Ethanol metabolism alters interferon gamma signaling in recombinant HepG2 cells

Authors

  • Natalia A. Osna,

    Corresponding author
    1. Liver Study Unit, The Omaha Veterans Affairs (VA) Medical Center, University of Nebraska Medical Center, Omaha, NE
    2. Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE
    • Liver Study Unit, Research Service (151), VA Medical Center, 4101 Woolworth Avenue, Omaha, NE 68105
    Search for more papers by this author
    • fax: 402-449-0604.

  • Dahn L. Clemens,

    1. Liver Study Unit, The Omaha Veterans Affairs (VA) Medical Center, University of Nebraska Medical Center, Omaha, NE
    2. Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE
    3. Department of Pathology/Microbiology, University of Nebraska Medical Center, Omaha, NE
    Search for more papers by this author
  • Terrence M. Donohue Jr.

    1. Liver Study Unit, The Omaha Veterans Affairs (VA) Medical Center, University of Nebraska Medical Center, Omaha, NE
    2. Department of Internal Medicine, University of Nebraska Medical Center, Omaha, NE
    3. Department of Pathology/Microbiology, University of Nebraska Medical Center, Omaha, NE
    4. Department of Biochemistry/Molecular Biology, University of Nebraska Medical Center, Omaha, NE
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Abstract

We previously showed that IFNγ signal transduction was suppressed by ethanol in recombinant HepG2 cells (VL-17A cells), which express alcohol dehydrogenase (ADH) and CYP2E1. We examined the mechanisms by which STAT1 phosphorylation is blocked by ethanol treatment in VL-17A cells. Cells were exposed to 0 or 100 mmol/L ethanol for 72 hours. STAT1 phosphorylation was determined by Western blot after 1 hour IFNγ exposure. Reduction of STAT1 phosphorylation by ethanol was prevented in the presence of 4MP, DAS, or uric acid, indicating that the oxidative products from ethanol metabolism were partly responsible for suppression of STAT1 phosphorylation. Ethanol exposure decreased STAT1 tyrosine phosphorylation, whereas serine phosphorylation on the protein was unchanged. These effects of ethanol were mimicked by the peroxynitrite (PN) donor, SIN-1, which also blocked tyrosine, but not serine phosphorylation, on STAT1. When cells expressing either ADH (VA-13 cells) or CYP2E1 (E-47 cells) were exposed to ethanol, both ADH- and CYP2E1-generated products reduced STAT1 phosphorylation. In addition, SOCS1, a negative regulator of IFNγ signaling and which is degraded by the proteasome, was stabilized by ethanol treatment, presumably because of inhibited proteasome activity. Furthermore, SIN-1 treatment elevated SOCS1 levels in VL-17A cells, indicating that PN has a role in SOCS1 elevation. In conclusion, under conditions of ethanol-elicited oxidative stress, PN prevents STAT1 phosphorylation by stabilization of SOCS1, and possibly by nitration of tyrosine residues in STAT1 protein. (HEPATOLOGY 2005;42:1109–1117.)

Interferon gamma (IFNγ) is a major pro-inflammatory cytokine that strongly contributes to innate and adaptive immune responses. In liver, hepatocytes are naturally exposed to IFNγ released by intrahepatic producers of this cytokine, including natural killers, natural killer T-lymphocytes, and activated T-cells. IFNγ promotes antigen presentation by activating peptide-trimming enzymes1, 2 and in this way regulates the ability of virus-infected hepatocytes (e.g., hepatitis C and B) to present viral peptides for recognition and elimination by cytotoxic T lymphocytes.3–5 IFNγ-induced proteasome-dependent breakdown of proteins/peptides and antigen presentation are only part of the possible IFNγ-mediated effects in liver cells. The cytokine also regulates cell proliferation, apoptosis, secretion of chemokines, and expression of adhesion molecules, and it has antifibrotic effects.6, 7 IFNγ plays a pivotal role in the induction of acetaminophen- and ConA-mediated toxicities in the liver.8, 9 IFNγ signals are mainly transduced via the Janus kinase I-signal transducer and activator of transcription 1 pathway (JAK-STAT1), which is responsible for most of the above-mentioned effects.1, 6, 10 STAT1 phosphorylation is a key step in the initiation of IFNγ signaling and precedes the activation of IFNγ-dependent genes. STAT1 can be phosphorylated on both tyrosine (Tyr) and serine (Ser) residues. Generally, IFNγ-induced phosphorylation on Tyr 701 of STAT1 activates protective functions of viable cells, whereas Ser 727 phosphorylation on the same protein transduces pro-apoptotic signals.11

