Absence of receptor interacting protein kinase 3 prevents ethanol-induced liver injury

Authors

  • Sanjoy Roychowdhury,

    Corresponding author
    1. Departments of Pathobiology, Center for Liver Disease Research, Cleveland Clinic, Cleveland, OH
    • Cleveland Clinic, Lerner Research Institute/NE40, 9500 Euclid Avenue, Cleveland, OH 44195
    Search for more papers by this author
    • fax: 216-636-1493

  • Megan R. McMullen,

    1. Departments of Pathobiology, Center for Liver Disease Research, Cleveland Clinic, Cleveland, OH
    Search for more papers by this author
  • Sorana G. Pisano,

    1. Departments of Pathobiology, Center for Liver Disease Research, Cleveland Clinic, Cleveland, OH
    Search for more papers by this author
  • Xiuli Liu,

    1. Departments of Anatomic Pathology, Center for Liver Disease Research, Cleveland Clinic, Cleveland, OH
    Search for more papers by this author
  • Laura E. Nagy

    1. Departments of Pathobiology, Center for Liver Disease Research, Cleveland Clinic, Cleveland, OH
    2. Departments of Gastroenterology, Center for Liver Disease Research, Cleveland Clinic, Cleveland, OH
    3. Department of Nutrition, Case Western Reserve University, Cleveland, OH
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Supported in part by ABMRF/The Foundation for Alcohol Research (S. R.) and grants RO1AA013868, P20 AA17069, from the National Institutes of Health and D.O.D. 10248754 (L. E. N.).

Abstract

Hepatocyte cell death via apoptosis and necrosis are major hallmarks of ethanol-induced liver injury. However, inhibition of apoptosis is not sufficient to prevent ethanol-induced hepatocyte injury or inflammation. Because receptor-interacting protein kinase (RIP) 3–mediated necroptosis, a nonapoptotic cell death pathway, is implicated in a variety of pathological conditions, we tested the hypothesis that ethanol-induced liver injury is RIP3-dependent and RIP1-independent. Increased expression of RIP3 was detected in livers of mice after chronic ethanol feeding, as well as in liver biopsies from patients with alcoholic liver disease. Chronic ethanol feeding failed to induce RIP3 in the livers of cytochrome P450 2E1 (CYP2E1)-deficient mice, indicating CYP2E1-mediated ethanol metabolism is critical for RIP3 expression in response to ethanol feeding. Mice lacking RIP3 were protected from ethanol-induced steatosis, hepatocyte injury, and expression of proinflammatory cytokines. In contrast, RIP1 expression in mouse liver remained unchanged following ethanol feeding, and inhibition of RIP1 kinase by necrostatin-1 did not attenuate ethanol-induced hepatocyte injury. Ethanol-induced apoptosis, assessed by terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling–positive nuclei and accumulation of cytokeratin-18 fragments in the liver, was independent of RIP3. Conclusion: CYP2E1-dependent RIP3 expression induces hepatocyte necroptosis during ethanol feeding. Ethanol-induced hepatocyte injury is RIP3-dependent, but independent of RIP1 kinase activity; intervention of this pathway could be targeted as a potential therapeutic strategy. (HEPATOLOGY 2013)

Multiple mechanisms of cell death, including apoptosis and necrosis, are activated during the progression of alcoholic liver disease (ALD).1, 2 Apoptosis, characterized by cell shrinkage and DNA condensation, minimizes the leakage of proinflammatory mediators.3 Necrosis, marked by cell swelling and membrane rupture, leads to inflammation via release of the intracellular danger signals. In addition to these well-studied cell death pathways, another mode of cell death, necroptosis, has recently been identified in various cell types.4 Necroptosis shares a common activation pathway with apoptosis, but morphologically resembles necrosis.5 Activation of prodeath ligands, including tumor necrosis factor-α (TNFα), CD95, or TNF-related apoptosis-inducing ligand, initiates necroptosis. If apoptosis is inhibited in cultured cells, necroptosis may be exacerbated in response to prodeath stimuli.6

