Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cellular resistance to cytotoxicity is conferred by various detoxification and antioxidant enzymes, including glutathione S-transferases (GSTs¹). GSTs plays an important role in protecting colonic epithelial cells against the effects of dietary mutagens and reactive oxygen species (Hayes and Pulford, 1995). Although the precise mechanisms are unclear, there is convincing evidence that the pathogenesis of inflammatory bowel disease is associated with increased oxidative damage due to a reduction in cytoprotection (Lih-Brody et al., 1996), including diminished GST activity (Bhaskar et al., 1995; Clapper and Szarka, 1998). Pro-inflammatory cytokines, notably tumor necrosis factor-, interleukin-1, and interleukin-6 are secreted by inflammatory cells and enterocytes during the intestinal inflammatory response (Stevens et al., 1992; Jung et al., 1995). There is substantial evidence that pro-inflammatory cytokines profoundly influence the constitutive expression of a variety of genes, including drug-metabolizing enzymes such as the cytochromes P450 (Ghezzi et al., 1986; Bertini et al., 1988; Trautwein et al., 1992). Relatively few studies, however, have investigated the effects of cytokines on GST gene expression. In cultured primary hepatocytes, the effects of cytokines on GST gene expression are variable, resulting in either down-regulation (Adams and Czuprynski, 1994; Maheo et al., 1997; Navasa et al., 1998) or up-regulation (Voss et al., 1996), depending on the cytokine treatment. There are no reports of the influence of cytokines on GST expression in cultured human enterocytes.

The human alpha class GST isoenzymes consist of homodimers or heterodimers of four major subunits (GSTA1, A2, A3, and A4) that are involved in detoxification of dietary carcinogens (Huber et al., 1997), and organic hydroperoxides (Hayes and Pulford, 1995). Induction of GSTs, including those of the alpha class, by various structurally unrelated electrophilic compounds and pro-oxidants, represents a major mechanism of protection against oxidative and chemical stress. The regulation of alpha class GSTs has been extensively studied in rodents, and a complex set of regulatory elements has been characterized in the 5'-flanking regions, which control basal and xenobiotic inducible expression (Daniel, 1993; Hayes and Pulford, 1995; Whalen and Boyer, 1998). In rats, for example, transcription of the GSTA2 gene is regulated through a "xenobiotic responsive element" that mediates induction by planar aromatic compounds and an "antioxidant responsive element", which mediates induction by phenolic antioxidants and pro-oxidants (Rushmore and Pickett, 1990; Daniel, 1993; Whalen and Boyer, 1998). A similar "electrophile responsive element" has been identified in murine GSTA1 (Daniel, 1993; Hayes and Pulford, 1995; Whalen and Boyer, 1998). Although the 5'-flanking regions of the human GSTA1 and GSTA2 genes are more than 95% homologous, they are very different from the promoter for rat GSTA and lack a functional antioxidant responsive element, suggesting that the mechanisms of transcriptional regulation are different in the human and rodent genes (Suzuki et al., 1994).

GST may have a particularly important cytoprotective role in the colonic mucosa since elevated GST levels contribute to increased resistance in colorectal tumors, and increased GST expression in differentiated human enterocytes may protect against oxidative damage (Vecchini et al., 1997). However, GST expression has not specifically been assessed in patients with colitis or correlated with cytokine levels in inflamed colon. We hypothesized that pro-inflammatory cytokines may have a negative regulatory influence on the expression of GST in human enterocytes. We assessed the influence of conditioned medium from LPS-activated macrophages and various pro-inflammatory cytokines (IL-1, IL-6, and TNF) on the expression of human GSTA1/A2 in Caco-2 cells. We present the first evidence that IL-1, a major pro-inflammatory cytokine secreted in the inflamed colonic mucosa, down-regulates the expression of GSTA1/A2 in Caco-2 cells by a transcriptional mechanism involving a hepatic nuclear factor 1 (HNF-1) site. Loss of GST expression in human enterocytes during colitis suggests that factors present in chronically inflamed colonic mucosa could have important modulatory effects on GST-mediated cytoprotection.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Recombinant Cytokines and Chemicals. Human recombinant IL-1, IL-6, and TNF were purchased from R & D Systems (Minneapolis, MN). Human GSTA1/A2 cDNA was kindly provided by Dr. C. P. Tu (Pennsylvania State University, University Park, PA). The cDNA for mouse 7S was obtained from Dr. A. Balmain (Onyx Pharmaceuticals, Richmond, CA). Antibodies against human alpha class GST were obtained from Biotrin International (Dublin, Ireland). All other compounds were readily available commercial products.

