Contribution of glutaredoxin-1 to Fas s-glutathionylation in ethanol-induced liver injury CURRENT

Background The reversible glutathionylation modification (PSSG) of Fas augments apoptosis, which can be reversed by the cytosolic deglutathionylation enzyme glutaredoxin-1 (Grx1), but its roles in alcoholic liver injury remain unknown. Therefore, the objective of this study was to investigate the impact of genetic ablation of Grx1 on Fas-SSG in regulating ethanol-induced injury. Methods The role of Grx1 in alcoholic liver injury was investigated in Grx1 knockout mice. Alcoholic liver injury was achieved by feeding mice with a liquid diet containing 5% ethanol for 2 weeks. Results We demonstrated that ethanol-fed mice had increased Grx1 activity and oxidative damage in the liver. On the other hand, Grx1-deficient mice had more serious liver damage when exposed to ethanol compared to that of wild-type mice, accompanied by increased alanine aminotransferase and aspartate aminotransferase levels, Fas-SSG, cleaved caspase-3 and hepatocyte apoptosis. Grx1 ablation resulted in the suppression of ethanol-induced nuclear factor-κB (NF-κB) signaling, its downstream signal, and Akt signaling cascades, which are required for protection against Fas-mediated apoptosis. Accordingly, blocking NK-κB prevented Fas-induced apoptosis in WT mice but not Grx1-/- mice. Furthermore, the number of Kupffer cells and related proinflammatory cytokines, including Akt, were lower in Grx1-/- livers than those of the controls. L also apoptosis TNF/FasL

Various studies support the critical role of the Fas death pathway in liver injury [12,13]. Fas is posttranslationally glutathionylated via the covalent attachment of glutathione (GSH), resulting in Fas-SSG. Fas-SSG enhances Fas recruitment and binding with FasL in lipid rafts, which was shown to be functionally important for promoting hepatocyte apoptosis [14,15,16]. While the mechanism of Fas-dependent cell death has clearly emerged, the detailed mechanism of redox-based sglutathionylation of Fas during alcoholic liver injury is still unknown.
In the sinusoidal vascular space of the liver, Kupffer cells account for approximately 25 percent of nonparenchymal cells and can be activated to generate a variety of soluble factors and cytokines by many endogenous and exogenous stimuli [17,18,19]. Tumor necrosis factor α (TNF-α) is considered the most critical cytokine during alcoholic liver injury. Previous research indicated that depletion of TNF-α signaling could reverse alcohol-induced liver injury in mice [20,21].
Here, we aimed to investigate the detailed mechanism by which redox-based s-glutathionylation of Fas promotes hepatic apoptosis in alcoholic liver injury. We also determined the contribution of Grx1 to Kupffer cell activation and inflammatory cytokine production in hepatic oxidative stress.

Materials And Methods Animal experiment
The animal studies were conducted in accordance with the criteria approved by the Institutional Review Board and the Animal Care and Use Committee of Chongqing Medical University. Eight-to tenweek-old female Grx1-/-mice (kindly provided by Dr. Hongbo Luo of Harvard University), and the genetically corresponding wild-type (WT) mice were subjected to either a Lieber-DeCarli liquid diet containing 1-5% ethanol (volume/volume) or isocaloric amounts of pair-fed (PF) Lieber-DeCarli control diets over a period of two weeks. Then, on the last day, the mice were intragastrically administered 50% (v/v) ethanol every 20 minutes with an accumulated dosage of 5 g/kg body weight (Fig. 1).
In one set of experiments, the mice were treated with anti-FasL Ab (0.1 mg/mouse) (MFL4, BD PharMingen) through tail vein injections 24 hours before the first ethanol administration and again every three days. In another set of experiments, the mice were intraperitoneally (i.p.) injected either 1 mg/kg of BAY 11-7082 (EMD Chemicals, Gibbstown, NJ) or the control (DMSO in PBS 20% Tween-80) 24 hours before the first ethanol administration. The above time points were optimized based on several preliminary experiments. There were 3-6 mice in each experimental group. Consumption was recorded daily, and animal body weights were measured weekly. At 6 hours following the last dose of ethanol, the mice were euthanized. Blood and liver tissues were collected under sterile conditions. Some of the liver tissues were fixed and sliced for hematoxylin and eosin (H&E) staining and morphological examination. Portions of the liver tissue were snap frozen in liquid nitrogen for quantitative real-time PCR (qPCR), Western blotting and biochemical assays. Tail snips were collected for genotyping.

