National Toxicology Program

V. TOXICOLOGICAL EFFECTS

Fumonisins are metabolites of Fusarium moniliforme, which has been linked with several diseases in humans and animals, including equine leukoencephalomalacia (ELEM), human esophageal cancer, and porcine pulmonary edema syndrome (PPE) [Ross et al., 1990; Voss et al., 1989]. In many cases, fumonisin B₁ and B₂ have been found at high levels in corn and feed samples contaminated with F. moniliforme that were obtained from areas with high incidences of esophageal cancer and outbreaks of PPE and ELEM (see section IV.C).

Although the databases were searched for toxicological information pertaining to the fumonisins A₁, A₂, B₁, B₂, B₃ and B₄, most of the investigations found in the published literature focus on the toxic effects fumonisin B₁, which is the major fumonisin produced in nature [Sydenham et al., 1991].

A. Chemical Disposition

1. Human Data

No data were found.

2. Animal Data

No data were found.

B. Acute

1. Human Data

No data were found.

2. Animal Data

·oral, rat

It was reported in an abstract that cultures of Fusarium moniliforme fed to rats of unspecified strain and sex, killed the animals in less than 24 hours. However, when pure fumonisin B₁ (21 mg/rat), a metabolite of F. moniliforme, was administered to rats by stomach intubation, no toxic effects were observed. No other data were reported [Mirocha et al., 1990].

C. Prechronic

1. Human Data

No data were found.

2. Animal Data

A summary of the Prechronic effects of Fumonisin in animals is presented at the end of this section in Tables 5 and 6; Table 5 describes the results of dietary fumonisin exposure, and Table 6 reports on the effects following intravenous and oral administration.

·oral, rat

To examine the relationship of dietary fumonisin concentration to hepatotoxicity, male Sprague Dawley rats were fed diets containing extracts of Fusarium moniliform (strain MRC 826) culture material (CM) and/or the extracted CM residues. Two experiments were conducted; one to assess the hepatotoxicity of chloroform/methanol (1:1) CM extractions and the CM residue after chloroform/methanol extraction, and the second to assess the hepatotoxicity of aqueous (using distilled-deionized water) CM extracts and the CM residue after water extraction. Control corn was also extracted in a similar manner for incorporation into the solvent control diets. Each experiment consisted of 4 groups of 5 animals; a solvent control group that was fed a diet containing the extract and the residue of control corn, a group fed the CM extract, a group fed the CM residue after extraction, and a group fed both the CM extract and residue (see Table 2 below). The amount of extract or residue per kilogram of the formulated test and solvent control diets was equivalent to 200 g of CM or control corn, respectively. In addition, positive and negative control groups (5 rats/group) were fed diets containing unextracted CM or unextracted control corn, respectively.

CM extracts and residues were analyzed for fumonisins B₁ and B₂ by hydrolysis followed by gas chromatography/mass spectroscopy, and by thin layer chromatography. Throughout each experiment, animals were observed daily for clinical signs, and body weight and food consumption were measured weekly. After two and four weeks, blood samples were taken for the determination of serum aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (AP), and bilirubin levels. At the end of the study (week 4), all animals were sacrificed and necropsied. Unless otherwise specified, statistical significance was judged at the level P<0.05. A summary of the dosing regimen, dietary fumonisin concentrations, and associated clinical, serum chemical, and histopathologic findings are presented below in Table 2.

All animals survived until the end of the study, and the behavior and appearance of animals in each group were similar. No significant differences were found in body weights, food consumption, relative liver weights, and histology of the liver between the solvent control groups and the negative control group. Throughout the study, animals fed the CM residue after chloroform/methanol extraction, the aqueous CM extract, or the unextracted CM (positive control) had significantly lower body weights than animals in the solvent or negative control groups (specific weights not reported). In each of these three groups, significantly decreased weight gains were found during weeks 1 and 2 only (data not reported). No significant differences in body weight were found between the groups fed the chloroform/methanol CM extract or the aqueous CM residue and their respective controls. When compared to the solvent controls, food consumption was significantly decreased in animals fed the chloroform/methanol CM extract plus extracted CM residue (during weeks 1-2 only), in animals fed the aqueous CM extract (during weeks 2-3 only), and animals fed the aqueous CM extract plus extracted CM residue (during weeks 2-4 only). Throughout the study (weeks 1-4), food consumption in the positive control group was significantly decreased compared to that in the solvent and negative control groups. After 2 and 4 weeks, serum ALT, AST, and AP activities were significantly increased in groups fed the CM residue after chloroform/methanol extraction, the chloroform/methanol CM extract plus extracted CM residue, the aqueous CM extract, the aqueous CM extract plus extracted CM residue, or the unextracted CM (positive control) compared to their respective solvent and negative controls.

Gross necropsy revealed that absolute and relative liver weights were significantly decreased in the groups fed the CM residue after chloroform/methanol extraction, the aqueous CM extract, and the unextracted CM compared to their respective control groups (see Table 3 below). No gross liver lesions were found. However, histological examination revealed liver lesions in 4-5 animals fed CM residue after chloroform/methanol extraction, chloroform/methanol CM extract plus extracted CM residue, aqueous CM extract, aqueous CM extract plus extracted CM residue, and the unextracted CM. These liver lesions were typically characterized by minimal to mild bile duct proliferation and hepatocellular hyperplasia. Other findings included hepatocellular degeneration and necrosis, apoptosis, pyknotic nuclei, mitotic figures, minimal fibrosis, and scant acute inflammatory infiltrates.

Fumonisins B₁ and B₂ were detected in all CM extracts and residues after extraction, and the highest fumonisin concentrations were present in those diets associated with toxic effects. There were no detectable fumonisins in the negative or solvent controls. The authors of this study point out that because the test diets were formulated with extracts and residues, the presence of other compounds in these materials having additive or synergistic effects cannot be dismissed; however, they feel that the data show a positive correlation between fumonisin concentration of the test diets and hepatotoxicity [Voss et al., 1990].

¹E = extract; R = residue after extraction; Corn = control corn; CM = culture material. Materials were added to basal feed at concentrations equivalent to 200 g CM per kg of formulated diet.
²Calculated dietary concentrations of fumonisin B1 and B2 based upon GC/MS analysis of CM extracts, CM residues and control corn; ND = none detected

______________

Reference: Voss et al., 1990

¹E = extract; R = residue after extraction; Corn = control corn; CM = culture material. Materials were added to basal feed at concentrations equivalent to 200 g CM per kg of formulated diet.
²Numbers in parantheses represent standard deviations; Groups with different letters (b or c) are significantly different; P<0.05.

