Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Fumonisins are mycotoxins produced by Fusarium moniliforme and related fungi that are common contaminants of corn, sorghum and related grains throughout the world (Marasas, 1994). Fumonisin B₁, the most abundant fumonisin, is nephrotoxic and hepatotoxic in many animals (Bondy et al., 1995; Voss et al., 1996), and it is hepatocarcinogenic in rats (Gelderblom et al., 1992). Recent studies linked fumonisins to high incidences of human esophageal cancer in the Transkei region of southern Africa (Sydenham et al., 1990). Fumonisins have been implicated as etiologic agents in the production of ELEM and PPE. Fumonisin B₁ alters cell morphology, cell-cell interactions, the behavior of cell surface proteins, protein kinases, metabolism of other lipids, cell growth, regeneration and viability (Merrill et al., 1996). Induction of apoptosis after fumonisin B₁ treatment has been demonstrated in cultured cells and animal tissues (Tolleson et al., 1996a, 1996b; Wang et al., 1996).

The mechanism of fumonisin B₁ toxicity is not fully understood. Fumonisin B₁ resembles the sphingoid base backbone of sphingolipids, a class of membrane lipids that play an important role in cell signal transduction pathways, cell growth, differentiation and cell death (Merrill et al., 1997). Fumonisin B₁ and other members of the fumonisin family are potent, competitive inhibitors of ceramide synthase, the enzyme that catalyzes the acylation of sphinganine in the de novo biosynthesis of sphingolipids and the reutilization of sphingosine derived from sphingolipid turnover (Wang et al., 1991; Merrill et al., 1993). Fumonisin-induced depletion of complex sphingolipids and accumulation of free sphingoid bases and their metabolites can, at least partially, account for the fumonisin B₁ effects on cell proliferation, differentiation and increased cell death (Harel and Futerman, 1993; Schroeder et al., 1994; Yoo et al., 1996).

Few studies have reported fumonisin B₁ effects on the immune response. Fumonisin reduced phagocytic activity in chicken macrophages and decreased the expression of CD3 receptors in mouse thymocytes in vivo and in vitro (Chatterjee and Mukherjee, 1994; Martinova et al., 1995; Qureshi et al., 1995). No significant effects on the chicken natural killer cell activity were observed (Qureshi et al., 1995). Varied results of suppression and stimulation of T-dependent antibody production in mice were observed when mice were exposed to a single or continuous fumonisin B₁ treatment, respectively, (Martinova and Merrill, 1995).

Cytokines are biological response modifiers produced by macrophages, T lymphocytes and other cell types when activated by specific antigens or other stimuli. They are crucial in the development of immune response, cell proliferation, differentiation, hematopoiesis and inflammation (Akira et al., 1990; Paul and Sedar, 1994). Sphingomyelin has emerged as a key participant in the signaling pathway of cytokines such as TNF-, IL-1 and IFN (for a review, see Merrill et al., 1997). The present study was based on the hypothesis that the fumonisin disruption of sphingolipid biosynthesis and turnover will alter lipid second messengers, resulting in altered cytokine production, and may be a factor in fumonisin B₁ toxicity.

In the current study, mRNA and protein levels for key cytokines were determined in peritoneal macrophages (IL-1 and TNF-) and splenic lymphocytes (IFN) after exposure of BALB/c mice to various doses (0-6.75 mg/kg/day subcutaneous for 5 days) of fumonisin B₁ in vivo. Selective induction of TNF- without an effect on other cytokines was observed. Pretreatment with anti-TNF- antibody with an acute exposure to the fumonisin B₁ (25 mg/kg subcutaneous) reversed the observed hematological effects of the mycotoxin. TNF- production also was observed when J774A.1 cells were cultured in the presence of fumonisin B₁. These results support the hypothesis that fumonisin B₁-increased TNF- expression is an important contributor to the fumonisin B₁ toxicity and increased apoptosis observed in liver and kidney of mice treated with fumonisin B₁ (Sharma et al., 1997).

