finding that. predominantly trimeric IgA is produced is important since larger molecular size IgA is considered more pathogenic in that it is more difficult to filter through the glomerulus and will more readily deposit in the kidney. In addition, while small IgA immune complexes can be cleared into the bile by the polymeric IgA receptor on epithelial cells and hepatocytes, this receptor is ‘unable to clear large IgA. immune complexes (Mr>1,000,000) (Daha et al., 1989). Thus further investigation of the molecular size of hyperelevated IgA is warranted. The production of hybridomas from PP of vomitoxin fed mice allowed a detailed analysis of the specificity, cross- reactivity and size of the IgA produced subsequent to 150 vomitoxin feeding. However, there remained the question of the validity of these results in an animal model. There was a possibility that the clones isolated were a population that does not truly represent the population of IgA secreting cells that is induced in the mouse. For this reason, IgA in the Balb/c mouse model was tested for similarities with monoclonal IgA. The results showed that the specificity of the IgA produced following vomitoxin exposure was very similar to that produced by the hybridomas. There was increased serum IgA level against PC, TNP, cardiolipin and sphingomyelin in vomitoxin-fed mice, as well as a trend for increased IgA secreting cells specific for the same antigens in both PP and spleen (Fig. 32). Serum IgA specificity to antigens could be inhibited by casein (Fig. 33) suggesting polyspecific reactivity of serum IgA analogous to monoclonal IgAs. Finally, the pathogenic IgA eluted from kidneys of vomitoxin- fed mice exhibited antigenic reactivity similar to monoclonal IgAs (Table 10). In conclusion, the findings in this thesis suggest a scenario whereby vomitoxin induces hyperelevation of serum IgA+ by altering expansion and/or differentiation of CD5+ B cells in the PP. These B cells produce polyspecific autoreactive and possibly polymeric IgA that apparently accumulates in the systemic compartment. This accumulation may be a result of increased IgA production as well as decreased catabolism and increased half-life of serum IgA. Due to their polyspecific reactivity, these IgA antibodies 151 have the capacity to form large immune complexes with both self-antigens (such as serum Igs and cellular components) as well.as bacterial and.dietary antigens that.penetrate into the systemic compartment. IgA.immune complexes accumulate in the serum and are trapped in the kidney because of their large size. Resident phagocytic cells may be activated either through binding to the antigen in the immune complexes or through binding to the Fc portion of the IgA through IgA Fc receptors on phagocytic cells. Following activation the phagocytic cells undergo a respiratory burst and release damaging metabolites and radicals which cause non-specific injury to kidney cells. 152 5.0 Further studies The appearance of CBS-derived B cells in the PP of vomitoxin-fed mice is unusual since CD5 B cells are mostly found in the peritoneal cavity and lamina propria. It would be valuable to obtain information on the presence of cytoplasmic CD5 in IgA+ cells in PP and spleen of vomitoxin- fed mice as compared to control animals. Another interesting subject is the pathogenic potential of monoclonal IgA antibodies. It seems important to find out whether the size is a major factor in pathogenesis. This could be achieved by digestion of the J chain that links the monomeric IgAs into trimers prior to injection into control mice. Inability of monomeric IgA to cause hematuria upon injection would suggest the importance of size in pathogenesis, possibly in combination with the antigenic specificity. Avidity to several antigens, most notably TNP- BSA, is expected to decrease as measured by ELISA following digestion of the J chain. An attempt at comparing the size of serum IgA to monoclonal IgA could be important. In this work, resolution of molecular weight of serum IgA was not successful due to the different charges of serum IgA. This caused serum IgA to appear as a smear on the‘Western.blot. Further attempts could be made at neutralizing the charges of serum IgA in order to measure the molecular size or at trying to measure the molecular size by HPLC. Alternatively, a