Culture and Treatment of Cells. Caco-2 cells were selected for this study because this cell line expresses high levels of alpha class GSTs and is highly suitable for the study of the molecular mechanisms of GST regulation in human enterocytes. Caco-2 cells were cultured under 5% CO₂ at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal bovine serum, and 100 µg/ml of penicillin and streptomycin with medium changes every 2 days. For the generation of conditioned medium (CM), murine monocyte-macrophage cells (RAW 264.7) were stimulated for 24 h with LPS (100 ng/ml) from Escherichia coli serotype 055:B5 (Sigma-Aldrich Canada Ltd., Oakville, ON, Canada) dissolved in RPMI 1640 in the presence or absence of 10% fetal bovine serum, denoted FBS+ or FBS, respectively. Conditioned media were then aspirated and frozen at 20°C until further use. Caco-2 cells were subsequently cultured in Dulbecco's modified Eagle's medium diluted with CM for 24 h at the following concentrations: 25%, 50%, and 100%. For experiments involving neutralization of IL-1, CM (FBS) was preincubated with two different concentrations (10 and 100 ng/ml) of anti-IL-1 polyclonal antibody (R & D Systems, Minneapolis, MN) for 1 h with gentle rocking prior to exposure to Caco-2 cells. For Northern analysis of GSTA1/A2 mRNA levels, Caco-2 cells were treated with various concentrations of pro-inflammatory cytokines in the following ranges: IL-1 (0.01-1.0 ng/ml); IL-6 (25-100 ng/ml), and TNF (25-100 ng/ml) for 24 h. For receptor antagonist studies, Caco-2 cells were preincubated with IL-1 receptor antagonist (IL-1ra) (800 ng/ml) for 1 h before treatment with IL-1 (1 ng/ml) for 24 h, after which total cellular RNA was collected. For studies of GSTA1/A2 mRNA stability, Caco-2 cells were treated with the polymerase II inhibitor 5,6-dichloro-1-ribofuranosyl-benzimidazole at a concentration of 30 µg/ml in the presence or absence of IL-1 (1 ng/ml). Cells were then harvested and total cellular RNA was collected at fixed time intervals (0, 1.5, 3, 6, 12, 24 h). For Western blot analysis of GST protein levels, the treatments were: IL-1 (0.01, 0.1, 1.0 ng/ml) and CM (25, 50, and 100%).

RNA Isolation and Northern Blot Analysis. Total RNA from cells was isolated using the TRIzol reagent method (Invitrogen, Carlsbad, CA). Northern blot analysis was performed with the extracted RNA as detailed previously (Benn et al., 1999). The blots were probed with -³²P-labeled cDNA for GSTA1 (808 bp) (Tu and Qian, 1986), which hybridizes with both A1 and A2 subunits and with a cDNA probe for the cytoplasmic RNA 7S as an internal load control. For quantitation of genes, phosphorimaging (GS-250 Molecular Imager; Bio-Rad Laboratories, Mississauga, ON, Canada) was used, and the intensity of the resultant bands was assessed by densitometry using Molecular Analyst software (Version 2.1; Bio-Rad Laboratories). Gene expression data were calculated by first normalizing for 7S, and values were expressed as a percentage of controls.

