Toxicology 223 (2006) 191–201

Amelioration of fumonisin B1 hepatotoxicity in miceby depletion of T cells with anti-Thy-1.2

Neelesh Sharma, Quanren He 1, Raghubir P. Sharma ∗Department of Physiology and Pharmacology, College of Veterinary Medicine, The University of Georgia,

Athens, GA 30602-7389, USA

Received 8 November 2005; received in revised form 17 March 2006; accepted 25 March 2006Available online 18 April 2006

Abstract

Fumonisin B1 is a mycotoxin produced by Fusarium verticillioides, frequently associated with corn. It produces species-specificand organ-specific toxicity, including equine leukoencephalomalacia, porcine pulmonary edema, and hepatic or renal damage inmost animal species. Fumonisin B1 perturbs sphingolipid metabolism by inhibiting ceramide synthase. Our previous studies in malemice indicated that fumonisin B1-induced hepatotoxicity is modulated by the localized activation of cytokines in liver macrophagesand other cell types. In the current study, male athymic nude mice and their wild type counterparts (WT), the latter with or withoutdepletion of T cells, were treated subcutaneously with fumonisin B1 at 2.25 mg/kg/day for 5 days and sampled 24 h after the lastinjection. Depletion of T cells in WT was achieved by a single intravenous injection of 50 �g monoclonal antibody against Thy-1.2surface antigen of mature peripheral T lymphocytes 24 h before the first fumonisin B1 treatment. The depletion of T cells nearlyabolished fumonisin B1-mediated liver toxicity as indicated by the near normal concentrations of circulating liver enzymes and byebikB©

K

1

vci

0

numeration of apoptotic hepatocytes. There was no difference in the fumonisin B1-induced elevation in circulating liver enzymesetween WT and nude mice. Fumonisin B1-induced mRNA expression of tumor necrosis factor � and interleukin-1� was observedn nude and WT mice but not in T cell-depleted mice. Hepatotoxic response to fumonisin B1 was unaltered in mice lacking naturaliller cells. This study suggested that T cells and corresponding proinflammatory cytokines have a vital role in mediating fumonisin1-induced hepatic toxicity.2006 Elsevier Ireland Ltd. All rights reserved.

eywords: Fumonisin B1; Hepatotoxicity; CD90.2 T cells; Tumor necrosis factor �; Interferon �; Anti-Thy-1.2

. Introduction

Fumonisin B1 is a mycotoxin produced by Fusariumerticillioides and related species, commonly found inorn. The toxicity of fumonisin B1 has been describedn a number of farm and laboratory animals and depends

∗ Corresponding author. Tel. +1 706 5422788; fax: +1 706 542 3015.E-mail address: rpsharma@vet.uga.edu (R.P. Sharma).

1 Current address: CyDex, Inc. Lenexa, KS 66214, USA.

on different factors such as animal species, gender, age,dose and route of administration (Diaz and Boermans,1994). It has been established as a cause of equineleukoencephalomalacia and porcine pulmonary edema.Fumonisin B1 has been shown to be hepatotoxic andnephrotoxic in mice and rats. The National ToxicologyProgram study reported renal carcinoma in male Fis-cher 344 rats and hepatic cancer in female B6C3F mice(Howard et al., 2001). In the Transkei region of SouthAfrica where incidence of esophageal cancer is high,an association between fumonisin B1 consumption and

300-483X/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.tox.2006.03.021

192 N. Sharma et al. / Toxicology 223 (2006) 191–201

esophageal cancer was reported (Sydenham et al., 1990).International Agency for Research on Cancer (1992)classified fumonisin B1 as a Group 2B carcinogen, i.e.,possibly carcinogenic to humans.

Fumonisin B1 is structurally similar to sphingoidbases and causes inhibition of ceramide synthase (sph-inganine or sphingosine N-acyltransferase) leading toaccumulation of free sphingoid bases, sphingoid basemetabolites, and depletion of more complex sphin-golipids (Riley et al., 1996). Studies in pig renal epithe-lial cells (LLC-PK1) and human colonic cells (HT29)have shown that fumonisin B1-induced apoptosis, necro-sis, and inhibition of proliferation are sphinganinedependent (Merrill et al., 2001). There is a close asso-ciation between increased apoptosis in liver and kidneyand the disruption of sphingolipid metabolism (Tsunodaet al., 1998).

Fumonisin B1 treatment caused an increased mRNAexpression of tumor necrosis factor � (TNF�), interferon� (IFN�) and interleukin (IL)-12, in liver, where cellsinvolved in TNF� production were identified as Kupffercells (Bhandari et al., 2002). Increase of proinflamma-tory cytokines from immune cells has been implicatedin fumonisin B1-induced hepatotoxicity. Depletion ofKupffer cells by gadolinium chloride attenuated fumon-isin B1 induced hepatotoxicity (He et al., 2005). Deletionof tumor necrosis factor receptors 1 and 2 and interferon� genes reduced fumonisin B1-induced hepatotoxicity inmice (Sharma et al., 2001, 2003).

The role of inflammatory cytokines in the initiation

row or total circulating immunoglobulin levels (Bondyet al., 1997). Recently, Taranu et al. (2005) reportedthat ingestion of fumonisin B1 altered the Th1/Th2 bal-ance leading to reduced vaccinal antibody response inpigs.