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Male BALB/c mice, 5 weeks old, were obtained from Charles River Laboratories (Wilmington, MA). Mice were acclimatized for 7 days in a temperature-controlled (22 ± 1°C, 50% relative humidity) and artificially illuminated room (12-hr light cycle). Pelleted feed (fumonisin-free)³ and fresh water were provided ad libitum. Food and water consumption were recorded daily.

Repeated fumonisin B₁ exposure. Fumonisin B₁ of 98+% purity (Sigma Chemical, St. Louis, MO) was dissolved in sterile phosphate-buffered saline just before administration. Fumonisin B₁ was tested for possible contamination with endotoxin, and no detectable level of endotoxin was found.⁴ Two sets of mice were separated into five groups and were injected subcutaneously with 0, 0.25, 0.75, 2.25 and 6.75 mg of fumonisin B₁/kg b.w./respectively, daily for 5 days. Body weights were recorded daily throughout the experimental period. Mice were killed with halothane 24 hr after the final dose. From one set of animals, the liver, thymus, kidney, brain and spleen were removed and weighed, and the blood was collected in heparinized tubes for red and white blood cell counts using an automated counter (Coulter, Hialeah, FL). Parts of the liver and kidney tissues were fixed in 10% neutral buffered formalin for 4 hr. The spleens were removed aseptically and used as a source of splenocytes. Macrophages were isolated from the second set of animals by intraperitoneal lavage (see below).

In situ TUNEL staining of the apoptotic bodies. To determine the presence of apoptosis in situ, fixed liver and kidney tissues were embedded in paraffin, and 5-µm sections were made and used for apoptosis analysis. Terminal UTP nucleotide transferase end-labeling was performed using ApopTag kit (Oncor, Gaithersburg, MD).

Isolation and activation of splenic lymphocytes. The aseptically collected spleens were maintained in cold complete RPMI medium (RPMI-1640; GIBCO, Grand Island, NY), containing 2 mM glutamine, 100 units/ml penicillin, 100 µg/ml streptomycin, 50 µM 2-mercaptoethanol and 10% heat-inactivated FBS (Life Technologies, Grand Island, NY). Monocellular cell suspensions were prepared using a Stomacher lab blender (STOM 80; Tekmar, Cincinnati, OH). Red blood cells were removed with ACK lysing buffer (0.15 M NH₄Cl, 1 mM KHCO₃, 0.1 mM Na₂EDTA, pH 7.4). The remaining cells were plated in 100-mm culture plates and incubated at 37°C for 1 hr to allow adherence of macrophages. The nonadherent cells were collected, counted with an automatic blood cell counter, adjusted to a concentration of 2 × 10⁶/ml, activated with 5 µg/ml ConA (Sigma) in complete RPMI and replated at 15 to 20 ml/plate in 100-mm Falcon culture dishes for 12 or 48 hr.

Isolation and activation of macrophages. For macrophages, animals were injected intraperitoneally with 3 ml of 3% BTG (GIBCO) on the fourth day before blood samples were taken. The fumonisin B₁ treatment continued as in other experiments. Four days after the BTG injection and 1 day after the last fumonisin B₁ treatment, peritoneal lavage was performed with 5 ml of ice-cold complete RPMI plus 10 units/ml heparin. Macrophages were counted with a hemocytometer and adjusted to 10⁶ cells/ml in complete RPMI medium, and 5 ml of the cell suspension was plated onto 60-mm dishes. To activate cells, LPS (10 µg/ml; Sigma) was added to each plate and cultured for 6 or 24 hr. These periods were optimized in prior experiments (Dugyala and Sharma, 1996).

Total RNA isolation and analysis. Total RNA was extracted from lymphocytes after 12-hr activation with Con A and from macrophages after 6-hr activation with LPS. The lymphocytes were harvested by centrifugation, and total RNA was extracted from the pellet with TRI reagent LS (Molecular Research Center, Cincinnati, OH) according to the manufacturer's protocol. RNA from macrophages was extracted by removing the medium and applying TRI reagent to the adherent cells on the plate according to the manufacturer's instructions. The purified RNAs were suspended in formazol (Molecular Research Center) and stored at 20°C.

