Potential conflict of interest: Nothing to report.
Peroxiredoxins (Prxs) are peroxidases that catalyze the reduction of reactive oxygen species (ROS). The active site cysteine residue of members of the 2-Cys Prx subgroup (Prx I to IV) of Prxs is hyperoxidized to cysteine sulfinic acid (Cys-SO2) during catalysis with concomitant loss of peroxidase activity. Reactivation of the hyperoxidized Prx is catalyzed by sulfiredoxin (Srx). Ethanol consumption induces the accumulation of cytochrome P450 2E1 (CYP2E1), a major contributor to ethanol-induced ROS production in the liver. We now show that chronic ethanol feeding markedly increased the expression of Srx in the liver of mice in a largely Nrf2-dependent manner. Among Prx I to IV, only Prx I was found to be hyperoxidized in the liver of ethanol-fed wildtype mice, and the level of Prx I-SO2 increased to ≈30% to 50% of total Prx I in the liver of ethanol-fed Srx−/− mice. This result suggests that Prx I is the most active 2-Cys Prx in elimination of ROS from the liver of ethanol-fed mice and that, despite the up-regulation of Srx expression by ethanol, the capacity of Srx is not sufficient to counteract the hyperoxidation of Prx I that occurs during ROS reduction. A protease protection assay revealed that a large fraction of Prx I is located together with CYP2E1 at the cytosolic side of the endoplasmic reticulum membrane. The selective role of Prx I in ROS removal is thus likely attributable to the proximity of Prx I and CYP2E1. Conclusion: The pivotal functions of Srx and Prx I in protection of the liver in ethanol-fed mice was evident from the severe oxidative damage observed in mice lacking either Srx or Prx I. (HEPATOLOGY 2011)
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.
Chronic ethanol consumption increases the production of a variety of reactive oxygen species (ROS), including superoxide, H2O2, lipid peroxides, and peroxynitrite in the liver, an effect that has been implicated as a major factor in the pathogenesis of alcohol-induced liver disease.1-4 Accumulation of ROS induces the expression of antioxidant enzyme genes through activation of a cis-acting element known as the antioxidant-responsive element (ARE). Transcription factors that transmit the oxidative stress signal to the ARE include nuclear factor erythroid 2-related factor 2 (Nrf2) and activator protein-1 (AP-1).5 Exposure of animals to chronic ethanol feeding or overexpression of cytochrome P450 2E1 (CYP2E1) in hepatocytes thus increases the expression of Mn2+-dependent superoxide dismutase (SOD) and heme oxygenase-1 by activating Nrf2 or AP-1 or both.3, 6, 7
Peroxiredoxins (Prxs) are a family of peroxidases that reduce peroxides and peroxynitrite with the use of reducing equivalents provided by thiol-containing proteins.8 Prxs contain a conserved cysteine residue (designated the peroxidatic cysteine, CP) in the NH2-terminal region that is the primary site of oxidation by H2O2. Mammalian tissues express six distinct Prx gene products (Prx I to VI), which can be divided into three subgroups: 2-Cys, atypical 2-Cys, and 1-Cys.8 Members of the 2-Cys subgroup (Prx I to IV) contain an additional conserved cysteine (designated the resolving cysteine, CR) in the COOH-terminal region, whereas Prx V and VI, members of the atypical 2-Cys and 1-Cys subgroups, respectively, do not contain this second conserved cysteine. Prx isoforms show distinct intracellular distributions, with Prx I and II being localized predominantly in the cytosol, Prx III being restricted to mitochondria, Prx IV being found mainly in the endoplasmic reticulum (ER), Prx V being detected in the cytosol and mitochondria, and Prx VI being present in the cytosol.8
In the catalytic cycle of 2-Cys Prx enzymes, which exist as homodimers, CP-SH is first converted to cysteine sulfenic acid (CP-SOH) by a peroxide. The unstable sulfenic intermediate then reacts with the CR-SH of the other subunit of the dimer to form a disulfide, which is subsequently reduced by thioredoxin to complete the catalytic cycle.8 As a result of the slow rate of its conversion to a disulfide, the sulfenic intermediate is occasionally oxidized further to cysteine sulfinic acid (CP-SO2H), leading to inactivation of peroxidase activity.9 This hyperoxidation is reversed by the ATP-dependent enzyme sulfiredoxin (Srx).10-12 Prx V and VI are less sensitive to hyperoxidation than are 2-Cys Prxs, and the hyperoxidized forms of Prx V and VI are not reduced by Srx.13 The Srx gene contains a functional ARE, which is activated via the AP-1 pathway in various cell types exposed to nitric oxide, 3′-5′-cyclic adenosine monophosphate (cAMP), or 12-O-tetradecanoylphorbol 13-acetate14, 15 or via the Nrf2 pathway in cortical neurons treated with a dithiolethione5 or in mouse lung exposed to hyperoxia.16 We now show that Srx is induced in the liver of ethanol-fed mice and demonstrate roles for both Srx and 2-Cys Prxs in protection of the liver from ethanol-induced oxidative damage.