Despite many reports confirming the role of proin-flammatory cytokines in fumonisin B1 induced liverdamage, little is known about the contribution of T cellsin this regard. We hypothesized that fumonisin B1 hep-atotoxicity in mice will be compromised by depletionof T cells. Athymic nude mice are deficient in T cells;therefore in the current study we employed nude miceand their immunocompetent wild-type counterparts inthis investigation. Depletion of T cells in wild-type micewas accomplished by injecting monoclonal antibodiesdirected against Thy-1.2 surface antigens; this approachhas become an established method to investigate the Tcell functions in mice (Opitz et al., 1982). The objec-tive of this study was to investigate the role of T cellsin fumonisin B1-induced hepatotoxicity in immunocom-petent mice by depleting T cells and comparing it withimmunodeficient athymic nude mice.

2. Materials and methods

2.1. Animals

Seven-week-old male athymic nude mice (Hsd:nu/nu, stock#6904M) and their wild-type (WT) counterparts (Hsd:nu/+,stock #7004M) were obtained from Harlan Laboratories (Indi-

and development of liver toxicity has been well doc-umented. Liver associated immune cells such as lym-phocytes, Kupffer cells and their cytokine products;especially TNF� and IFN�, are involved in liver injuryduring alcoholic hepatitis (Batey and Wang, 2002).TNF� and IL-1� are responsible for certain patho-logical manifestation of acetaminophen induced hep-atotoxicity in mice (Blazka et al., 1995). IFN� andTNF� released from T lymphocytes and macrophages,respectively, are involved in concanavalin A (con A)-induced hepatitis and in liver toxicity in response torepeated challenges of lipopolysaccharide (Wolf et al.,2001).

Reports on the effects of fumonisin B1 on the immunesystem are conflicting. Fumonisin- B1 was found tobe both immunosuppressive and immunostimulatory inBALB/c mice, depending on the dose administered andinduced an antigenic response (Martinova and Merrill,1995). Fumonisin B1 was shown to alter the expressionof the CD3 receptor, which is integral to T lymphocyteactivation. Gavage studies in mice indicated no changesin myeloid, erythroid or other cell types in bone mar-

anapolis, IN). The nu/+ mice are heterozygous between nudeand normal mice and are fully immunocompetent; these wereused as controls (henceforth referred to as WT) as our aimwas also to compare their response to fumonisin B1 withcorresponding immunodeficient nude mice. The mice wereacclimated for one week before dosing under environmentalconditions at 23 ◦C and 65% relative humidity with a 12-hlight/dark cycle. Animals had free access to fumonisin freefeed and water at their own discretion. All experiments wereconducted according to the Public Health Service Policy onHumane Care and Use of Laboratory Animals and protocolswere approved by the Institutional Animal Care and Use Com-mittee.

2.2. Treatments

Animals were randomly divided in four groups of WT miceand two groups of nude mice (five mice per group). Twogroups of WT mice received iv injections of 50 �g mono-clonal endotoxin-free anti-mouse Thy-1.2 (CD90.2) antibody(Southern Biotech, Birmingham, AL) per mouse 24 h beforethe first fumonisin B1 treatment; these mice from now on arereferred as T cell-depleted (TCD) mice. Dose of antibody toachieve depletion of T cells was based on previous studies

N. Sharma et al. / Toxicology 223 (2006) 191–201 193

and adjusted for the purity of the product (Opitz et al., 1982;Tiegs et al., 1992). One group each of WT, TCD and nudemice were treated with five daily subcutaneous injections ofeither phosphate-buffered saline (PBS) or 2.25 mg/kg/day offumonisin B1 (>98% purity, from Programme on Mycotoxinsand Experimental Carcinogenesis, Tygerberg, South Africa) inPBS. The dose of fumonisin B1 and duration of treatment wasbased on our previous studies in which it produced a consistenthepatotoxic damage in male mice (Sharma et al., 2001, 2003).The dose employed here was optimized for hepatotoxicity andsphinganine increase in mouse liver (Tsunoda et al., 1998);the time of sampling was based on our previous observationsthat effects of fumonisin B1 decline rapidly after cessation oftreatment and are considerably less after 24 h (Enongene et al.,2002). The dose used in this study corresponds to 14 mg/kg offumonisin B1 in rodent feed, well within the range of expo-sures encountered with toxicity episodes of natural exposures(Dutton, 1996). All treatments and samplings were performedat the same time of the day (mornings).

Twenty-four hours after the fifth and final injection, micewere euthanized by halothane and blood was drawn by heartpuncture in heparinized tubes for hematological evaluation;plasma was isolated for determination of alanine aminotrans-ferase (ALT) and aspartate aminotransferase (AST). Activitiesof plasma ALT and AST were determined using a Hitachi912 Automatic Analyzer (Roche Diagnostics, Indianapolis,IN). Total and differential leukocyte counts were performedon blood.

Total white and red cells in blood were counted usingan electronic counter (Coulter Electronics Inc., Hialeah, FL).Liver, kidney and spleen were removed, weighed, and aliquotsof liver were fixed immediately in neutral 10% formalin orq ◦

a

2

amshSmwmss

2h

pc(k

ously (Sharma et al., 2003). The stained apoptotic cells werecounted in the whole section and normalized to the unit area.In addition hematoxylin and eosin stained sections were exam-ined under a light microscope.