IFNγ signaling is negatively regulated by several factors. The most potent of these are the suppressors of cytokine signaling 1 and 3 (SOCS1 and 3), which block STAT1 phosphorylation at the level of JAKs upstream from STAT1 phosphorylation.12 In addition, SOCS1 functions as an E3 ligase for phosphorylated JAK2, catalyzing its ubiquitylation for subsequent degradation by the proteasome.13

Previously, we showed that a 3-day exposure to ethanol impairs IFNγ signaling by diminishing STAT1 phosphorylation in VL-17A cells that express alcohol dehydrogenase (ADH) and cytochrome P450 2E1 (CYP2E1).14 Ethanol-elicited alteration of IFNγ signaling in VL-17A cells also prevented proteasome activation by the cytokine.14 Similar effects on IFNγ signal transduction were obtained with freshly isolated hepatocytes treated acutely with ethanol.15, 16

This study was conducted to ascertain the mechanisms by which STAT1 phosphorylation is blocked in ethanol-metabolizing VL-17A cells. The results indicated that ethanol metabolism reduced STAT1 phosphorylation, and that peroxynitrite and acetaldehyde may each have a role in this process.

Abbreviations

FNγ, interferon gamma; STAT1, signal transducer and activator of transcription 1; SOCS1-suppressors of cytokine signaling 1; ADH, alcohol dehydrogenase; CYP2E1, cytochrome P450 2E1; DMEM, Dulbecco's modified Eagle's medium; 4MP, 4-methylpyrazole; DAS, diallyl sulfide; UA, uric acid; PBS, phosphate-buffered saline; GSH, glutathione; TCA, trichloracetic acid; 4-NC, 4-nitrocatechol; BSO, L-buthionine sulfoximine; PN, peroxynitrite.

Materials and Methods

Reagents and Media.

High-glucose Dulbecco's modified Eagle medium (DMEM) and fetal bovine serum were purchased from Invitrogen (Carlsbad, CA). Human recombinant interferon gamma (IFNγ) was from PeproTech, Inc. (Rocky Hill, NJ). Antibodies to phosphorylated STAT1 (one specific for Tyr 701 and the other for Ser 727) were from Cell Signaling (Beverly, MA); antibodies to the STAT1 protein were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Antibody to SOCS1 was from US Biological (Swampscott, MA). Other reagents (all of analytical grade quality) were from Sigma (St. Louis, MO).

Cell Lines.

For this study, we used recombinant HepG2 cells, which constitutively express ethanol-metabolizing enzymes: VA-13 cells are ADH-expressing cells17; E-47 cells are CYP2E1-expressing cells (a gift of Dr. Arthur Cederbaum); VL-17A cells are both ADH- and CYP2E1-expressing cells.14 All cells were cultured in DMEM in the presence of selective antibiotics (Zeocin-for VA-13 and VL-17A cells and G418-for E-47 and VL-17A cells, each at 400 μg/mL). Cell culture conditions were the same as described elsewhere.14, 17, 18

Cell Treatments.

Cells were plated onto 6-well plates in DMEM with 10% fetal bovine serum. After overnight attachment, the medium was changed to serum-free DMEM with or without 100 mmol/L ethanol, and cells were incubated for up to 72 hours with or without 100 mmol/L ethanol in the presence or absence of 5 mmol/L 4-methylpyrazole (4MP), 10 μmol/L diallyl sulfide (DAS), or 100 μmol/L uric acid (UA). An adherent sealing membrane was overlaid onto each plate to minimize ethanol evaporation. After the 72-hour ethanol treatment, cells were exposed to IFNγ (10 ng/mL) for 1 hour before harvest. In some experiments, interferon alpha (IFNα), 250 U/mL, was used to activate signal transduction.

Hepatocyte Preparation.

Hepatocytes were obtained from male Wistar rats fed the Lieber-DeCarli control or ethanol liquid diet for 5 weeks. Cells were prepared by the collagenese perfusion method.19 Hepatocytes were incubated in suspension in serum-free DMEM for 1 hour in 6-well plates before use in experiments. Experiments consisted of an hour-long incubation of cells in the presence or absence of 25 mmol/L ethanol. Cells were then collected, pelleted by centrifugation, washed in phosphate-buffered saline (PBS), and lysed, as described below, for detection of pSTAT1/STAT1.

Detection of STAT1 and PhosphoSTAT1.

