An overview of rodent toxicities: liver and kidney effects of fumonisins and Fusarium moniliforme.

Fumonisins are produced by Fusarium moniliforme F. verticillioides) and other Fusarium that grow on corn worldwide. They cause fatal toxicoses of horses and swine. Their effects in humans are unclear, but epidemiologic evidence suggests that consumption of fumonisin-contaminated corn contributes to human esophageal cancer in southern Africa and China. Much has been learned from rodent studies about fumonisin B1(FB1), the most common homologue. FB1 is poorly absorbed and rapidly eliminated in feces. Minor amounts are retained in liver and kidneys. Unlike other mycotoxins, fumonisins cause the same liver cancer promotion and subchronic (studies (3/4) 90 days) liver and kidney effects as (italic)F. moniliforme. FB 1 induces apoptosis of hepatocytes and of proximal tubule epithelial cells. More advanced lesions in both organs are characterized by simultaneous cell loss (apoptosis and necrosis) and proliferation (mitosis). Microscopic and other findings suggest that an imbalance between cell loss and replacement develops, a condition favorable for carcinogenesis. On the molecular level, fumonisins inhibit ceramide synthase, and disrupt sphingolipid metabolism and, theoretically, sphingolipid-mediated regulatory processes that influence apoptosis and mitosis. Liver sphingolipid effects and toxicity are correlated, and ceramide synthase inhibition occurs in liver and kidney at doses below their respective no-observed-effect levels. FB1 does not cross the placenta and is not teratogenic in vivoin rats, mice, or rabbits, but is embryotoxic at high, maternally toxic doses. These data have contributed to preliminary risk evaluation and to protocol development for carcinogenicity and chronic toxicity studies of FB1 in rats and mice.

The impact of fumonisins on human health remains unclear but is of concern. Consumption of F. moniliforme-molded, home-grown corn has been correlated with high esophageal cancer rates in areas of southern Africa (19)(20)(21) and central China (22), and comparatively high fumonisin concentrations are found in the corn from these high esophageal cancer areas (4,(23)(24)(25)(26)(27)(28). It has also been suggested that fumonisins are a risk factor for liver cancer (29), and FB 1 , like some F. moniliforme isolates (30,31), was hepatocarcinogenic when fed (50 ppm) to male BD IX rats (32). Studies are limited, but hepatotoxicity and atherogenic serum lipid profiles, perhaps secondary to liver dysfunction, were found in nonhuman primates fed diets containing fumonisin (13,33,34). Finally, a possible link between fumonisin exposure and neural tube defects in humans has been proposed (35).
In vivo investigations using rodents and rabbits have contributed significantly to F. moniliforme and fumonisin research. Fumonisins were discovered using an in vivo liver bioassay (5). Data from subchronic toxicity studies have been used in preliminary risk evaluations (36,37), have been useful for developing protocols for chronic studies, and have otherwise increased our understanding of these compounds. An overview of toxicity and other important data obtained during in vivo investigations follows.

Absorption, Biodistribution, and Pharmacokinetics
Fumonisins are poorly absorbed and rapidly eliminated; small amounts accumulate in liver and kidneys (Table 1). Norred et al. (38) recovered 80% of the radiolabel from feces within 48 hr and ≤ 3% from urine within 96 hr after giving a single oral dose of [ 14 C]FB 1 (0.045 µCi) to rats. Small but relatively constant amounts of radiolabel were found in liver (about 0.4% of the dose) and kidney (about 0.1% of the dose) up to 96 hr postdosing. After administering the same dose for 3 consecutive days to rats, more than 75% of the [ 14 C] was excreted in feces and about 4% in urine within 72 hr of the last dose. Liver-and kidney-specific activities peaked 24 hr after the last dose, but persisted for another 48 hr. Like FB 1 , fumonisin B 2 (FB 2 ) is also rapidly cleared from plasma and excreted (82% within 72 hr, mostly during the first 24 hr), predominantly in the feces (39). Only about 1% of the dose was recovered in the urine.