Oil red O (ORO) staining
Oil red O (ORO) staining was performed as described previously [22]. Briefly, snap-frozen slides of liver sections were processed and stained with ORO solution. After washing in distilled water, the slides were subjected to H&E staining. Hepatic lipid droplets in H&E-and ORO-stained slides were observed under a light microscope.

Biochemical analysis
Abdominal aortic blood was collected, and serum ALT and AST levels were measured by commercial enzymatic assay kits (Diagnostic Chemicals, LTD, Oxford, CT) as previously described to detect liver damage [2]. Liver total cholesterol (TC) and triglyceride (TG) levels were assessed using commercially available reagents (California, United States).

GSH and GSSG measurement
The measurement protocol for reduced and oxidized glutathione was conducted as previously reported [22]. Glutathione (GSH) and oxidized glutathione (GSSG) were determined spectrophotometrically using GSH and GSSG assay kits (S0053, Beyotime). The ratio of GSH to GSSG (GSH/GSSG) is presented as nmol/mg protein and is an indicator of oxidative stress.
Dual confocal immunofluorescence detection of S-glutathionylated Fas Dual confocal immunofluorescence staining and evaluation was performed as previously reported [22]. Briefly, the liver sections were incubated with the primary antibodies, (rabbit anti-GSH [Virogen] and goat anti-Fas antibody] [BD Biosciences]). The sections were then incubated with donkey antimouse Alexa Fluor 488 (Invitrogen) and donkey anti-rabbit Alexa Fluor 594 (Invitrogen) to develop immunofluorescence [22]. Finally, the nuclei were stained with DAPI (Sigma). During the experiment, the primary anti-GSH or anti-Fas antibodies were omitted as negative controls. The slides were observed via fluorescence confocal microscopy (Leica Microsystems, Heidelberg, Germany).

Superoxide Detection by Confocal Microscopy
The level of intracellular reactive oxygen species (ROS) was determined in liver slides through staining with the ROS-sensitive dye dihydroethidium (DHE) as previously described [22]. The liver

Immunohistochemistry
Immunohistochemical assays were conducted according to previously described protocols [23].
Briefly, the slides were incubated with primary antibodies, including F4/80 (Abcam), at room temperature for 2 h. The expression of target proteins was visualized with a DAB peroxidase substrate kit (Vector Laboratories, Burlingame, CA). Quantitative real-time PCR RNA was extracted from snap-frozen liver tissues using TRIzol (Invitrogen, Carlsbad, CA) and then reverse transcribed (Omniscript; Qiagen, Hilden, Germany). Quantitative real-time PCR was performed on an ABI 7500 Real-Time PCR System (Applied Biosystems) using SYBR Green (Bio-Rad). The primer pairs and probes were used as previously reported [22]. The primers that were used to analyze the mRNA levels of IL-10, IL-12, MCP-1, IFN-γ and β-actin were designed using the Primer-BLAST system (www.ncbi.nlm.nih.gov/tools/primer-blast/) [5].

TUNEL assay
The paraffin-embedded sections were subjected to a terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick-end labeling (TUNEL) assay using the APO-BrdU™TUNEL assay kit (Invitrogen) following the manufacturer's instructions. The TUNEL-positive cells were visualized under fluorescence microscopy.

ELISA analysis of active TNF-α in liver homogenate
Snap-frozen liver tissue samples (30-50 mg) were homogenized and centrifuged. The supernatants were used to detect the protein concentrations with a Pierce BCA kit (Thermo Scientific, Rockford, IL).
The TNF-α concentration in liver homogenates was measured using a TNF-α Quantikine ELISA Kit (R&D Systems, Minneapolis, MN) according to the manufacturer's instructions.
Afterwards, the membranes were incubated with specific diluted primary antibodies against the following proteins: RelA (Cell Signaling Technology, Boston, MA), RelB (Cell Signaling), IkBa (Santa Cruz), P-Akt (Santa Cruz), P-4E-BP1 (Santa Cruz), and β-Actin (Santa Cruz). After washing, the membranes were incubated with the appropriate horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. The densitometric analysis was conducted using ImageJ software (National Institutes of Health).