2.5. Analyses of sphingolipids

Levels of free sphinganine and sphingosine in hepatic tis-sues were determined after base treatment of lipid extractsby high performance liquid chromatography according to themethods reported previously (Tsunoda et al., 1998). The con-centration of sphingoid bases, sphinganine and sphingosine,was quantified based on the recovery of the C20-sphinganine(D-erythro-C20-dihydro-sphingosine; Matreya, Pleasant Gap,PA) as the internal standard.

2.6. Fumonisin hepatotoxicity in natural killer (NK)cell-depleted mice

Male mice severely deficient in NK cells, C57BL/6J-Lystbg-j/J (stock #000629 from Jackson Laboratory, Bar Har-bor, ME), and their wild-type counterparts (C57BL/6J, stock#00064, also form Jackson Laboratory) were similarly treatedwith fumonisin B1 as described above. Depletion of NK cellswas also accomplished in C57BL/6J mice by iv injection of50 �g low endotoxin and azide-free anti-mouse NK1.1 (South-ern Biotech), 24 h before the first fumonisin B1 injection.Animals (5 animals/group) were sampled 24 h after the lastfumonisin injection as indicated above and their circulatingliver enzyme activities was determined.

2.7. RNA isolation and ribonuclease protection assay

uickly frozen in liquid nitrogen and stored at –85 C untilnalysis.

.3. Detection of Thy-1.2 bearing T cells in liver

Liver sections were obtained 6 days after the iv injection ofnti-Thy-1.2 (Thy-1.2 injected on day 0, fumonisin B1 treat-ents on days 1–5, sampling of tissues on day 6). Frozen

ections (5 �M) were thawed in the presence of 10% normalorse serum in PBS and blocked with the same solution for 1 h.ections were incubated overnight with 1:100 dilution of anti-ouse Thy-1.2 antibody in a humidified chamber at 4 ◦C, thenashed with PBS and incubated with peroxidase-labeled anti-ouse immunoglobulin G for 1 h. After washing with PBS, the

ections were stained with 3,3-diaminobenzidine and counter-tained with hematoxylin and eosin.

.4. Determination of apoptosis in liver andistopathology

Liver sections (5 �m) were prepared from formalin-fixedaraffinized liver samples and subjected to terminal deoxynu-leotidyl transferase (TdT)-mediated dUTP nick end labelingTUNEL) with a peroxidase-based in situ Cell Death Detectionit (Roche Diagnostics, Indianapolis, IN) as described previ-

(RPA)

Total RNA from liver tissue was isolated using TRI® reagent(Molecular Research Center, Cincinnati, OH) and used for RPAusing RiboQuant TM RPA starter kit (BD Biosciences, SanDiego, CA), as described earlier (Sharma et al., 2003). Highspecific activity �32P-UTP-labeled anti-sense RNA probeswere synthesized using T7 RNA polymerase in vitro tran-scription kit according to the manufacturer’s protocol (BDBiosciences, San Diego, CA). The synthesized probes werehybridized with 50 �g RNA overnight at 65 ◦C and digestedwith RNase A and RNase T1, followed by treatment withprotease K. The RNase-protected products were purified andresolved on gels containing 5% polyacrylamide/7 M urea. The�32P labeled bands were exposed to an FX Imaging ScreenK-HD ® (Bio-Rad Laboratories, Hercules, CA) for 6–24 h andscanned by Bio-Rad Molecular Imager ® FX. The relative geneexpression was digitized using the Quantity One® software(Bio-Rad Laboratories) and normalized against ribosomal pro-tein L32, the housekeeping gene (Zhang et al., 2005). Since thephosphoimager provides band densities (radioactivity) that arelinear for six orders of magnitude, the radioactivity counts forL32 were divided by 10 before obtaining ratios for differentcytokines.

194 N. Sharma et al. / Toxicology 223 (2006) 191–201

2.8. Statistics

Data from these experiments were analyzed by two-wayanalysis of variance (ANOVA) followed by Duncan’s multi-ple range test. Results are presented as mean ± standard error(S.E.). All tests were performed using an SAS computer pro-gram (SAS Institute, Cary, NC). The level of p ≤ 0.05 wasconsidered statistically significant.

3. Results

3.1. Injection of antibody against Thy-1.2 antigendepleted T cells in WT mice

Treatment of WT mice with the monoclonal Thy-1.2 antibody resulted in reduction of peripheral leuko-cyte count to nearly half of the WT mice, measured6 days after the iv injection of the antibody (Fig. 1a).Athymic nude mice also had a lower white blood countcompared to WT mice. Differential leukocyte countsindicated that the decrease in TCD and nude micewas primarily due to a reduction in lymphocytes; theamounts of other leukocytes were not affected. Thelymphocyte numbers were 2.4 ± 0.3 × 103/mm3 in theWT strain, and 1.0 ± 0.4 × 103/mm3 in TCD or nude

mice, whereas the numbers of neutrophils in all groupswere 0.9–1.1 ± 0.5 × 103/mm3. No effect of antibodyinjection or fumonisin B1 treatment was observed onperipheral red blood cell counts in any group (data notpresented). Since athymic nude mice were sensitive tofumonisin B1 hepatotoxicity, we examined the presenceof T cells in their livers by immunochemical staining ofThy-1.2 antigen bearing cells. Results indicated moder-ate staining of T cells in WT or nude mice but no stainingin TCD mice, further confirming a nearly total depletionof T cells by anti-Thy-1.2 antibody, and suggesting thepresence of residual Thy-1.2 antigen-bearing cells in liv-ers of nude mice (Fig. 1b–d).