______________

Reference: Voss et al., 1990

·oral, horse

Leukoencephalomalacia (LEM) was induced in two horses (unspecified strain) by the oral administration of fumonisin B₁ (FB₁). In a pilot trial, a filly received 59.5 mg/kg of a 50% preparation of FB₁, administered in 21 doses of 1.25-4 mg/kg over 33 days (the other 50% was inorganic matter that co-eluted during purification). In the second experiment, a colt received 44.3 mg/kg of 95% pure FB₁ in 20 doses of 1-4 mg/kg in 29 days. The FB₁ used in both experiments was isolated from corn cultures of Fusarium moniliforme MRC 826. The horses were closely observed, and serum samples were collected periodically for the determination of aspartate transaminase (AST), gamma glutamyl transferase (GGT), lactate dehydrogenase (LD), and total bilirubin. When dosing was complete, the animals were sacrificed and necropsied.

In the filly, clinical signs became apparent on days 22-27 and consisted of apathy, changes in temperament, lack of coordination, walking into objects, and paralysis of the lips and tongue. However, the filly improved progressively and by day 28 had apparently recovered. The colt exhibited clinical signs from days 24-26 and again from days 31-33. The symptoms consisted of apathy, docility, tremors, pawing motions, bumping into objects, inability to eat or drink, and soporiferousness. Chemical analyses of serum samples showed that the filly had elevated AST activity between days 22-31 (maximum of 365 U/l on day 23), while the colt had elevated GGT activity between days 20-33 (maximum of 52 U/l on day 33).

Gross necropsy of the filly revealed a sunken area (2 cm in diameter) in the lateral part of the anterior frontal lobe of the left cerebral hemisphere. There was slightly more cerebrospinal fluid in this area, and the fluid was tinged yellowish-brown. In addition, the white matter on the cut section of this focus was softer than normal and reddish-brown. Microscopic examination of the lesion revealed necrosis of the white matter, numerous macrophages, aggregates of mineralization, and small hemorrhages. At the periphery of the necrotic area, the blood vessels showed hypertrophy and hyperplasia of endothelial cells, fibrinoid changes of their cell walls, and perivascular mononuclear cell infiltration. The white matter close to the focal lesion had mild status spongiosis and mild to moderate proliferation of astrocytes. No other lesions were evident in other tissues, except for diffuse cloudy swelling and hydropic degeneration of hepatocytes.

Necropsy of the colt showed swelling of the cerebral hemisphere and flattening of the gyri. A yellowish-brown focus was seen in the subcortical white matter of the left dorsal frontal lobe, and extended posteriorly to the occipital lobe. A smaller, gelatinous focus was found in the white matter of the right occipital lobe. In addition, the kidneys were moderately swollen and appeared grayish-yellow. No other macroscopic lesions were seen in any tissues. Microscopic examination of the lesions revealed rarefaction of the neuropil, partial loss of cellular detail of the white matter, swelling and proliferation of the astrocytes, infiltration of macrophages, and swelling of the axons. As seen in the filly, the blood vessels around the foci had hyperplasia and hypertrophy of endothelial cells, as well as perivascular edema. The white and grey matter of the rest of the left side of the brain showed moderate edema, and the right side showed only a mild edema. Evaluation of the proximal convoluted tubules in the kidneys revealed cloudy swelling and hydropic degeneration.

The lesions seen in both horses are characteristic of LEM, and the authors concluded that these results unequivocally prove that fumonisin B₁ can induce LEM in horses [Kellerman et al., 1990].

oral, horse

During the fall of 1989, 18 of 66 purebred Arabian horses at a breeding/training stable in Arizona became ill over a 7-day period with equine leukoencephalomalacia (ELEM). Of the 18 horses affected, 14 died from the condition and 4 partially recovered, but were mildly affected with impaired vision and deviated lips and noses. All of the animals had been fed a diet containing a substantial amount of white corn screenings (1:1 with sweet feed) for 26 days. The animals also received 0.2 kg/day of a protein supplement and free choice of alfalfa or grass hay. Gross examination of the two batches of screenings used in the feed did not reveal any obvious mold, and both batches contained cob parts, damaged kernels, and undamaged kernels. Necropsies were performed on 10 animals, and tissues were collected for histological examination. In addition, several feed samples (corn screenings, sweet feed, protein supplement, alfalfa pellets) were collected and chemically analyzed for the presence of fumonisin B₁ (FB₁) and B₂ (FB₂).

Concentrations of FB₁ in the single subsample from batch one of the corn screenings and the two subsamples from batch two were 37, 58, and 122 ppm, respectively. The respective levels of FB₂ in these samples were 2, 11, and 23 ppm. Subsamples of the protein supplement, alfalfa pellets, and sweet feed contained little if any FB₁ (<5 ppm). A subsample from batch two was then separated into undamaged kernels, damaged kernels, and cob parts, and the levels of fumonisins were measured in each component. In damaged kernels and cob parts, the concentrations of FB₁ were 148 and 144 ppm, respectively, and the levels of FB₂ were 41 and 31 ppm, respectively. In the sample of undamaged kernels the levels of FB₁ and FB₂ were less than 5 ppm.

Gross examination of all horses necropsied showed focal to diffuse unilateral areas of liquefactive necrosis in areas of the cerebral white matter. In some animals, portions of the cerebrum disintegrated when removed from the cranial vault. Also, hemorrhagic foci were often present in the brain stem. Histopathological findings included rarefied white matter with pyknotic nuclei and eosinophilic cytoplasm. Tissue structures were unidentifiable in some sections, while other sections often had hemorrhagic foci located in a distinct perivascular pattern. Microscopic lesions were present mostly in the cerebrum, but were also observed in the brain stem. The authors of this study using information on diet, animal weights, and feeding practices, estimated the total FB₁ dosage for 13 of the 14 horses that died during the outbreak of ELEM; the doses ranged from 0.6-2.1 mg/kg/day. This was the first definitive report on ELEM and associated fumonisin concentrations [Wilson et al., 1990].