Liver and kidney accumulated relatively high amounts of [ 14 C]FB 1 following intraperitoneal (ip) or intravenous (iv) dosing to rats (Table 1). Up to 66% of the radiolabel appeared in feces, suggesting that FB 1 (or possibly a metabolite) is excreted in bile. This was confirmed by Shephard et al. (40) who, within 4 hr of giving 7.5 mg/kg body weight (bw) [ 14 C]FB 1 ip to cannulated rats, recovered about 67% of the dose as unchanged FB 1 in the bile. In contrast, only 0.2% of the radiolabel was recovered in bile following oral administration of the same dose (7.5 mg/kg bw [ 14 C]FB 1 ), further suggesting that gastrointestinal absorption of FB 1 by rats is low.
Little pharmacokinetic data is available. Shephard et al. calculated time of plasma maximum concentration (T max ) of about 20 min, peak plasma concentration (C max ) of 8.6 µg/mL, and a serum elimination half-life Fumonisins are produced by Fusarium moniliforme (= F. verticillioides) and other Fusarium that grow on corn worldwide. They cause fatal toxicoses of horses and swine. Their effects in humans are unclear, but epidemiologic evidence suggests that consumption of fumonisin-contaminated corn contributes to human esophageal cancer in southern Africa and China. Much has been learned from rodent studies about fumonisin B 1 (FB 1 ), the most common homologue. FB 1 is poorly absorbed and rapidly eliminated in feces. Minor amounts are retained in liver and kidneys. Unlike other mycotoxins, fumonisins cause the same liver cancer promotion and subchronic (studies ≤ 90 days) liver and kidney effects as F. moniliforme. FB 1 induces apoptosis of hepatocytes and of proximal tubule epithelial cells. More advanced lesions in both organs are characterized by simultaneous cell loss (apoptosis and necrosis) and proliferation (mitosis). Microscopic and other findings suggest that an imbalance between cell loss and replacement develops, a condition favorable for carcinogenesis. On the molecular level, fumonisins inhibit ceramide synthase, and disrupt sphingolipid metabolism and, theoretically, sphingolipid-mediated regulatory processes that influence apoptosis and mitosis. Liver sphingolipid effects and toxicity are correlated, and ceramide synthase inhibition occurs in liver and kidney at doses below their respective no-observed-effect levels. FB 1 does not cross the placenta and is not teratogenic in vivo in rats, mice, or rabbits, but is embryotoxic at high, maternally toxic doses. These data have contributed to preliminary risk evaluation and to protocol development for carcinogenicity and chronic toxicity studies of FB 1 in rats and mice.  (41). T max , C max , and T 1/2 of FB 2 are similar ( Table 1). There have been no reports of fumonisin metabolism by the liver, kidney, or other tissues, although the intestinal flora of the rat, like that of the nonhuman primate (42), hydrolytically removes the tricarballylic acid groups from fumonisin's hydrocarbon backbone (43).

Fumonisins, F. moniliforme, and in Vivo Toxicity
Fumonisins are fungal products. Therefore, the relationship between fumonisin toxicity and toxicity of the fungi should be kept in perspective. Not all fungi identified as F. moniliforme are toxic. For example, culture material (0, 4, 8, or 16% w/w in the diet) of isolate MRC 826 caused significant dose-related toxicity and liver pathology (kidney was not examined) when fed to rats for 4 weeks, whereas isolate RRC 415 culture material was without effect (44). Similarly, only 3 of 11 F. moniliforme isolates induced γ-glutamyl transpeptidase (GGT)-positive liver foci in rats (45). MRC 826 culture material (0.5% in the diet), but not a 10-fold higher dietary concentration of isolate MRC 1069 culture material, caused hepatocarcinomas in rats (30). It has been shown that MRC 826 is a fumonisin producer (46), but that both MRC 1069 (30) and RRC 415 (47) produce predominantly fusarin C.