Grx1 Activity Assay
A glutaredoxin activity assay was conducted using a glutaredoxin fluorescent kit (11536, Cayman) according to the manufacturer's instructions.

Immunoprecipitation Assay
Tissue lysates were prepared from snap-frozen liver tissues according to the manufacturer's instructions (KeyGEN BioTECH). The protein concentration was determined by using the BCA method (P0010, Beyotime) and equalized for the immunoprecipitation assay [22]. The tissue lysate was then incubated with anti-glutathione antibody (Virogen) (2 µg/ml) overnight at 4 °C and immunoprecipitated by protein G agarose beads. The isolated immunoprecipitated proteins were then incubated with the anti-Fas antibody (Santa Cruz). For the control, dithiothreitol (DTT, 50 mM) was added to reduce S-glutathionylated proteins.

Statistical analysis
Statistical analysis was conducted with GraphPad Prism version 4 (GraphPad Software, Inc., La Jolla, CA). The results are expressed as the mean ± SEM. One-way ANOVA was performed for comparisons between groups in at least three independent experiments. Significance was defined as P < 0.05 and is indicated in the figure legends.

Ethanol administration induced Grx1 activity in hepatic tissue
We first examined GSH and GSSG levels, which reflect ethanol-induced hepatic oxidative stress. The results showed significantly decreased GSH levels in ethanol-fed mice compared with those of pair-fed mice ( Fig. 2A). Conversely, the GSSG level presented a marked increase (Fig. 2B). Correspondingly, the GSSG/GSH ratio increased from approximately 0.2 in the pair-fed mice to 1.8 in ethanol-fed mice, indicating hepatic oxidative stress (Fig. 2C). Grx1 depletion exacerbated this oxidative stress. We further evaluated oxidative production by detecting oxidized dihydroethidium (DHE) in situ. As Grx1 cleaves protein-GSH adducts, which constitute an important regulatory switch. We next investigated whether ethanol feeding influenced Grx1 activity. The results shown in Fig. 2E demonstrated that ethanol increased Grx1 activity following the last dose of ethanol(day 1), which persisted for 7 days compared with that of controls. As expected, in agreement with the results of Grx1 activity, increased Grx1 protein levels were detected by Western blot analysis in hepatic homogenates (Fig. 2F).

Grx1 depletion increased susceptibility to ethanol-induced hepatic injury
Next, we compared hepatic injury in Grx1-/-and WT mice with ethanol administration. As shown in levels increased by 25%, TC levels increased by 15%, and TG levels increased by 35% in Grx1-/-mice compared with those of their WT counterparts (Fig. 3A, C, D). Although the AST level increased in response to alcohol, no difference between ethanol-fed Grx1-/-and WT mice was found (Fig. 3B). More steatosis (fat deposition) was present in the livers of ethanol-fed Grx1-/-mice compared with that of their WT counterparts, as indicated by H&E and Oil Red O staining (Fig. 3E, F), which was consistent with the biochemical analyses.

Grx1 ablation-induced apoptosis was Fas dependent
We then assessed ethanol-induced hepatic apoptosis using a terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay. As expected, ethanol-exposed mice exhibited higher rates of positive staining in hepatocytes, with marked increases in Grx1-/-mice compared to those of their WT counterparts (Fig. 5A), indicating severe hepatic injury in Grx1-depleted mice. Caspase-3 activation was higher in Grx1 ablation liver tissues than in their WT counterparts (Fig. 5B), which was similar to the results of the TUNEL assay. To demonstrate whether the TNF/FasL system is responsible for hepatic apoptosis, Fas L was blocked to confirm this cause-and-effect relationship. As shown in Fig. 5A and Fig. 5B, Fas L neutralization not only attenuated caspase-3 activation but also suppressed ethanol-induced apoptosis in Grx1-/-and WT mice. These results suggest that the TNF/FasL system plays a crucial role in ethanol-induced acute hepatic injury, which was modulated by Grx1.
Grx1 ablation enhanced Fas-induced apoptosis by NF-κB/Akt signaling We next analyzed NF-κB/Akt signaling, which are key molecules implicated in Fas-induced liver injury.
As shown in Fig. 6A, when exposed to ethanol, the expression of NF-kB family members and NF-κBdependent cytokines increased in the Grx1-/-liver tissues compared with those of the WT counterparts. Consistent with the NF-κB signal, the p-Akt level (Ser473) in Grx1-/-livers was lower compared with that of the WT counterparts. Accordingly, the downstream target of p-Akt, eukaryotic translation initiation factor 4E-binding protein 1, was significantly increased in ethanol-exposed Grx1-/-mice compared with that of ethanol-exposed WT mice.