·oral, horse

An abstract of an unpublished paper presented at the Fumonisin Symposium held in Raleigh, North Carolina (April 24-25, 1991) describes the results of a study done to determine the minimal dose of contaminated corn screenings needed to reproduce equine leukoencephalomalacia (ELEM) in ponies. Groups of 4-5 ponies were fed formulated diets containing naturally contaminated corn screenings with fumonisin B₁ concentrations of 8, 22, or 44 ppm. Two of the ponies fed 44 ppm fumonisin B₁ died of moderate to severe liver disease and mild encephalopathy. The remaining two ponies in this group died of classic ELEM. Only one pony in the 22 ppm dose group died of ELEM; nine days prior to death, this animals developed elevated liver enzyme levels. The other three ponies fed 22 ppm fumonisin B₁ showed mild behavioral problems, but did not have acute signs of ELEM or elevated liver enzyme levels. In the group given feed containing 8 ppm fumonisin B₁, one pony showed behavioral changes, but no significant gross lesions were found upon necropsy. The ponies fed 8 ppm did show minor, nonspecific lesions in the liver, kidney and brain stem. The authors of this abstract concluded that further evaluation of diets at 8 ppm fumonisin B₁ are needed [Wilson et al., 1991, as reported in FDA, 1991].

·oral, swine

An abstract of an unpublished paper presented at the Fumonisin Symposium held in Raleigh, North Carolina on April 24-25, 1991 reports that pigs fed naturally contaminated corn screenings containing 166 ppm fumonisin B₁ and 48 ppm fumonisin B₂ developed pulmonary edema, pancreatic lesions, and liver damage. Respiratory problems that were observed were not, according to the authors, due to cardiac injury since cardiac dysfunction was not seen. Elevated serum cholesterol and liver enzyme levels were seen in pigs with lung injury. However, a progressive increase in these levels was also observed in pigs that did not die of pulmonary edema. Electron microscopy of tissue sections revealed that hepatocytes, pulmonary type II epithelial cells, and pancreatic acinar cells had intracellular membrane degeneration and plasma membrane changes. According to the authors of this abstract, these findings suggest that cell membranes might be an early target of fumonisins. In addition, Kupffer cells and intravascular macrophages contained myelin figures, suggesting that these cells might also be involved in pathogenesis. The authors speculated that fumonisins induce abnormalities in membrane lipid turnover activated processes in the affected cells, which culminate in pulmonary edema [Haschek et al., 1991, as reported in FDA, 1991].

·oral/intravenous, swine

On 2 southwest Georgia farms, pulmonary edema and hydrothorax were observed in mature swine that died approximately 5 days after consuming corn screenings. An experimental feeding study was conducted in conjunction with a fumonisin injection study to investigate the possible relationship between the deaths and the presence of fumonisins, toxic metabolites of the fungus Fusarium moniliforme. Corn screenings from each farm (1.5 kg samples) were analyzed for the presence of fumonisins, and preliminary data indicated that the concentrations of FB₁ in Feed A and Feed B were 105 mg/kg and 155 mg/kg, respectively. No data were reported on the concentrations of FB₂.

For the feeding study, two groups (Group A or Group B) of 3 swine were fed corn screenings collected from each farm (Feed A or Feed B) for 28 days; a control pig was fed a commercially available grower ration. All pigs were weighed on days 0, 14, and 28, and were observed twice daily for any clinical abnormalities. For the injection study, fumonisins B₁ and B₂ (FB₁ and FB₂; 98% pure) were dissolved in saline with 5% ethanol and injected into swine according to the following dosing regimes: swine 1 received 4 daily injections of 0.4 mg FB₁/kg body weight; swine 2 was given 7 daily injections 0.174 mg FB₁/kg; a third pig received 5 daily injections of 0.3 mg FB₂/kg; and a fourth pig was injected for 7 days with 1.0 ml of saline with 5% ethanol (solvent control). Swine that died during the feeding study or as a result of the injections were necropsied. All other animals were sacrificed and necropsied at the end of the study. Tissue samples were taken from each animal for histological evaluation.

Animals in the feeding study were unable to maintain body weight; the data are reported below in Table 4. On the seventh day of the feeding study, one pig in Group B (fed Feed B) was found dead, and a second, severely dyspneic pig, was euthanized. Necropsy of these animals revealed marked pulmonary edema and hydrothorax. The remaining pig in Group B was sacrificed and necropsied on day 28, and no signs of pulmonary edema were found. In Group A, a severely anorectic pig was euthanized on day 14, and the other two animals were sacrificed at the end of the study. Necropsy showed that none of these animals had developed pulmonary edema or hydrothorax. In the second part of the study, the pig injected with 0.4 mg FB₁/kg/day died on day 5, after receiving a total of 11.3 mg of FB₁. Necropsy of this animal revealed lesions similar to field cases and other experimental cases of pulmonary edema. The other two animals (one receiving a total of 8.65 mg FB₁; the other receiving a total of 10 mg FB₂) survived until the end of the study, and were sacrificed and necropsied 24 hours after their last injection. Neither of these animals had developed pulmonary edema.

The pathological abnormalities found in the animals that developed pulmonary edema after feeding or after injection were essentially the same. The trachea and bronchi contained a clear, foamy liquid, and a golden-yellow liquid filled the thoracic cavities. Interlobular edema was marked, and was most pronounced in the hilus area. Lobular atelectasis was also seen. Microscopically, the alveoli contained only a few cells (mostly macrophages), and had focal to diffuse areas of alveolar septal congestions with capillary thromboses (indicating thrombostasis). In addition, pancreatic lesions were present in all pigs with pulmonary edema/hydrothorax, and consisted of focal to massive necrosis, acinar cell dissociation, and rounded individual acinar cells. In pigs from the feeding study, liver changes were also found; these changes were characterized by centrolobular and random hepatocellular cytoplasmic vacuolar change, hepatocellular cytomegaly, disorganized hepatic cords, and early pirolobular fibrosis. No pulmonary, pancreatic, or liver pathology was noted in the control pig from either study.

The authors of this study conclude that FB₁ affects the pancreas and the lungs, and produces distinct lesions that should not be confused with other conditions that induce pulmonary and/or thoracic effusion. Also, they state that since only swine in the feeding study developed liver lesions, the damage may have been related to nutrient availability, and additional research should be conducted to determine the hepatotoxicity of FB₁ [Harrison et al., 1990].

*Deceased or removed from study.