measurement of the J chain amount in both serum IgA and supernatant IgA from PP 153 and spleen of vomitoxin-fed mice will give an indication of increased polymeric IgA. In this case IgM, which also has J chain, is not expected to cause misinterpretation of the results since IgM levels decrease in vomitoxin-fed mice, while an increase in polymeric IgA is the expected result. The activation of B cells in vomitoxin-fed mice through T cells is of major importance. Studies on the effect of vomitoxin on the T cells are ongoing and will shed light on the mechanism of IgA hyperelevation during vomitoxin feeding. Elevation of serum IgE in vomitoxin-fed mice was previously reported (Pestka and Dong, 1992) . It would be interesting to stain for IgE in the kidneys of vomitoxin-fed mice. Glomerular IgE accumulation is likely to be observed due to both the increased levels of serum IgE as well as the high affinity of monoclonal IgA to IgE which might indicate deposition of IgA-IgE immune complexes in the kidney. The presence of IgE in the kidney will have implications in both phagocyte activation and possibly in allergic reactions resulting in kidney damage. Finally, one major aspect of vomitoxin feeding has not been studied, namely, the penetration of vomitoxin from the intestinal tract into the PP. Studies of gavaging with radioactively labeled vomitoxin and measurement of radioactivity in different tissues and organs in the mouse should be done at several time intervals after gavage. These studies will supply important information on the concentrations of vomitoxin in the different tissues, most 154 importantly the PP, as well as give an estimate of the time it takes for vomitoxin to penetrate from the intestine into the internal organs. This information is especially relevant in view of the ongoing use of different vomitoxin concentrations in tissue culture to study its effect on T and B cells. LIST OF REFERENCE 6.0 LIST 0? REFERENCES Abbas H.K., Shier W.T. and Mirocha C.J. (1984) Sensitivity of cultured human and mouse fibroblasts to trichothecenes. J; Aeeeei.effi Anali enema 67. 607-610. Abbas H.K., Mirocha C.J., Pawlosky R.J. and Pusch D.J. (1985) Effect of cleaning, milling and baking on deoxynivalenol in wheat- 12211 Enzireni uierenieli 50. 482-486- Abbas K.A., Lichtman A.H. and Pober J.S. (1991) Cellular and Molecular Immunology (ed. M.J. Wonsiewicz), W.B. Saunders Company, Philadelphia, PA. Abouzied M.M. , Asghar A. , Pearson A.M. , Gray J.I. , Miller E.R. and Pestka J.J. (1990) Monoclonal anibody-based enzyme-linked immunosorbent assay for C19-delta-16-steroids in sera of boar pigs. Mane]. ef Annealing]. and Feed enemiein 38. 331-335. Abouzied M.M., Azcona J.I., Braselton W.E. and Pestka J.J. (1991) Immunochemical assessment of mycotoxins in 1989 grain foods: evidence for deoxynivalenol (vomitoxin) contamination. Annie Enxi.niereei 57. 672-677. Alamartine E., Sabatier J.C. and Berthoux F.C. (1990) Comparison.of pathological lesions on repeated renal biopsies in 73 patients with primary IgA glomerulonephritis: value of quantitative scoring and approach to final prognosis. QM Neenreleex 34. 45-51. Arima S., Nakayama M., Naito M., Sato T. and Takahashi K. (1991) Significance of mononuclear phagocytes in IgA nephrOPathy- Kidnex Internaeienal 39. 684-692- Ausubel F.A., Brent R., Kingston R.E., Moore D.D., Seidman J.G., Smith J.A. and Struhl K. (eds.) (1990) Current protocols in molecular biology. Greene Publishing and Wiley- Interscience, New York. Bamburg J.R. (1983) Biological and biochemical actions of trichothecene myotoxins, in "Progress in molecular and subcellular biology" (F.E. Hahn) vol 8 (pp. 41-110) . Springer- Verlag, Berlin. 155 156 Bamburg J .R. and Strong F.M. ( 1971) 12,13-Epoxitrichothecenes, in "Microbial toxins" (S. Kadis, A. Ciegler and S.J. Ajl) vol. VII (pp. 207-292) Academic Press, London, New York. Bene M.C. and Faure G. (1987) IgA nephropathy. M m 1W 9: 337-394- Bigazzi P.E. (1988) Autoimmunity induced by chemicals. L nglgglg: glig. onigol, 26, 125-156. Bohinski R.C. (1973) Modern concepts in biochemistry. Allyn and Bacon, Inc., Boston. Bondy G.S. and Pestka J.J. (1991) Dietary exposure to the trichothecene vomitoxin (deoxynivalenol) stimulates terminal differentiation of peyer's patch B cells to IgA secreting plasma cells. lenieelegx _and Amelia Pearmaeeleex 108. 