Hepatotoxicity
Histopathologic effects in rats, which have been referred to by various terms such as hepatopathy, hepatosis, or toxic hepatitis, have been reported by several research groups. Their descriptions are consistent, differing only in detail and nomenclature (5,56-58, 61,65-67). The initial finding is small, rounded, eosinophilic hepatocytes that appear to have pulled away from neighboring cells. The chromatin of these cells is irregularly condensed and marginated or may be fragmented. Inflammatory response is absent to minimal. Although their appearance is consistent with apoptosis, these cells were commonly described as single-cell necrosis until their apoptotic nature was histochemically confirmed by Tolleson et al. (61) and Howard et al. (68,69). This does not mean, however, that necrosis (oncotic necrosis in the traditional sense, as opposed to programmed cell death or apoptosis) does not play a role in fumonisin-induced hepatotoxicity. Necrotic hepatocytes are also present early-on. Serum chemical indications of hepaocellular injury, including increased alanine and aspartate transaminase, alkaline phosphatase, and lactate dehydrogenase activities, as well as increased cholesterol and triglyceride concentrations ( Figure 2), are routine, early findings. As tissue injury progresses, both apoptotic and necrotic cells increase in number, mitotic figures appear with increasing frequency, hepatocellular cytoplasm becomes increasingly vacuolated, and cytomegaly with variability in cell and nuclear size becomes obvious. Bile duct and oval cell proliferation, foci of cellular alteration, cholangiomatous lesions, and fibrosis occur in long-standing or advanced lesions, giving a picture of nodular regeneration or cirrhosis (32). Females are generally more sensitive than males (Table 2). Along with the aformentioned clinical chemical indicators, which continue to rise, serum GGT activity and bilirubin concentration increase as liver injury becomes more severe.
Fumonisins induced both GGT and the placental form of glutathione S-transferase (GSTP)-positive foci in BD IX and Fischer rats (5,62,70,71). Foci induction in diethylnitrosamine-pretreated Fischer rats was dose related, and GSTP was a more sensitive marker than GGT ( Figure 3) (70). From these and other data, Gelderblom et al. (72) proposed that FB 1 was a tumor promoter at doses not causing significant liver pathology, but when given at overtly hepatotoxic doses, it was also a weak initiator. FB 1 (50 ppm) caused marked nonneoplastic changes (cirrhosis) and hepatocellular carcinomas when fed to BD IX males for 20 months or more (32). Relatively   (73), showed that FB 1 was a kidney carcinogen in rats and a liver carcinogen in mice, respectively.

Nephrotoxicity
The kidney was the most sensitive target organ in Sprague-Dawley and Fischer 344 rats fed FB 1 for up to 90 days (Table 2) (57,58,61,74) or given FB 1 by gavage or ip injection for 4-11 days (65-67,75). Males were more sensitive than females. In contrast, Gelderblom et al. (5,32) described hydropic degeneration, occasional necrosis, and a few other renal abnormalities in their studies in BD IX rats, but did not refer to the kidney as a target organ. This suggests that significant differences in response to fumonisins may exist among various rat strains. As in liver, apoptosis is the initial microscopic finding in kidney. Apoptotic cells are initially found almost exclusively in tubules of the outer medulla (designated "corticomedullary junction" in some publications). Many of the apoptotic cells appear rounded and detached from adjacent cells and the basement membrane. Mitotic figures appear, and the number of apoptotic cells increases in the tubule epithelium as injury progresses. Cytoplasmic vacuolation and basophilia, decreased cellular height, and alterations in nuclear size and staining become evident. At this point, lesions may extend deeper into the medulla or into the cortex, and epithelial cells are sloughed into the tubular lumina. Thus, there is simultaneous cell loss and replacement revealed on three levels: on the cellular level by apoptosis and mitosis, on the histologic level by tubular atrophy and hyperplasia (regeneration), and grossly by decreased kidney weight ( Figure 4). The failure of regeneration to keep pace with cell loss may be quite important, as imbalances between cell loss and replacement in tissues may be a significant contributor to carcinogenesis (76)(77)(78).
Increased serum creatinine, decreased CO 2 and, less often, increased urea nitrogen occur in FB 1 -treated rats (57,58,79). Increased activities of N-acetyl-β-glucosaminidase, GGT, and lactate dehydrogenase have been found in urine, giving further clinical evidence of tubular injury (65,75). Proteinuria may also occur and, because of the presence of high-molecular-weight proteins in the urine, it has been suggested that FB 1 may also cause glomerular injury (75). Alternatively, sloughed tubule cells, which can be numerous in urine, are a more likely cause of proteinuria (80).