Plasmid Construction. A LUC reporter gene plasmid construct containing the entire 5'-flanking region of the GSTA1 gene and various truncated constructs were prepared by polymerase chain reaction (PCR) with the following specific oligonucleotide primers (the length of the resultant PCR product from the transcriptional start site is indicated in parentheses):

Site-Directed Mutagenesis. Activator protein 1 (AP-1) (CGTCTTCT; 886-779) and AP-2 (ACCCGGGCA; 1171-1163) sites were deleted from 1614 LUC and 1195 LUC, respectively, and an HNF-1 (GGAACACATTAAC; 182-170) site was deleted from 286 LUC by site-directed mutagenesis using a two-stage PCR protocol as described (Wang and Malcolm, 1999). The following oligonucleotide primers were designed to span and delete these consensus sequences:

Transfection and Luciferase Assay. Twenty-four hours after reaching confluence, Caco-2 cells (2 × 10⁵/35-mm dish) were transfected with a DNA mixture containing 2.5 µg/ml GST-LUC reporter plasmid and 1 µg/ml -gal plasmid using LipofectAMINE (Invitrogen) according to the manufacturer's recommendations. After 24 h, medium containing IL-1 at the indicated concentrations was added. Twenty-four hours later, cells were harvested and washed twice with phosphate-buffered saline, cell extracts were prepared, and aliquots of the lysates were assayed for luciferase and -galactosidase enzyme activities as described (Brasier et al., 1989). Luciferase activities were normalized to - galactosidase activity to correct for differences in transfection efficiency. The plasmid, pSV40-Luc, in which the LUC gene is under the control of a SV40 promoter and enhancer, and the promoterless plasmid pLuc-0 served as positive and negative controls, respectively. All transfection studies were repeated at least three times.

SDS-PAGE and Western Blot Analysis. Alpha class GST protein was identified by Western blot analysis as previously described (Kirby et al., 1993). Briefly, protein (50 µg) from cell extracts was separated by SDS-PAGE on a 12% polyacrylamide gel, transferred to nitrocellulose, and incubated for 2 h with rabbit anti-human GST polyclonal antibody diluted to 1:500. After incubation with goat anti-rabbit peroxidase secondary antibody (Vector Laboratories, Burlington, ON, Canada), bands on blots were detected by chemiluminescence (ECL; Amersham Biosciences, Piscataway, NJ) and visualized by autoradiography. The relative concentration of GST protein levels was then determined by densitometry. Three independent experiments were performed, and Western blots are presented from one representative experiment.

GST Activity. GST has high substrate specificity toward cumene hydroperoxide (Hayes and Pulford, 1995). Catalytic activity toward cumene hydroperoxide (CuOOH) was measured in extracts of Caco-2 cells using modifications of a method described by Habig et al. (1974). Reactions were performed in 0.1 M potassium phosphate (pH 7.0) at 25°C in a total volume of 1 ml. Cell extracts (400 µg) were incubated with CuOOH (1.5 mM), GSH (1 mM), NADPH (0.1 mM), and glutathione reductase (0.3 unit), and the time-dependent disappearance of NADPH was monitored at 340 nm with  = 6.2 mM¹ cm¹. Protein concentration was estimated by the Bradford method (Bradford, 1976). Specific activities were standardized to protein and expressed as nanomoles per minute per milligram.

Statistical Analysis. Data are presented as mean ± S.D. Groups of data were compared by one-way analysis of variance and, if significant, were further evaluated by Fisher's least significant difference test. Significance was established at p values <0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effect of CM from LPS-Stimulated Macrophages on GSTA1/A2 mRNA Levels. Previous studies have shown, by reverse transcription-PCR analysis and cytokine bioassays, that LPS stimulation of the murine monocyte-macrophage cell line (RAW 264.7) results in increased mRNA levels and bioactivities of IL-1, IL-6, and TNF (Adams and Czuprynski, 1994). A similar study revealed, by Northern blot analysis, that levels of IL-1 mRNA were 4-fold higher in RAW 264.7 cells stimulated with LPS in the complete absence of FBS compared with mRNA levels in the presence of 2% FBS (Perera et al., 1998). To determine whether pro-inflammatory cytokine(s) in CM from LPS-activated macrophages affect GSTA1/A2 mRNA levels in Caco-2 cells and to determine which cytokine may contribute to an effect, we compared GSTA1/A2 mRNA levels in Caco-2 cells incubated with CM that was either FBS+ or FBS. Increasing concentrations of CM (25, 50, and 100%) resulted in reductions in GSTA1/A2 mRNA levels that were significantly greater in cells incubated with FBS CM (reductions of 30, 40, and 49%) (p < 0.05) compared with cells cultured in FBS+ CM (reductions of 3, 15, and 18%) (Fig. 1). Treatment of Caco-2 cells with LPS alone (50 ng/ml) had no effect on GSTA1/A2 mRNA levels (data not shown).