Cytokine quantification. Supernatants from macrophage cultures activated for 24 hr were used to quantify IL-1 and TNF-. Supernatants from splenocyte cultures activated for 48 hr were used to quantify IFN. All supernatants were frozen and maintained at 70°C until assayed. All cytokines were quantified by ELISA using a commercial kit (Genzyme Diagnostics, Cambridge, MA) according to the manufacturer's instructions.

Assay for measuring cytotoxic activity. WEHI-164 (fibrosarcoma, mouse) cells⁵ were cultured in complete RPMI-1640 with 10% FBS. Cell line was passed twice weekly, and cells in their mid log phase were used in the experiments after assessment of the cell viability using trypan blue exclusion. The MTT (Sigma) cytotoxicity assay (Mosmann, 1983) was used to measure percentage of dead cells. WEHI-164 cells were seeded at a concentration of 7.5 × 10⁴ cells/well in 3072 Falcon 96-well microplates (Becton Dickinson, Franklin Lakes, NJ) after sensitizing with actinomycin D (1 µg/ml). Appropriate dilutions of the supernatants from the samples were made in RPMI-1640 and added to the cells. After 18 hr of incubation at 37°C in a 5% CO₂ chamber, 20 µl of MTT (5 mg/ml) per well was added and incubated for 4 hr. After incubation, 150 µl was removed, and 100 µl of warm isopropyl alcohol with 0.04 N HCl was added to dissolve the dark-blue formazan crystals. The plates were read in a microwell plate reader (SLT Labinstruments, Salzburg, Austria) using an absorbance wave length of 570 nm and a reference wavelength of 630 nm. Percentage of dead target cells was determined as reported previously (Mosmann, 1983).

Acute exposure to fumonisin B₁. In another set of experiments, five animals each were administered a single dose of 25 mg/kg subcutaneous of fumonisin B₁ and the same dose of fumonisin B₁ with a 15-min prior intravenous injection of 50 µg of polyclonal anti-mouse TNF- antibody raised in rabbit (Endogen, Woburn, MA). The control group was treated with saline. All mice were killed after 4 hr, and blood samples were collected; total red and white and percent differential cell counts were obtained. Neutrophil and lymphocyte numbers were calculated from the total leukocyte counts. Reticulocytes were counted from three random fields of blood smears of each mouse and averaged.

In vitro fumonisin B₁ exposure. BALB/c macrophage cell line (J774A.1, ATCC TIB-67) was grown in Dulbecco's modified Eagle's medium (2 mM glutamine, 100 units/ml penicillin, 100 µg/ml streptomycin) with 10% FBS (Life Technologies). Cells were passed twice weekly, and cells in their mid log phase were used in the experiments after assessment of cell viability using trypan blue exclusion. The cells (3.5 × 10⁶/ml) were cultured in triplicate in 35-mm Petri dishes in the presence of varying concentrations of fumonisin B₁ (0-10 µM) for 24 hr. Supernatants were collected and tested for TNF- bioactivity using the MTT cytotoxicity assay described earlier.

Statistical methods. Data from these studies were analyzed by one-way analysis of variance followed by Duncan's multiple range test unless mentioned otherwise in Results. Statistical calculations used the SAS computer program (SAS Institute, Cary, NC). A value of P < .05 was used to indicate significant differences, except where indicated otherwise.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Fumonisin B₁ treatment selectively affects kidney weights and leukocyte counts. Kidney weights were significantly decreased in all the fumonisin B₁-treated (0.25-6.75 mg/kg/day for 5 days) groups compared with the control mice. No significant differences were observed in body weights, liver, thymus, spleen and brain weights (table 1). There were no differences observed in food and water intake for any of the experimental groups. White blood cell counts were significantly increased for the two highest doses and decreased significantly in the low-dose group. No effects were observed in red cell count, recovered splenocyte and peritoneal macrophage numbers (table 2).