There is little data on renal function in fumonisin-treated animals. Urinary output and water consumption were increased in male rats fed F. moniliforme culture material (64). These observations were consistent with those of Bondy et al. (65,67) and Suzuki et al. (75), who studied renal function in rats given daily ip doses (7.5-10 mg/kg for 4 days) or oral doses (1-75 mg/kg for 11 days) of FB 1 . Signs of renal dysfunction in these animals included increased output of hypoosmotic urine, increased urinary enzyme levels, proteinuria, increased serum Mg 2+ and Ca 2+ concentrations, and decreased anion (p-aminohippurate, up to 80% reduction) and cation (tetraethylammonium, up to 40% reduction) transport by renal cortex slices. Interestingly, some indications of renal dysfunction peaked on days 6-8 of the 11-day exposure period (67), suggesting that fumonisin-damaged kidneys have some functional adaptive capacity.

Other in Vivo Toxicologic Findings in Rats
There are indications that the immune system may be a target. Bondy et al. (66) found disseminated thymic necrosis with decreased thymic weight and increased serum IgM concentrations in FB 1 -exposed rats. Others found that the immune responses to sheep red blood cells and to splenic clearance of Listeria monocytogenes were slightly decreased in rats given 15 (L. monocytogenes) or 25 (sheep red blood cells) mg/kg bw FB 1 for 14 days (81).
Testicular tubule epithelial degeneration (59), decreased heart weight (54), adrenal cortex hypertrophy, and cytoplasmic vacuolation (consistent with Zona fasciculata lipoidosis and probably a nonspecific stress response) (64), cytoplasmic vacuolation of myeloid precursor cells in bone marrow, and other hematologic findings (66) have been found in rats given fumonisins or F. moniliforme culture materials. Little toxicologic importance has yet been given to any of these observations.
Because of the human health implications, esophageal effects of fumonisins are of special interest. Marasas et al. (31) found basal cell hyperplasia of the esophagus in rats fed F. moniliforme culture material. Others have noted a transient increase in the 5-bromo-2´deoxyuridine labeling index in esophageal epithelium 3 days following an iv injection of 1.25 mg/kg bw (82), which suggests that FB 1 may be mitogenic under some conditions. However, there is no evidence from subchronic (5,57,58,61) or chronic (32,73) feeding studies of purified FB 1 that the esophagus is a target organ. Furthermore, FB 1 (5 mg/kg bw/day for 5 weeks) had no effect on the Environmental Health Perspectives • VOLUME 109 | SUPPLEMENT 2 | May 2001 261  number of esophageal papillomas produced in rats concurrently given the esophageal carcinogen N-methylbenzylnitrosamine (83). Thus, there is no evidence from rodent studies that FB 1 is an esophageal carcinogen, and the possibility that F. moniliforme produces other potentially (esophageal) carcinogenic compounds should not be dismissed.

Hepatic and Renal Toxicity in Other Laboratory Species
Liver and kidney are also targets in mice (74,84,85). The pathology is similar to that seen in rats, and females are more sensitive to hepatotoxicity than males. Apoptosis, hepatocellular hyperplasia, bile canaliculi hyperplasia, and Kupffer cell hyperplasia were the principal findings in females (B6C3F 1 ) fed ≥ 99 ppm FB 1 and males fed 484 ppm FB 1 for 4 weeks (68,73). A hepatocellular cytoplasmic alteration described as reduced cytoplasm, basophilia, and loss of cytoplasmic vacuoles was found in both sexes fed ≥ 99 ppm. Histopathology findings (hepatopathy) in female B6C3F 1 mice fed 81 ppm FB 1 for 90 days (57) were characterized by apoptosis (= single-cell necrosis), cytomegaly, increased mitotic figures, scant inflammatory infiltrates, and pigmented macrophages. Serum chemical indications of hepatic injury of the same type found in rats also occurred. Compared to rats, mice are resistant to nephrotoxicity. No kidney lesions were found in mice fed diets with 81 ppm FB 1 for 90 days (57) or diets with 484 ppm FB 1 for 28 days (68). However, when given at relatively high doses by oral or parenteral routes, FB 1 is nephrotoxic, as illustrated by the findings of Sharma et al. (84). They found terminal deoxynucleotidyl transferase-mediated nickend labeling (TUNEL)-positive cells in the renal tubules of males (females were not examined) given subcutaneous injections of FB 1 for 5 consecutive days ( Figure 5). Findings were corroborated by the dose-related increase in apoptotic cells, which was found during routine microscopic examinations (hematoxylin and eosin sections). Bondy et al. (85) found slight increases in single-cell necrosis (= apoptosis) in the renal tubules of female but not male B6C3F 1 mice given 15-75 mg/kg FB 1 by gavage for 14 days.