3.2. Depletion of T cells reduced fumonisinB1-induced hepatotoxicity

After 5 days of treatment with fumonisin B1, therewere no significant differences in body, liver, kidney orspleen weights related to fumonisin B1 treatment (datanot shown). Hepatotoxicity produced by fumonisin B1 inWT or nude mice was similar to that described earlier andprimarily included apoptotic hepatocytes and occasionalmitotic figures scattered throughout the liver tissue withinflammatory changes characterized by anisocytosis and

ce aftere (WT)esenceice. (d)

Fig. 1. (a) Peripheral white blood cell count in different groups of mi* indicates a significant difference from the similarly treated wild-typliver section after pretreatment with anti Thy-1.2 antibody showing prWT. (c) No stained areas were observed in T cell depleted (TCD) mrepresents 25 �m.

5 days of saline or fumonisin B1 treatment; mean ± S.E. (n = 5). Thegroup (p < 0.05). (b) Immunohistochemical staining of frozen mouseof Thy-1.2 antigen bearing T cells in the periportal regions of liver inOccurrence of T cells in nude mice. The bar in the lower left in “b”

N. Sharma et al. / Toxicology 223 (2006) 191–201 195

Fig. 2. Activities of circulating liver enzymes, alanine aminotransferase (ALT) and aspartate aminotransferase (AST) in plasma of wild type (WT),T cell depleted (TCD) and nude mice after treatment with fumonisin B1. Results are expressed as mean ± S.E. (n = 5). Asterisks on bars indicate asignificant difference from the corresponding saline treated group at p < 0.05. The letter “a” on bar indicates significant difference (p < 0.05) betweenthe fumonisin B1-treated WT and TCD groups.

cytoplasmic vacuolation (Sharma et al., 1997; Voss etal., 1995). Fumonisin B1 treatment of WT mice causeda marked increase in plasma ALT and AST activities,32-fold and 14-fold, respectively, compared to corre-sponding saline-treated WT group (Fig. 2). The fumon-isin B1-induced elevation of ALT and AST activity wasnot observed in TCD mice. Plasma ALT and AST activ-ities increased in fumonisin B1-treated nude mice to asimilar extent as in WT mice indicating that nude miceresponded similarly as the wild type to fumonisin B1hepatotoxicity.

3.3. Depletion of T cells diminished fumonisinB1-induced hepatocyte apoptosis

Histological observations were consistent withplasma ALT and AST findings, i.e., after treatment withfumonisin B1 the WT mice had an increased number ofTUNEL-positive cells in liver (Fig. 3). Few apoptoticcells were seen in any of the saline-treated groups. Theincrease of TUNEL-positive cells was not observed inthe TCD mice (pretreated with anti-Thy-1.2 antibody)after fumonisin B1 treatment. The athymic nude micewere susceptible to hepatotoxic effects of fumonisin B1similar to WT mice, as evident by the presence of apop-totic cells after fumonisin B1 treatment.

3.4. Fumonisin B1-induced accumulation of freesphingoid bases is not affected by T cell depletion

gfi

free sphinganine and sphingosine to the same extent inlivers of WT or TCD mice, the differences in fumonisintreated groups of these two strains were not significantlydifferent. The concentration of sphinganine in the liver offumonisin B1-treated nude mice was somewhat higherthan that observed in fumonisin B1-treated WT mice;however, the difference was not statistically significant(Fig. 4).

3.5. NK cells are not important for fumonisin B1

hepatotoxicity in mice

A five-day fumonisin B1 treatment in NK cell defi-cient mice or in wild-type mice with depleted NK cells(after injection of anti-NK1.1 antibody) showed hep-atic response to fumonisin similar to that in respectivecontrol groups (Fig. 5). Fumonisin B1-induced increasein plasma ALT was similar in wild-type controls, NKcell deficient mutants, or in mice pre-treated with anti-NK1.1 antibody. Effects on AST were similar (notshown).

3.6. T cell depletion decreased the mRNAexpression of proinflammatory cytokines afterfumonisin B1 treatment

It has been established that fumonisin B1 causeslocalized activation of proinflammatory cytokines inmice liver produced by Kupffer cells, T cells, NK cells

Concentrations of the free sphingoid bases, sphin-osine and sphinganine, in liver with or without theumonisin B1 treatment of different groups are shownn Fig. 4. Fumonisin B1 treatment increased the levels of

or epithelial cells. In the current study fumonisin B1induced the mRNA expression for cytokines producedby helper T (Th1) cells such as IL-2, IFN�, lymphotoxin� (LT-�) in both WT and nude mice. A representative gelfrom RPA is presented (Fig. 6). The respective mRNA