Activation of the receptor interacting protein (RIP) kinases and subsequent necroptosis is implicated in a variety of pathological conditions, including ischemia/reperfusion injury,7 viral infection,8 acute pancreatitis,9 and ileitis.10 RIP1 and RIP3, members of the serine-threonine kinase family, are central mediators of necroptosis.11 Genetic ablation of RIP3 or blockade of RIP1 kinase activity with necrostatin-1 prevents necroptosis in various injury models.12, 13 Although a majority of studies indicate that an interaction between RIP1 and RIP3 is critical for necroptosis, RIP3 alone can trigger necroptosis in the absence of RIP1 in specific cell types.14

The intricate balance between cell death and prosurvival pathways is critical for regulating liver injury and inflammation during progression of ALD. Deaciuc et al15 demonstrated that ethanol-induced apoptosis sensitizes rat hepatocytes to lipopolysaccharide-mediated cytotoxicity. However, using both genetic and pharmacological approaches, we have shown that inhibition of apoptosis is not sufficient to prevent hepatic inflammation and hepatocyte injury in mouse models of ethanol-induced steatohepatitis.16 Similarly, in a mouse model of methyl-choline-deficient diet-induced nonalcoholic steatohepatitis, inhibition of apoptosis ameliorates fibrosis and hepatic inflammation, but fails to attenuate hepatocyte injury,17, 18 suggesting that nonapoptotic cell death pathways are also critical for hepatocyte injury in a variety of liver diseases. Interestingly, RIP3, the central mediator of necroptosis, is induced in mouse liver concomitant with hepatic injury following chronic, heavy ethanol feeding16; expression of RIP3 is increased in response to ethanol even when apoptosis is inhibited.16 Taken together, these results suggest that RIP3-driven necroptosis could be a key component of ethanol-induced liver disease.

Because abrogation of apoptotic cell death fails to prevent ethanol-induced hepatocyte injury and inflammation,16 we hypothesized that RIP3-driven cell death contributes to liver injury following ethanol feeding. Increased RIP3 expression was detected in livers of ethanol-fed mice, as well as in liver biopsies from patients with alcoholic liver disease. In mice, induction of RIP3 expression in response to ethanol feeding was CYP2E1-dependent. In parallel, RIP3-deficient C57BL/6J female mice were protected against ethanol-induced liver injury and inflammation. However, when mice were treated with necrostatin-1 to inhibit RIP1 kinase activity, ethanol-induced hepatocyte injury was not reduced. Taken together, these results provide evidence that RIP3-driven necroptosis, independent of RIP1, is critical for the progression of ethanol-induced liver injury and this pathway could be a potential therapeutic target to control liver injury and inflammation following ethanol consumption.

Abbreviations

4-HNE, 4-hydroxy-2-nonenal; ALD, alcoholic liver disease; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CK18, cytokeratin 18; CYP2E1, cytochrome P450 2E1; FADD, Fas-associated death domain; IL-6, interleukin-6; JNK, c-jun N-terminal kinase; MCP-1, macrophage chemoattractant protein-1; mRNA, messenger RNA; pJNK, phosphorylated JNK; PLA, proximity ligation assay; RIP, receptor-interacting protein kinase; ROS, reactive oxygen species; TNF-α, tumor necrosis factor-α; TNFR1, tumor necrosis factor receptor 1; TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling; WT, wild-type.

Materials and Methods

Details regarding materials and basic biochemical methods are given in the Supporting Information.

Mouse Models.

All procedures using animals were approved by the Cleveland Clinic Institutional Animal Care and Use Committee. Female mice were housed in shoe box cages (two animals per cage) with microisolator lids. Standard microisolator handling procedures were used throughout the study. Mice were age-matched, randomized into ethanol-fed and pair-fed groups, and then adapted to a control liquid diet for 2 days.

The ethanol-fed groups were allowed free access to an ethanol-containing diet. Control mice were pair-fed a control diet that isocalorically substituted maltose dextrins for ethanol over the entire feeding period. Two models of ethanol-induced liver injury were used: (1) The short-term ethanol-induced liver injury model consisted of 1% (vol/vol) ethanol for 2 days followed by 6% ethanol for 2 more days. The 6% (vol/vol) diet provided ethanol as 32% of total calories in the diet (4d,32%) and is a model of binge ethanol consumption.19 (2) The chronic ethanol-induced liver injury model consisted of 1% (vol/vol) ethanol for 2 days followed by 2% ethanol (11% of total calories) for 2 days (4d,11%), 4% ethanol for 1 week, 5% ethanol for 1 week, followed by 6% ethanol for the final week (25d,32%). Feeding for 4d/11% models the early response to ethanol, while 25d,32% is a model of chronic ethanol consumption.19

Necrostatin-1 Treatment.