Repeated fumonisin B₁ treatment induces apoptosis in mouse liver. Apoptotic bodies were identified in the liver (fig. 1) and kidneys of all fumonisin B₁-treated animals (not shown). Results indicated that the 5-day treatment of fumonisin B₁ was sufficient to cause apoptosis in these organs. The induction of apoptosis was not due to caloric restriction because food and water consumption were not influenced by the fumonisin B₁ treatment. There were no other degenerative changes observed in liver sections (data not presented). A detailed quantitative account of pathogenesis of fumonisin B₁ in liver and kidney has been presented elsewhere (Sharma et al., 1997).

View larger version (183K): [in this window] [in a new window]   Fig. 1.   Micrograph of a liver section showing apoptotic TUNEL-positive cells (arrow) in mice treated with fumonisin B1 (2.25 mg/kg/day) for 5 days. Tissue was fixed in 10% neutral buffered formalin for 4 hr and paraffin embedded. Sections were cut (5 µm), and TUNEL was performed. Digoxigenin-UTP-labeled DNA was detected with anti-digoxigenin-peroxidase antibody followed by peroxidase detection with 0.05% diaminobenzidine and 0.02% H2O2. Tissues were counterstained with hematoxylin for 45 sec. TUNEL-positive round cells (arrow) were dark brown. Apoptotic cells were either absent in control livers or were very few (Sharma et al., 1997).

Fumonisin B₁ induces TNF- expression in peritoneal macrophages. In the LPS-induced peritoneal macrophages, fumonisin B₁ produced a dose-dependent increase in TNF- mRNA levels (fig. 2a); however, the changes were not statistically significant. Levels of TNF- protein in the LPS-treated peritoneal macrophages were significantly increased in all the fumonisin B₁-treated groups except the highest dose group (fig. 2b), reaching nearly 2-fold in the 2.25 mg/kg/day group. To confirm the biological activity of TNF- in supernatants, TNF- sensitive WEHI-164 cells were used for cytotoxicity assays (fig. 3). A significant increase in the cytotoxicity of medium on WEHI-164 cells in the three highest fumonisin B₁-dose groups was observed; the magnitude of change was similar to that of TNF- measured by ELISA. Cytotoxicity was reversed when polyclonal anti-TNF- antibody was added, indicating that the effects were due to the TNF- levels in the supernatants. Data show that 5-day FB₁ treatment increased the TNF- induction when the peritoneal macrophages were activated with LPS in vitro.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Effect of fumonisin B1 on the TNF- induction in peritoneal macrophages. a, Total RNA (8 µg) of the LPS-activated (10 µg/l × 106 cells for 6 hr) BTG-stimulated peritoneal macrophages from control and fumonisin B1-treated mice were analyzed by Northern blotting using respective murine cytokine and human -actin digoxigenin-labeled riboprobes. Bars on bottom, band densities of six different animals from each dose group, measured in pixels with UN-SCAN-IT software program. Values of the total RNA per lane were measured and equalized using -actin standards, and the total cytokine mRNA in the treated groups was calculated as percentage of -actin mRNA in the same sample. b, Supernatants collected from LPS (10 µg/ml) activated peritoneal macrophages after 24 hr were used to detect TNF-. Data are mean ± S.E.M. (n = 6). *Significant difference at P < .05 between treatment and control groups.

View larger version (33K): [in this window] [in a new window]   Fig. 3.   Effect of fumonisin B1 on the TNF- bioactivity. Peritoneal macrophage culture supernatants (100 µl of 50-fold dilution/well) from control and fumonisin B1-treated mice were added to actinomycin (1 µg/ml)-sensitized WEHI-164 cells (7.5 × 104/100 µl/well). The percentage of dead cells were quantified after 22 hr of incubation. Recombinant mouse TNF- of known concentration was tested for bioactivity and used as validation for the experimental results. Polyclonal anti-mouse anti-TNF- antibody (6 µg/ml) was used for neutralization of the cytotoxicity. Data are mean ± S.E.M. (n = 6). *Significant difference at P < .05 between treatment and control groups.