196 N. Sharma et al. / Toxicology 223 (2006) 191–201

Fig. 3. Enumeration of apoptotic cells in liver by TUNEL assay. (a) Liver from fumonisin B1-WT mouse. (b) Liver from a fumonisin B1-treatedTCD mouse. (c) Liver from fumonisin B1-treated nude mouse. Bar in the lower right in “c” equals 10 �m. Arrows indicate TUNEL positive nucleiin apoptotic cells. (d) The number of TUNEL-positive cells in tissue sections (mean ± S.E., n = 5). The asterisk (*) indicates a significant differencefrom the corresponding saline-treated group (p < 0.05). The letter “a” on the bar indicates a significant difference (p < 0.05) between the fumonisinB1-treated WT and TCD groups.

expression for cytokines produced by Kupffer cells ornatural killer T cells, e.g., TNF�, IL-1�, IL-1�, IL-6and interleukin-1 receptor antagonist (IL-1Ra), were alsoincreased in these groups (Figs. 7 and 8). Such effectof fumonisin B1 on mRNA expression of cytokines pro-duced by both T cells and Kupffer cells was not observedin TCD mice (Figs. 7 and 8).

4. Discussion

Our results showed that in vivo depletion of T cells byantibodies directed against the Thy-1.2 surface antigenof T lymphocytes protected mice from fumonisin B1-induced liver damage. Induction of an immune deficientstate in experimental animals by anti-lymphocyte anti-

Fig. 4. Free sphingosine and sphinganine levels in the liver of saline-treated or fumonisin B1-treated mice. Results are expressed as mean ± S.E.(n = 5). Asterisk on bars indicates a significant difference from the corresponding saline-treated group (p < 0.05).

N. Sharma et al. / Toxicology 223 (2006) 191–201 197

Fig. 5. Alanine aminotransferase (ALT) activity after treatment withfumonisin B1 in plasma of wild-type C56BL/6, NK cell knockout (NKcell KO), and C56BL/6 mice pretreated with anti-NK1.1 antibody.Results are expressed as mean ± S.E. (n = 5). Asterisks (*) on barsindicate a significant difference from the corresponding saline treatedgroup at p < 0.05.

bodies has been intensively studied. Opitz et al. (1982)showed that within 8 h after administration of Thy-1.2antibody helper function of T cells in vivo was decreasedand reached to undetectable level within 24 h. No Thy-1.2 bearing cells could be detected in the peripherallymphoid organs by fluorescein-labeled antibody 24 hlater. Antibody-coated Thy-1.2 bearing cells are rapidlyphagocytized by macrophages and eliminated from tis-sues and circulation. Antibody against Thy-1.2 is there-fore suitable to use for detection of any leftover antigenbearing T cells. Despite the fact that the anti-Thy-1.2antibody induced a transient and complete immune defi-cient state, the animals tolerated this treatment withoutany clinical signs of adverse effects.

Subtle inflammatory changes in the protocol used incurrent study are reflected by increase of plasma ALTand AST activities. Fumonisin B1-induced hepatotoxi-city was not altered in athymic nude mice. This can beexplained by the observation that residual functional Tcells were present in livers of nude mice and their liversshowed induced mRNA expression of proinflammatorycytokines after fumonisin B1 treatment. Others have alsodemonstrated the presence of functional T lymphocytesin athymic nude mice (Ikehara et al., 1984; Shimamuraet al., 1997).

It appears that fumonisin B1-induced hepatotoxicityin mice does not involve NK cells. The NK cell-deficientstrain or normal mice with NK cells depleted by anti-body had a similar response as their respective con-trols. Although NK cells produce cytokines similar toTc

c

Fig. 6. A representative gel from RNase protection assay. A customtemplate set from BD Biosciences was employed with non-hybridizedprobe set (P) run as a size marker. The respective lanes for protectedmRNA from individual animals are labeled as 1, saline-treated WT;2, fumonisin B1-treated WT; 3, saline-treated TCD; 4, fumonisin B1-treated TCD; 5, saline-treated nude, and 6, fumonisin B1-treated nudemouse. Averages of 5 animals per group from gels exposed on the samephosphoimaging screen are shown in Figs. 7 and 8.

1992) and Pseudomonas exotoxin (Schumann et al.,1998). In these models activation of T lymphocytes fol-lowed by Kupffer cell activation and subsequent induc-tion of cytokines are critical factors in the developmentof hepatotoxicity. Lymphocytes can damage hepatocyteseither by releasing TNF-related apoptosis-inducing lig-and (TRAIL), as well as Fas ligand or by inducing proin-flammatory cytokines from Kupffer cells. The Fas ligandis expressed in activated T cells and induces apoptosisin Fas-bearing hepatocytes, as described for alcohol-induced hepatitis (Batey and Wang, 2002) and hepatitiscaused by hepatitis B virus (Kondo et al., 1997). Devel-opment of con A-induced hepatitis was reduced signifi-cantly in IFN� knockout mice, suggesting a central rolefor cytokines produced by activated T cells (Streetz et

h1 lymphocytes, exact involvement of NK cell-secretedytokines in hepatic damage is not well understood.

Participation of T cell in liver cell destruction is aommon mechanism as shown for con A (Tiegs et al.,

198 N. Sharma et al. / Toxicology 223 (2006) 191–201

Fig. 7. Fumonisin B1 induced alteration in gene expression of inflammatory cytokines, TNFα, LT�, IL-2 and IFN�, in mice treated with saline orfumonisin. The relative mRNA expression was normalized against L-32, a house-keeping gene. Mean ± S.E. (n = 5). Asterisks on bars indicate asignificant difference from the saline treated group of same strain p < 0.05. Letter “a” on bars indicates a significant decrease from the similarlytreated WT (p < 0.05).

al., 2001). Hepatocytes bearing Fas are also susceptibleto damage by activated lymphocytes (Galle et al., 1995).