For harvesting, hepatoma cells were washed in PBS and then were lysed directly on the plates by freezing and thawing in 0.5 mL lysis buffer [40 mmol/L Tris-HCl (pH 7.4), 1 mmol/L EDTA, 10 mmol/L sodium pyrophosphate, 2 mmol/L sodium orthovanadate, 1 mmol/L phenylmethylsulfonyl fluoride, 100 mmol/L sodium fluoride, 1% Triton X-100, 2 μg aprotinin/mL). Detection of phospho (p) STAT1 and STAT1 was performed by Western blot as described by Osna et al.14 For pSTAT1, 2 distinct antibodies were used, 1 specific for phosphotyrosine 701, the other for phosphoserine 727 . The immunoreactive protein bands were quantified by densitometry, and the pSTAT1/STAT1 densitometric ratio was calculated.

Assay of Total Glutathione (Total GSH).

VL-17A cells, parental HepG2 cells (incubated in culture for 24-48 hours), and freshly isolated hepatocytes (1 × 106) were harvested and lysed by sonication. A portion of each lysate was treated with an equal volume of 10% (w/v) trichloroacetic acid (TCA) to precipitate proteins. A separate portion of each lysate was used for protein quantification. Total GSH (GSH+GSSG) levels were detected in TCA-soluble fractions of cell lysates by the enzymatic recycling method using glutathione reductase and Ellman's reagent.20

CYP2E1 Catalytic Activity.

The microsome pellet derived from the 105,000g centrifugation of cell lysates (prepared by sonication in PBS and subjected to low-speed centrifugation) was re-suspended in 250 μL PBS containing 20% (v/v) glycerol. CYP2E1 activity was measured by hydroxylation of p-nitrophenol, as described originally by Koop21 and modified by Chen and Cederbaum.22 In this micro-assay, the reaction mixture had a final volume of 100 μL and was incubated for 1 hour at 37°C. The reaction was terminated by addition of 30% TCA. Insoluble proteins were removed by centrifugation, and color development was initiated by addition of 10 μL 10 N NaOH to the supernatant. Formation of 4-nitrocatechol (4-NC), the product of the reaction, was measured by its absorbance at 515 nm. This was compared with the absorbance obtained using known quantities of 4-NC. One CYP2E1 enzyme unit catalyzed the formation of 1 nanomole 4-NC per hour. Specific activity is expressed as units per milligram protein. Protein concentration was measured by the dye binding technique of Bradford,23 using reagents from Bio Rad (Hercules, CA). CYP2E1 activity in VL-17A cells was associated with the presence of a 52-kd immunoreactive protein that was detected on Western blots with an anti-CYP2E1 antibody using lysates from these cells. These same cells also contained the 42-kd subunit corresponding to alcohol dehydrogenase, which was detected with anti-ADH (courtesy of Dr. Michael Felder, University of South Carolina) (data not shown).

Statistical Analyses.

Data are expressed as mean values ± standard deviation. Comparisons for significance between 2 groups were determined by Student t test. Multiple comparisons were determined by one-way ANOVA, using a Tukey post-hoc test. A probability value of .05 or less was considered significant.

Results

Effects of Ethanol Treatment on IFNγ-Induced STAT1 Phosphorylation in VL-17A Cells.

In this study, we further addressed the question of whether reduction in STAT1 phosphorylation can be attributed to the effects of ethanol alone or its metabolism in VL-17A cells. Exposure of ADH+/CYP2E1+VL-17A cells to 100 mmol/L ethanol for 3 days suppressed STAT1 phosphorylation on Tyr 701 by 50% (Fig. 1). Only attached cells were lysed after all experimental treatments and used for further analysis. Viability of these cells remained the same in control and ethanol-treated samples. This suppression by ethanol was prevented when 4MP was present, suggesting that ethanol metabolism causes the reduction in STAT1 phosphorylation (Fig. 1). We observed the same pattern when VL-17A cells were treated with IFNα after a 72-hour exposure to 100 mmol/L ethanol. STAT1 phosphorylation was 61.2% ± 10.3% of control (P = .0003, 4 determinations in 2 experiments).

Figure 1.

Effects of ethanol metabolism on STAT1 phosphorylation in VL-17A cells. VL-17A cells treated or not with 100 mmol/L ethanol in the presence or absence of 5 mmol/L 4MP, 10 μmol/L DAS, or 100 μmol/L UA for 72 hours and then exposed to IFNγ for 1 hour before harvest. Cells were lysed, immunoblotted, and probed for pSTAT1 and STAT1, as described in Methods. The results are presented as percent of control (IFNγ alone) of pSTAT1/STAT1 ratio. The data represent the mean ± SD from four independent experiments. * indicates a significant difference (P < .05) between IFNγ-treated samples and other treatments. Inset: Representative blot of immunoreactive pSTAT1 and STAT1. STAT1, signal transducer and activator of transcription 1; DAS, diallyl sulfide; IFNγ, interferon gamma.