Materials and Methods
(See Supporting Information for Materials and Methods.)
Chronic Ethanol Feeding Induces Srx Expression in Mouse Liver.
Enzymes responsible for the elimination of ROS in mammalian cells include SODs, catalase, glutathione peroxidases, and Prxs. Ethanol feeding increases the expression of MnSOD in rat liver.3 We investigated the effect of chronic ethanol feeding on the expression of Prx I to VI and Srx in the liver of male mice. Mice were maintained on a control diet or an ethanol-containing diet for 2 weeks, after which the expression of Srx and Prx I to VI at the protein and messenger RNA (mRNA) levels was measured by immunoblot analysis and quantitative reverse transcription (RT) polymerase chain reaction (PCR) analysis, respectively. Ethanol feeding increased the abundance of both Srx protein (≈10-fold) (Fig. 1A,B) and Srx mRNA (≈6-fold) (Fig. 1C), but it had no substantial effect (changes of <30%) on the amounts of the six Prx proteins or mRNAs (Fig. 1A-C). The abundance of Prx VI protein and mRNA was previously shown to be reduced by factors of 1.5 and 1.9, respectively, in the liver of ethanol-fed mice.17 Consistent with previous observations,7 the amount of CYP2E1 was increased (Fig. 1D) and oxidative damage was evident from an increased level of 4-hydroxy-2-nonenal (4-HNE) protein adduct (Fig. 1E) in the liver of mice subjected to chronic ethanol treatment. To examine the effect of acute ethanol exposure on the expression of Srx we administered a single oral dose of ethanol (5 g/kg)18 to mice. The amounts of Srx protein and mRNA in the liver remained largely unchanged at 6 and 72 hours after alcohol treatment. The acute ethanol exposure also had a minimal effect on the levels of CYP2E1 and no effect on the levels of sulfinic Prx I, 4-HNE protein adduct, and protein 3-nitrotyrosine (3-NT) (Supporting Information Fig. 1C,D).
We also examined the effect of ethanol on Srx expression in Hepa1c1c7 (mouse hepatoma) cells, HepG2 (human hepatoma) cells, E47 cells (HepG2 cells that constitutively express CYP2E1), and C34 cells (control for E47 cells). The abundance of Srx protein was not affected by exposure of any of these cells to 100 mM ethanol for 18 hours, whereas the protein levels of Srx were increased slightly in E47 cells (Supporting Information Fig. 2B). Similar treatment of primary mouse hepatocytes also showed no significant effect of ethanol on Srx and CYP2E1 expression (Supporting Information Fig. 2C). It was shown previously that HepG2 cells resist the adverse effect of ethanol because the cells contain a very low amount of CYP2E1.39
Ethanol-Induced Srx Expression in the Liver Is Mediated by Nrf2.