_____________

Reference: Harrison et al., 1990

·oral/intravenous, swine

As described in an abstract, the hepatotoxicity of fumonisin B₁ was examined in female crossbred swine. In the first part of the study, two pigs were given daily intravenous injections of FB₁ (70% pure); one pig received 7.9 mg/kg/day for 9 days and the other received 4.5 mg/kg/day for 4 days (for a total of 72 and 77 mg, respectively). A third control pig was given daily intravenous injections of saline. Clinical signs and gross lesions were not observed in any of the three pigs. However, necropsy revealed that the hepatic lobules were disorganized with scattered hepatocyte necrosis and mitosis. In the second part of the study, corn screenings contaminated with FB₁ (222 ppm) were fed to three pigs, and uncontaminated corn was fed to two control pigs. All three pigs fed contaminated corn developed respiratory distress within 3-5 days; one was killed on day 4 and one was found dead on day 6. These animals had severe pulmonary interstitial edema, pleural effusion, and individual pancreatic acinar necrosis. Clinical signs in the third pig regressed, and the animal was sacrificed on day 15. The two control pigs were sacrificed on days 4 and 15. Pigs fed FB₁ had liver lesions identical to pigs given FB₁ intravenously, and in both groups, liver enzymes were elevated. The authors concluded that this mycotoxin, given orally and intravenously, is hepatotoxic to pigs [Ness et al., 1991].

·intravenous, horse

Fumonisin B₁ (FB₁) was extracted and purified from the culture material of Fusarium moniliforme MRC 826; the culture material contained approximately 1 g/kg of FB₁. A mare (unspecified strain) was given seven intravenous injections of 0.125 mg FB₁/kg body wight/day on days 0-4, 7, and 9. Serum samples were taken periodically for the determination of aspartate transaminase (AST), gamma glutamyl transferase (GGT), lactate dehydrogenase (LD), and total bilirubin. The horse was sacrificed on day 10 and necropsied.

Clinical signs became apparent on day 8 and consisted of transient nervousness followed by apathy, reluctance to move, loss of coordination, inability to eat, paralysis of the lower lip and tongue, watery, green discharge from the nostrils, and dyspnea. The horse fell down in a convulsive seizure and was euthanasized (day 10). Chemical analysis of serum samples revealed mild elevations of the AST (229 U/l) and GGT (222 U/l) levels on days 8-10. Gross necropsy of the animal revealed severe edema of the brain, and grayish-brown foci (5 mm in diameter) in the medulla oblongata. Other lesions that were noted were congestion and edema of the diaphragmatic lobe of the left lung, mild perirenal edema, and petechiae in the mucosa, and a mild edema of the submucosa in the cecum. Microscopic examination of the medulla oblongata revealed distinct areas of severe necrosis of the gray and white matter that were characterized by rarefaction of the neuropil, necrosis of neurons and glial cells, swelling of glial cells and axons, infiltration by neutrophils and macrophages, and small perivascular hemorrhages. The white and gray matter around these necrotic areas showed evidence of severe edema. Other abnormalities included congestion of the spinal cord, mild edematous changes in the gray matter of the lumbar region, mild nephrosis and edema of the submucosa in the large intestine, and mild congestion and edema of the lungs. No other significant changes were seen in the other tissues examined. The authors stated that these changes represented early, bilaterally distributed leukoencephalomalacia in the brain stem. They also concluded that fumonisin B₁, produced by F. moniliforme, causes equine leukoencephalomalacia [Marasas et al., 1988].

¹FB1 and FB2 = fumonisin B1 and B2, respectively
²NS = Not specified
³The authors estimated that the daily dose of fumonisin B1 was 0.6-2.1 mg/kg

¹FB1 and FB2 = fumonisin B1 and B2, respectively
²NS = Not Specified

D. Chronic/Carcinogenicity

1. Human Data/Case Reports

Although there is no data directly linking fumonisins to cases of human cancer, several reports indicate that ingestion of Fusarium moniliforme contaminated grains by humans is linked to relatively high incidences of human esophageal cancer. In South Africa, human esophageal cancer has a high occurrence in the southwestern districts of the Transkei (Kentani), while in the northeastern region (Bizana) the rate is low. Corn is the main dietary staple in both areas; however corn in the southwestern districts contains a higher percentage of F. moniliforme. During 1985, moldy and healthy corn samples were collected from the two areas, and were analyzed for the presence of several Fusarium mycotoxins, including fumonisins B₁ (FB₁) and B₂ (FB₂). Both fumonisin species were detected in all samples of moldy corn. In the northeast area, the mean concentrations of FB₁ and FB₂ were 6.5 ± 5.3 and 2.5 ± 2.2 µg/g, respectively. In the area with high rates of esophageal cancer, the levels were significantly higher (P<0.01), with mean concentrations of FB₁ and FB₂ of 23.9 ±14.6 and 7.6 ± 4.6 µg/g, respectively. The fumonisin levels determined in the healthy corn samples taken from the Bizana were low, with a mean concentration of FB₁ and FB₂ of 0.06 ± 0.2 and <0.05 ± 0.05 µg/g, respectively (only 3 of the 12 samples analyzed were positive for fumonisins). In Kentani, however, the mean concentrations of FB₁ and FB₂ were significantly (P<0.001) higher (1.6 ± 2.1 and 0.5 ± 0.7 µg/g, respectively), and were recorded in 100% and 83% of the samples, respectively [Sydenham et al., 1990b].

In addition, both fumonisin B₁ and fumonisin B₂ were detected in commercial corn-based samples obtained from Charleston, South Carolina in 1989. This city has one of the highest incidences and mortality rates of esophogeal cancer in the United States. The mean levels of fumonisin B₁ (detected in 7/7 samples) and B₂ (detected in 6/7 samples) were 635 and 182 ng/g, respectively. Even though these levels were lower than those determined in home-grown corn samples from the Transkei, the authors stated that the presence of these mycotoxins in commercial foodstuffs is cause for further investigation of the role of fumonisins in the etiology of esophogeal cancer [Sydenham et al., 1991].