Fumonisin B₁ has no effect on IL-1 and IFN induction. There were no significant differences in the IL-1 mRNA levels and protein levels as measured by Northern blotting and ELISA, respectively, when peritoneal macrophages from 5-day fumonisin B₁-treated animals were cultured with LPS (fig. 4, a and b). The mRNA and protein levels for IFN were not altered in the splenocytes cultured with Con A in vitro (fig. 5, a and b). Data presented in figures 4 and 5 do not show differences related to the treatment and are therefore important in suggesting a TNF--specific effect of fumonisin B₁.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Effect of fumonisin B1 on the IL-1 induction in peritoneal macrophages. a, IL-1-specific mRNA in macrophages. Experimental conditions and data presentation are same as in fig. 2. b, Supernatants collected from LPS (10 µg/ml)-activated peritoneal macrophages after 24 hr were used to detect IL-1. Data are mean ± S.E.M. (n = 6).

View larger version (36K): [in this window] [in a new window]   Fig. 5.   Effect of fumonisin B1 on the Con A-induced IFN- mRNA expression in mice splenocytes. a, Total RNA (8 µg) of the Con A-activated (5 µg/2 × 106 cells for 12 hr) splenocytes from control and fumonisin B1-treated mice were analyzed using murine IFN and human -actin digoxigenin-labeled riboprobes. Data are presented similar to that in fig. 2. b, Splenocytes (2 × 106/ml) from control and fumonisin B1-treated mice were activated with 5 µg/ml Con A for 48 hr, and IFN in culture supernatants was analyzed. Data are mean ± S.E.M. (n = 6).

Hematological changes can be reversed by anti-TNF- antibody. A single fumonisin B₁ dose (25 mg/kg subcutaneous) significantly increased the total and differential leukocytes and the presence of reticulocytes in peripheral blood (table 3). These elevated cell numbers were not observed when the fumonisin B₁-treated animals were also administered 50 µg of the IgG fraction of polyclonal antibody raised against recombinant murine TNF-.

Fumonisin B₁ induces TNF- production in J774A.1 cells in vitro. A significant increase in the TNF- levels was observed in J774A.1 cells when cultured with 0.1 and 1 µM fumonisin B₁ in vitro for 24 hr as measured by the MTT cytotoxicity assay (fig. 6). These observations further indicate that fumonisin B₁ can induce the production of TNF- by macrophages in vitro. A decrease in TNF- production in supernatants of macrophages treated in vitro with the higher concentrations of fumonisin B₁ (particularly at 10 µM) can be explained on the basis of the cytotoxic response of fumonisin B₁ in J744A.1 cells. Fumonisin B₁ produced 23% and 26% cytotoxicity in J774A.1 cells at 1 and 10 µM, respectively.

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Effect of varying concentrations of fumonisin B1 on TNF- production by J774A.1 cells. Cells (3.5 × 106/ml) were plated as 1 ml/dish. Supernatants were separated 24 hr after fumonisin B1 exposure, and their cytotoxicity was determined on WEHI-64 (TNF--sensitive) cells as described in Materials and Methods. Data are mean ± S.E.M. (n = 3). *Significant difference at P < .05 between treatment and control groups.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The results of this study are consistent with the hypothesis that TNF- is a contributor to fumonisin B₁ toxicity in mice. The TNF- involvement, either as a mediator or as a consequence of initial biochemical effects of fumonisin B₁, is supported by the results showing that (1) macrophages derived from animals treated with fumonisin B₁ produce higher amounts of TNF- (2) increased apoptosis in liver and kidney was seen in response to fumonisin B₁ treatment (Sharma et al., 1997), (3) the hematological effects of acute fumonisin B₁ exposure were prevented by anti-TNF- administration and (4) murine macrophages cultured in vitro produced higher amounts of TNF- when incubated with the toxin. The results of the current study suggest that fumonisin B₁ exposure induced TNF-, which plays at least an immunomodulatory role.

TNF- is a known initiator of the intracellular events leading to apoptosis. A significant reduction in the kidney weights at all doses of fumonisin B₁ in the present study indicated that kidney is sensitive to the effect of this toxin (table 1). Apoptosis was significantly increased in both liver and kidney of the fumonisin B₁-treated mice (Sharma et al., 1997). Several studies on rodents show that fumonisin B₁ affects liver and kidney. Fumonisin B₁-fed male fisher F344 rats (0-484 ppm for 4 weeks) showed a significant dose-dependent decrease in kidney weights (Tolleson et al., 1996a). Liver and kidney effects were also observed when rats were fed 15 ppm of fumonisin B₁ for 4 weeks (Voss et al., 1993) or injected intraperitoneally with 7.5 or 10 mg of fumonisin B₁/kg for 4 days (Bondy et al., 1995). Microscopic lesions were seen in both liver and kidney but probably because of the higher regeneration ability of liver, weights of this organ were not different from the controls.