During a 4d,32% ethanol exposure protocol, C57BL/6J wild-type (WT) mice received a once-daily intraperitoneal injection of necrostatin-1 (1.65 mg/kg) or vehicle (2% dimethyl sulfoxide in phosphate-buffered saline) before ethanol (4d,32%) feeding. This concentration/dose regimen inhibits RIP1 kinase activity in vivo.10, 20, 21

Human Liver Samples.

Deidentified human liver biopsy samples from 20 ALD patients and eight patients with minimal liver pathology were obtained from the Cleveland Clinic surgical pathology database. The selection criteria are described in the Supporting Information. All procedures using deidentified human liver tissue were approved by the Cleveland Clinic Institutional Research Board.

Histology and Immunohistochemistry.

For human liver biopsies, paraffin-embedded livers were deparaffinized and stained for RIP3 as described above, except that 3-amino-9-ethylcarbazole was used instead of 3,3′-diaminobenzidine as the horseradish peroxidase–specific chromogen. Omitting the primary antibody (no primary immunoglobulin G control) in this protocol abrogated the staining, demonstrating the specificity of the immunoreactive staining (data not shown). Images were acquired in a blinded manner using a 20× objective. RIP3 immunostaining was then scored by an experienced pathologist (X. Liu) taking into consideration staining intensity and percentage of positive cells using a scale of 0-3 (0, lack of any staining; 1, faint staining in <10% of cells; 2, fine granular staining in 10%-50% cells or coarse granular staining in 10%-20% of cells; 3, fine granular staining in >50% cells or coarse granular staining in >20% cells). A subset of these subjects was analyzed via morphometric semiquantitation analysis using Image-Pro Plus software (n = 6 for control and n = 11 for ALD cases).

Details regarding mouse liver biopsies and in situ proximity ligation assay (PLA) are given in the Supporting Information.

Statistical Analysis.

Values shown in all figures represent the mean ± SEM (n ≥ 4 for pair-fed, n ≥ 6 for ethanol-fed). Analysis of variance was performed using the general linear models procedure (SAS, Carey, IN). Data were log-transformed as necessary to obtain a normal distribution. Follow-up comparisons were made by least square means testing. A Student t test was used for comparing values obtained from two groups (for Fig. 2 only).

Results

Ethanol Induces RIP3 in Mouse Liver.

If RIP3-dependent necroptosis contributes to ethanol-induced liver injury, then RIP3 expression should be increased in response to ethanol feeding. To test this hypothesis, RIP3 expression was evaluated by immunohistochemistry in livers from C57BL/6 mice following 4d,11% or 4d,32% ethanol feeding and 25d,32% ethanol feeding. RIP3 expression increased following 4d,32% and 25d,32% ethanol feeding, but not after 4d,11% ethanol feeding (Fig. 1A). Although weak RIP3 staining was visible around the central veins in livers of pair-fed mice, robust RIP3 staining extending beyond the pericentral area was detected in livers from 25d,32% ethanol-fed mice (Fig. 1A). In contrast, hepatic expression of RIP1 was not affected by ethanol feeding (Supporting Fig. 1).

Figure 1.

RIP3 was induced in mouse liver following ethanol feeding. (A) C57BL/6J WT mice were allowed free access to ethanol (4d,11%; 4d,32%; or 25d,32%) or pair-fed control diets as described in Materials and Methods. Paraffin-embedded livers were deparaffinized followed by immunodetection for RIP3. The area under the black box was cropped and enlarged. (B) 129S1/Svlmj (WT for CYP2E1−/−) and CYP2E1-deficient mice were allowed free access to ethanol (25d,32%) or pair-fed control diets. (A,B) Paraffin-embedded livers were deparaffinized followed by RIP3 immunostaining. Nuclei were counterstained with hematoxylin. All images were acquired using a 20× objective. The black arrows indicate RIP3-positive staining. (C) RIP3-stained areas were quantified using Image-Pro Plus software and analyzed. (D) AST activity was determined in plasma samples from ethanol- and pair-fed mice. (C,D) Values with different alphabetical superscripts are significantly different from each other (P < 0.05). All images were acquired using a 20× objective. (E) RIP3 and FADD association was detected using DuoLink proximity ligation assay. Images were acquired using a 40× objective.