2. Animal Data

·oral, rat

A short-term cancer promotion-initiation bioassay was used as a monitoring system to isolate cancer-promoting compounds from cultures of Fusarium moniliforme MRC 826. Fractions isolated from culture material were screened for cancer-promoting activity by incorporating them in rat mash (at a concentration of 5%) and feeding them to groups of 5 male BDIX rats that had been initiated with a 200 mg/kg intraperitoneal dose of diethylnitrosamine (DEN). The feeding period lasted 4 weeks. All rats were killed after the 4-week promotion treatment, and their livers were analyzed histologically for gamma-glutamyl-transpeptidase-positive (GGT+) foci; the induction of GGT was used as the indicator of cancer-promoting activity. In the first part of the study, culture material was successively extracted with ethyl acetate and aqueous methanol (CH₃OH-H₂O; 3:1). Two samples of culture material, both extracts, and the remaining residues were tested for cancer-promoting activity. In addition, two control groups were included; one initiated with DEN and given feed without culture material, and one receiving treated feed without initiation (given dimethyl sulfoxide (DMSO) instead of DEN). In the second part of the study, the aqueous methanol extract was successively partitioned and fractionated to purify the cancer-promoting compounds; fractions obtained at each step were tested for cancer-promoting activity as described above.

Exposure to diets containing 5% of culture material for 4 weeks significantly induced (P<0.001) the formation of GGT+ foci in DEN-initiated rats. Induction of the foci was not seen in either control group. Following extraction, the bulk of the cancer-promoting activity was recovered in the aqueous methanol extract. When this extract was dried and partitioned between CH₃OH-H₂O (1:3) and CHCl₃, all of the cancer-promoting activity (induction of GGT+ foci) was again found in the aqueous phase, and none was detected in the CHCl₃ phase. This CH₃OH-H₂O fraction was chomatographed on an Amberlite XAD-2 column, and the column was successively eluted with H₂O, CH₃OH-H₂O (1:3 and 1:1), and CH₃OH. Although the majority of the fraction eluted from the column with the CH₃OH-H₂O eluent, the cancer-promoting compound(s) were eluted with CH₃OH. The CH₃OH eluate was further purified on Amberlite XAD-2; the column was successively eluted with CH₃OH-H₂O (3:1) and CH₃OH. This time the active compound was recovered with the CH₃OH-H₂O, and it was fractionated on a Sephadex LH-20 column; fractions obtained from this column were tested for cancer-promoting activity. Two compounds induced GGT+ foci; they were purified on a C18 reverse-phase column, chemically characterized and given the names fumonisin B₁ and B₂. Fumonisin B₁ was the main compound obtained (10 times more than fumonisin B₂), with approximately 2 g purified from 1 kg of the cultured corn in later bulk extractions. The purity of fumonisin B₁ obtained from these later extractions (determined by high performance liquid chromatography) was 92% [Gelderblom et al., 1988].

_______________________

In the study described above, the cancer-promoting activity and toxicity of fumonisin B₁ (FB₁) was tested in male BD IX rats. The cancer-promoting activity was tested using the same 4-week promotion-initiation bioassay described above for the culture material of Fusarium moniliforme, except FB₁ was incorporated into the diet at a concentration of only 0.1%. To examine the toxicity of the compound, four rats were given daily oral doses of 0.95 g FB₁/10 ml of dimethyl sulfoxide (DMSO). However, since three of the four rats died within 3 days, another experiment was run in which four rats were given 12 daily doses of 0.19 g FB₁/10 ml DMSO, followed by 9 daily doses of 0.28 g FB₁/10 ml DMSO. Rats were sacrificed after 21 (toxicity study) or 33 days (cancer-promotion study), and examined histologically. In addition, rats given dietary FB₁ were weighed twice weekly.

A dietary level of 0.1% FB₁ "markedly" induced the formation of GGT+ foci in both DEN-initiated rats and the noninitiated (DMSO control) rats; however, induction was significantly higher (P<0.005) in the initiated group than in the control group. Both of these groups also had a reduction in weight gain in the first week of treatment. By the end of the 4-week feeding period, the mean body weights of the rats treated with FB₁ (initiated and noninitiated) were significantly lower (P<0.0001) than those of the nontreated rats (exact weights not reported).

Necropsy of rats that died after 3 days of dosing with FB₁ revealed toxic hepatitis, characterized by single-cell necrosis with mild fatty changes, hydropic degeneration, and hyaline droplet degeneration. Also, Kupffer cells were increased and enlarged, and a few hepatocellular nuclei were enlarged. Less severe lesions occurred in some of the other organs. These included fatty changes and scant necrosis in the proximal convoluted tubules of the kidney, and lymphoid necrosis in the Peyer's patches and scattered focal epithelial necrosis in the mucosa of the stomach. In addition, two of the rats had severe, disseminated acute myocardial necrosis and severe pulmonary edema. Similar chronic toxic hepatitis was seen in the rats killed after 21 days of oral dosing with FB₁ and rats killed after 33 days of receiving dietary FB₁. In the latter group, the changes in the liver were more advanced and caused distortion of the lobular structure. These animals also developed hyperplastic nodules containing hepatocytes with vesicular nuclei, foamy cytoplasm, and mitotic figures, and kidney lesions similar to those seen in the rats that died after three days of oral dosing. No lesions occurred in the liver or the kidneys of rats in the control groups.

The authors of this study concluded that fumonisin B₁ is a complete carcinogen, and might be responsible for the hepatocarcinogenicity of F. moniliforme MRC 826. They also state that the subacute pathological changes in the liver of rats caused by FB₁ were similar to those caused by culture material of F. moniliforme, indicating this fumonisin may also be responsible for the toxicity of the fungus in rats [Gelderblom et al., 1988].

·oral, rat

As described in an abstract, the carcinogenic potential of an alcohol:water (1:1) extract of Fusarium moniliforme (FUSX), containing 20 ppm fumonisin B₁ (FB₁) was examined in female F344/N rats. Five groups of 6 animals were fed a semipurified diet, with or without FUSX and with or without a 30 mg/kg oral dose of diethylnitrosamine (DEN) as an initiation agent. The dosing scheme was as follows: group 1 received the control diet (diet without FUSX) for 13 weeks; group 2 also received the control diet for 13 weeks, but was given the dose of DEN after one week; group 3 was given the FUSX diet for 13 weeks; group 4 was given the FUSX diet for 1 week, the dose of DEN, and then the control diet for 12 weeks; and group 5 received the FUSX diet for 13 weeks, with the dose of DEN after week 1. To assess the early stages of carcinogenesis, placental glutathione S-transferase-positive (PGST[+]) hepatocytes were counted in 5 frozen hepatic sections/rat using immuohistochemistry. The results show that groups 4 and 5 had significantly more (P<0.05) PGST[+] hepatocytes than the other three groups; animals in both these groups were fed diets containing FUSX and were given the dose of DEN. The authors of this study concluded that FUSX had significant co-initiating activity, and F. moniliforme may pose a co-carcinogenic risk even during short-term, low-level exposure [Lebepe and Hendrich, 1991].