520- 530. Bradlaw J.A., Swentzel K.C. and Alterman E. (1985) Evaluation of purified 4-deoxynivalenol (vomitoxin) for unscheduled DNA synthesis in the primary rat hepatocyte-DNA repair essay. Egg M Toxic, 23, 1063-1067. Brian P.W. and McGowan J.C. (1946) Biologically active metabolic products of the mold Mill-1.5.129 W S. Pope. Nature (Lond.) 157, 334. Bruijn J.A., van Leer E.H.G., Baelde H.J.J., Corver W.E., Hogendoorn P.C.W. , Jan Fleuren G. (1990) Characterization and in vivo transfer of nephritogenic autoantibodies directed against dipeptidyl peptidase IV and laminin in experimental lupus nephritis. ngg lnyggt. 63, 350-359. Budavari S., O'Neil M.J., Smith A. and Heckelman P.E. (1989) The Heck Index. An encyclopedia of chemicals, drugs and biologicals. Eleventh edition. Published by Merck and Co., Inc. Rahway, N.J. USA. Coligan J.E. , Kruisbeek A.M. , Margulies D.H. , Shevach E.M. and Strober W. (1992) Current Protocols in Immunology. Creene Publishing and Wiley-Interscience, New York. Cooray R. and Lindahl-Kiessling K. (1987) Effect of T-2 toxin on the spontaneous antibody-secreting cells and other non- lymphoid cells in the murine spleen. Eggd m nglgglg 25, 25-29. Coppock R.W. , Swanson S.P. , Gelberg H.B. , Koritz G.D. , Hoffman W.E. , Buck W.B. and Vesonder R.F. (1985) Preliminary study of the pharmacokinetics and toxicopathy of deoxynivalenol (vomitoxin) in swine. Am; J_. Vet, Reg; 46, 169-174. b'r‘l'nm #0.me :m..t:!m 157 Corrier D.E. (1991) Mycotoxins: mechanisms of immunosuppression. Vgt. lmmugol. 1% 30, 73-87. Cote L.M., Reynolds J .D., Vesonder R.F., Buck W.B., Swanson S.P., Coffey R.T. and Brown D.C. (1984) Survey of vomitoxin- contaminated feed grains in midwestern United States and associated health problems in swine. L A1_I_1g;|;_L Veg. AM 198, 184. ' Cote L.M., Beasley V.R., Bratich P.M., Swanson S.P., Shivaprasad H.L. and Buck W.B. (1985) Sex-related reduced weight gains in growing swine fed diets containing deoxynivalenol. Mal g; filing], Scieggg 61, 942-950. Daha M.R., Gorter A., Rits M., van Es L.A. and Hiemstra P.S. (1989) Interaction of immunoglobulin A with complement and phagocytic cells. Biochemistry of the acute allergic reactions: fifth international symposium, pages 247-261. D'Amico G. (1987) The commonest glomerulonephritis in the world: 19A nephrOPathy- enameru alennnal ef Medieine 245. 709-727. Dong W., Sell J.E. and Pestka J.J. (1991) Quantitative assessment of mesangial immunoglobulin A (IgA) accumulation, elevated circulating IgA immune complexes, and hematuria during vomitoxin-induced IgA nephropathy. m Aggl_,_ Tgxic. 17, 197-207. Eisen H.N., Simms E.S. and Potter M. (1968) Mouse myelome proteins with antihapten antibody activity. The protein produced by plasma cell tumor MOPC-315. Bigghgmlm 7, 4126- 4134. El-Banna A.A. , Lau P.-Y. and Scott P.M. (1983) Fate of mycotoxins during processing of foodstuffs II - deoxynivalenol (vomitoxin) during making of egyptian bread. L Logd Ereieeiien 46. 484-486. Emancipator S.N. and Lamm M.E. (1989) Biology of disease. IgA nephropathy: pathogenesis of the most common form of glomerulonephritis. nggrgtgzy lgvgstigatigg 60, 168-182. Forsell J.H., Witt M.E., Tai J.-H., Jensen R. and Pestka J.J. (1986) Effects of 8-week exposure of the B6C3F1 mouse to dietary deoxynivalenol (vomitoxin) and zearalenone. 25L thmg 19812.2. 24, 213-219. Forsell J.H., Jensen R., Tai J.-H., Witt M., Lin W.S. and Pestka J.J. (1987) Comparison of acute toxicities of deoxynivalenol (vomitoxin) and 15-acetyldeoxynivalenol in the B6C3F1 mouse. 2;}. 9151; M2... 25, 155-162. Forsytk Emetic Enviro: Friend Hartin wheat Fyfe G and l: germi: Galfr antih 3-46. Gear! and anti 158 Forsyth D.M., Yoshizawa T., morooka N. and Tuite J. (1977) Emetic and refusal activity of deoxynivalenol to swine. Agglg Egyirggg Microbiol. 