Grx1 depletion alters Kupffer cells after ethanol exposure
Previous research suggested that Kupffer cells are the essential effectors in response to ethanol stress. We then addressed the involvement of Grx1 in Kupffer cells from pair-and ethanol-fed mice to explore the mechanism of monocyte or macrophage infiltration or apoptosis. As shown in Fig. 7A and 7B, the expression of F4/80, a marker of Kupffer cells, increased following ethanol exposure in WT mice, and in Grx1-/-mice, this increase was more pronounced, suggesting that the inflammatory cell infiltration induced by ethanol is exacerbated by Grx1 depletion. Ethanol also induced the expression of several proinflammatory cytokines (TNF-α, IL-10, and IL-12) in liver tissue from WT mice, and this increase was more pronounced in Grx1-/-mice (Fig. 7C). However, ethanol inhibited IFN-γ mRNA expression in hepatic tissues from both WT and Grx1-/-mice and had no effect on MCP1 mRNA expression in WT and Grx1-/-mice. Taken together, these results indicated that Grx1 was involved in ethanol-induced Kupffer cells activation and alterations in innate hepatic cytokines.

Discussion
Current research defines the mechanism of hepatocyte apoptosis as being controlled by the redox modulation system, which accounts for ethanol-induced liver injury. We demonstrated in this study that initial depletion of Grx1 brings about Fas s-glutathionylation during ethanol exposure, which promotes the aggregation and binding of Fas with FasL, and further activates caspases, resulting in apoptosis; these effects all occur via inactivation of the NF-κB/Akt pathway. To our knowledge, we observed the involvement of Grx1 in Fas-induced hepatocyte apoptosis for the first time. Moreover, our results showed that depletion of Grx1 in mice plays a critical role in ethanol-induced liver injury.
Oxidative stress is the main pathogenesis of ethanol-induced liver injury and exhibits a significant effect on the regulation of signaling proteins [21]. Oxidative stress modulates glutathione to the disulfide form [22,23], and some reports have documented that ethanol exposure decreases GSH concentration [24] and restores GSH to normalize ALT in vivo [25,26], suggesting that the imbalance in GSH due to oxidative stress contributes to hepatocyte injury. In the present study, we found a decrease in the GSH level and an increase in the GSSG-to-GSH ratio during ethanol-induced liver injury in conjunction with oxidative stress. Furthermore, the increase in GSSG favored sglutathionylation through catalysis by the protein thiol oxidoreductase Grx1, which reduces mixed disulfide bonds, especially protein s-glutathionylation (PSSG) [27,28]. Here, we reported that ethanol exposure induced Grx1 activity in liver tissue that specifically reduced GSH mixed disulfides, whereas Grx1 depletion promoted PSSGs, which might be associated with increases in apoptosis.
Grx1 has been implicated in various diseases, including chronic pulmonary disease, lung inflammation, Pseudomonas aeruginosa pneumonia, and hepatic warm I/R injury [29,30,31,32]. The reducing power of Grx1 reverses protein oxidation, thereby potentially inhibiting the subsequent deleterious influence of the protein. We previously demonstrated enhanced Grx1 activity in intestinal tissue after the induction of NEC [22,23] and that ablation of Grx1 resulted in enhanced intestinal injury [22]. Collectively, the current evidence [11,33] demonstrated that Grx1 and s-glutathionylation function in diverse settings and play complex roles in disease pathogenesis. Because Grx1 has antioxidant and anti-inflammatory functions, it would be expected that Grx1 depletion would promote the susceptibility of the liver to injury. Herein, we demonstrated that Grx1 deletion promoted ethanolinduced liver injury in mice that manifested as increased plasma ASL and ALT levels and increased caspase activities.