TNF- is known to stimulate the production of cytokines such as GM-CSF, M-CSF, IL-8 and other cytokines (Fiers, 1991). There was a significant increase in the number of white blood cell counts in the blood obtained from the two highest dose groups of repeated mycotoxin exposure, indicating that fumonisin B₁ induced the hematopoiesis/chemotaxis directly or indirectly via activation of colony-stimulating factors, growth factors or both. These factors, GM-CSF and M-CSF, induce the production of granulocytes from stem cells and migration of granulocytes, especially neutrophils, into the circulation (Gasson et al., 1984; Whetton and Dexter, 1989).

A dose-dependent increase in TNF- mRNA levels and no effect on IL-1 mRNA levels were observed in the fumonisin B₁-treated peritoneal macrophages when cultured in vitro with LPS (fig. 2). Fumonisin B₁ appears to induce specifically TNF- mRNA transcription, even though the increase was not statistically different. The levels of TNF secreted in the medium were significantly increased at all dose levels. The TNF- produced in the supernatant was biologically active. The fact that anti-TNF- antibody neutralized the TNF- effect confirms that cytotoxicity was due to the TNF- and not other cytotoxic factors in the supernatants of the peritoneal macrophages from fumonisin B₁-dosed mice.

A single dose of fumonisin B₁ (25 mg/kg) also significantly increased the leukocyte counts similar to the effects seen in the 5-day fumonisin B₁ exposures. The neutrophil counts increased more than the lymphocytes after mice were injected with fumonisin B₁. The normal leukocyte counts in the fumonisin B₁-dosed animals that were pretreated with anti-TNF- implied that TNF- was responsible for the leukocyte increase in this experiment. A significant increase in reticulocyte counts after a single exposure of fumonisin B₁ was prevented by anti-TNF-. The anti-TNF- was produced against recombinant murine protein, unlikely to cross-react with other factors like lymphotoxin. An increase in granulocyte counts was observed in broiler chicks with fumonisin B₁ treatment (Dombrink-Kurtzman et al., 1993). The appearance of reticulocytes can be explained by the effect of fumonisin B₁ on hemopoietic cells during development in the bone marrow.

The TNF- induction in response to fumonisin B₁ may be related to its other reported biochemical effects. Fumonisin B₁ specifically inhibits sphinganine N-acyl transferase (ceramide synthase), resulting in disruption of sphingolipid metabolism. Increase in free sphinganine, sphingosine and sphingoid base metabolites and depletion of more complex sphingolipids have been reported after fumonisin B₁ exposure (Wang et al., 1991; Yoo et al., 1996; Smith and Merrill, 1995; Riley et al., 1996). Sphingosine and its phosphorylated product S-1-P and ceramide can act as second messengers during cytokine expression, cell activation, differentiation and apoptosis (Ballou et al., 1996; Merrill et al., 1997). These effects are very complex depending on the type of cells. Sphingosine and S-1-P activate PLD and phosphatidic acid and increase mitogenesis, whereas ceramide inhibits PLD and mitogenesis (Yamada and Sakane, 1993; Gomez et al., 1994). S-1-P also activates the MAP kinase cascade, which in turn may activate transcription factors (AP-1 and NF-B) essential for TNF induction (Müller et al., 1993).