Chronic ethanol feeding of mice was previously shown to increase Nrf2 expression ≈2-fold in the liver.6 The role of Nrf2 in ethanol-induced Srx expression in the liver was investigated with the use of Nrf2-deficient mice. The amount of Srx protein in the liver was increased ≈9-fold by ethanol feeding in Nrf2+/+ mice but only ≈2-fold in Nrf2−/− mice (Fig. 2B,C). Ethanol feeding also induced similar changes in the hepatic abundance of Srx mRNA (Fig. 2D). In addition, the basal level of Srx mRNA was reduced by ≈50% in Nrf2−/− mice compared with that in Nrf2+/+ animals (Fig. 2D). These results suggested that the Nrf2-ARE pathway plays a key role in the induction of Srx in the liver of ethanol-fed mice. The observation that ethanol still induced an ≈2-fold increase in Srx expression in the liver of Nrf2−/− mice, however, suggested that the AP-1-ARE pathway might also contribute to this effect.
Ethanol Feeding Induces Hyperoxidation of Prx I and III, but Not That of Prx II or IV, in the Liver of Srx−/− Mice.
Srx is responsible for reduction of the hyperoxidized forms of 2-Cys Prx enzymes (Prx I to IV) generated during elimination of peroxides. Hyperoxidized 2-Cys Prxs can be detected by immunoblot analysis with antibodies generated in response to a sulfonylated peptide modeled on the conserved peroxidatic cysteine residue (CP). Given that the amino acid sequences surrounding CP are identical for 2-Cys Prx enzymes, the antibodies react with all of these hyperoxidized proteins.13 To investigate the role of Srx in ethanol-fed mice, we generated Srx−/− mice (Supporting Information Fig. 3). Srx+/+ and Srx−/− mice were then subjected to chronic ethanol feeding, after which liver proteins were subjected to immunoblot analysis with antibodies to Srx, to sulfinic forms of 2-Cys Prxs (Prx-SO2), and to Prx I to IV (Fig. 3). Prx I and Prx II, which differ by only one amino acid residue in size, cannot be separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), whereas the molecular sizes of Prx I/II, III, and IV differ sufficiently to allow such separation. In NIH 3T3 cells that had been exposed to 100 μM H2O2 for 10 minutes the cytosolic enzymes Prx I and II as well as the mitochondrial enzyme Prx III were found to be completely hyperoxidized (see below), whereas hyperoxidation of the ER-localized Prx IV was not detected (not shown). Extracts of the H2O2-treated cells are included as a standard for Prx I/II-SO2 and Prx III-SO2 in Figure 3. The intensity of the immunoreactive band that migrated at a position corresponding to that of Prx I/II-SO2 was increased slightly in the liver of ethanol-fed Srx+/+ mice compared with that in control Srx+/+ mice, whereas it was increased markedly in ethanol-fed Srx−/− mice relative to that in Srx−/− mice fed the control diet (Fig. 3). The intensity of the band corresponding to Prx III-SO2 was also increased slightly in ethanol-fed Srx−/− mice compared with those in the other three groups. Prx IV-SO2 was not detected in any of the four groups of mice. Neither ablation of Srx nor ethanol feeding substantially affected the levels of 2-Cys Prxs (Prx I to IV) in the liver (Fig. 3).