CYP2E1-mediated ethanol metabolism is critical for ethanol-induced lipid peroxidation and hepatocyte injury. Mice deficient in CYP2E1 are protected from lipid peroxidation and hepatocyte injury following both short-term and chronic ethanol feeding.22 Making use of CYP2E1-deficient mice, we next investigated if ethanol-induced RIP3 expression is CYP2E1-dependent. CYP2E1-deficiency blunted ethanol-induced RIP3 expression (Fig. 1B-C), as well as prevented the ethanol-induced increase in plasma AST, a marker of hepatocyte injury (Fig. 1D), indicating that CYP2E1 contributes to ethanol-induced RIP3 expression and liver injury.

Ethanol-Induced RIP3 Is Associated With Fas-Associated Death Domain in Mouse Liver.

Upon activation, RIP3 is known to form a complex with RIP1, Fas-associated death domain (FADD), TRAD or caspase-8.23 Making use of Duolink in situ PLA, the interaction between RIP3 and FADD was assessed in mouse liver following chronic ethanol feeding. This PLA assay is able to detect two proteins within a close proximity.24 Chronic ethanol feeding induced RIP3-FADD association (Fig. 1E). Although, ethanol feeding induced RIP3 around the central veins over a wide range of area, the ethanol-induced RIP3-FADD interaction was not as broadly distributed.

RIP3 Induction in ALD.

Apoptosis and necrosis are associated with the progression of ALD.3 Apoptotic bodies are found in liver biopsies from patients with ALD.25 However, the role of necroptosis in ALD has not been investigated. Liver biopsies from patients with ALD were stained for RIP3. Higher RIP3 expression in livers from ALD patients compared with controls (Fig. 2). In the livers from the control group, weak RIP3 staining was visible. Out of 20 liver biopsies from ALD patients, 16 scored positive and the mean score of RIP3 expression in ALD patients was higher than that in controls (Fig. 2B). As in the mouse models of ethanol-induced liver injury, RIP3 expression in the livers of ALD patients was primarily restricted to hepatocytes. Semiquantification using morphometric analysis also showed increased expression of RIP3.

Figure 2.

RIP3 was detected in liver biopsies from patients with ALD. Paraffin-embedded human liver biopsies were deparaffinized followed by RIP3 immunostaining using a standard immunohistochemistry technique. Nuclei were counterstained with hematoxylin. A selected RIP3-stained area was cropped and enlarged. The black arrows indicate RIP3-positive staining. (A) Images were acquired using a 20× objective. (B) RIP3 staining were scored on a scale of 0-3 (n = 8 for control, n = 20 for ALD). (C) RIP3-stained areas were quantified using Image-Pro Plus software and analyzed. *P < 0.05 (n = 6 for control, n = 11 for ALD).

Absence of RIP3 Prevents Liver Injury in Mice Following Ethanol Feeding.

To examine the contribution of RIP3-driven cell death in ethanol-mediated hepatocellular injury, C57BL/6J WT and RIP3-deficient mice were allowed free access to Lieber-DeCarli ethanol-diet for 4d,32% or pair-fed control diet. Ethanol feeding increased alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activity in plasma (Fig. 3A), as well as hepatic triglyceride content in WT mice (Fig. 3A,B). If RIP3 contributes to ethanol-induced hepatocyte injury, deletion of RIP3 should ameliorate the increase in plasma ALT/AST following ethanol feeding. Consistent with this hypothesis, ethanol-induced ALT/AST were attenuated in RIP3-deficient mice (Fig. 3A). In addition, mice lacking RIP3 showed reduced ethanol-induced hepatic lipid accumulation (Fig. 3A,B). This protection was not due to differences in ethanol intake, as WT and RIP3-deficient mice consumed equal amounts of ethanol (Supporting Table 1).

Figure 3.

Ethanol-induced liver injury and steatosis were attenuated in RIP3-deficient mouse livers. C57BL/6J WT and RIP3-deficient mice were allowed free access to ethanol (4d,32%) or pair-fed control diets. (A) ALT and AST activities were measured in plasma. Hepatic triglyceride content was measured in whole liver homogenate. Values with different alphabetical superscripts are significantly different from each other (P < 0.05). (B) Frozen liver sections were subjected to Oil Red O staining. All images were acquired using a 20× objective.

