Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
Potential conflict of interest: Nothing to report.
Processing of peptides for antigen presentation is catalyzed by antigen-trimming enzymes, including the proteasome and leucine aminopeptidase. Oxidative stress suppresses proteasome function. We hypothesized that in liver cells, processing of antigenic peptides is altered by ethanol metabolism. To address this issue, soluble extracts of ethanol-metabolizing VL-17A cells treated with 100 mM ethanol or left untreated were incubated with C-extended or N-extended 18-27 HBV core peptides. Peptide cleavage was measured by recovery after HPLC. Ethanol exposure to VL-17A cells increased CYP2E1 and decreased proteasome peptidase activities. The latter effect was prevented by treatment of cells with inhibitors, 4-methylpyrazole and diallyl sulfide. Ethanol treatment of VL-17A cells also reduced the activity of leucine aminopeptidase (LAP). Consequently, cleavage of both C-extended and N-extended peptides by cytosolic extracts was suppressed by pretreatment of cells with ethanol. Treatment of cells with interferon gamma, which enhances proteasome activity, did not reverse the effects of ethanol. Ethanol exerted similar effects on WIFB cells, indicating that its effects are not unique to one cell type. Conclusion: Ethanol metabolism suppresses activities of antigen-trimming enzymes, thereby decreasing the cleavage of C-extended and N-extended peptides. This defect may potentially result in decreased MHC class I–restricted antigen presentation on virally infected liver cells. (HEPATOLOGY 2007;45:53–61.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
The proteasome is a multicatalytic enzyme crucially important for proteolysis of oxidized and misfolded proteins, peptides and short-lived signal transduction factors. In vivo, the proteasome exists in equilibrium as 2 particles: the 20S proteasome degrades substrate proteins with no prior ubiquitylation and the 26S form degrades ubiquitylated proteins in an ATP-dependent manner.
Both 20S and 26S proteasomes generate peptides for MHC class I-restricted antigen presentation.1, 2 Processing of antigens is a necessary step for recognition by cytotoxic T lymphocytes (CTLs) of virally infected cells. To be recognized by CTLs, “nonself” antigens are processed to peptides of 8–10 amino acids, which then bind to MHC class I molecules in the endoplasmic reticulum.3 Initial processing of antigenic proteins in the cytoplasm is catalyzed by the proteasome and/or leucine aminopeptidase (LAP).4 Both enzymes are activated by the proinflammatory cytokine interferon gamma (IFNγ),5 which induces formation of the immunoproteasome that cleaves antigenic proteins to uniformly sized peptides for presentation.
The foregoing series of events occurs after hepatitis B or C infections during which CTLs eliminate HBV-infected or HCV-infected hepatocytes are based on ability of immune system to clear infected hepatocytes. Epidemiological studies have provided supporting evidence of severe and rapid progression of chronic hepatitis B and C in alcoholics.6 Moreover, animal studies demonstrate that ethanol administration to mice prevents activation of CTLs in response to immunization with NS5 HCV protein.7 Thus, the link between prolongation of viral hepatitis and ethanol consumption may be explained by inefficient presentation of viral peptides–MHC class I complex on hepatocytes.
Because processing of peptides for antigen presentation depends, in part, on proteasome function, the effects of oxidative stress inducers, including ethanol, on the peptidase activities of the proteasome were addressed in the present study. Proteasome function is apparently susceptible to oxidative stress because it forms adducts with protein carbonyls, 4-hydroxynonenal (4-HNE) and 3-nitrotyrosine derived from peroxynitrite.8–10 Other studies have revealed a reduction of proteasome activity by ethanol metabolism in liver and in recombinant hepatoma (VL-17A) cells.11 Furthermore, IFNγ signaling, which activates proteasome function via the JAK-STAT1 pathway, is blocked by exposure of VL-17A cells to ethanol.12, 13
The latter findings led us to hypothesize that ethanol-induced oxidative stress impairs the function of antigen-trimming enzymes, thereby limiting their ability to process antigenic peptides, which potentially can be presented by liver cells. Here, we sought to determine whether prior exposure of cells to ethanol affected the subsequent cleavage of extended peptides to produce an HBV core peptide. Our experiments were conducted using soluble fractions of ethanol metabolizing VL-17A hepatoma cells with the HLA-A2 phenotype and the 18–27 AA HBV core peptide. Here, we report that in VL-17A cells, ethanol metabolism suppresses the major peptidase activities of 20S and 26S proteasomes and of LAP, thereby altering proteasomal cleavage of C-extended 18–27 core peptide (FLPSDFFPSVRDL), and of N-extended 18–27 core peptide (ELLSFLPSDFFPSV) cleavage by LAP. Neither enzyme activity was completely restored by IFNγ treatment after cells were exposed to ethanol.