The extent of Prx hyperoxidation was evaluated further by two-dimensional (2D) PAGE followed by immunoblot analysis (Supporting Information Fig. 5). 2D-PAGE separates not only Prx I and II but also their covalently modified forms. Hyperoxidation of a cysteine residue induces an acidic (leftward) shift in the position of proteins on 2D gels. 2D-PAGE and subsequent immunoblot analysis with antibodies to Prx I revealed two major spots of Prx I for Srx+/+ mice fed the ethanol-containing or control diets as well as for Srx−/− mice fed the control diet, whereas three major spots were detected for ethanol-fed Srx−/− mice (Supporting Information Fig. 5A). Alignment with the Prx I-SO2 spot of H2O2-treated NIH 3T3 cells revealed that only the leftmost spot for ethanol-fed Srx−/− mice corresponded to hyperoxidized Prx I. The middle spot that was observed in all four groups of mice appeared to be due to another type of modification, with 2-Cys Prx proteins being known to undergo phosphorylation, acetylation, and COOH-terminal truncation.19-22 The intensities of the three spots suggested that 30% to 50% of Prx I was hyperoxidized in the liver of ethanol-fed Srx−/− mice. A faint spot of Prx I-SO2 at the position corresponding to the leftmost Prx I spot of ethanol-fed Srx−/− mice was also apparent for Srx+/+ mice fed the ethanol-containing diet. These results thus indicated that ethanol feeding results in a slight accumulation of Prx I-SO2 in the liver of Srx+/+ mice and that Srx ablation markedly potentiates this effect. Prx II appeared as a single spot in the liver of all four groups of mice, with no spot corresponding to that of Prx II-SO2 in H2O2-treated NIH 3T3 cells being detected (Supporting Information Fig. 5B), suggesting that Prx II was not susceptible to hyperoxidation by ethanol-induced ROS. Prx III appeared as one major and two minor spots, of which the leftmost spot could not be seen in some samples (Supporting Information Fig. 5C). The immunoblot of H2O2-treated NIH 3T3 cells with antibodies specific for Prx-SO2 indicated that Prx III-SO2 can be found in the positions of both minor spots. A faint spot of Prx III-SO2 was detected only in the liver of ethanol-fed Srx−/− mice, suggesting that Prx III is vulnerable to hyperoxidation by ethanol-induced ROS but that Prx III-SO2 accumulates only in the absence of Srx. These data also suggested that hyperoxidation was not the only cause of the acidic shift of Prx III in 2D gels.
Prx I Is Selectively Hyperoxidized Because It Is Localized Together with CYP2E1 on the Cytosolic Side of the ER Membrane.
Ethanol-induced ROS production is mediated in part by CYP2E1, most of which is anchored at the ER membrane through its hydrophobic NH2-terminal domain, with the catalytic domain being exposed to the cytosol.4 Although both Prx I and II are cytosolic proteins and Prx II is more sensitive to hyperoxidation by ROS produced extracellulary or intracellulary than is Prx I,20 our present data indicated that ethanol-induced ROS in the liver mediate hyperoxidation of Prx I but not that of Prx II. We therefore investigated the possibility that Prx I might be associated with the ER membrane. Immunoblot analysis of cytosolic and microsomal fractions prepared from mouse liver revealed that Prx I was present in both fractions, whereas Prx II and Prx IV were exclusively found in the cytosolic and microsomal fractions, respectively (Fig. 4A). The amounts of Prxs in the two fractions were estimated with the use of glutathione S-transferase (GST)-Prx fusion proteins as standards: Prx I was estimated to be present at ≈4 μg per milligram of cytosolic protein and ≈3 μg per milligram of microsomal protein, whereas Prx II was detected at ≈2 μg per milligram of cytosolic protein and Prx IV at ≈3 μg per milligram of microsomal protein (Fig. 4A).
To determine the topology of the ER-associated Prx I, we treated microsomes isolated from mouse liver with proteinase K in the absence or presence of Triton X-100 and then subjected the microsomal proteins to immunoblot analysis with antibodies to Prx I, to CYP2E1, to protein disulfide isomerase (PDI), and to ERP72. PDI and ERP72 are known to be localized to the ER lumen. Incubation of the microsomes with proteinase K in the absence of Triton X-100 resulted in the proteolysis of Prx I and CYP2E1, whereas PDI and ERP72 remained intact (Fig. 4B). However, in the presence of Triton X-100 all four proteins underwent proteolysis. These results indicated that, like CYP2E1, Prx I is localized at the cytosolic side of the ER membrane.
Srx was readily detected in the microsomal fraction of the liver from ethanol-fed Srx+/+ mice (Fig. 4C), suggesting that Srx also translocates to the site of Prx I hyperoxidation. Srx is a cytosolic protein but translocates into mitochondria under conditions that result in Prx III hyperoxidation.23 The mitochondrial translocation of Srx was also observed in the liver of ethanol-fed mice (Fig. 4D).
Ethanol-Induced Liver Damage Is Exacerbated by Ablation of Srx or Prx I.