LPS-induced production of TNF- has been suggested through DAG-independent PKC activation (Shabira et al., 1994). Sphinganine inhibits Ca⁺⁺-activated phospholipid-dependent phophorylating enzyme PKC (Hannun et al., 1986; Hannun and Bell, 1987). Sphingosine pretreatment reversed the phorbol myristate acetate-induced down modulation of TNF- production in response to zymosan (Sanguedolce et al., 1992). These studies suggest that DAG-independent PKCs that are not suppressed by free sphingoid bases such as sphingosine (increased in response to fumonisin B₁) may play a role in the TNF- induction. Ganglioside (GM3) depletion and membrane disruption may also induce TNF- production (Wood et al., 1992; Zieglerheitbrock et al., 1992). Fumonisin B₁ causes complex sphingolipid depletion (Yoo et al., 1996) and thus may lead to events involved in TNF- production.

The effects of fumonisin B₁ are selective on TNF- without affecting the production of IL-1 by macrophages. No significant differences in mRNA and protein levels of IFN were observed when splenocytes from fumonisin B₁-treated animals were cultured in vitro with Con A. The mechanism by which fumonisin B₁ induces apoptosis is not known; however, TNF- is a well-established signaling messenger in the induction of apoptosis (Wallach, 1996).

One pathway in apoptosis signaling is the sphingomyelin cycle (Hakomori and Igarashi, 1995; Hannun, 1995). It involves activation of a neutral sphingomyelinase (SMase) initiating SM hydrolysis to form ceramide, which acts as second messenger in signaling apoptosis. Ceramide is induced by extracellular signals such as TNF- via p60 TNF receptor (TNFR-I, p55) (Dbaibo et al., 1993; Kolesnick and Golde, 1994; Darney and Agarwal., 1997). Studies show that DAG-dependent PKC activation and ceramide production have an antagonistic relationship, and sphinganine and sphingosine inhibit PKC, thereby increasing the potency and efficacy of the ceramide in inducing apoptosis in cell lines (Jarvis et al., 1996). PKC inhibition activates the neutral SMase and induces apoptosis via increased ceramide production (Chmura et al., 1996). Thus, one explanation for the increased apoptosis in liver and kidney observed in several fumonisin studies, including this one, is that TNF--induced ceramide production is potentiated by fumonisin-induced increase in free sphingoid bases, resulting in inhibition of PKC and stimulation of neutral SMase. Fumonisin-inhibition of PKC has been demonstrated in cell lines (Huang et al., 1995). In addition, free sphingoid bases or inhibition of their de novo synthesis have been shown to induce apoptosis in cell lines (Nakamura et al., 1996). We recently reported that ceramide levels were not depleted in mouse tissues after treatment with the same doses of fumonisin B₁, as in our current study (Tsunoda et al., in press). It is likely that a possible depletion of ceramide due to inhibition of ceramide synthase was compensated by its increased formation from sphingomyelin via TNF--induced activation of SMase.

The possible production and involvement of cytotoxic factors and changes in the macrophage structures in fumonisin B₁ toxicity have been speculated on earlier (Qureshi and Hagler, 1992; Guzman et al., 1997). An ultramicroscopic examination of pulmonary intravascular macrophages from pigs treated with fumonisin B₁ and B₂ containing cultures indicated the formation of multilamellar bodies inside the cytoplasm (Colvin et al., 1993). Pulmonary intravascular macrophages account for nearly 25% of the capillary volume in pigs (Warner and Brain, 1990; Colvin et al., 1993), a unique feature in this species. The authors suggested the possibility of activated macrophages at the endothelial sites and release of vasoactive amines that may be involved in lung damage.

Fumonisin B₁ therefore appears to potentiate TNF- production in peritoneal macrophages when activated with LPS in vitro, and TNF- may play a role in the mechanism of fumonisin B₁-induced apoptosis and toxicity. Macrophages are ubiquitous in the body, and tissue macrophages are 500 to 1000 times more in number than those in the blood circulation (Cohn, 1983). Present data suggest a strong involvement of TNF- in fumonisin B₁ toxicity. Future studies to (1) demonstrate the presence of TNF- in target organs, (2) investigate the role of free sphingoid bases and ceramide levels due to fumonisin B₁ exposure in the activation of TNF- and its family, (3) determine whether TNF is involved in toxic effects of this mycotoxin other than the hematological ones and (4) study the role of TNF- in the cell proliferation and cell death or apoptosis caused by fumonisin B₁ and similar toxins are under way at our laboratory.