4MP, 4-methylpyrazole; AA, antimycin A; BSO, L-buthionine sulfoximine; Cht-L, chymotrypsin-like activity; CTL, cytotoxic T lymphocyte; CYP2E1, cytochrome P450 2E1; DAS, diallyl sulfide; HBV, hepatitis B virus; IFNγ, interferon gamma; LAP, leucine aminopeptidase; MHC, major histocompatibility complex; SIN1, 3-morpholinosydnonimine hydrochloride; SOCS1, suppressors of cytokine signaling 1; STAT1, signal transducer and activator of transcription 1; T-L, trypsin-like activity.
Materials and Methods
Reagents and Media.
High-glucose Dulbecco's modified Eagle medium (DMEM) and fetal bovine serum (FBS) were purchased from Invitrogen (Carlsbad, CA). Human recombinant IFNγ was from PeproTech Inc. (Rocky Hill, NJ). Extended HBV core peptides, FLPSDFFPSVRDL and ELLSFLPSDFFPSV(98% pure) were synthesized at VA Medical Center Peptide Facility, Omaha, NE. Antibody to phosphorylated STAT1 (Tyr 701) was from Cell Signaling (Beverly, MA); antibody to the STAT1 protein was from Santa Cruz Biotechnology Inc. (Santa Cruz, CA). Other reagents, all of analytical grade quality, were from Sigma (St. Louis, MO).
For this study, we used a recombinant cell line VL-17A, (derived from HepG2 cells), which constitutively expresses the ethanol-metabolizing enzymes, alcohol dehydrogenase (ADH) and cytochrome P450 2E1 (CYP2E1). Cell culture conditions have been described elsewhere.11, 13 In some experiments, we also used the ethanol-metabolizing hybrid cell line, WIF-B, using the culture conditions already described.14
VL-17A cells were plated onto 75-cm flasks, at 2 × 106 cells per flask, in DMEM containing 10% FBS. After overnight attachment, the DMEM with or without 100 mM ethanol was applied, and cells were incubated for up to 72 hours in the presence or absence of IFNγ (10 ng/ml). After incubation, cells were detached by trypsinization. Lysates were prepared by sonication in PBS and cytosols were obtained by centrifugation for 1 hour at 105,000g and supplemented with 20% (wt/vol) glycerol.
WIF-B cells were cultured in Coon's modified medium supplemented with 3.75% FBS as previously described. The cells were cultured for 7–10 days to obtain maximal polarized phenotype followed by 48 hours of treatment with media containing 10 mM HEPES in the presence or absence of ethanol (25, 50, and 100 mM) in sealed dishes, with or without IFNγ. After incubation, cells were detached, as before, sonicated in PBS and cytosols were obtained as described above.
Detection of Peptide Cleavage.
Cytosol protein (100 μg/mL) derived from cell lysates was mixed with 5 nM N- or C-extended peptide in 50 mM Tris-HCl (pH 8.5) and 5 mM MgCl2 in a total volume of 100 μl The reaction was stopped by 20% trichloracetic acid. A 50 μl aliquot of each supernatant was subjected to reverse-phase HPLC on a Vydac C18 monomeric column equilibrated with 0.1% trifluoroacetic acid, with a flow rate of 1 ml/minute. Peptides were eluted with a linear acetonitrile gradient ranging from 20% to 40% (vol/vol). Peptides were detected by their absorbance at 214 nm and their quantities calculated by integration of the peptide peaks on chromatograms. The percent of remaining (uncleaved) peptide was calculated as integration units obtained from the intact peptide peak after incubation with cytosol, divided by integration units obtained from identically treated unincubated samples and multiplied by 100.
Cytochrome P4502E1 (CYP2E1) Catalytic Activity.
CYP2E1 activity was detected in microsome fractions of cell lysates by the formation of 4-nitrocatechol detected spectrophotometrically, as described.12 Specific activity is expressed as units per milligram protein.