There is considerable evidence that cytokine sig-naling pathways play an important role in fumonisinB1-induced hepatotoxicity in vitro and in vivo. Fumon-isin B1 toxicity was reduced in mice lacking eitherTNF� receptor (TNFR) 1 or TNFR 2 (Sharma et al.,2001). Jones et al. (2001) reported that fumonisin B1toxicity was effectively prevented by inhibition of cas-pases, which are downstream of TNF� cellular signal-ing. Bhandari et al. (2002) reported that interaction ofdifferent types of cell types in liver may amplify thecytokine production induced in fumonisin B1-treatedmice. Sharma et al. (2004) demonstrated similar resultsin an in vitro system of macrophages and nonparenchy-mal epithelial cells. The ultimate extent of liver injury infumonisin B1 hepatotoxicity may therefore involve thehepatic cytokine network rather than a single cytokine(Bhandari et al., 2002).

The immunoregulatory cytokines IL-2, IFN�, andTNF� constitute the type 1 (Th1) cytokine response,whereas the cytokines IL-4, IL-5, IL-6, IL-8, IL-10,

and IL-13 constitute the type 2 (Th2) cytokine response(Abbas et al., 1996). Taranu et al. (2005) have shownthat fumonisin B1 can alter the Th1/Th2 balance in pigstowards activation of Th1 cytokines such as IFN andreduction of Th2 cytokines such as IL-4.

At microscopic level fumonisin B1 produced mini-mal inflammatory response in mice livers, the changesnoticed were primarily apoptotic. It is possible thatcytokines released by activated T cells in the liversinusoid reach Kupffer cells in the periportal area andsurrounding hepatocytes by diffusion through fenes-trae. Liver sinusoidal endothelial cells pose a barrierfor T cells and widespread T cell infiltration into theparenchyma is only observed in acute viral hepatitis. Inthe present study a possible damage to Kupffer cells byantigen-antibody complex in anti-Thy-1.2 injected ani-mals is not likely as a similar response was not observedwhen mice were injected with anti-NK1.1 antibody.

Fumonisin B1 induced hepatotoxicity in severalstrains of mice or with different doses of fumonisin B1has been shown to correlate well with intracellular accu-mulation of sphinganine and sphingosine (Tsunoda et al.,

N. Sharma et al. / Toxicology 223 (2006) 191–201 199

Fig. 8. Fumonisin B1 induced alteration in gene expression of inflammatory cytokines, IL-1�, IL-1�, and IL-6, and anti-inflammatory factor IL-1Ra,in mice treated with saline or fumonisin. The relative mRNA expression was normalized against L32, a house-keeping gene. Mean ± S.E. (n = 5).Asterisks on bar indicate a significant difference from the saline treated group of same strain (p < 0.05). Letter “a” on bars indicates a significantdecrease from the similarly treated WT (p < 0.05).

1998). In our study, sphinganine and sphingosine accu-mulated to a similar extent in control and TCD mice,which seems to indicate that, their accumulation is notrelated to the observed protection from fumonisin B1-induced hepatotoxicity in TCD mice. It appears thataccumulation of sphingolipids is responsible for acti-vation of T cells leading to the production of variouscytokines. Such effect of fumonisin B1 and associatedaccumulation of free sphingoid bases on specific lym-phocyte subpopulations has not been investigated.

Fumonisin B1 increased the accumulation of intracel-lular sphingosine 1-phosphate (S1P) by phosphorylationof sphingosine by sphingosine kinase (Smith and Merrill,1995). Higher concentrations of intracellularly producedS1P increase IFN� and IL-2 secretion in peripheral bloodlymphocytes (Payne et al., 2004). Physiological con-centrations of extracellular S1P are required for opti-mal activity of T cells (Graeler and Goetzl, 2002) andincrease the response of T cells to chemokines by bind-ing to the S1P receptor on T cells (Wang et al., 2004).Although we did not observed a difference in liver sph-ingosine levels of fumonisin-treated normal versus TCD

mice, a lack of T lymphocytes in TCD mice wouldeliminate the possible activating function of sphingosine1-phosphate in lymphocytes.

Liver-associated lymphocyte population is verydiverse and includes heterogeneous population of cellssuch as CD4+ T cells, CD8+ T cells, natural killer Tcells and a thymus independent natural killer (NK) cells(Doherty et al., 1999). Unlike con A-induced hepato-toxicity, fumonisin B1-induced hepatotoxicity does notdepend on NK cells as deletion of NK cells by inducedmutation did not influence the fumonisin B1-inducedhepatotoxicity, as observed in our preliminary experi-ments with NK cell-deficient mice or after deletion ofNK cells with anti-NK1.1 antibody.

The role of T lymphocyte-dependent inflammatorycytokines was also demonstrated in our recent obser-vations where lupus-prone NZB/NZW-F1 mice wererefractory to hepatotoxicity against fumonisin B1 despiteaccumulation of sphingoid bases in liver (Sharma et al.,2005). These animals are defective in cytokine produc-tion and their regulation. Results obtained in this studyindicate that T cell activation followed by production of

200 N. Sharma et al. / Toxicology 223 (2006) 191–201

proinflammatory cytokines is an important mechanismfor fumonisin B1-induced hepatotoxicity in mice.