·oral, rat

The toxic and carcinogenic effects of the Fusarium moniliforme metabolite, fumonisin B₁ (FB₁), were examined in 50 male BD IX rats. For over 26 months, a group of 25 animals was fed a semi-purified corn-based diet containing 50 mg/kg of pure (not < 90%) FB₁, isolated from a culture material of F. moniliforme strain MRC 826. The FB₁ was dissolved in methanol and evaporated into a subsample of the diet, which was then mixed into the diet to obtain the desired fumonisin concentration. A control group of 25 animals received the same diet without FB₁, and with an equal volume of methanol. Five rats from each group were sacrificed at 6, 12, 20 and 26 months. When rats in the experimental group died during the course of the study, an equal number of rats from the control group were killed. During the experiment, the rats were observed daily for clinical signs and weighed weekly. After sacrifice or death, the animals were necropsied, and the organs were examined macroscopically for abnormalities; the liver was also examined histologically. In addition, blood was collected from each animal, and the serum samples were analyzed for alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), gamma-glutamyltranspeptidase (GGT), bilirubin, total protein, globulin, albumin, cholesterol, urea, and creatinine.

Both control and experimental animals became obese after 20 months; however, from 12 months onward, the weight gain of the control group was significantly more (P<0.01) than that seen in the FB₁ group (see Table 7 below). Five rats from the FB₁ group died, mainly from pneumonia, between months 18 and 24, while the survival rate of the control group was 96%. No other clinical signs were reported. Analysis of the serum samples showed that there was a "marked increase" in levels of AST, GGT, ALP, creatinine and bilirubin in the FB₁-treated rats killed at 20 and 26 months (data not reported). However, only the levels of AST, GST and bilirubin were significantly higher in the experimental group compared to controls (P values not reported). In addition, the albumin:globulin ratio was significantly reduced (P<0.005) in the experimental group at 26 months compared to controls. This reduction was due to an increase in serum globulin levels (data not reported). Serum cholesterol levels were significantly higher (P<0.005) in the FB₁-treated animals killed at 20 months, but not in the groups killed at 26 months. No differences were seen in the total protein content, and no results were reported on urea.

Pathological changes are summarized below in Table 7. All FB₁-treated animals that died or were killed from 18 months onward (n=15) suffered from a macro- and micronodular cirrhosis, had large expansive nodules of cholangiofibrosis at the hilus of the liver, and had a multitude of hepatic regenerative nodules. The changes that would evolve into cirrhosis and cholangiofibrosis were present in the livers of FB₁ rats killed after only 6 months; these changes consisted of scattered areas of fibrosis, bile duct hyperplasia, and lobular distortion. In treated rats killed from 12 months onward, the liver was distorted and had fatty changes, necrosis, hemorrhage, and irregular blood supply. Fully developed regenerative nodules and cholangiofibrosis were present in the liver of treated rats as early as 6 months. However, both types of lesions increased in severity and size, and changed histologically as the study progressed. By the terminal stages of the study, some of the regenerative nodules manifested cellular characteristics of preneoplastic changes, and a few transformed into hepatocellular carcinoma. Cholangiofibrosis, from 18 months onward, was characterized by irregular, duct-like structures with an epithelial lining that contained necrotic cells and lacked mitotic figures. The large amounts of cellular debris and mucus produced by the epithelium caused distention and rupture of the tubules. According to the authors, some of the larger lesions (up to 3 cm) may have progressed to cholangiocarcinoma. Ten of the 15 rats in the experimental group that were killed or died after 18 months, developed primary hepatocellular carcinoma with varying histological differentiation. Four of these animals also developed metastases of the heart, lungs, or kidneys. No neoplastic lesions were seen in any of the control animals during the course of the experiment.

Animals from both the experimental and the control groups consistently developed lesions in the kidneys and the lungs. A majority of the rats (specific numbers not reported) that died or were killed after 18 months had a mild to moderately severe incidental interstitial pneumonia and lymphocytic bronchitis that did not differ significantly between the two groups. The lesions in the kidneys, which consisted of focal to diffuse interstitial lymphocytic nephritis and mild membranoproliferative glomerulonephritis, did not differ significantly between the experimental and control group up to 20 months. However, chronic lesions (interstitial nephritis) occurred in the kidneys of FB₁-treated rats killed at 26 months. The authors point out that no lesions were observed in the esophagus, heart, or forestomach of treated rats, which is contrary to previous findings when F. moniliforme was fed to rats (see Gelderblom et al., 1988 above). They concluded that FB₁ is responsible for the hepatocarcinogenic and the hepatotoxic, but not all the toxic effects induced by culture material of F. moniliforme MRC 826 in rats [Gelderblom et al., 1991].

¹Means in a column followed by the same letter do not differ significantly (P<0.05). If the letters differ but the cases do not, then P<0.01; if the letters and cases differ, then P<0.001.
²RN = regenerative nodules; CF = cholangiofibrosis; Cirrh. = cirrhosis; HCC = hepatocellular carcinoma.
³Lung, heart and/or kidney metastases - 4/15 of the FB1-treated rats that were killed or died between 18-26 months.
⁴ND = not determined
⁵Survival rate: controls, 96%; treated, 80%

_________________

Reference: Gelderblom et al., 1991

E. Reproductive Effects and Teratogenicity

1. Human Data

No data were found.

2. Animal Data

No data were found.

F. Genetic Toxicology

1. Human Data

No data were found.

2. Prokaryotic Data

·In an abstract of an unpublished study presented at the Fumonisin Symposium held in Raleigh, North Carolina on April 24-25, 199, it was stated that fumonisins (B's) are not mutagenic and had no effect on unscheduled DNA synthesis. No other data were reported [Gelderblom et al., 1991, as reported in FDA, 1991].

3. Eukaryotic Data

No data were found.