Ethanol-induced oxidative damage in the liver includes protein modifications such as carbonylation, formation of 4-HNE adducts, and nitration of tyrosine to give 3-nitrotyrosine (3-NT).1 Ethanol feeding induced an ≈2-fold increase in protein carbonylation in the liver of Srx+/+ mice. Such carbonylation was increased ≈1.7-fold by ablation of Srx and was increased an additional ≈1.5-fold by ethanol feeding in Srx−/− mice (Supporting Information Fig. 6A). Immunohistochemical and immunoblot analyses also revealed that the amounts of 4-HNE and 3-NT adducts in the centrilobular areas of the liver were increased by ethanol feeding or Srx depletion and were increased further by ethanol feeding in Srx−/− mice (Fig. 5, Supporting Information Fig. 6C). The levels of 4-HNE or 3-NT were increased ≈1.2-fold by ablation of Srx and was increased an additional ≈1.5-fold by ethanol feeding in Srx−/− mice (Supporting Information Fig. 6B,D). Given that the damaging effect of Srx ablation was likely attributable to Prx I hyperoxidation, we investigated the role of Prx I in ethanol-induced liver damage by generating Prx I−/− mice (Supporting Information Fig. 4). As expected, immunohistochemical and immunoblot analyses of 4-HNE and 3-NT revealed that the extent of ethanol-induced liver damage was substantially greater in Prx I−/− mice than in Prx I+/+ mice (Fig. 6, Supporting Information Fig. 7B). The levels of 4-HNE or 3-NT were increased ≈1.2-fold by ablation of Prx I and was increased an additional ≈1.2- to ≈1.5-fold by ethanol feeding in Prx I−/− mice (Supporting Information Fig. 7A, C).
We have shown that chronic ethanol feeding induces Srx expression at both the mRNA and protein levels in the liver of wildtype mice. Chronic exposure of Srx−/− mice to ethanol resulted in the hyperoxidation of a substantial proportion of Prx I and a smaller proportion of Prx III in the liver, whereas hyperoxidized Prx II was not detected. Given that all 2-Cys Prxs undergo unavoidable hyperoxidation during catalytic function and that the inactivated (hyperoxidized) enzymes can be reactivated only by the action of Srx, the observed formation of Prx I-SO2 and Prx III-SO2 suggests that, among the four 2-Cys Prxs, Prx I and III participate in the reduction of ethanol-induced ROS. The key antioxidant functions of Srx, the guardian of 2-Cys Prxs, and of Prx I were evident from the marked oxidative damage induced by chronic ethanol feeding in the liver of Srx−/− or Prx I−/− mice.
The pathways contributing to ethanol-induced ROS production in the liver include the induction of CYP2E1,24-27 inhibition of mitochondrial function,28-30 stimulation of Kupffer cells,31 and activation of NADPH oxidase at the plasma membrane,32 and of xanthine oxidase in the cytosol.33 The generated ROS trigger the dissociation of Nrf2 from Kelch-like ECH-associated protein 1 (Keap1), and the dissociated Nrf2 then translocates to the nucleus, where it associates with other nuclear proteins and binds to the ARE of the Srx gene and activates its transcription (Fig. 7). The important role of Nrf2 in this process was shown by our observation that ethanol-induced up-regulation of Srx expression was greatly attenuated in the liver of Nrf2−/− mice. The expression of Nrf2 itself was previously shown to be increased by ethanol-induced CYP2E1 and to play a key role in prevention of alcohol-induced liver injury.6, 34 The ethanol-induced up-regulation of Srx expression in the liver of Nrf2−/− mice was not completely abolished in the present study, however, with this Nrf2-independent increase in Srx expression likely being attributable to the action of AP-1 or other transcription factors that are able to transmit the oxidative signal to the ARE of the Srx gene.