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Reduction of GSTA1/A2 mRNA levels in Caco-2 cells exposed to CM from LPS-stimulated murine macrophages cultured in the presence or absence of serum. A, Caco-2 cells were exposed for 24 h to increasing concentrations of CM (25, 50, and 100%) from murine macrophages (RAW 264.7) stimulated with LPS (50 ng/ml) either in the absence (FBS) or presence (FBS+) of 10% fetal bovine serum. CM from unstimulated RAW 264.7 cells was included as a control. Total RNA was isolated and Northern blot analysis was performed with 32P-labeled cDNA that cross-hybridizes with both human GSTA1 and A2 as described under Materials and Methods. B, after autoradiography, signals were quantitated by densitometry, normalized to 7S RNA, and expressed as relative intensity as a percentage of control. Results are the mean of three independent experiments. , significantly different from control; , significantly different from FBS+, p < 0.05.

Effect of CM on GST Protein Levels and Catalytic Activity. To determine whether CM effects also occur at the protein level, we determined GST protein levels in cell extracts of Caco-2 cells cultured in various concentrations of FBS CM by Western blot analysis. Polyclonal antibodies recognized a single 25-kDa band representing human GST protein. Exposure of Caco-2 cells to FBS CM (25, 50, and 100%) for 24 h resulted in a dose-related reduction in GST protein levels to 65, 54, and 50% of control values, respectively (Fig. 2, A and B). The influence of CM on GST activity toward the alpha class substrate cumene hydroperoxide (CuOOH) was also assessed in cell extracts from Caco-2 cells (Fig. 2C). GSTA1/A2 activity was reduced to approximately 58% and 55% (p < 0.05) of control values following exposure to 50% and 100% CM, respectively (Fig. 2C).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effect of CM on GST protein and activity in Caco-2 cells. A, Caco-2 cells were cultured with LPS-stimulated FBS CM at the doses indicated for 24 h. Control cells were cultured with FBS CM (100%) that was not stimulated with LPS. GST protein levels in cell extracts were determined by Western blot analysis as described under Materials and Methods. B, signal intensities of three independent experiments were assessed by densitometry and expressed as a percentage of control intensity. C, GST activity toward cumene hydroperoxide was determined in cell extracts as described under Materials and Methods. Significantly different from control, , p < 0.05, , p < 0.01.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Effect of anti-IL-1 antibody on the ability of CM to reduce GST protein levels in Caco-2 cells. A, Caco-2 cells were exposed to 100% CM (FBS) for 24 h in the presence or absence of neutralizing anti-IL-1 antibody (10 and 100 ng/ml). GST protein levels were determined by Western blot analysis as described under Materials and Methods. B, signal intensity was quantitated by densitometry and expressed as relative densitometry units. , significantly different from cells treated with CM in the absence of neutralizing antibody (p < 0.01).

Effect of IL-1 on GSTA1/A2 mRNA Levels. To confirm the role of IL- in down-regulating GSTA1/A2 expression and to rule out a regulatory role for IL-6 and TNF, we examined the effect of various concentrations of these pro-inflammatory cytokines on GSTA1/A2 mRNA levels in Caco-2 cells treated for 24 h. Northern analysis revealed that treatment with IL-1 resulted in dose-dependent reductions in GSTA1/A2 mRNA levels that were significantly different from controls (Fig. 4), whereas treatment with IL-6 and TNF had no effect (data not shown).