34, 547-552. Friend D.W., Trenholm H.L., Elliot J.I., Thompson B.K. and Hartin K.E. (1982) Effect of feeding vomitoxin-contaminated wheat to pigs. Cag, J; Agimg Sci. 62, 1211-1222. Fyfe G. , Cebra-Thomas J.A. , Mustain E. , Davie J.H. , Alley C.D. and Nahm M.H. (1987) Subpopulations of B lymphocytes in germinal centers. Ihg Journal gf lmmggglggy 139, 2187-2194. Galfre G. and Milstein C. (1983) Preparation of minoclonal antibodies: Strategies and.procedures. Metngggunnzymglg 1983, 3-46. Gearhart P.J. and Cebra J.J. (1979) Differentiated B lymphocytes potential to express particular antibody variable and constant regions depends on site of lymphoid tissue and antigen load. J; Exp‘ Mggg 149, 216-227. Gonzalez-Cabrero J., Egido J., Mampaso F., Rivas M.C. and Hernando L. (1989) Characterization of circulating idiotypes containing immune complexes and their presence in the glomerualr mesangium in patients with IgA nephropathy. Cling Erna Immnneli 7‘: 204-209- Gonzalez-Cabrero J., de Nicolas R., Hernando L. and Egido J. (1991) Expression and synthesis of cross-reactive idiotypic antibodies by peripheral blood lymphocytes and their suppression. by’ anti-idiotypes in. ‘patients with IgA nephrOPathy- The Jeannal ef Immnneleex 147. 4162-4166- Good A.H., Wofsy L., Henry L. and Kimora J. (1980) In: Selected Methods in Cellular Immunology, eds. B.B. Mishell and S.M. Shiigi (Freeman, San Francisco, CA) pp. 345-347. Greene D.M., Bondy G.S., Azcona-Olivera J.I. and Pestka J.J. (1992) The role of strain and gender in susceptibility to vomitoxin-induced IgA.dysregulation. Meeting of the Institute of Food Technologists, June 20-25 1992, New Orleans, LA. Guery J.C., Druet E., Glotz D., Hirsch F., Mandet C., De Heer E. and Druet P. (1990) Specificity and cross-reactive idiotypes of anti-glomerular’basement membrane.autoantibodies in HgCl sub(2)-induced autoimmune.glomerulonephritis. flung,gg Immnneli 20: 93-100- Guery J.C., Tournade H., Pelletier L., Druet E. and Druet P. (1990a) Rat anti-glomerular basement membrane antibodies in toxin-induced autoimmunity and in chronic graft-vs.-host reaction share recurrent idiotypes. Egrg,igulmmgnglg 20, 101- 105. 159 Hagler W.M., Tyczkowska K. and Hamilton P.B. (1984) Simultaneous occurrence of deoxynivalenol, zearalenone, and aflatoxin in 1982 scabby wheat from the midwestern United States. Anglg Envinon, Microbigl, 47, 151-154. Harlow E. and Lane D. (1988) Antibodies. A laboratory Manual. Cold Spring Harbor Laboratory. Hashimoto Y., Kawamura M., Ichikawa K., Suzuki T., Sumida T., Yoshida S., Matsuura E., Ikehara S. and Koike T. (1992) Anticardiolipin antibodies in NZW x BXSB F1 mice. A model of antiphospholipid syndrome. The Jounnal g£_lnnnnglgg1§;§ 149, 1063-1068. Herbert W.J. (1978) Passive haemagglutination with special reference to the tanned cell technique. In: Hnndbggk gt Experinentnl lnngnglggy. Edited by D.M. Weir, pp. 20.1-20.2. Hietaniemi V. and Kumpulainen J. (1991) Contents of Ensalinn toxins in Finnish and imported grains and feeds. Eggd Addiniygg nng Qontamingnts 8, 171-182. House C. (1991) Vomitoxin reported in several states. Egedgnnffg May 27, 1991 pp. 5 8 46. Jahn S., Schwab J ., Hansen A., Heider H., Schroeder C., Lukowsky A., Achtman.M., Matthes H., Kiessig S.T., Volk H.D., Krueger D.H. and Von Baehr R. (1991) Human hybridomas derived from CD5+ B lymphocytes of patients with chronic lymphocytic leukemia (B-CLL) produce multi-specific natural IgM (kappa) antibodies. Cling Exp, lnnnngl 83, 413-417. Jarvis M.R. and Voss E.W. (1983) Determination of dissociation constants and ligand specificity of detergent solubilized surface membrane immunoglobulin A from MOPC-315. Mglggglg; Immuneleex 20. 125-136- Jelinek C.F., Pohland A.E. and Wood G.E. (1989) Worldwide occurence of mycotoxins in foods and feeds - an update. J; Aeaee. ere. Anal... Chem... 72. 223-230. Joffe A.z. (1971) Alimentary toxic aleukia, in "Microbial toxins” (S. Kadis, A. Ciegler and.S.J. Ajl) vol. VII (pp. 139- 189). Academic Press, London, New York. Jones C.L., Powell H.R., Kincaid-Smith P. and Roberton D.M. (1990) Polymeric IgA and immune complex concentrations in IgA- related renal disease. Kinney Inngnnnnlgnnl 38, 323-331. Kasaian M.T., Ikematsu H. and Casali P. (1991) Minireview. (205+ B lymphocyteS- Ereeeeeinea ef the seeietx fer. Exeerimenial.Bielegx.ane.ueeieine 197. 226-240- 160 Kallenberg C.G.M, Cohen-Tervaert J.W., van der woude F.J., Goldschmeding R., von dem Borne A.E.G.K. and Weening J.J. (1991) Autoimmunity to lysosomal enzymes: New clues to vasculitis and glomerulonephritis? lnnnnglg nggy 12, 61-64. Khera K.S., Whalen C., Angers G., Vesonder R.F. and Kuiper- Goodman T. (1982) Embryotoxicity of 4-deoxynivalenol (vomitoxin) in mice. Bull, Enyironn. annang TgnngL 29, 487- 491. Khera K.S., Arnold D.L., Whalen C., Angers G. and Scott P.M. (1984) Vomitoxin (4-deoxynivalenol): effects on reproduction of mice and rats- Texieem ans! Anelies! W 74. 