A lower concentration of ethanol (25 mmol/L) did not affect STAT1 phosphorylation in parental HepG2 or in VL-17A cells, whereas in isolated rat hepatocytes, this ethanol concentration efficiently suppressed STAT1 phosphorylation. The pSTAT1/STAT1 ratio in IFNγ-exposed control hepatocytes derived from 3 pair-fed rats was 0.65 ± 0.09 compared with 0.33 ± 0.11 ratio obtained after 1 hour 25 mmol/L ethanol treatment (P = .007). The pSTAT1/STAT1 ratio after IFNγ treatment of hepatocytes from 3 ethanol-fed rats was 0.026 ± 0.008 (P = .0002 vs. pair-fed controls). However, in our hands, spontaneous STAT1 phosphorylation in hepatocytes was also evident, even in the absence of IFNγ (up to 50 % of IFNγ-induced levels), which blurred the differences between experimental groups. In addition, with continued culture in vitro for more than 18 hours, dedifferentiation of hepatocytes occurred, resulting in the loss of ethanol metabolizing ability, as indicated by significant declines in ADH and CYP2E1 activities. These results tempered our enthusiasm for using freshly isolated hepatocytes for these studies.

In VL-17A cells, we observed little-to-no effect of 100 mmol/L ethanol on STAT1 phosphorylation when the duration of ethanol treatment was less than 3 days (Fig. 2A), even if that was accompanied by an ethanol-elicited induction of CYP2E1 (Fig. 2B). The apparent resistance of STAT1 phosphorylation to the suppressive effects of ethanol may have been due to total glutathione levels in VL-17A cells, which were 3.4-fold higher than in rat hepatocytes and 2.5-fold higher than in parental HepG2 cells (Fig. 2C). Furthermore, there was no indication that ethanol treatment decreased total GSH in VL-17A cells (data not shown). To determine whether glutathione protects STAT1 phosphorylation, VL-17A cells were treated with 100 mmol/L ethanol in the presence or absence of the GSH depleting agent, L-buthionine sulfoximine (BSO).24 Whereas a 2-day treatment with ethanol did not normally affect STAT1 phosphorylation, a significant reduction of STAT1 phosphorylation in VL-17A cells became evident after 2 days of ethanol exposure when the total GSH level was decreased threefold (compared with controls) by treatment with BSO (Fig. 3A-B).

Figure 2.

Time course of ethanol effects on IFNγ-induced STAT1 phosphorylation (A), CYP2E1 activity (B), and comparison of total GSH levels in VL-17A cells, HepG2 cells, and hepatocytes (C). (A) Cells were treated or not with 100 mmol/L EtOH for 24, 48, and 72 hours and then exposed to IFNγ (IFNγ bar shows no exposure to ethanol, E 24 h+IFNγ shows exposure to ethanol for 24 hours, E48h+IFNγ is 48 hours' ethanol exposure and 72h+IFNγ is 72 hours' ethanol exposure before treatment with IFNγ). pSTAT1 and STAT1 were determined as described in Methods. Data presented are from 3 independent experiments as mean ± SD. * indicated a significant difference (P < .05) between IFNγ-treated samples and IFNγ+EtOH-treated samples. (B) Two sets of VL-17A cells were treated with 100 mmol/L EtOH for 0, 24, 48, and 72 hours followed by 1 hour exposure to IFNγ. CYP2E1 activity was determined in 1 set of cells. In the other set of cells, the pSTAT1/STAT1 ratio was measured in lysates at each time point. Data are presented as CYP2E1 specific activity and the percent control (IFNγ alone) of the pSTAT1/STAT1 densitometric ratio. (C) Cells (HepG2 and VL-17A cells) were plated at 1 × 10 6/well in 6-well plates for 24 to 48 hours, then harvested, lysed, and precipitated with 5% TCA (4 experiments with duplicate determinations). Hepatocytes were freshly isolated from 9 rats, each maintained on the Lieber-DeCarli control diet. Total glutathione was determined, as described in Methods. *Significant (P < .05) difference between VL-17A cells and other cells. STAT1, signal transducer and activator of transcription 1; GSH, glutathione; IFNγ, interferon gamma; TCA, trichloracetic acid.

Figure 3.