Rabbits are quite sensitive to FB 1 (86). As in male rats, the kidney is a more sensitive target organ than liver. Morphologic, serum chemical, and tissue sphingolipid findings were similar to those seen in rodents.

Fumonisins and Sphingolipids: Mechanistic Considerations
The preponderance of experimental evidence, including evaluations of unscheduled DNA synthesis, Salmonella typhimurium mutagenicity, SOS response in Escherichia coli, mitotic index, and micronucleus formation [reviewed by Howard (87)], suggests that fumonisins exert their effects through a nongenotoxic mode of action. A significant breakthrough in understanding these compounds occurred when Wang et al. (88) discovered that fumonisins inhibit the enzyme ceramide synthase [sphinganine (sphingosine)-N-acetyltransferase], leading to disruption of de novo sphingolipid biosynthesis. The immediate consequences thereof are accumulation of the sphingoid bases sphinganine (Sa) and sphingosine (So), an increase in the Sa to So ratio (Sa/So), and depletion of complex sphingolipids (CSLs) in tissues ( Figure 6). Recognizing the importance of sphingolipids in cell regulatory processes, including those related to proliferation and apoptosis (76,(89)(90)(91), investigators have proposed that ceramide synthase inhibition is the critical mechanistic step in fumonisin toxicity, starting a cascade of molecular events eventually leading to cytotoxicity or neoplasia.
Fumonisin exposure, sphingolipid effects, and toxicity are correlated in vivo. Liver and kidney Sa, So, and Sa/So were increased in rats fed 15-150 ppm FB 1 for 4 weeks ( Figure  7) (80), and increases occurred at doses equal to or less than those causing microscopic lesions (apoptosis). Importantly, liver and kidney Sa/So increases were correlated with the severity of hepatopathy and nephropathy in rats fed F. moniliforme culture material (71 ppm FB 1 ), water-extracted culture material (11 ppm FB 1 ), or an alkali-treated (nixtamalized) culture material containing 58 ppm HFB 1 but no measureable FB 1 (92,93).
The role of sphingolipids as mediators of fumonisin toxicity has not yet been proven, and other mechanisms may come into play.
For example, several groups have presented data suggesting that fumonisins cause compositional or oxidative damage to cellular lipids, which in turn causes molecular events culminating in oxidative damage to DNA and other critical macromolecules (94)(95)(96)(97)(98)(99). A more detailed discussion of molecular mechanism is beyond the purpose of this review but can be found elsewhere in this issue (87,(100)(101)(102).

Reproduction and Teratology Studies
Javed et al. (103) found that FB 1 was embryotoxic and caused malformations when injected into chicken eggs. FB 1 and HFB 1 also inhibited growth of rat embryos exposed in vitro on gestation day (GD) 9.5 (104). Under similar conditions, 100 or 300 µM HFB 1 caused neural tube and other malformations (105). Although useful as screens, in vitro methods allow direct fetal exposure and avoid maternal gastrointestinal absorption, pharmacokinetics, placental transfer, and other potential barriers of in utero exposure.