G. Other Toxicological Effects

1. Immunotoxicity

·in vitro, chicken

As described in an abstract, chicken peritoneal macrophages (PM) and a chicken macrophage cell line MQ-NCSU were exposed to fumonisin B₁ at concentrations of 0.5, 5.0, 10.0, 20.0, 40.0, or 100.0 µg/ml for 2 and 4 hours. The three lowest concentrations of FB₁ caused significant (P value not reported) cytotoxicity in PM after 2 and 4 hours of exposure. Morphological changes in these cells included cytoplasmic blebing and nuclear disintegration. After 4 hours of exposure to 20, 40, and 100 µg/ml of FB₁, significant depression was seen in the phagocytic potential of PM. Exposure to FB₁ alone and after stimulation with lipopolysaccharide induced cytolytic factor secretion by MQ-NCSU cells. According to the authors, these findings, and the fact that FB₁ is a metabolite of Fusarium moniloforme, imply that FB₁ might be a cause of the immunosuppression widely observed in poultry production [Qureshi and Hagler, 1991].

2. Neurotoxicity

No data were found.

3. Biochemical Toxicology

·in vitro, rat hepatocytes and rat liver microsomes

The potential for fumonisins to inhibit de novo sphingolipid biosynthesis by disrupting the metabolism of sphingosine (base backbone of sphingolipids) was examined in cultures of hepatocytes obtained from male Sprague-Dawley rats and in rat liver microsomes. First, the effects of fumonisin B₁ and B₂ on the incorporation of [¹⁴C]serine (serene palmitoyltransferase is a biochemical intermediate to sphingosine) into [¹⁴C]sphingosine were examined with respect to time of incubation and concentration of fumonisin. Cell cultures were incubated for 2 hours with 1 µM fumonisin B₁, for 2 or 16 hours with 1 µM fumonisin B₁ and [¹⁴C] serine (25 mCi/mmol), or for 16 hours with 1 µM fumonisin B₁, followed by a 2-hour incubation with [¹⁴C] serine. Control cultures were incubated without fumonisin B₁.

Fumonisin B₁ caused almost complete inhibition of [¹⁴C]sphingosine formation by the hepatocytes. Similar inhibition occurred when [¹⁴C]serine and fumonisin B₁ were added together for 2 or 16 hours, and when the cells were incubated for 16 hours with fumonisin B₁ before the addition of [¹⁴C]serine. The IC₅₀ for this inhibition was approximately 0.1 µM for fumonisin B₁. Fumonisin B₂ produced a comparable degree of inhibition (IC₅₀ not reported). In contrast, incubation with fumonisin B₁ increased the amount of the biosynthetic intermediate sphinganine in the cultures; hepatocytes incubated with 1µM fumonisin B₁ for 4 days, had a 110-fold increase in sphinganine. According to the authors, this suggests that fumonisins inhibit the conversion of [¹⁴C]sphinganine to N-acyl-[¹⁴C]sphinganine; a step that may precede the introduction of the 4,5-trans double bond of sphingosine.

To demonstrate the inhibition at this step of the sphingosine metabolic pathway, the authors conducted in vitro assays of sphingosine N-acyltransferase (reported to acylate both sphinganine and sphingosine) using rat liver microsomes, and followed the conversion of [³H]sphingosine to [³H]ceramide by rat hepatocytes. Results from these tests showed that at 0.1 µM of fumonisin B₁ caused 50% inhibition in sphingosine N-acyltransferase activity. Also, when rat hepatocytes were incubated with 1 µM fumonisin B₁ and 1 µCi of [³H]sphingosine for one hour, the conversion of [³H]sphingosine to ceramides was significantly inhibited (P<0.05) compared that of untreated control cultures, with an IC₅₀ of 0.1 µM. The authors concluded that these results provide identification of a biochemical target for the action of fumonisins and imply that inhibition of de novo sphingolipid biosynthesis in vitro may underlie the hepatotoxicity and hepatocarcinogenicity of this mycotoxin in vivo [Wang et al., 1991].

·in vitro, frog

The effects of fumonisin on transmembrane potentials and currents of the frog (Rana esculenta) heart muscle (fine atrial trabeculae) were studied using the double sucrose gap technique. Fumonisin (280 µM) shortened the duration of the plateau and the repolarization phase of the action potential and, under voltage clamp conditions, inbited the calcium current by 44%. The block occurred without alteration of either the kinetic parameters or the apparent reversal potential of the current, suggesting that fumonisin blocked the maximal calcium conductance. Fumonisin mainly reduced the phasic component of the peak tension. The time to peak tension was unchanged, while its relaxation phase was accelerated at positive membrane potentials. This finding suggests that sodium-calcium exchange was accelerated by fumonisin. Dose-response curves for the calcium current and the peak tension indicated half inhibition at about 100 µM fumonisin and a stochiometric parameter of 1.7. The authors state that this may indicate that more than one toxin molecule interacted with the calcium channel. The authors also conclude that these data suggest that the effect of fumonisin on the calcium current and the tension might account for cardiac failure reported in cases of Fusarium moniliforme-induced animal intoxication [Sauviet et al., 1991].

4. Cytotoxicity

·in vitro, human epithelial/baby hamster kidney cells

Human epithelial (Hep-2) and baby hamster kidney (BHK) cells were used to determine the toxicity of fumonisin B₁. Cell cultures were incubated for 24 hours with 1, 2, 3, or 4 µl of fumonisin extract (2 mg) dissolved in methanol (1 ml). These volumes were added to 0.2 ml of cell culture media and resulted in fumonisin concentrations of 5, 10, 15, and 20 µg/ml. A known toxin standard (T-2) and a methanol control were also included. Cytotoxicity was evaluated by staining (Giemsa) for cell viability. The results of this assay showed that BHK cells were not sensitive to fumonisin B₁; none of the concentrations tested caused significant cell death (data and P values not reported). However, fumonisin B₁ did effect Hep-2 cells, with even the lowest concentration decreasing cell viability [La Grenade, 1990].