Ethanol increases the abundance of CYP2E1 in the liver largely by preventing its proteolysis.1 CYP2E1, which exhibits a high rate of NADPH oxidase activity even in the absence of substrate, reduces molecular oxygen to superoxide. Superoxide is converted to H2O2 and peroxynitrite, both of which generate other types of ROS such as lipid peroxides. Given the location of CYP2E1 at the cytosolic side of the ER membrane, ROS production by this enzyme must also be localized to the outside surface of the ER, where we have now shown a large proportion of Prx I is also found. Prx I is therefore likely preferentially engaged in the reduction of ROS produced by CYP2E1 and becomes hyperoxidized (Fig. 7). The preferential hyperoxidation of Prx I occurs despite the fact that both Prx I and II are cytosolic proteins and that Prx II is more prone to hyperoxidation than is Prx I in most cell types.20 In addition to inducing the expression of Srx in the liver, ethanol feeding elicited the translocation of some Srx molecules to microsomes. However, the capacity of Srx located near the surface of the ER was not sufficient to fully counteract the hyperoxidation of Prx I, with consequent accumulation of a small amount of Prx I-SO2. Proteome analysis identified Prx I (but not other Prxs) among the many oxidatively damaged (hyperoxidized or carbonylated) proteins in the liver of ethanol-fed rats,35, 36 indicative of the proximity of Prx I to the ROS source.
Ethanol feeding increases the production of ROS in mitochondria,4, 28, 30 with this effect likely resulting in Prx III hyperoxidation and the translocation of Srx into mitochondria.23 Our failure to detect Prx III-SO2 in the liver of ethanol-fed wildtype mice suggests that the capacity of mitochondrial Srx is sufficient to counteract the hyperoxidation of Prx III in mitochondria (Fig. 4D). Chronic ethanol feeding in Srx−/− mice increased the amount of Prx I-SO2 to 30 to 50% of total Prx I, which likely corresponds to virtually all ER-bound Prx I molecules. We did not detect Prx II-SO2 in the liver of ethanol-fed wildtype or Srx−/− mice, however, suggesting that Prx II was not engaged in ROS elimination rather than that the capacity of Srx in the cytosol was sufficient to counteract its hyperoxidation. Kupffer cells produce H2O2, which can diffuse across biological membranes and impose oxidative stress on hepatocytes. The apparent absence of Prx II-SO2 in the liver of Srx−/− mice, however, suggests that H2O2 molecules produced by Kupffer cells do not generate a high level of stress in hepatocytes. We did detect Prx III-SO2 in the liver of ethanol-fed Srx−/− mice, corroborating the notion that Prx III-SO2 was not detected in ethanol-fed Srx+/+ mice because mitochondrial Srx was sufficient to counteract Prx III hyperoxidation. Only a small proportion of Prx III was hyperoxidized, however, probably because the mitochondrial production of ROS is not as substantial as that mediated by the microsomal CYP2E1. It was shown previously that ethanol could also increase CYP2E1 within mitochondria, although to a much smaller extent than in the microsomes.38 Alternatively, glutathione peroxidase might be the major enzyme responsible for the elimination of mitochondrial ROS. Glutathione peroxidase and glutathione homeostasis in mitochondria have been shown to be important for prevention of ethanol-induced oxidative injury in the liver.4, 33 We did not detect Prx IV-SO2 in the liver of ethanol-fed Srx−/− mice. Prx IV is a luminal ER protein, and, given that thioredoxin is not present inside the ER, Prx IV does not function as a peroxidase in this organelle but rather serves as a peroxide sensor for other proteins.37 It is thus possible that ROS produced by CYP2E1 are rapidly removed by Prx I before they can cross the ER membrane.
Together, our results suggest that, among the 2-Cys Prxs, Prx I is largely responsible for the reduction of ROS generated in the liver in response to ethanol exposure because of its proximity to CYP2E1. Prx I molecules are thus converted to the inactive, sulfinylated form (Prx I-SO2). Reactivation of such hyperoxidized Prx I requires the action of Srx, the expression of which is greatly increased in the liver of ethanol-fed mice. The pivotal roles of Srx and Prx I in protection of the liver against ethanol-induced oxidative stress were apparent in Srx−/− and Prx I−/− mice. Subjection of such mice to chronic ethanol feeding thus resulted in more severe oxidative damage to the liver, as revealed by carbonylation of soluble proteins and by the formation of 4-HNE and 3-NT adducts in the centrilobular regions, compared with that observed in ethanol-fed wildtype mice.