Effects of BSO on total GSH levels and on STAT1 phosphorylation in control and ethanol-treated VL-17A cells. (A) VL-17A cells were treated with 0, 0.1, or 1 mmol/L BSO for 48 hours. Data from 3 independent experiments with duplicate determinations are presented. *Significant difference (P < .05) between control and BSO-treated samples. (B) VL-17A cells were exposed or not to either 100 mmol/L EtOH or 0.1 mmol/L BSO or to both agents for 48 hours and then treated with IFNγ. Data from 3 experiments are presented. *Significant (P < .05) difference between IFNγ-treated samples compared with other samples in set. Inset: representative blot. BSO, L-buthionine sulfoximine; GSH, glutathione; STAT1, signal transducer and activator of transcription 1; IFNγ, interferon gamma.

STAT1 Phoshorylation in VA-13 (ADH+) Cells and in E-47 (CYP2E1+) Cells.

These experiments were designed to determine whether ethanol exposure affected IFNγ signaling in cells that metabolize ethanol via either ADH (VA-13 cells) or CYP2E1 (E-47 cells) and to determine whether ADH- or CYP2E1-generated ethanol metabolites alone can affect STAT1 phosphorylation.

In VA-13 cells, 72-hour ethanol treatment reduced STAT1 phosphorylation by 85%, and this reduction was prevented by incubation with 4MP (Fig. 4A). In E-47 cells, ethanol-elicited suppression of STAT1 phosphorylation depended on the basal CYP2E1 activity. We observed a reduction in STAT1 phosphorylation when basal CYP2E1 activity (before ethanol exposure) was 17 ± 5 units/mgP (n = 3). However, in some experiments (n = 3), when E-47 cells had significantly lower CYP2E1 activity (7.3 ± 3 units/mgP, P < .05), no effect of ethanol on STAT1 phosphorylation was seen (Fig. 4B).

Figure 4.

Effects of ethanol on STAT1 phosphorylation in VA-13 cells and in E-47 cells. (A) VA-13 cells were treated or not with 100 mmol/L ethanol for 72 hours and then exposed to IFNγ for 1 hour. pSTAT1/STAT1 ratio was determined by Western blot. Data from 3 experiments are presented as percentage of control (IFNγ-treated) pSTAT1/STAT1 ratio. *Significant (P < .05) difference between IFNγ-treated samples and ethanol treatments. Inset: Representative blot. (B) E-47 cells with higher (17 ± 5.2 units/mgP) or lower (7.3 ± 2.5 units/mgP) CYP2E1 specific activity were treated or not with 100 mmol/L ethanol for 72 hours and then exposed to IFNγ for 1 hour. pSTAT1/STAT1 ratio was determined by Western blot. Data from 6 experiments (3 for cells with high CYP2E1 activity and 3 for cells with low CYP2E1 activity) are presented as the percentage of control of pSTAT1/STAT1 ratio. *Significant (P < .05) difference between the indicated value and IFNγ-treated samples. Inset: representative blot. STAT1, signal transducer and activator of transcription 1; IFNγ, interferon gamma.

Effects of Ethanol Treatment on SOCS1 Expression in VL-17A, E-47, and VA-13 Cells.

We tested whether ethanol metabolites induce SOCS1, a negative regulator of IFNγ-mediated STAT1 phosphorylation. In VL-17A cells, ethanol exposure enhanced SOCS1 expression, and this effect was numerically higher when ethanol was combined with IFNγ (Fig. 5A). Elevation of SOCS1 was blocked by inclusion of 5 mmol/L 4MP or 10 μmol/L DAS in the medium. A similar elevation in SOCS1 levels was observed in E-47 cells (Fig. 5B). However, VA-13 cells exhibited no change in SOCS1 levels after treatment with both IFNγ and ethanol (Fig. 5B).

Figure 5.

Effects of ethanol on SOCS1 levels in VL-17A (A), E-47, and VA-13 (B) cells. (A) VL-17A cells were treated as indicated and then all (except controls) were exposed to IFNγ for 1 hour. SOCS1/beta-actin ratio was determined by Western blot. (B) E-47 and VA-13 cells. Cells were treated or not with 100 mmol/L EtOH for 72 hours and then exposed to IFNγ for 1 hour. SOCS1/beta-actin ratio was determined by Western blot. Data from 3 experiments are presented as fold increase of SOCS1/beta-actin ratio. *Significant (P < .05) difference between the control samples (c) and the specific treatments. Inset: Representative blot. SOCS1, suppressors of cytokine signaling 1; IFNγ, interferon gamma.

The JAK2–SOCS1 complex is degraded by the proteasome.25, 26 To determine whether SOCS1 induction by ethanol involved stabilization of SOCS1, VL-17A cells were treated or not with ethanol and IFNγ in the presence or absence of the proteasome inhibitor MG132. Treatment with ethanol and IFNγ or IFNγ and MG132 elevated SOCS1 levels over those of controls, but no further enhancement occurred by the combination of these agents (Fig. 6).