Acknowledgments

Studies reported in this paper were supported in partby US Public Health Service grant No. ES09403 fromthe National Institute of Environmental Health Sciencesand the Center for Academic Excellence in Toxicologyat the University of Georgia.

References

Abbas, A.K., Murphy, K.M., Sher, A., 1996. Functional diversity ofhelper T lymphocytes. Nature 383, 787–793.

Batey, R.G., Wang, J., 2002. Molecular pathogenesis of T lymphocyte-induced liver injury in alcoholic hepatitis. Front. Biosci. 7,d1662–d1675.

Bhandari, N., Brown, C.C., Sharma, R.P., 2002. Fumonisin B1-inducedlocalized activation of cytokine network in mouse liver. FoodChem. Toxicol. 40, 1483–1491.

Blazka, M.E., Wilmer, J.L., Holladay, S.D., Wilson, R.E., Luster, M.I.,1995. Role of proinflammatory cytokines in acetaminophen hepa-totoxicity. Toxicol. Appl. Pharmacol. 133, 43–52.

Bondy, G.S., Suzuki, C.A., Fernie, S.M., Armstrong, C.L., Hierlihy,S.L., Savard, M.E., Barker, M.G., 1997. Toxicity of fumonisin B1

to B6C3F1 mice: a 14-day gavage study. Food Chem. Toxicol. 35,981–989.

Diaz, G.J., Boermans, H.J., 1994. Fumonisin toxicosis in domesticanimals: a review. Vet. Hum. Toxicol. 36, 548–555.

Doherty, D.G., Norris, S., Madrigal-Estebas, L., McEntee, G., Traynor,

International Agency for research on Cancer, 1992. Fumonicin B1.IARC Monograph on the Evaluation of carcinogenic Risks toHumans. IARC 822, 301–366.

Jones, C., Ciacci-Zanella, J.R., Zhang, Y., Henderson, G., Dickman,M., 2001. Analysis of fumonisin B1-induced apoptosis. Environ.Health Perspect. 109 (Suppl 2), 315–320.

Kondo, T., Suda, T., Fukuyama, H., Adachi, M., Nagata, S., 1997.Essential roles of the Fas ligand in the development of hepatitis.Nat. Med. 3, 409–413.

Martinova, E.A., Merrill Jr., A.H., 1995. Fumonisin B1 alters sph-ingolipid metabolism and immune function in BALB/c mice:immunological responses to fumonisin B1. Mycopathologia 130,163–170.

Merrill Jr., A.H., Sullards, M.C., Wang, E., Voss, K.A., Riley, R.T.,2001. Sphingolipid metabolism: roles in signal transduction anddisruption by fumonisins. Environ. Health Perspect. 109 (Suppl2), 283–289.

Opitz, H.G., Opitz, U., Hewlett, G., Schlumberger, H.D., 1982. A newmodel for investigations of T-cell functions in mice: differentialimmunosuppressive effects of two monoclonal anti-Thy-1, 2 anti-bodies. Immunobiology 160, 438–453.

Payne, S.G., Milstien, S., Barbour, S.E., Spiegel, S., 2004. Modulationof adaptive immune responses by sphingosine-1-phosphate. Semin.Cell Dev. Biol. 15, 521–527.

Riley, R.T., Wang, E., Schroeder, J.J., Smith, E.R., Plattner, R.D.,Abbas, H., Yoo, H.S., Merrill Jr., A.H., 1996. Evidence for dis-ruption of sphingolipid metabolism as a contributing factor in thetoxicity and carcinogenicity of fumonisins. Nat. Toxins 4, 3–15.

Schumann, J., Angermuller, S., Bang, R., Lohoff, M., Tiegs, G., 1998.Acute hepatotoxicity of Pseudomonas aeruginosa exotoxin A inmice depends on T cells and TNF. J. Immunol. 161, 5745–5754.

Sharma, N., He, Q., Sharma, R.P., 2004. Augmented fumonisin B1 tox-icity in co-cultures: evidence for crosstalk between macrophagesand non-parenchymatous liver epithelial cells involving proinflam-matory cytokines. Toxicology 203, 239–251.

O., Hegarty, J.E., O’Farrelly, C., 1999. The human liver containsmultiple populations of NK cells, T cells, and CD3+CD56+ naturalT cells with distinct cytotoxic activities and Th1, Th2, and Th0cytokine secretion patterns. J. Immunol. 163, 2314–2321.

Dutton, M.F., 1996. Fumonisins, mycotoxins of increasing importance,their nature and their effects. Pharmacol. Therap. 70, 137–161.

Enongene, E.N., Sharma, R.P., Bhandari, N., Miller, J.D., Meredith,F.I., Voss, K.A., Riley, R.T., 2002. Persistence and reversibility ofthe elevation in free sphingoid bases induced by fumonisin inhibi-tion of ceramide synthase. Toxicol. Sci. 67, 173–181.

Galle, P.R., Hofmann, W.J., Walczak, H., Schaller, H., Otto, G., Strem-mel, W., Krammer, P.H., Runkel, L., 1995. Involvement of theCD95 (APO-1/Fas) receptor and ligand in liver damage. J. Exp.Med. 182, 1223–1230.