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Dose-related reduction of GSTA1/A2 mRNA levels by IL-1 in Caco-2 cells. A, cells were incubated for 24 h in the presence of recombinant human IL-1 at the doses indicated. GSTA1/A2 mRNA levels were determined by Northern blot analysis as described under Materials and Methods. The Northern blot shown is representative of three experiments showing similar results. B, densitometric analysis of three independent experiments indicated that GSTA1/A2 mRNA levels from all IL-1-treated cells were significantly lower than control cells, , p < 0.05 and , p < 0.01.

View larger version (28K): [in this window] [in a new window]   Fig. 5.   Effect of IL-1ra on the ability of IL-1 to reduce GSTA1/A2 mRNA levels in Caco-2 cells. Caco-2 cells were preincubated for 1 h in the presence or absence of IL-1 receptor antagonist prior to treatment with IL-1 for 24 h as described under Materials and Methods. A, GSTA1/A2 mRNA levels were determined by Northern blot analysis. B, densitometric analysis of signal intensities of three independent experiments. , significantly different from control; , significantly different from IL-1-treated cells (p < 0.05).

Effect of IL-1 on GST Protein Levels and Catalytic Activity. GST expression was significantly reduced by IL-1 to a maximum of approximately 50% of control values at a dose of 0.1 ng/ml (p < 0.05; Fig. 6, A and B) and was not altered further at concentrations as high as 50 ng/ml (data not shown). IL-1-mediated reductions in GST expression were also reflected at the level of catalytic activity where a maximal reduction of 50% was observed at a dose of 0.1 ng/ml (Fig. 6C). Treatment of another human enterocyte cell line (HT-29) with IL-1 (10 ng/ml) also resulted in a 40% reduction in GST protein levels (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Effect of IL-1 on GST protein and catalytic activity in Caco-2 cells. Caco-2 cells were exposed to IL-1 at the doses indicated for 24 h. GST protein levels were reduced in extracts of Caco-2 cells as determined by Western blot analysis as described under Materials and Methods. B, signal intensities of three independent experiments were assessed by densitometry. C, GST activity toward cumene hydroperoxide was determined as described under Materials and Methods. , significantly different from control (p < 0.05).

Effect of IL-1 on GSTA1/A2 mRNA Stability. To determine whether changes in mRNA stability contributed to the altered expression of GSTA1/A2 in IL-1-treated Caco-2, mRNA decay experiments were performed. Northern analysis revealed similar half-life values of GSTA1/A2 mRNA in control (16.5 h) and IL-1-treated cells (17.1 h). This experiment was repeated twice with similar results.