Proteasome chymotrypsin-like (Cht-L) and trypsin-like (T-L) peptidase activities were detected by in vitro fluorometric assay and the fluorometric substrates Suc-LLVY-AMC or LSTR-AMC, respectively.15 Chymotrypsin-like activity was also detected in situ by measuring hydrolysis of the membrane-permeable substrate, methoxysuccinyl-phe-leu-phe-7-amido-4-trifluoromethyl coumarin (MeO-Suc-FLF-AFC).16
Leucine Aminopeptidase Activity.
LAP activity was measured in cytosols, which were subjected to an additional 5 hours centrifugation at 105,000g (to remove the proteasome) followed by a 15 minute pretreatment with the proteasome inhibitor MG132 (100 μM). Catalytic assay of LAP used the fluorogenic LAP-AMC substrate (Bachem) at 200 μM in 1 ml of 50 mM Tris-HCl (pH 8.5), 5 mM MgCl2.17 After 75 minutes incubation at 37°C, the reaction was stopped with 500 μl of 2% SDS, and fluorescence was measured at 389 nm (excitation) and 440 nm (emission).
Detection of STAT1 and PhosphoSTAT1.
Cells were lysed on the plate by freezing and thawing and were subjected to SDS-PAGE on 12% gels. Proteins were transferred from gels to nitrocellulose membranes. Phospho-STAT1 and STAT1 were then detected as described.13
Detection of Immunoproteasome Subunits and PA28.
Immunoproteasome subunits and a 20S proteasome activator, PA28, were detected by Western blot, using specific antibodies (Affiniti, UK). The densitometric ratios of immunoproteasome subunit or PA28 to that of β-actin (loading control) were calculated.
Data are expressed as mean values ± standard deviation. Multiple comparisons for significance were determined by 1-way ANOVA, using a Tukey post-hoc test. Comparison between 2 groups was performed by Student t test. A P ≤ 0.05 was considered significant.
Effects of Ethanol on Proteasome Activity in VL-17A Cells.
The generation of peptides for antigen presentation requires proteasome Cht-L activity and T-L activity, which cleaves peptides on the C-termini of hydrophobic or basic residues, respectively.4 We therefore examined whether prior ethanol exposure affected either Cht-L (Suc-LLVY-AMC hydrolysis) or T-L (LSTR-AMC hydrolysis) peptidase activities. Cytosols were incubated in the presence and absence of 0.5 mM ATP, in order to detect both 26S and 20S proteasome activities, respectively. Prior treatment with ethanol, antimycin A (AA), and L-buthionine sulfoximine (BSO) suppressed ChT-L and T-L activities of both the 20S and 26S forms of the proteasome (Fig. 1A). The patterns of ethanol-induced suppressive effects were similar for both forms of the enzyme, but the Cht-L activity exhibited greater sensitivity to ethanol than the T-L activity. Furthermore, ethanol exposure decreased the Cht-L activity of the 26S proteasome to a greater extent (77% reduction) than that of the 20S proteasome (50% reduction). BSO and AA similarly affected proteasome activities. Suppression of Cht-L activity by ethanol was blocked by simultaneous incubation with 4-methylpyrazole (4MP), an inhibitor of ethanol metabolism. Similar results were obtained when cells were treated with diallyl sulfide (DAS), a specific inhibitor of CYP2E1 activity (Fig. 1B). Additionally, suppression of proteasome activity coincided with the ethanol-elicited elevation of CYP2E1 activity (Fig. 1C).
Immunoproteasome Subunits and PA28 in VL-17A Cells.
Both the 20S proteasome activator, PA28, and the immunoproteasome subunit, LMP2, were enhanced by treatment of cells with IFNγ, but this effect was blocked in the presence of 100 mM ethanol (Fig. 2A,B). Treatment with 4MP prevented this ethanol-elicited blockade on PA28 and LMP2 expression.
C-Extended Peptide Cleavage in VL-17A Cells.
We measured the cleavage of the C-extended peptide FLPSDFFPSVRDL after incubation with cytosol fractions from lysates of VL-17A cells. The parent C-extended peptide had a retention time of 32 minutes (Fig. 3A). Incubation of this peptide with cytosol for 1 hour caused a decrease in the recovery of the latter peak and the appearance of a new peak with a retention time of ∼24 minutes, indicating partial hydrolysis of the parent peptide and its conversion to a more hydrophilic form (Fig. 3B). The LAP inhibitor, bestatin (100 μM) inhibited C-extended peptide cleavage by about 20%. Inclusion of the proteasome inhibitor MG132 (10 μM) prevented disappearance of the peptide that eluted at 32 minutes. Thus, cleavage of C-extended peptide was principally catalyzed by the proteasome (Fig. 3C,D).