To assess the reproductive effects of F. moniliforme, male and female Sprague-Dawley rats were fed culture material of isolate MRC 826 providing 0, 1, 10 or 55 ppm FB 1 (106). The culture material was minimally toxic to males (≥ 10 ppm) and females   (55 ppm), as indicated by serum chemistry findings and kidney pathology. All reproductive end points in males, which included testicular morphology, sperm morphology, and sperm motility, were unaffected. One half of the mated females from each group were examined on GD 15, at which time no differences among groups were found in the number of corpora lutea, implantation sites, resorptions, dead fetuses, and live fetuses per dam. The remaining females gave birth and were observed, along with their litters, for 21 days postpartum. Other than a slight decrease in weight gain (but not absolute weight) of litters from groups given 10 and 55 ppm FB 1 , no differences were found in maternal reproductive or offspring development variables. Although liver Sa/So of the dams fed 55 ppm FB 1 was significantly increased, no differences in the Sa/So of control and high-dose (55 ppm) fetuses were found on GD 15 (abdominal slices containing liver and kidney), indicating that fumonisins did not cross the placenta. This was corroborated in a second study in which no radiolabel (< 0.02% of the dose) was found in the fetuses following iv injection of [ 14 C]FB 1 to pregnant females on GD 15 (106). Pregnant rats were given 1.875-15 mg/kg bw purified FB 1 on GD 3-16 (107) and examined on GD 17 and GD 20. FB 1 had no effect on maternal reproductive variables. The high dose (15 mg/kg FB 1 ) was maternally toxic, causing decreased weight gain on GD 17, decreased kidney weights, and increased Sa/So of liver, kidney, and serum. Sa/So was also increased in livers of dams given 7.5 mg/kg, kidneys of dams given ≥ 1.875 mg/kg, and serum of dams given 7.5 mg/kg FB 1 . Decreased length and weight of female fetuses were noted on GD 20. Otherwise, there was no evidence of fetal toxicity or teratogenicity. The study was repeated at doses ranging from 6.25 to 50 mg/kg FB 1 (108). Apoptosis and other microscopic findings typical of fumonisins were found in kidney (≥ 6.25 mg/kg) and liver (25 or 50 mg/kg). Increased Sa/So was found in maternal liver (≥ 25 mg/kg), kidney (≥ 6.25 mg/kg), and serum (≥ 25 mg/kg). Fetal deaths were increased at 25 and 50 mg/kg, and the number of viable fetuses/dam (12.0 ± 1.1 vs control value of 14.0 ± 0.5), fetal length, and fetal weight were decreased at 50 mg/kg. The incidence of hydrocephalic fetuses and skeletal anomalies such as wavy ribs and reduced ossification was increased somewhat at 50 mg/kg, but no teratogenic effects were found. Sa/So of fetal tissues were unchanged at any of the doses studied, suggesting that the fetal effects were indirect and secondary to maternal toxicity.
Results of other developmental toxicity studies (Table 3) generally agree and likewise suggest that FB 1 is not teratogenic but may be embryotoxic at maternally toxic doses (109,110). In contrast, Floss et al. (111,112) concluded that FB 1 was a developmental toxin in hamsters at dosages that were not maternally toxic. It is possible that there are species-related differences in maternal response. However, detailed serum chemical, histopathologic, or fetal and maternal tissue sphingolipid evaluations, which may have revealed maternal toxicity in the hamsters, were not undertaken. Collins et al. (107,108), LaBorde et al. (110), and Voss et al. (106) have shown significant organ weight, pathology, and sphingolipid effects in dams that otherwise appear unaffected by FB 1 .

Applied Studies Using Culture Materials
FB 1 does not occur alone. Co-exposure undoubtedly occurs with other mycotoxins and mycotoxin products formed during grain handling or food preparation. As illustrated below, F. moniliforme culture material can be a useful, cost-effective tool for studying how coexposure to other mycotoxins or mycotoxin products influences FB 1 toxicity. However, such experiments should be carefully designed and results interpreted with caution. Not all F. moniliforme strains are toxic, and culture materials are complex mixtures containing biologically active compounds (both known and unknown) that may confound results.
To determine the potential in vivo toxicity of FB 2 and FB 3 , Voss et al. (64) studied three genetically related F. moniliforme isolates. Isolate M3125 produced FB 1 , FB 2 , and FB 3 in the approximate ratio of 1:0.35:0.15. Isolate 107-R-7 produced FB 2 (no detectable FB 1 or FB 3 ), and isolate 397-R-74 produced FB 3 (no FB 1 or FB 2 ). Low (4.6-6.9 ppm), mid (32-53 ppm), and high (219-303 ppm) levels of culture materials of each isolate were fed to rats for 3 weeks. All were toxic. Their effects were qualitatively indistinguishable, consisting of decreased weight gain, decreased kidney weights, increased serum chemical indications of hepatotoxicity, increased Sa/So, and apoptosis in the liver and kidneys. All findings were consistent with the effects of FB 1 (57,58), and elevated tissue Sa/So was correlated with various toxicologic end points. Thus, hepato-and nephrotoxicities can be induced by FB 1 nonproducing fungi, and toxicity studies of purified FB 2 and FB 3 are warranted.