Fas-induced apoptosis has been proposed to function in various liver diseases, such as ischemia/reperfusion injury, viral hepatitis, fulminant hepatic liver failure, and nonalcoholic and alcoholic steatohepatitis [34,35,36,37]. Previous research suggested that s-glutathionylation of Fas results in robust Fas activation, which activates FasL-induced signaling and apoptosis [16,38]. To date, the role of Grx1 in modulating Fas protein s-glutathionylation during ethanol stress has not been studied. In the present study, we demonstrated that after Grx1 ablation, increased s-glutathionylation of Fas overlapped with accelerated apoptosis that was consistent with ethanol-induced liver injury.
We also demonstrated that these increases were normalized by FasL antibody, indicating an association with Fas s-glutathionylation and suggesting the involvement of the Grx1/Fas sglutathionylation signaling axis in ethanol-induced liver injury.
In addition to the antioxidant effect, the Grx1/Fas axis might regulate the inflammatory signaling cascades that facilitate the function of Grx1 in ethanol-induced liver injury. Previous work demonstrated that nuclear factor-κB (NF-κB) was also dampened by s-glutathionylation, and Grx1 ablation further inhibited NF-κB activation [39,40,41]. In the current setting, NF-κB activation was reduced in ethanol-treated Grx1-ablated mice compared with that of WT mice, supporting the effect of Grx1 on NF-κB activity in the pathogenesis of ethanol-induced liver injury. NF-κB is well known as a proinflammatory transcription factor [42,43,44], and it could be inferred that NF-κB sglutathionylation is responsible for the phenotype and is involved in the molecular modulation observed in the present study. We observed that NF-κB blockade promoted Fas-induced hepatocyte apoptosis in WT mice but was unable to further enhance the effect of Grx1 ablation, which supported the role of the NF-κB redox state in mediating the antiapoptotic effects of Grx1. In addition, Akt has also been linked to the regulation of Grx1 [45,46], which was also observed in the present study. TNFα is a downstream proinflammatory factor of NF-κB, which was confirmed to be necessary for liver injury induced by ethanol [47]. In the current study, Grx1 ablation further consolidated ethanolinduced increases in TNF-α mRNA and protein levels, which was normalized by blocking NF-κB, confirming the current hypothesis. Given the observations that the various inflammatory mediators mentioned above were regulated by Grx1 under ethanol exposure, we hypothesized that Grx1 normalizes rather than eliminates inflammatory cytokine activation, which may be a useful therapeutic target for ethanol exposure.
Under ethanol exposure, injured apoptotic cells are mostly hepatocytes, and there is the possibility that other immune cells within the liver may also undergo activation, including macrophages, infiltrating immune cells, and activated stellate cells. Here, we evaluated this hypothesis and demonstrated the hepatic macrophages accumulation and activation in ethanol-treated Grx1-/-mice.
A previous report suggested that macrophage alternation in Fas-induced apoptosis accounted for increased inflammatory cytokines [48]. A low level of early proinflammatory cytokines (i.e., TNF-α) exhibited here was in accordance with the above scenario. Furthermore, inhibition of inflammatory cell apoptosis would result in an exacerbated inflammatory or fibrotic response [49,50]. To our knowledge, this is the first report to explore Grx1 ablation-modulation of macrophage enrichment in the liver. However, our current study was limited in that we did not address the cellular specificity of immune cell apoptosis. Additional studies are required to assess these effects, which might be beyond the scope of the current research.
In conclusion, our study illuminated a new dimension of knowledge regarding Grx1, which plays an essential role in modulating ethanol-induced liver injury. The underlying mechanisms are related to redox-based regulation of Fas/FasL-induced apoptosis by multiple roles of Grx1 in the pathogenesis of ethanol exposure. Correspondingly, Grx1 might be a critical therapeutic target for treatment of ethanol-induced liver injury.