Figure 4 summarizes the results of multiple cleavage experiments using cytosol fractions derived from cells treated with 100 mM ethanol or other inducers of oxidative stress, or left untreated. Cytosols derived from ethanol-treated cells stabilized uncleaved peptide 5-fold compared with cytosols from untreated cells (Fig. 4A). This effect was similar to that of BSO, which decreases GSH synthesis, 3-morpholinosydnonimine hydrochloride (SIN1), which generates peroxynitrite, and antimycin A (AA), which inhibits respiratory activity and generates peroxides in mitochondria (Fig. 4B).
Treatment with IFNγ, by itself, did not affect C-extended peptide cleavage (Fig. 4A). When cells were exposed to both ethanol and IFNγ, the peptide hydrolysis was partially accelerated, compared with fractions of cells exposed to ethanol alone.
N-Extended Peptide Cleavage by Cytosols of VL-17A Cells.
Cleavage of ELLSFLPSDFFPSV was measured in vitro using cytosols from VL-17A cells. The retention time of the intact chromatographed peptide was 43 minutes (Fig. 5A). Incubation with cytosol for 1.5 hrs caused a substantial decline in the amount of the parent peptide, which was inhibited partially (30%) by 10 μM MG132 and completely with 100 μM bestatin (Fig. 5B, C, D). The latter results indicate that LAP catalyzed hydrolysis of this peptide. When cells were exposed to ethanol, the effect on subsequent cleavage of N-extended peptide was less prominent than that on the C-extended peptide. The former results were highly variable, and did not achieve statistical significance when control and ethanol-treated cells were compared (Fig. 6A).
LAP Activity in VL-17A Cells.
LAP activity in cytosols of VL-17A cells was increased by IFNγ treatment, but this effect was abrogated by ethanol exposure (Fig. 6B). In parental HepG2 cells, LAP responded neither to ethanol, nor to IFNγ (Fig. 6C).
Proteasome, CYP2E1, and IFNγ Signaling Studies in Ethanol-Treated WIFB Cells.
Using WIFB cells, we confirmed the effects of ethanol on proteasome activity and C-extended peptide cleavage previously seen in VL-17A cells. Ethanol treatment enhanced CYP2E1 activity and suppressed proteasome function, decreasing C-extended peptide cleavage (Fig. 7A). Additionally, in cytosols from these cells, we observed an ethanol-induced dose-dependent decrease in pSTAT1/STAT1 ratio, which was reversed by incubation with DAS (Fig. 7B).
This study addressed the effects of ethanol and other inducers of oxidative stress on the peptide-processing machinery in liver cells. To our knowledge, this investigation represents the first attempt to examine whether oxidative stress impairs the initial steps of peptide processing required for MHC class I–restricted antigen presentation. To mimic core antigen cleavage in HBV-infected hepatocytes, hepatoma cells were used as the source of trimming enzymes by incubating soluble fractions derived from these cells with C-extended HBV core peptide, FLPSDFFPSVRDL or its N-extended ortholog, ELLSFLPSDFFPSV.