Graeler, M., Goetzl, E.J., 2002. Activation-regulated expression andchemotactic function of sphingosine 1-phosphate receptors inmouse splenic T cells. FASEB J. 16, 1874–1878.

He, Q., Kim, J., Sharma, R.P., 2005. Fumonisin B1 hepatotoxicity inmice is attenuated by depletion of Kupffer cells by gadoliniumchloride. Toxicology 207, 137–147.

Howard, P.C., Eppley, R.M., Stack, M.E., Warbritton, A., Voss, K.A.,Lorentzen, R.J., Kovach, R.M., Bucci, T.J., 2001. Fumonisin B1

carcinogenicity in a two-year feeding study using F344 rats andB6C3F1 mice. Environ. Health Perspect. 109 (Suppl 2), 277–282.

Ikehara, S., Pahwa, R.N., Fernandes, G., Hansen, C.T., Good, R.A.,1984. Functional T cells in athymic nude mice. Proc. Natl. Acad.Sci. U.S.A. 81, 886–888.

Sharma, R.P., Bhandari, N., He, Q., Riley, R.T., Voss, K.A., 2001.Decreased fumonisin hepatotoxicity in mice with a targeted dele-tion of tumor necrosis factor receptor 1. Toxicology 159, 69–79.

Sharma, R.P., Dugyala, R.R., Voss, K.A., 1997. Demonstration of insitu apoptosis in mouse liver and kidney after short-term repeatedexposure to fumonisin B1. J. Comp. Pathol. 117, 371–381.

Sharma, R.P., He, Q., Johnson, V.J., 2003. Deletion of IFN-� reducesfumonisin-induced hepatotoxicity in mice via alterations in inflam-matory cytokines and apoptotic factors. J. Interferon Cytokine Res.23, 13–23.

Sharma, R.P., He, Q., Riley, R.T., 2005. Lupus-prone NZBWF1/J mice,defective in cytokine signaling, are resistant to fumonisin hepato-toxicity despite accumulation of liver sphinganine. Toxicology 216,59–71.

Shimamura, M., Ohteki, T., Launois, P., Garcia, A.M., MacDonald,H.R., 1997. Thymus-independent generation of NK1+ T cells invitro from fetal liver precursors. J. Immunol. 158, 3682–3689.

Smith, E.R., Merrill Jr., A.H., 1995. Differential roles of de novo sphin-golipid biosynthesis and turnover in the “burst” of free sphingosineand sphinganine, and their 1-phosphates and N-acyl-derivatives,that occurs upon changing the medium of cells in culture. J. Biol.Chem. 270, 18749–18758.

Streetz, K., Fregien, B., Plumpe, J., Korber, K., Kubicka, S., Sass,G., Bischoff, S.C., Manns, M.P., Tiegs, G., Trautwein, C., 2001.Dissection of the intracellular pathways in hepatocytes suggests arole for Jun kinase and IFN regulatory factor-1 in con A-inducedliver failure. J. Immunol. 167, 514–523.

N. Sharma et al. / Toxicology 223 (2006) 191–201 201

Sydenham, E.W., Thiel, P.G., Marasas, W.F.O., Shephard, G.S., Van-schalkwyk, D.J., Koch, K.R., 1990. Natural occurrence of someFusarium mycotoxins in corn from low and high esophageal can-cer prevalence areas of the Transkei, Southern Africa. J. Agric.Food Chem. 38, 1900–1903.

Taranu, I., Marin, D.E., Bouhet, S., Pascale, F., Bailly, J.D., Miller,J.D., Pinton, P., Oswald, I.P., 2005. Mycotoxin fumonisin B1 altersthe cytokine profile and decreases the vaccinal antibody titer inpigs. Toxicol. Sci. 84, 301–307.

Tiegs, G., Hentschel, J., Wendel, A., 1992. A T cell-dependent exper-imental liver injury in mice inducible by concanavalin A. J. Clin.Invest. 90, 196–203.

Tsunoda, M., Sharma, R.P., Riley, R.T., 1998. Early fumonisin B1

toxicity in relation to disrupted sphingolipid metabolism in maleBALB/c mice. J. Biochem. Mol. Toxicol. 12, 281–289.

Voss, K.A., Chamberlain, W.J., Bacon, C.W., Herbert, R.A., Walters,D.B., Norred, W.P., 1995. Subchronic feeding study of the myco-toxin fumonisin B1 in B6CF1 mice and Fischer 344 rats. Fundam.Appl. Toxicol. 24, 102–110.

Wang, W., Graeler, M.H., Goetzl, E.J., 2004. Physiological sphingo-sine 1-phosphate requirement for optimal activity of mouse CD4+regulatory T Cells. FASEB J. 18, 1043–1045.

Wolf, D., Hallmann, R., Sass, G., Sixt, M., Kusters, S., Fregien,B., Trautwein, C., Tiegs, G., 2001. TNF-�-induced expression ofadhesion molecules in the liver is under the control of TNFR1—relevance for concanavalin A-induced hepatitis. J. Immunol. 166,1300–1307.

Zhang, X., Ding, L., Sandford, A.J., 2005. Selection of reference genesfor gene expression studies in human neutrophils by real-time PCR.BMC Mol. Biol. 6, 4, doi:10.1186/1471-2199-6-4.