Environmental Health Perspectives • VOLUME 109 | SUPPLEMENT 2 | May 2001 It has been suggested that fusaric acid, another mycotoxin commonly produced by F. moniliforme (51,113), exacerbates fumonisin toxicity (114,115). Bacon et al. (116) demonstrated synergistic embryotoxicity by simultaneous injection of FB 1 and fusaric acid in ovo. Diets containing F. moniliforme MRC 826 culture material providing low (3.4 ppm), slightly higher (18 ppm), or very high (437 ppm) amounts of FB 1 and, at each fumonisin level, 0, 20, 100, or 400 ppm fusaric acid were fed to rats for 4 weeks (54). Dose-related body weight, serum chemical, liver and kidney pathologies, and renal sphingolipid effects typical of fumonisins were caused by the culture material. No evidence of synergism was found. Fusaric acid alone up to 400 ppm in the diet was not toxic, and its presence did not modify the response of the animals to the culture material.
Masa flour is made from nixtamalized corn. During nixtamalization, corn is boiled under alkaline conditions sufficient to convert fumonisins to their hydrolyzed forms (117,118). A study by Hendrich et al. (63) showed that nixtamalization did not reduce hepatotoxicity or cancer-promoting activity of F. proliferatum culture material, even though FB 1 and FB 2 were converted to HFB 1 and HFB 2 . However, others reported that although cytotoxic in vitro, purified HFB 1 had no effect in vivo, and they proposed that it was not gastrointestinally absorbed (62). To further study nixtamalization and in vivo toxicity, Voss (92) fed rats F. moniliforme culture material providing 71 ppm FB 1 , water-extracted culture material providing about 11 ppm FB 1 , or a nixtamalized culture material providing 58 ppm HFB 1 , but no measurable FB 1 . After 4 weeks the culture material and the nixtamalized culture material caused the hepatic and renal lesions typical of fumonisins, though the nixtamalized material was somewhat less potent. The water-extracted culture material elicited a noticeably lesser nephrotoxic response and was not hepatotoxic. Sa and Sa/So increases in liver and kidney were increased in all three groups, and the increases were correlated with the severity of liver and kidney injury (93). These results agree with the in vitro findings of Norred et al. (119), who reported that HFB 1 inhibited ceramide synthase in precision-cut rat liver slices, but less potently than FB 1 . The consequences of chronic HFB 1 exposure remain unknown, and, given the popularity of masa-based food products, additional investigations on its occurrence in foods and its toxicity are needed.

Summary and Conclusions
In vivo studies of F. moniliforme and fumonisins in rodents have shown that FB 1 and probably FB 2 , FB 3 , and HFB 1 cause the toxic and pathologic effects of F. moniliforme. These studies have provided other important data including the following: a) Gastrointestinal absorption is low, absorbed fumonisins are rapidly eliminated, and only minor amounts are retained in liver and kidney. b) Liver and kidneys are the two major target organs, although differences in response occur between sexes, strains, and species. c) Fumonisins may have other, more subtle organ-specific effects; however, there is no compelling evidence that the esophagus is a target organ. d ) Apoptosis is the initial and presumably critical event in the pathogenesis of liver and kidney lesions characterized by simultaneous cell loss and regeneration. e) A key molecular event in fumonisin cytotoxicity is inhibition of ceramide synthase, leading to disruption of sphingolipid metabolism and probably of sphingolipid regulatory function. f ) FB 1 does not cross the placenta and is not teratogenic in laboratory species; however, fumonisins may be embryotoxic at maternally toxic doses. g) The use of culture materials provides a cost-effective means of studying F. moniliforme and fumonisins, as long as caution is exercised in study design and data interpretation. The relationships between F. moniliforme, fumonisins, and human health remain unresolved. Undoubtedly, in vivo investigations in rodents will continue to provide insight into the effects and modes of action of these important mycotoxins.