RIP3 Is Critical for Hepatic Inflammation in Mice During Ethanol Exposure.

Ethanol-induced liver injury is associated with hepatic inflammation.26, 27 Dying hepatocytes release proinflammatory mediators and damage-associated molecular pattern proteins during a variety of hepato-pathological conditions, triggering more cell death and further aggravating liver inflammation.28

If release of damage-associated molecular pattern proteins from the necrotic cells activates ethanol-induced hepatic inflammation, the reduction in hepatocyte injury markers, ALT/AST, in RIP3-deficient mice should be associated with decreased expression of inflammatory mediators following ethanol exposure. Ethanol feeding increased the number of inflammatory foci containing monocytes, in the livers of WT mice (Fig. 4A). The appearance of ethanol-induced inflammatory foci was attenuated in RIP3-deficient mice (Fig. 4A). Similarly, ethanol feeding for 4d,32% also induced proinflammatory mediators, as assessed by the expression of macrophage chemoattractant protein-1 (MCP-1), interleukin-6 (IL-6) and TNFα messenger RNA (mRNA) (Fig. 4B). RIP3-deficiency ameliorated ethanol-induced expression of MCP-1, IL-6, and TNFα mRNA (Fig. 4B). TNFα protein, measured by immunohistochemistry, increased with ethanol feeding in the livers from WT but not in RIP3-deficient mice (Fig. 4C). Ethanol-induced immunoreactive TNFα was detected in CD68-positive macrophages, residing in the hepatic sinusoid, as well as in hepatocytes (Fig. 4C and Supporting Fig. 4). Consistent with ethanol-induced increase of immunoreactive TNFα in liver, 4d,32% ethanol feeding also elevated TNFα concentration in plasma, as detected by enzyme-linked immunosorbent assay. RIP3 deficiency blunted the ethanol-induced increase of TNFα concentration in plasma (Fig. 4E). Taken together, these results indicate that RIP3 contributes to increased expression of proinflammatory mediators in mouse liver following ethanol exposure.

Figure 4.

RIP3 deficiency reduced ethanol-mediated hepatic proinflammatory cytokine expression in mouse liver. C57BL/6J WT and RIP3-deficient mice were allowed free access to ethanol (4d,32%) or pair-fed control diets. (A) Paraffin-embedded liver sections were stained with hematoxylin and eosin. All images were acquired using a 10× objective. The areas within the gray boxes are enlarged to visualize an inflammatory focus. The area under the box was cropped and enlarged. (B) Expression of MCP-1, IL-6, and TNFα mRNA was detected in mouse livers using quantitative reverse-transcription polymerase chain reaction. *P < 0.05 for ethanol-fed versus pair-fed within each genotype. (C) Immunoreactive TNFα protein was evaluated by immunohistochemistry in frozen livers. All images were acquired using a 40× objective. The area under the box was cropped and enlarged. (D) TNFα-positive areas were quantified using Image-Pro Plus software and analyzed. (E) Plasma TNFα concentrations were measured via enzyme-linked immunosorbent assay. *P < 0.05 for ethanol-fed versus pair-fed within each genotype.

Using the chronic ethanol feeding model, RIP3 deficiency also attenuated hepatocyte injury, measured by ALT/AST, and hepatic TG accumulation after 25d,32% ethanol feeding (Fig. 5A). CYP2E1 was induced after this chronic ethanol exposure independent of genotype (Supporting Fig. 1A). Ethanol-induced expression of MCP-1, IL-6, and TNFα mRNA, as well as immunoreactive TNFα, were also blunted in livers of RIP3-deficient mice (Fig. 5B,C). Accumulation of 4-hydroxy-2-nonenal (4-HNE) adducts in the liver, an indicator of oxidative stress, was also reduced in RIP3-deficient mice following ethanol feeding for 25d,32% (Supporting Fig. 2)

Figure 5.