345- 356. Khera K.S., Whalen C. and Angers G. (1986) A teratology study on vomitoxin (4-deoxynivalenol) in rabbits. Egg Qneng Tgxigg 24, 421-424. King R.R., McQueen R.E., Levesque D. and Greenhalgh R. (1984) Transformation of deoxynivalenol (vomitoxin) by rumen microorganisms. J; Agzigg Eggg gngng 32, 1181-1183. Klinman D.M. (1990) Polyclonal B cell activation in Lupus- prone mice precedes and predicts the development of autoimmune disease. Cling lnvest. 86, 1249-1254. Klinman D.M. (1992) Similarities in B cell repertoire development between autoimmune and aging normal mice. The lemma]. ef Immuneleex 148. 1353-1358. Klinman D.M. and Holmes K.L. (1990) Differences in the repertoire expressed by peritoneal and splenic Ly-1 (CD5)+ B cells. The lemma]. ef Immunelm 144. 4520-4525- Klinman D.M. and Stenberg A.D. (1987) Novel ELISA and ELISA- spot assays used to quantitate B cells and serum antibodies specific for T cell and bromelated mouse red blood cell autoantigens. lemma]. ef Immunelmieal Meeheee 102. 157-164. Kroese F.G.M., Butcher E.C., Stall A.M., Lalor P.A., Adams S. and Herzenberg L.A. (1989) Many of the IgA producing plasma cells in murine gut are derived from self-replenishing precursors in the peritoneal cavity. lnng.1nngnglg 1, 75-84. Kumar V., Bartholomew W., Shieh M.-T., Pierce C. and Kaul N. (1991) Standardization of ELISA for the detection of anti- cardiolipin antibodies - effect of non-specific IgG binding. Imaeleqieal Inxeseieatiene 20. 583-593. Kuus-Reichel K., Beebe A. and Knott C. (1991) Isolation of hybridoma forming cells from immune spleens by unit gravity sedimentation. Hynziggnn 10, 529-538. Lafarge P., Mo Immunos tricho1 thymic for eX' Lalor Ly-l Autoit Lalor cell 22;; Lee I Ueno hiVaI harm Lee‘ Y.( mycc whee Li 1 Rae 355 can 161 Lafarge-Frayssinet C., Lespinats G., Lafont P., Louisillier F., Mousset S., Rosenstein ‘Y. and. C. Frayssinet (1979) Immunosuppressive effects of Enggninn extracts and trichothecenes: blastogenic response of murine splenic and thymic cells to mitogens (40439). Proceedings of the society for experimental biology and medicine 160, 302-311. Lalor P.A. (1991) An evolutionarily-conserved role for murine Ly-1 B cells in protection against bacterial infections. Anteimmnniix 10. 71-76- Lalor P.A. and Morahan G. (1990) The peritoneal Ly-1 (CD5) B cell repertoire is unique among murine B cell repertoires. Earl £1.1mnnneli 20. 485-492. Lee U.-S., Jang H.S., Tanaka T., Hasegawa A., Oh Y.-J. and Ueno Y. (1985) The coexistance of the m mycotoxins nivalenol, deoxynivalenol and zearalenone in Korean cereals harvested in 1983. Eggg Aiding M 2, 185-192. Lee U.-S., Jang H. S., Tanaka T., Oh Y.-J., Cho C.-M. and Ueno Y. (1987) Effect of milling on decontamination of W mycotoxins nivalenol, deoxynivalenol and zearalenone in Korean wheat. L mg m Qnem, 35, 126-129. Li P.K.-T., Burns A.P., So A.K.L., Pusey C.D., Feehally J. and Rees A.J. (1991) The DQw7 allele at the HLA-DQB locus is associated with susceptibility to IgA nephropathy in caucasians. Kidney lntgnngnigngl 39, 961-965. Lun A.K., Young L.G. and Lumsden J.H. (1985) The effect of vomitoxin and feed intake on the performance and blood characteristics of young pigs. m1 gt Animal figiengg 61, 1178-1185. Luo Y. , Yoshizawa T. and Katayama T. (1990) Comparative study on the natural occurrence of mm mycotoxins (Trichothecenes and Zearalenone) in corn and wheat from high- and low-risk areas for human esophageal cancer in China. ABEL. £111.. hiereh... 5‘: 3723-3726- Magyarlaki T., Davin J .C., Szabados E. , Kocsis B. and Nagy J. (1990) Peripheral B-lymphocyte markers and function in IgA nephropathy. Slinieal Henhreleex 33. 123-129- Marasas W.F.O., van Rensburg S.J. and Mirocha C.J. (1979) Incidence of mm species and the mycotoxins, deoxynivalenol and zearalenone, in corn produced in esophageal cancer areas in Transkei. L M1121. EM m 27, 1108-1112. Matsiota P., Dosquet P., Louzir H., Druet E., Druet P. and Avrameas S. (1990) IgA polyspecific autoantibodies in IgA nephropathy- Cline Exni.Immnneli 79. 