Effect of IL-1 on Transcriptional Regulation of the GSTA1 Gene. To locate the regulatory element within the GSTA1 promoter region that is responsible for the decrease in gene expression by IL-1, luciferase activities were determined in IL-1-treated and untreated cells transfected with various truncated or deleted constructs (Fig. 7A). A clear dose-dependent decrease in LUC activity was observed in cells transfected with 1614 LUC that were subsequently treated with increasing concentrations of IL-1 (Fig. 7A), thereby paralleling the effects on GSTA1 mRNA levels in Caco-2 cells. A 60% decrease in luciferase activity occurred with an IL-1 concentration of 1 ng/ml which was not reduced further with concentrations as high as 50 ng/ml (data not shown). Because of the role of AP-1, AP-2, and HNF-1 sites in mediating IL-1 effects on gene transcription, these sites were targeted for deletion by site-directed mutagenesis. IL-1 caused a significant reduction in luciferase activity in all truncated constructs to a maximal reduction of 80% of untreated controls observed with the 286 LUC construct (p < 0.01; Fig. 7C), suggesting that this region contains a sequence that is negatively regulated by IL-1. Deletions of AP-1 and AP-2 sites from 1614 LUC and 1195 LUC had no effect on IL-1 responsiveness. However, deletion of the HNF-1 site from the 286 LUC abrogated the IL-1 effect, suggesting that HNF-1 is the key site for IL-1 repression of GSTA1 gene expression.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effect of IL-1 on expression of various GSTA1-LUC gene constructs transfected in Caco-2 cells. A, diagrammatic representation of GSTA1-LUC constructs and several known cis-acting elements in the 5'-flanking region of GSTA1. B, relative LUC activities in 1614 LUC transfectants treated with various concentrations of IL-1 as indicated. C, relative LUC activities in cells transfected with truncation and deletion constructs and treated with IL-1 (1 ng/ml). Caco-2 cells were transiently transfected with plasmid constructs containing various lengths of the 5'-flanking region of GSTA1 and constructs in which AP-1, AP-2, and HNF-1 cis-acting sequences had been deleted (i.e., AP-1, AP-2, and HNF-1) as well as cytomegalovirus -galactosidase constructs. Cells were then cultured for 24 h in the presence or absence of IL-1 (10 ng/ml) as described under Materials and Methods. Cells were harvested, and LUC activities were measured and corrected for differences in transfection efficiency based on -galactosidase activity. LUC activities (mean ± S.D.) are expressed as a percentage relative to the 100% control values obtained from corresponding untreated cells transfected with the same plasmid. Results are derived from at least three separate experiments. Statistically significant differences in GSTA1/A2 mRNA levels in cytokine-treated cells and control cells are indicated as , p< 0.05 and , p < 0.01.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The regulation of GST gene expression by inflammatory cytokines has profound biological implications, particularly in tissues such as the colonic epithelium, where GSTs play a critical cytoprotective role against electrophilic metabolites of xenobiotics and lipid peroxidation. Whereas cytokine-mediated down-regulation of cytochromes P450 has been well documented, relatively limited attention has been focused on the detailed regulatory mechanisms of GST isoenzymes by cytokines, and no studies have been conducted in cultured human enterocytes. The results of the present study provide the first evidence of marked down-regulation of human GSTA1/A2 in Caco-2 cells by the pro-inflammatory cytokine IL-1. This IL-1-mediated down-regulation occurs at the level of transcription and involves an HNF-1 response element.

Our finding of reduced GSTA1/A2 mRNA, protein, and catalytic activity in Caco-2 cells exposed to CM from LPS-stimulated macrophages suggests that products of activated phagocytic cells, notably pro-inflammatory cytokines, are important regulators of alpha class GSTs in human enterocytes. Previous studies have shown that exposure of cultured mouse macrophages to LPS results in the production of various pro-inflammatory cytokines including IL-1, IL-6, and TNF (Adams and Czuprynski, 1994). Moreover, IL-1 transcript levels are much higher in LPS-stimulated macrophages cultured in the absence of serum (Perera et al., 1998). In this regard, our finding of marked reduction in GSTA1/A2 mRNA levels in Caco-2 cells exposed to serum-free CM suggests involvement of IL-1 in GSTA1/A2 down-regulation. The repressive role of IL-1 was further established when the effects of CM on GSTA1/A2 expression were attenuated with anti-IL-1 neutralizing antibodies. Furthermore, dose-related reduction of GSTA1/A2 mRNA, protein, and activity levels in Caco-2 cells by human recombinant IL-1 indicates that this cytokine is an important repressor of GST expression. Finally, we demonstrated that effects of IL-1 on GSTA1/A2 expression are mediated via the IL-1 receptor since pretreatment of Caco-2 cells with IL-1ra significantly inhibited repression of transcription by IL-1.