·in vitro, rat hepatocytes

In a study examining the cytotoxicity of aqueous and organic extracts of 10 phytopathogenic isolates of Fusarium moniliforme, the possible correlation between the toxic effects and the ability of the isolates to produce fumonisin B₁ was investigated. Cultures of rat hepatocytes obtained from male Sprague-Dawley rats were exposed for two hours to aliquots of the organic extracts (dissolved in dimethyl sulfoxide) or aqueous extracts (dissolved in water) at doses equivalent to 50, 100, or 200 mg of freeze dried culture material. The fumonisin B₁ concentrations were determined in each of the 10 culture materials; the mycotoxin was produced by 8 of the 10 isolates at concentrations ranging from 3-1090 µg/g dry weight. Cells were exposed for two hours to purified fumonisin B₁ (isolated from F. moniliforme MRC 826 culture material) at concentrations ranging from 10-7 to 10-2 M. Cytotoxicity was determined by the measurement of [³H] valine uptake and lactate dehydrogenase (LDH) release. Valine incorporation was expressed as a percentage of the values determined in control cells dosed with only solvent, and LDH release was expressed as a percentage of the release measured after the lysis of control cells.

The results of LDH analysis show that aqueous extracts of the cultures caused no hepatocyte lethality at the doses used in this study (did not cause an LDH release comparable to control cultures). However, organic extracts were more toxic at the higher dose levels (100 and 200 mg equivalents/ml), and the extract of the isolate RRC 415 was the most potent. Despite the absence of significant cell death, the uptake of valine by hepatocytes was reduced or completely blocked by aqueous extracts of the culture material when compared to that of control cultures (treated with solvent only). Organic extracts had less effect on valine incorporation, with the observed reductions coinciding with cell death. Fumonisin B₁ did not cause a release of LDH and only partially inhibited valine uptake at the highest dose tested (10-2 M). The authors point out that the aqueous extracts of cultures that produced little or no fumonisin B₁ had severe effects on the ability of hepatocytes to incorporate valine; therefore, fumonisin B₁ was not responsible for the cytotoxicity seen in these cultures [Norred et al., 1991].

· According to an abstract of a study by Shier, fumonisins B₁ and B₂ were toxic to 7/9 rat hepatoma cell lines, with approximate LD50 values for the most sensitive cell line (H4TG) of 5 and 2 µg/ml, respectively. Of 15 other mammalian cell lines examined, only MDCK dog kidney epithelial cells were sensitive to these mycotoxins; the LD50 values for both fumonisin B₁ and B₂ in this cell line were 3 µg/ml [Shier, 1990].

H. Federal Research In Progress

The U. S. Department of Agriculture (USDA) is currently sponsoring numerous on-going federal research programs on fumonisin mycotoxins. The objectives of those research programs pertaining specifically to animal and human toxicity have been summarized below. Following each summary, the performing organization, the project number, and the research contract dates have been listed [Federal Research in Progress Database, 1991].

·Occurrence of Mycotoxins and the Implications to Animal and Human Health Determine the effect of fumonisin B₁ on immunocompetent cells in vitro; determine the toxic effects of fumonisin B₁ during a 90-day subchronic exposure in rats; determine the systemic and local gut level immunologic effects of fumonisin B₁ [University of Idaho, Project No. IDA00995, July 1991-June 1994].

·Fumonisins and Other Mycotoxins Produced by Fusarium Moniliforme

Establish the toxicity of mycotoxins produced in F. moniliforme infected corn fed to swine and the residue levels of mycotoxins, especially the fumonisins in corn and animal tissues. Intermediate objectives include isolation of fumonisin B₁ and B₂ in quantities suitable for feeding studies to demonstrate the toxicity of these mycotoxins in rats and swine. Develop improved analytical methods for determination of fumonisins and other F. moniliforme mycotoxins and mycotoxin metabolites in corn and animal tissues. Survey the mycotoxins and mycotoxin metabolites in corn and animal tissues. Survey the mycotoxin profile of contaminated corn used as swine in feed [Iowa State University, Project No. IOW02955, June, 1990-June 1995].

·Chemical Isolation and Toxicologic Characterization of Fumonisins

Develop and validate analytical procedures for detection of fumonisins in grain. Characterize the potential for hepatic, neurologic, and reproductive effects of fumonisin toxicosis in horses and swine. Attempt to detect toxicokinetic characteristics of fumonisins in food animals [Iowa State University, Project Number IOWV-410-2390, January, 1990-September, 1992].

·Reduction of Mycotoxin Hazards Through Assessment of Their Toxicological Properties

Assess the toxic properties of mycotoxins from fungal species such as the Fusaria, Aspergilli and Penicillia in order to develop strategies to alleviate the hazards associated with mycotoxin contamination of crops including corn and wheat. Study the interaction of mycotoxins with critical biochemical/metabolic pathways to delineate those toxins with carcinogenic, mutagenic, or other toxic properties [Agricultural Research Service, Athens, Georgia, Project No. 6612-42000-014-000, March 1991-March 1996].

·Occurrence, Biosynthesis and Regulation of Toxic Secondary Metabolites in Fungal-Infected Plants

Determine structures of the fumonisin mycotoxins from Fusarium moniliforme, produce sufficient quantities for collaborative studies of analytical methodology and cellular toxicity, and explore methodology to inactivate the biological activity of these toxins [Northern Regional Research Center, Project No. 3620-42000-006-00D, July 1987- June 1992].

·Occurrence of Mycotoxins and the Implications to Animals and Human Health

Identify in corn, small grains, and forages new fungal metabolites and/or new mycotoxin combinations associated with deleterious biological effects on animals and characterize the fungi, and environmental conditions affecting their production, with special emphasis on the fumonisins. Develop new methods and improve existing methods for detection, identification, and quantification of specific mycotoxins and their metabolites in feeds and food. Improve methods for controlling formation and detoxification of mycotoxins in feeds and goods and develop recommendations for utilization or disposal of contaminated feeds [Georgia Coastal Plain Experimental Station, Tifton Georgia, Project No., GE000122, January, 1990-March, 1995].

·Molds and Mycotoxin Hazards in Foods, Feeds, and the Environment

Study the incidence of Fusarium graminearum and Fusarium moniliforme and their toxins in grain and grain products; develop needed methods for detection, enumeration, and quantification of Fusarium species and their toxins in grains and grain products. Evaluate the potential toxicity of Fusarium species isolated from grains and grain products; study the fate of Fusarium molds and mycotoxins in several types of food processes and processes food products; study the levels of molds and mycotoxins in grain dusts; evaluate mold and mycotoxin problems in foods and feeds as they arise due to mold invasion [University of Nebraska, Project No. NEB-16-056, October, 1990-September, 1995].

Return to Table of Contents page