361-366- Matst diarl from Mats diar from vari nous 11:00.3 rn" ‘ 162 Matsuoka Y. and Kubota K. (1981) Studies on mechanisms of diarrhea indued by fusarenone-X, a trichothecene mycotoxin from m species. onicol. AngL Pnannacol, 57, 293-301. Matsuoka Y. and Kubota K. (1985) Studies on mechanisms of diarrhea induced by fusarenone-X, a trichothecene mycotoxin from m species; the effeccts of fusarenone-X and various cathartics on the digestion and the absorption in the mouse intestine. Xagakg Zasshi 105, 77-82. McGhee J.R., Mestecky J., Elson C.O. and Kiyono H. (1989) Regulation of IgA synthesis and immune response by T cells and interleukins- Jenrnal ef elinieal Immnneleex 9. 175-199. Mestecky J. and McGhee J.R. (1987) Immunoglobulin A (IgA): molecular and cellular interactions involved in IgA biosynthesis and immune response. Adynnggn in munglggy 40, 153-245. Michaelsen T.E., Lofsgaard.M.F., Aase A. and Heyman B. (1992) Unexpected interaction of some anti-TNP hybridoma antibodies with Superose HPLC gel filtration resins. lgnnnnl g: Immnneleeieal.hetheee 146. 9-16- Montinaro V., Esparza R., Cavallo T. and Rifai A. (1991) Antigen as mediator of glomerular injury in experimental IgA nephropathy- Laheraierx.Inxeetieatien 64. 508-519. Morrissey R. E . and Vesonder R. F . (1985) Effect of deoxynivalenol (vomitoxin) on fertility, pregnancy, and postnatal development of Sprague-Dawley rats . Aan Emu m 49, 1062-1066 . Muso E., Yoshida H., Takeuchi E., Shimada T., Yashiro M., Sugiyama T. and Kawai C. (1991) 91111,, Em M 84, 459- 465. Naito I. and Sado Y. (1989) Early changes in rat experimental autoimmune glomerulonephritis induced with the nephritogenic antigen from bovine renal basement membrane. L Qlin_,_ 1.92,, mm... 33: 187-193 . Pelletier L., Rossert J., Pasquier R. , Vial M.C. and Druet P. (1990) Role of CD8 super (+) T cells in mercury-induced autoimmunity or immunosuppression in the rat. m L We. 31: 55'74 . Pestka J.J. and Bondy G.S. (1990) Alteration of immune function following dietary mycotoxin exposure. gun L Ehxeieli.2harmaeeli 68. 1009-1016. Pestka J.J. and Chu F.S. (1984) Aflatoxin B1 dihydrodiol antibody: production and specificity . Angling m 163 Enyizonmental Microbiology 47, 472-477. Pestka J.J. and Dong W. (1992) Serum IgE hyperelevation in B6C3F1 mice following pulsed dietary exposure to the trichothecene ‘vomitoxin. FASEB, .April 1992, .Anaheim, California. Pestka J.J. and Forsell J.H. (1988) Inhibition of human lymphocyte transformation by the macrocyclic trichothecenes roridin A and verrucarin A. Tgnigglggy Lenteng 41, 215-222. Pestka J.J., Tai J.-H., Witt M.F., Dixon.D.E. and Forsell J.H. (1987) Suppression of immune response in the B6C3F1 mouse after dietary exposure to the fusarium mycotoxins deoxynivalenol (vomitoxin) and zearalenone. Edi Cheni_1gxigi 25, 297-304. Pestka J .J. , Moorman M.A. and Warner R.L. (1989) Dysregulation of IgA production and IgA nephropathy induced by the trichothecene vomitoxin. Edi Chen; ngigi 27, 361-368. Pestka J.J., Dong W., Warner R.L., Rasooly L. and Bondy G.S. (1990) Effect of dietary administration of the trichothecene vomitoxin (deoxynivalenol) on IgA and IgG secretion by Peyer's patch and splenic lymphocytes. 2g; Cngni,1gxigi 28, 693-699. Pestka J.J., Dong‘W., Warner R.L., Rasooly L., Bondy G.S. and Brooks K.H. (1990a) Elevated membrane IgA+ and CD4+ (T helper) populations in murine peyer's patch and splenic lymphocytes during dietary administration of the trichothecene vomitoxin (deoxynivalenol). Egg ghgni ngigg 28, 409-420. Pestka J .J ., Moorman M.A. and Warner R.L. (1990b) Altered serum immunoglobulin response to model intestinal antigens during dietary exposure to vomitoxin (deoxynivalenol). Texieeleex Letters 50. 75-84. Pollmann D.S., Koch B.A., Seitz L.M., Mohr H.E. and Kennedy G.A. (1985) Deoxynivalenol-contaminated wheat in swine diets. leernal ef Animal aeienee 60. 239-247. Porcher J.M., Lafarge-Frayssinet C. and Frayssinet C. (1987) Determination of cytotoxic trichothecenes in corn by cell culture toxicity assay. 2; Aggggi 9111 Anni; gngni 70, 844- 849. Rittenberg M.B. and Pratt K.L. (1969) Antitrinitrophenyl (TNP) plaque essay. Primary response of Balb/c mice to soluble and particulate immunogen. 2:22... 5.29... Expi Big; 15% 132, 575-581. Ramakrishna Y., Bhat R.V. and Vasanthi S. (1990) Natural occurence of mycotoxins in staple foods in India. 