These extended peptides were chosen based on the premise that a 18–27 core decapeptide, FLPSDFFPSV, is generated in a proteasome-dependent manner according to proteasomal cleavage predictions (PAProC program, www.paproc.de) and that the core peptide is a known T cell epitope, recognized in the context of HLA-A2 by CTLs.18 In this study, the major focus was on the proteasome and LAP, which catalyze the upstream steps of peptide cleavage. Here, we observed that cleavage of C-extended peptide was catalyzed principally by the proteasome and that hydrolysis of N-extended peptide was predominantly catalyzed by LAP in cytosolic fractions of VL-17A cells, confirming results obtained with other C-extended and N-extended peptides in HeLa cell extracts.1, 5 Furthermore, prior treatment of VL-17A cells with ethanol inhibited the subsequent hydrolysis of C-extended peptide. We noted that both 20S and 26S proteasomes showed a similar response to oxidative stress (but with different magnitude of effects). However, in our cell-free system, the substrate peptides were not ubiquitylated prior to cleavage, because the assay buffer contained no ATP and used HBV core peptides had no lysine residues available for isopeptide linkage with ubiquitin, which is a prerequisite for protein degradation by 26S proteasome. Ethanol-induced suppression of 20S proteasome function by products of ethanol metabolism correlated with the inhibition of C-extended peptide cleavage. Such inhibition was not observed when ethanol metabolism was blocked by 4MP or by DAS, indicating that the products of CYP2E1 catalysis suppressed proteasome function. A CYP2E1-dependent decrease in proteasome activity, that we demonstrated in 2 different liver-derived cell lines, has also been identified in other studies.9, 19 Generation of peroxides by mitochondria due to AA treatment20 and reduced protection from intracellular oxidants due to BSO-induced glutathione depletion also decreased proteasome function, indicating that the suppression of proteasome activity is a common occurrence following generation of intracellular oxidants.
Suppression of LAP activity by ethanol was demonstrated in VL-17A, but not in HepG2 cells, indicating that this effect is exerted by ethanol metabolism. However, cleavage of N-extended peptide was not significantly affected by ethanol treatment of VL-17A cells, showing that other enzymes may be involved in N-extended peptide cleavage.
IFNγ induces proteasome activity. However, when cells were exposed to ethanol, IFNγ treatment neither fully reversed the inhibitory effects of ethanol on C-extended peptide cleavage and proteasome function, nor did it enhance N-extended peptide cleavage and LAP activity. Furthermore, PA28, a 20S proteasome activator and an LMP2 immunoproteasome subunit, both of which are normally induced by IFNγ, failed to respond to the cytokine in ethanol-treated VL-17A cells. Concurrently, blockade of ethanol metabolism by 4MP prevented ethanol-induced suppression of PA28 and LMP2. Similar suppressive effects of ethanol on the immunoproteasome in macrophages were reported recently.21, 22
Previously, we reported that ethanol metabolism suppresses STAT1 phosphorylation in VL-17A cells, resembling the effects of ethanol in freshly isolated hepatocytes.12, 13 Here, similar effects of ethanol metabolism on IFNγ signaling were confirmed in WIFB cells, which were sensitive to even lower (25 mM) concentrations of ethanol than VL-17A cells. The products of ethanol metabolism, generated by CYP2E1 catalysis, are responsible for impaired signal transduction because suppressive effects on STAT1 phosphorylation were blocked by DAS treatment. The blockade of IFNγ signaling by ethanol and other inducers of oxidative stress may be partially explained by elevation of SOCS1, which is a proteasome substrate that is apparently stabilized due to compromised proteasome function. Thus, oxidative stress prevents effective IFNγ signal transduction, thereby blocking activation of antigen-trimming enzymes in liver cells by IFNγ.
The mechanism we propose to account for our findings is illustrated in Fig. 8. Normally, C-extended and N-extended peptides are cleaved by either proteasome or LAP in the cytosol and can bind to the HLA-A2 groove. However, CYP2E1-generated products of ethanol metabolism suppress proteasome activity, causing stabilization of a negative regulator of STAT1 phosphorylation, SOCS1, which prevents IFNγ signaling. Consequently, induction by IFNγ of PA28 and the immunoproteasome subunit LMP2 and further peptide cleavage is suppressed. Similarly, ethanol metabolism decreases the activation of LAP function by IFNγ bringing about further impairment of antigen processing. Finally, faulty trimming of antigenic peptides and their presentation may be compromised.
In summary, we conclude that: (1) C-extended peptide (FLPSDFFPSVRDL) is cleaved by the proteasome and N-extended peptide (ELLSFLPSDFFPSV) is cleaved by LAP in cytosols of VL-17A cells; (2) Ethanol and other inducers of oxidative stress partially suppress peptide cleavage by decreasing proteasome and LAP activities; and (3) ethanol metabolism prevents activation by IFNγ of immunoproteasome, proteasome, and LAP due to suppression of IFNγ signaling.
Overall, the results of this work lead us to suggest that the impaired function of the peptide-trimming enzymes proteasome and LAP hypothetically may account for faulty MHC class I–restricted antigen presentation by liver cells.
We thank John Evans for excellent technical assistance. We are also grateful to Drs. Frederick Hamel and Thomas Freeman for help with HPLC.