Figure 6.

Effects of ethanol and MG132 on SOCS1 levels. VL-17A cells were pretreated or not with 100 mmol/L ethanol for 3 days and then were further exposed or not to 10 μmol/L MG132 for 5 hours or to IFNγ for 1 hour before harvest. SOCS1/beta-actin ratio was determined by Western blot. Data from 3 experiments are presented as the mean ± SD fold increase of SOCS1/beta-actin ratio. *Significant (P < .05) difference between control samples and other treatments. Inset: Representative blot. SOCS1, suppressors of cytokine signaling 1; IFNγ, interferon gamma.

Influence of CYP2E1-Generated Products and Peroxynitrite on STAT1 Phosphorylation in VL-17A Cells.

The ethanol-elicited blockade of STAT1 phosphorylation was prevented by the CYP2E1 inhibitor, DAS, or the antioxidant UA, which scavenges peroxynitrite (PN) (Fig. 1), indicating the contribution of CYP2E1-generated products (ROS and RNS) in the regulation of IFNγ signaling. By themselves, neither DAS nor UA affected STAT1 phosphorylation (data not shown). Although ethanol metabolism blocked phosphorylation of STAT1 on Tyr 701, it did not affect phosphorylation of Ser 727 on the same protein (Fig. 7A). PN, a reaction product of superoxide and nitric oxide, causes protein nitration on Tyr residues but does not affect Ser residues.27 To determine whether PN affects IFNγ signaling, we tested the effects of SIN-1, a PN donor, on IFNγ-induced STAT1 phosphorylation in VL-17A cells. Incubation for 24 hours with either 0.1 or 1 mmol/L SIN-1 blocked STAT1 phosphorylation on Tyr but not on Ser residues (Fig. 7B-C). These results with SIN-1 mimicked the pattern obtained in ethanol-treated VL-17A cells.

Figure 7.

Effects of ethanol and SIN-1 on IFNγ-induced STAT1 phosphorylation on Tyr and Ser residues in VL-17A cells. (A) VL-17A cells were treated or not with 100 mmol/L EtOH for 72 hours and then exposed to IFNγ for 1 hour before harvest. pSTAT1 Tyr, pSTAT1 Ser, and STAT1 were each determined by Western blot. Data from 3 independent experiments are presented as the mean ± StDev percent of control of pSTAT1/STAT1 ratio *Significant difference (P < .05) between IFNγ-treated and IFNγ+ethanol-treated samples. Inset: Representative blot. (B) and (C) VL-17A cells were treated or not for 24 hours with 0.1 and 1 mmol/L SIN-1 and then exposed to IFNγ for 1 hour. STAT1 phosphorylation on Tyr 701 (A) and Ser 751 (B) was determined by Western blot. Data from 3 experiments are presented as percent of control of pSTAT1/STAT1. *Significant (P < .05) difference between IFNγ treated and SIN-1+IFNγ treated samples. Inset: Representative blot. IFNγ, interferon gamma; STAT1, signal transducer and activator of transcription 1.

Discussion

We have shown previously that IFNγ-activated STAT1 phosphorylation is suppressed by ethanol in ADH+/CYP2E1+ VL-17A cells, but not in HepG2 cells or in VI-R2 cells (the latter are HepG2 cells transfected with empty vectors).14 Here, we confirmed that STAT1 phosphorylation was blocked by ethanol metabolism, because STAT1 phosphorylation proceeded normally when 4MP, an inhibitor of ethanol metabolism by ADH and CYP2E1, was included in the medium. Furthermore, DAS, a CYP2E1 inhibitor, as well as UA, an antioxidant and PN scavenger, also prevented ethanol-elicited suppression of STAT1 phosphorylation. These data indicate that CYP2E1- or ADH-generated ethanol metabolites and PN are involved in suppression of STAT1 phosphorylation. The suppressive effects of ethanol metabolism on IFNγ signaling were confirmed by the results of some successful experiments on freshly isolated hepatocytes, where we observed up to a 50% reduction in STAT1 phosphorylation after a 1-hour exposure to 25 mmol/L ethanol. Similar results were reported by Chen et al.15 However, these authors were unable to demonstrate the ethanol-elicited reduction in STAT1 phosphorylation in hepatocytes cultured for longer periods or in ADH+ HepG2 (VA-13) cells. This may have been due to the shorter time of ethanol exposure (1 hour) in the recombinant VA-13 cells they used.