1‘ Agzigl Eggfi thmi 38, 1857-1859. flH-U) In. ”In 164 Richardson K.E. , Hagler W.M. , Campbell C.L. and Hamilton P.B. (1985) Production of zearalenone, T-2 toxin and deoxynivalenol by msani um spp. isolated from plant materials grown in North Carolina. Myggnathologin 90, 155-160. Robbana-Barnat S., Loridon-Rosa B., Cohen H., Lafarge- Frayssinet C., Neish G.A. and Frayssinet C. (1987) Protein synthesis inhibition and cardiac lesions associated with deoxynivalenol ingestion in mice. Egg: W and sentaminante 4. 49-55- Robbana-Barnat S. , Lafarge-Frayssinet C. , Cohen H. , Neish G.A. and Frayssinet C. (1988) Immunosuppressive properties of deoxynivalenol. ngigglggy 48, 155-166. Rosen J.D., Cohen H., Mirocha C.J. and Schiefer H.B. (1985) Yellow rain and the bee faeces theory. Ngnnng 313, 271. Sakai H., Miyazaki M., Endoh.M. and Nomoto Y. (1989) Increase in IgA-specific switch T cells in patients with IgA nephropathy- elini em... Immunel... 78. 378-382- Schena F.P. (1990) A retrospective analysis of the natural history of primary IgA nephropathy worldwide. 111g Angniggn lenrnal ef heeieine 89. 209-215. Schiefer H.B. (1988) The differential diagnosis: Are there natural explanations for what is called ”yellow rain" and its alleged effects? ggnngnng Iggiggli 2, 51-62. Schmid I. , Uittenbogaart C.H. and Giorgi J.V. (1991) A gentle fixation and permeabilization method for combined cell surface and intracellular staining with improved precision in DNA quantification. W 12, 279-285. Scott P.M., Harwig J. and Blanchfield B.J. (1980) Screening Fusarium strains isolated from overwintered Canadian grains for trichothecenes. uyggpnthglggin 72, 175-180. Scott P.M., Lau P.-Y. and Kanhere S.R. (1981) Gas chromatography with electron capture and mass spectrometric detection of deoxynivalenol in wheat and other grains. L Aesec... eff... Mal... Chem... 64. 1364-1370. Scott P.M., Nelson K., Kanhere S.R., Karpinski S.P., Hayward S. , Neish G.A. and Teich A.H. (1984) Decline in deoxynivalenol (vomitoxin) concentrations in 1983 Ontario winter wheat before harvest- Anal... Enxiren... Mierehiel... 48. 884-886- Sheu C.W., Moreland F.M., Lee J.K. and Dunkel V.C. (1988) Morphological transformation of Balb/ 3T3 mouse embryo cells in mtg by vomitoxin. EL 31% 19312... 26, 243-245. Shirt- (199 vari Shoi Kwo red £30 935 St: lawn ’73 5.: 165 Shinkai Y., Karai M., Osawa G., Sato M. and Koshikawa S. (1990) Antimouse laminin antibodies in IgA nephropathy and various glomerular diseases. Nennrgn 56, 285-296. Shotwell O.L., Bennett G.A., Stubblefield.R.D., Shannon G.M., Kwolek W.F. and Plattner R.D. (1985) Deoxynivalenol in hard red.'winter"wheat: relationship Ibetween ‘toxin levels and factors that could be used in grading. in Assog, gig; Anni; gngni 68, 954-957. Stryer, L. (1988) Biochemistry. Third edition. W.H. Freeman and Company, New York. Sydenham E.W. , Thiel P.G. , Marasas W.F.O. and Nieuwenhuis J.J. (1989) Occurence of deoxynivalenol and nivalenol in Engaging minegmn infected undergrade wheat in South Africa. L 5911.9... 1129.4 Chem... 37: 922-925- Tanaka T., Hasegawa A., Matsuki Y., Lee U.-S. and Ueno Y. (1986) A limited survey of m mycotoxins nivalenol, deoxynivalenol and zearalenone in 1984 UK-harvested wheat and barley. Feed Agni;- Qeniam. 3. 247-252- Tanaka T. , Hasegawa A. , Yamamoto S. , Sugiura Y. and Ueno Y. (1988a) A case report on a minor contamination of nivalenol in cereals harvested in Canada. Mygggnnnglggin 101, 157-160. - Tanaka T. , Hasegawa A. , Yamamoto s. , Lee U.-S. , Sugiura Y. and Ueno Y. (1988b) Worldwide contamination of cereals by the W mycotoxins nivalenol , deoxynivalenol , and zearalenone. 1. Survey of 19 countries. L Agnigl Egg: Chem, 36, 979-983. Tanaka T., Yamamoto S., Hasegawa A., Aoki N., Besling J.R., Sugiura Y. and Ueno Y. (1990) A survey of the natural occurence of mm mycotoxins, deoxynivalenol, nivalenol and zearalenone, in cereals harvested in the Netherlands. 11292263021931! 110: 19-22- Telford W.G., King L.E. and Fraker P.J. (1991) Evaluation of glucocorticoid-induced DNA fragmentation in mouse thymocytes by flow cytometry. gall 2129.113... 24, 447-459. Thompson W.L. and Wannemacher Jr. R.R. (1984) Detection and quantitation of T-2 mycotoxin with a simplified protein synthesis inhibition essay. Angl,_ 311111221]... W 40, 1027-1031. Trenholm H.L. , Cochrane W.F. , Cohen H. , Elliot J .I. , Farnworth E.R. , Friend D.W. , Hamilton R.M.G. , Standish J.F. and Thompson B.K. (1983) Survey of vomitoxin contamination of 1980 Ontario white winter wheat crop: results of survey and feeding trials. 