To date, three studies have examined the influence of cytokines on the expression of GST isoenzymes, all of which have been conducted in hepatocytes in primary culture (Langouet et al., 1995; Voss et al., 1996; Maheo et al., 1997). In rat hepatocytes IL-1 resulted in down-regulation of rGSTA2 and M1 due to mRNA destabilization rather than altered rates of gene transcription. Our finding of IL-1-mediated reduction of LUC reporter activity indicates that IL-1 decreases GSTA1 transcription. Moreover, we have excluded the involvement of post-transcriptional events since IL-1 treatment does not alter GSTA1/A2 mRNA half-life in RNA decay experiments. Both positive and negative regulatory regions have been identified in the GSTA1 promoter, including several transcription factor recognition sites such as AP-1, AP-2, HNF-1, a glucocorticoid-responsive element, and several, as yet uncharacterized, negative regulatory and enhancer elements (Suzuki et al., 1994; Lorper et al., 1996; Whalen and Boyer, 1998). Although few functional assays have been performed, there is accumulating evidence to suggest that AP-1 sites may play a role in the regulation of GSTA1 and GSTA2. For example, a study of human hepatocytes recently demonstrated that induction of GSTA1 and GSTA2 by IL-4 paralleled an increase in AP-1 binding activity (Langouet et al., 1995). In mice, up-regulation of GSTA1 by oxidants and phenolic antioxidants occurs via dual AP-1-like sites present in the electrophile-responsive element (Friling et al., 1992). Moreover, it has been demonstrated that the negative influence of IL-1 on microsomal triglyceride transfer protein is mediated via AP-1 and HNF-1 sites (Navasa et al., 1998). For these reasons, we decided to rule out the involvement of AP-1, AP-2, and HNF-1 in IL-1-mediated down-regulation of GSTA1. We eliminated the involvement of AP-1 and AP-2 binding sequences in IL-1-mediated effects since deletion of these sites in luciferase reporter constructs did not alter IL-1 responsiveness. Further mapping of the GSTA1 promoter region with progressive 5' deletion constructs localized the IL-1-responsive region to a sequence within a region 286 bp upstream to the transcriptional start site. This region of the promoter contains an HNF-1 site the deletion of which abrogated IL-1 effects in reporter assays. In the liver, HNF-1 levels are reduced during inflammatory responses (Burke et al., 1994), and numerous genes that contain functional HNF-1 binding sequences are down-regulated under these conditions (Memon et al., 2001). Our findings are supported by results of reporter assays that have shown that binding of HNF-1 to its consensus sequence in the 5'-flanking region of human GSTA2 increases promoter activity (Klone et al., 1990; Clairmont et al., 1994). The possibility that other factors induced by IL-1 may bind to the HNF-1 site and down-regulate GSTA1/A2 remains to be investigated.

IL-1 plays a central role in the development of inflammatory bowel disease (IBD) (Dinarello, 1994). Studies in animal models of colitis (Rachmilewitz et al., 1989) and in human patients with IBD have shown that enhanced production of IL-1 correlates with the degree of mucosal inflammation and necrosis (Ligumsky et al., 1990). The importance of IL-1ra in limiting the clinical severity of IBD has been demonstrated in rodent models of colitis (Cominelli et al., 1990) as well as in human patients (Casini-Raggi et al., 1995). Thus, the balance between IL-1 and IL-1ra may play a critical role in host defenses to mucosal inflammation. It is unclear whether the relationship of IL-1 tissue levels and the degree of mucosal inflammation and necrosis during IBD can be attributed to lack of GST-mediated defense since correlative studies have not yet been performed. However, studies of colonic mucosal antioxidant enzymes in patients with IBD and in animal models of colitis have revealed significant reductions in total glutathione S-transferase activity (Bhaskar et al., 1995; Clapper et al., 1999) and superoxide dismutase activity (Lih-Brody et al., 1996), indicating that antioxidant defenses are compromised during colitis.

In conclusion, we have shown that exposure of cultured human enterocytes to the pro-inflammatory cytokine IL-1 markedly down-regulates the expression of alpha class GSTs in culture by a transcriptional mechanism. We have delineated a region 286 bp upstream to the coding region through which IL-1 reduces GSTA1 transcriptional activity and have implicated an HNF-1 site in this down-regulation. We have also shown that expression of GSTA1/A2 transcripts is reduced by human recombinant IL-1 and secretion of this cytokine by activated mouse macrophages. Further studies are required to fully elucidate the detailed regulatory mechanisms and the biological implications of down-regulation of alpha class GSTs during colitis.