VL-17A cells were resistant to the suppressive effects of ethanol on IFNγ signal transduction and required 3 days of exposure to 100 mmol/L ethanol to decrease STAT1 phosphorylation. One could argue that 100 mmol/L ethanol treatment does not mimic pharmacological in vivo conditions in alcohol-consuming humans or during oral ethanol feeding of rodents, in which the blood ethanol concentration is generally reported in the range from 25 to 60 mmol/L (115-276 mg/dL).28, 29 However, the rodent intragastric feeding model achieves peak blood and urinary alcohol concentrations of 76 to 109 mmol/L (350-500 mg/dL).30, 31 Also, dysregulation of signal transduction mechanisms and cell damage most likely depends on the level of oxidative stress, which is an imbalance between the generation of cellular oxidants and the ability of the cell to neutralize them.32 Here, we observed that VL-17A cells had threefold higher levels of total GSH than in rat hepatocytes. GSH is the most abundant and a crucial intracellular antioxidant and an efficient PN scavenger.33 Thus, to effect a change in STAT1 phosphorylation, VL-17A cells required treatment with high concentrations of ethanol. Our other studies with VL-17A cells have revealed that acetaldehyde levels are increased by twofold in cells exposed to 100 mmol/L ethanol compared with those exposed to 25 mmol/L ethanol (Donohue et al., unpublished observations). These results, and the facts that our observed changes in STAT1 phosphorylation were attributed to ethanol metabolism, and that reduction of GSH by BSO treatment hastened the ethanol-elicited reduction of STAT1 phosphorylation, indicate that ethanol-induced oxidative stress was responsible for impaired signaling of IFNγ in CYP2E1-expressing cells. Similar effects of arachidonic acid on p38 MAPK signaling, aggravated by BSO, were recently reported in E-47 cells.24

VL-17A cells express both ADH and CYP2E1 enzymes. To clarify whether only CYP2E1-related products by themselves cause a reduction in STAT1 phosphorylation or whether ADH-generated metabolites (e.g., acetaldehyde) also efficiently suppress STAT1 phosphorylation, we examined the effects of ethanol on signal transduction in VA-13 (ADH+) cells and E-47 (CYP2E1+) cells. In E-47 cells, ethanol exposure reduced STAT1 phosphorylation when basal CYP2E1 catalytic activity was comparatively high, whereas in E-47 cells with low basal CYP2E1 activity, STAT1 phosphorylation proceeded normally in the presence of 100 mmol/L ethanol, indicating that a relatively high level of CYP2E1-generated products is necessary to suppress IFNγ signaling. However, in VA-13 cells, STAT1 phosphorylation was suppressed by ethanol even in the absence of CYP2E1, indicating that acetaldehyde may suppress STAT1 phosphorylation. This observation has been further supported by recent data (not shown) indicating that STAT1 phosphorylation was decreased by 38% or 43% after 3 days of exposure of VL-17A cells to either 200 or 400 μmol/L acetaldehyde, respectively. In VA-13 cells, even in the absence of CYP2E1, PN still may have been generated in mitochondria during ethanol treatment; elevated 3NT adducts were observed in lysates of ethanol-treated VA-13 cells compared with unexposed controls (data not shown). Furthermore, acetaldehyde contributes to mitochondrial damage by decreasing mitochondrial GSH and inducing of ROS/RNS formation.34, 35 Mitochondria are primary targets of ethanol-induced oxidative stress.36–39 Thus, acetaldehyde may cause increased superoxide production in mitochondria, which is followed by enhanced PN production, because iNOS activity and nitric oxide release are also upregulated by ethanol treatment.40, 41 However, the amount of PN generated in VA-13 cells in the absence of CYP2E1 may have been insufficient to suppress proteasome function. An alternative mechanism for suppression of STAT1 phosphorylation in ADH-expressing cells is the binding of acetaldehyde or malondialdehyde-acetaldehyde to lysine residues, one of which, Lys 705, may influence phosphorylation of neighboring Tyr 701 in STAT1. In addition, acetaldehyde may have directly bound to tyrosine residues, as this interaction has been reported with other proteins.42