RIP3 deficiency prevented chronic ethanol-induced liver injury and hepatic proinflammatory cytokine expression in mice. C57BL/6J WT and RIP3-deficient mice were allowed free access to ethanol (25d,32%) or pair-fed control diets. (A) ALT and AST activities were measured in plasma. Hepatic triglyceride content was measured in whole liver homogenates. Values with different alphabetical superscripts are significantly different from each other (P < 0.05). (B) Expression of MCP-1, IL-6, and TNFα mRNA was detected in mouse livers using quantitative reverse-transcription polymerase chain reaction. *P < 0.05 for ethanol-fed between genotypes. (C) Immunoreactive TNFα protein was evaluated by immunohistochemistry in frozen livers. All images were acquired using a 40× objective. TNF-positive areas were quantified using Image-Pro Plus software and analyzed. Values with different alphabetical superscripts are significantly different from each other (P < 0.05). (D,E) Infiltration of CD45- and Ly6c-positive cells was evaluated via immunohistochemistry in frozen livers. Numbers of foci, as well as the total number of CD45+ or Ly6C+ cells, were counted. *P < 0.05 for comparison between pair-fed and ethanol-fed mice. The area under the box was cropped and enlarged.

Infiltration of immune cells in the liver following chronic ethanol feeding was assessed by immunohistochemistry for CD45, a leukocyte marker,29 and Ly6c, a monocyte marker30 (Fig. 5D,E). In response to chronic ethanol feeding, the number of Ly6c+ cells increased in the liver of WT mice. In contrast, ethanol feeding did not increase the Ly6c+ cell numbers in RIP3−/− mice. While the total number of CD45+ cells was not influenced by ethanol feeding, the number of foci containing CD45+ cells increased after chronic ethanol feeding. This ethanol-induced increase in CD45+ cells containing foci was blunted in the livers of RIP3-deficient mice (Fig. 5D,E).

RIP3 Deficiency Does Not Prevent Ethanol-Induced Apoptosis in Mouse Liver.

In cell culture models, down-regulation of one cell death pathway often results in an increased activation of alternative death cascades.6 However, in mouse models of ethanol-induced liver injury, inhibition of apoptosis using Bid-deficient mice or the pan-caspase inhibitor VX166 did not exacerbate expression of RIP3 after ethanol exposure.16 Making use of RIP3-deficient mice, we were able to test the parallel hypothesis to assess whether loss of the necroptotic cell death pathway would influence ethanol-induced hepatocyte apoptosis. Ethanol feeding increased the number of terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling (TUNEL)-positive nuclei (Fig. 6 A,B) and the number of cytokeratin 18 (CK18)-positive cells (Fig. 6 C,D) in livers of WT mice. However, RIP3 deficiency did not attenuate this apoptotic response (Fig. 6A-D).

Figure 6.

RIP3 deficiency did not prevent markers of ethanol-induced apoptosis in mouse liver. C57BL/6J WT and RIP3-deficient mice were allowed free access to ethanol (4d,32% or 25d,32%) or pair-fed control diets. Paraffin-embedded livers were deparaffinized followed by (A,B) TUNEL or (C,D) CK18 staining. (A,B) Images were acquired using 40× objective for TUNEL and 20× for CK18 staining. TUNEL-positive cells were counted and are expressed as the percent positive of total number of cells per 40× frame. (C,D) CK18-positive cells were counted and are expressed as the total number of cells per 20× frame. Values with different alphabetical superscripts are significantly different from each other (P < 0.05).

Necrostatin-1 Treatment Does Not Attenuate Ethanol-Induced Liver Injury.

Although inhibition of RIP1 kinase activity with necrostatin-1 prevents cell death and improves pathology following ischemic injury in brain,7 RIP3 can also execute necroptotic cell death in an RIP1-independent manner.14 If ethanol-induced hepatocyte injury is RIP1 kinase–dependent, necrostatin-1 treatment should ameliorate ethanol-induced increases in plasma ALT/AST. Treatment of mice with necrostatin-1 did not attenuate the ethanol (4d,32%)-induced increase in ALT/AST or hepatic triglyceride accumulation (Fig. 7). Moreover, RIP1 protein expression in mouse liver remained unchanged following ethanol feeding (Supporting Fig. 1B).

Figure 7.

Necrostatin-1 treatment did not reduce ethanol-induced liver injury. C57BL/6J WT mice were allowed free access to ethanol (4d,32%) or pair-fed control diets. Mice received a once-daily injection of necrostatin-1 (1.65 mg/kg) intraperitoneally just prior to the daily provision of ethanol and control diet. ALT and AST activities were measured in plasma. Hepatic triglyceride content was measured in whole liver homogenates. Values with different alphabetical superscripts are significantly different from each other (P < 0.05).

RIP3 Is Critical for c-jun N-terminal Kinase Activation in Mice During Ethanol Exposure.