1.. Aeeee. eta. Anal... Chem. 66. 92-97. Tran] Hart (den pou] 0‘0 dec inn 166 Trenholm H.L. , Hamilton R.M.G. , Friend D.W. , Thompson B.K. and Hartin K.E. (1984) Feeding trials with vomitoxin (deoxynivalenol)-contaminated wheat: effects on swine, poultry, and dairy cattle. JAyMA 185, 527-531. Tryphonas H., Iverson F., So Y., Nera B.A., McGuire P.F., O'Grady L., Clayson D.B. and Scott P.M. (1986) Effects of deoxynivalenol (vomitoxin) on the humoral and cellular immunity of mice. Ignigglggy Letteng 30, 137-150. Tutelyan V.A., Eller K.I., Sobolev V.S., Pimenova V.V., Zasharova L.P. and Muzychenko N.I. (1990) A survey of the occurence of deoxynivalenol in wheat from 1986-1988 harvests in the USSR. Eggg Agdit, $211523... 7, 521-525. Ueno Y. (1983) Trichothecenes: Chemical, biological and toxicological aspects. In: Developments in Food Science vol. 4, ed. Y. Ueno (Elsevier). Ueno Y. (1984) Toxicological features of T-2 toxin and related trichothecenes. EnnL Anglg Tngggh 4, 8124-8132. Ueno Y. (1987) Trichothecenes in food. In: Mycotoxins in food (P. Krogh ed.), pp. 123-147. Academic Press. Ueno Y., Lee U.-S., Tanaka T., Hasegawa A. and Matsuki Y. (1986) Examination of Chinese and U.S.S.R. Cereals for the W mycotoxins , nivalenol , deoxynivalenol and zearalenone. miggn 24, 618-621. Vesonder R.F. , Ciegler A. , Jensen A.H. ( 1973) Isolation of the emetic principle from maxim-infected corn. AggL, Migrghigh 26, 1008-1010. Vesonder R.F., Ciegler A., Rogers R.F., Burbridge K.A., Borhast R.J. and Jensen A.H. (1978) Survey of 1977 crop year preharvest corn for vomitoxin. Angling gnd hierehieleex 36. 885-888. Vidal D.R. (1990) Proprietes immunosuppressives des mycotoxines dy groupe des trichothecenes. 3141.13. m m 88, 159-192. Voss K.A., Plattner R.D., Bacon C.W. and Norred W.P. (1990) Comparative studies of hepatotoxicity and fumonisin B1 and B2 content of water and chloroform/methanol extracts of Fusarium moniliforme strain MRC 826 culture material. W 112, 81-92. Wang E., Norred W.P., Bacon C.W., Riley R.T. and Merrill A.H. Jr. (1991) Inhibition of sphingolipid biosynthesis by fumonisins: Implications for diseases associated with Fusarium moniliforme. L 5.19.]... m 266, 14486-490. 167 Warner R., Brooks K. and Pestka J.J. (1992) IDLYILIQ exposure to vomitoxin (deoxynivalenol) enhances T cell-mediated IgA secretion by splenic B cells. FASEB meeting, April 1992, Anaheim, California. Watson S.A., Mirocha C.J. and Hayes A.W. (1984) Analysis for trichothecenes in samples from southeast asia associated with "yellow rain". Enng, 52211 Ignigi 4, 700-717. Whitlow L.W., Nebel R.L. and Behlow R.F. (1985) Mycotoxin survey results. Proceedings of the dairy fieldmen and sanitarians conference, February 18-19, 1985, pp. 61-62 Whitlow L.W., Nebel R.L., Behlow R.F., Nagler W.M. and Brownie C.F-G. (1986) A survey of North Carolina dairy feeds for mycotoxins. Dairy Fieldmen and sanitarians conference proceedings, February 17-18, 1986, pp. 34-35. Whitlow' L.W. and. Hagler ‘W.M. (1987) The association. of productivity losses in dairy cows with deoxynivalenol. Proceedings from a paper presented at a symposium on Recent Developments in the Study of mycotoxins, Rosemont, Illinois, December 17, 1987. Witt M., Hart L.P. and Pestka J.J. (1985) Purification of deoxynivalenol by water-saturated silica gel chromatography. 1.. m Edi Chem... 33. 745-748. Wood G.E. and Lawrence C. (1989) Limited survey of deoxynivalenol in wheat and corn in the United States. L Aasee... 52ft. Anal... Chem... 72. 38-40. Yoshizawa T. , Matsuura Y. , Tsuchiya Y. , Morooka N. , Kitani K. , Ichinoe M. and Kurata H. (1979) 1‘ Eggg nygi figgi ingnn 20, 21. Yoshizawa T., Swanson S.P. and Mirocha C.J. (1980) T-2 toxin metabolites in the excreta of broiler chickens administered 3H-labelled T-2 toxin. Anni; Enyingni Migzgnigli 39, 1172- 1177. Young J.C., Fulcher R.G., Hayhoe J.H., Scott P.M. and Dexter J.E. (1984) Effect of milling and baking on deoxynivalenol (vomitoxin) content of easter Canadianuwheats. J; Agnig‘,£ggg thmi 32, 659-664. Zoller M. and Achtnich M. (1991) Evidence for regulation of naturally activated autoreactive B cells. m L m 21, 305-312. NSTATE UNI V. LIB llllllLllllllllllHl m Ill! Hjlflllfflfllflzllll1|!"ng