The negative regulator of IFNγ signaling, SOCS1, was induced by ethanol treatment in VL-17A cells. SOCS1 levels depend on proteasome activity because the JAK2-SOCS1 complex is degraded by both 20S and 26S forms of the proteasome.43–45 SOCS1 level was indeed higher in VL-17A cells treated with both IFNγ and ethanol, whereas STAT1 phosphorylation was decreased. Ethanol treatment also increased SOCS1 in E-47 cells with high CYP2E1 activity, but not in VA-13 cells that express no CYP2E1, indicating that CYP2E1-generated products are primarily responsible for SOCS1 elevation, presumably due to its stabilization. We previously demonstrated that, in VL-17A cells, CYP2E1-generated products contributed to ethanol-mediated suppression of proteasome function.46 We have also shown that proteasome activity is sensitive to PN. Higher PN concentrations suppress proteasome function.47 Thus, we propose that PN (derived in VL-17A cells during ethanol metabolism) suppressed proteasome activity, thereby stabilizing SOCS1 from degradation. When we treated VL-17A cells with IFNγ and MG132, SOCS1 was elevated to a level comparable to that in cells exposed to IFNγ and ethanol for 3 days. Furthermore, no additive effects of both MG132 and ethanol on SOCS1 levels were seen, indicating that ethanol enhances SOCS1 by the same mechanism as MG132. Furthermore, in VL-17A cells, PN, generated by SIN-1, increased SOCS1 levels and suppressed STAT1 phosphorylation. We have simultaneously observed the reduction of STAT1 phosphorylation and induction of SOCS1 levels by the same dose of SIN-1 (not shown). Thus, PN and possibly other CYP2E1-generated products stabilized SOCS1 by preventing its proteasome-dependent degradation, thereby blocking STAT1 phosphorylation. However, ethanol-induced SOCS1 elevation is not the sole reason for suppression of STAT1 phosphorylation; we have evidence that transfection of VL-17A cells with SOCS1 siRNAs only partially reversed the ethanol-elicited suppression on STAT1 phosphorylation (N. Osna, unpublished results).

It is noteworthy that ethanol metabolism suppressed phosphorylation on Tyr 701, but did not affect phosphorylation on Ser 751 of STAT1. Differential regulation of Tyr and Ser phosphorylation by ethanol is pathogenically important because it is believed that in most cells the protective effects of IFNγ are activated via Tyr STAT1 phosphorylation, whereas harmful effects (e.g., apoptosis induction) are related to Ser STAT1 phosphorylation.11, 48 Moreover, as a powerful oxidant and nitrating agent, PN modifies Tyr residues by reacting to form 3NT adducts without affecting Ser residues.49 The inhibition of STAT1 phosphorylation obtained in VL-17A cells after treatment with the PN donor SIN-1 resembled that elicited by ethanol, namely, a reduction in tyrosine but not in serine phosphorylation. We have preliminary evidence that accumulation of 3NT adducts occurred in STAT1 from ethanol- and IFNγ-treated VL-17A cells and that the formation of 3NT on this protein may prevent its phosphorylation. In macrophages, nitration of tyrosine residues on STAT1 impairs the response to IFNγ.50 In addition to induction of SOCS1 levels and possible nitration of STAT1 tyrosine residues, other reasons for reduction of STAT1 phosphorylation by ethanol may be elevation of SOCS3 or SH2-containing phosphatases that also alter IFNγ signaling. That remains to be investigated.

Based on the data presented here, a proposed scheme of the regulation of IFNγ-induced STAT1 phosphorylation by ethanol metabolism is presented in Fig. 8.

Figure 8.

The proposed effects of EtOH metabolism on IFNγ-induced STAT1 phosphorylation in VL-17A cells. Gray arrows show positive effects, black arrows show negative effects, X shows blockade. Ethanol (EtOH) is metabolized by ADH and CYP2E1 to acetaldehyde (Ach) and ROS/RNS, respectively. These reactive intermediates can damage mitochondria, thereby increasing superoxide production. Because ethanol also induces iNOS, causing nitric oxide (NO) formation, there is increased a formation of PN (ONOO−), a reaction product of Omath imageγ-induced STAT1 phosphorylation. Stabilization of SOCS1 is one reason for impaired IFNγ signaling. In addition, PN reacts with STAT1 on Tyr residues, making them unavailable for phosphorylation. IFNγ, interferon gamma; STAT1, signal transducer, and activator of transcription 1. IFNγ, interferon gamma; STAT1, signal transducer and activator of transcription 1.

We conclude that IFNγ-induced STAT1 phosphorylation on Tyr 701 is reduced by ethanol metabolism in VL-17A cells. Both ADH- and CYP2E1-generated products can independently block STAT1 phosphorylation, but only CYP2E1-generated products can elevate SOCS1 levels. When CYP2E1 catalytic activity is high and cells are exposed to high doses of ethanol (100 mmol/L), mimicking severe oxidative stress in vivo, PN suppresses STAT1 phosphorylation by stabilizing SOCS1, a negative regulator of IFNγ signaling.

Acknowledgements

The authors thank Sandy Todero, Ronda White, Andrea Stieren, and John Evans for excellent technical assistance. We also wish to thank Dr. A. Cederbaum for providing E-47 cells and Dr. Carol Casey for providing freshly isolated rat hepatocytes.

Ancillary

Advertisement