Activation of c-jun N-terminal kinase (JNK) is implicated to ethanol-induced steatosis and oxidative stress in mouse liver.31 If RIP3 is required for JNK activation, RIP3-deficiency should attenuate ethanol-induced phosphorylated JNK (pJNK). To test this hypothesis, we next assessed JNK activation using immunohistochemistry for pJNK. Ethanol feeding (4d,32%) induced pJNK-positive cells in the liver. Interestingly, most of the pJNK staining was restricted within the nuclei, with low cytosolic expression. RIP3 deficiency reduced the numbers of pJNK-positive cells in the liver (Fig. 8).

Figure 8.

JNK activation was reduced in RIP3-deficient mouse livers following chronic ethanol feeding. C57BL/6J WT and RIP3-deficient mice were allowed free access to ethanol (4d,32%) or pair-fed control diets. Paraffin-embedded livers were deparaffinized followed by pJNK immunodetection. (B) Enlargement of the black box in panel A. Black arrows indicate positive staining for pJNK; open arrows indicate cells void of pJNK-positive staining. (C) pJNK-positive cells were counted and are expressed as the total number of cells per 20× frame. All images were acquired using a 20× objective. Values with different alphabetical superscripts are significantly different from each other (P < 0.05).

Discussion

There is a direct association between cell death and progression of alcoholic liver disease, however, differential contributions of specific cell death pathways to hepatocyte injury during alcohol exposure is still not understood. When apoptosis is prevented either by genetic manipulation or pharmacological intervention, ethanol-induced liver injury still progresses in mouse models.16 These data suggest that nonapoptotic cell death pathways are critical for hepatocyte death following ethanol feeding. Ethanol also induces RIP3, a central molecule of the necroptosis pathway, concomitantly with the hepatocyte injury markers ALT and AST16 (Figs. 3A and 5A). We report for the first time that increased RIP3 expression was also detected in human liver biopsies from ALD patients. Mice deficient in RIP3 were protected from ethanol-induced steatosis, hepatocyte injury, and inflammation following both short-term and chronic ethanol feeding. In contrast to RIP3, expression of RIP1 remained unchanged following ethanol feeding. Moreover, treatment with a RIP1 kinase inhibitor, necrostatin-1, could not prevent ethanol-induced hepatocyte injury, indicating that ethanol-induced hepatocyte injury is RIP3-dependent but RIP1-independent.

Interaction of TNFα with its receptor tumor necrosis factor receptor 1 (TNFR1) initiates both apoptosis and necroptosis; cellular fate depends on a variety of intracellular factors, including energy status of the cell.12 Upon activation of TNFα-mediated signaling, RIP1 interacts with either caspase 8 to induce apoptosis or binds to RIP3, resulting in mitochondrial dysfunction and cell death via necroptosis.13 RIP3 can also execute cell death in an RIP1-independent manner following interaction of TNFα with TNFR1 via JNK activation and reactive oxygen species (ROS) overproduction. Ethanol induces TNFα expression in mouse liver as early as 4 days after feeding.32 Mice lacking TNFR1 are protected from ethanol-induced liver injury and inflammation, demonstrating a central role of TNFα-mediated signaling during progression of ethanol-induced liver injury.19, 33 Because TNFα is central to ethanol-induced liver damage, we hypothesized that during ethanol exposure, TNFα-driven prodeath signals converge at RIP3 to activate the necroptosis pathway, leading to hepatocyte injury and inflammation.

Excessive alcohol consumption for a short time span (binge consumption) or extended period of time (chronic consumption) leads to hepatic pathology. In recent years, binge drinking is becoming more frequent, particularly among young people. Therefore, to study RIP3-driven cell death pathway during progression of ethanol-induced liver injury, we used two different models of ethanol exposure. As a model of chronic alcohol consumption, mice were fed a diet with increasing concentrations of ethanol for 25 days. Alternatively, mice were fed a 4d,32% ethanol diet to mimic a binge drinking pattern. In both binge and chronic models, ethanol exposure increased RIP3 expression in WT mouse liver concomitant with ALT and AST, two markers of hepatocyte injury. RIP3 expression was also higher in livers of ALD patients compared with livers with normal pathology, indicating that RIP3 may also mediate human ALD.

If increased RIP3 results in activation of necroptosis, deletion of RIP3 should prevent hepatocyte death following ethanol exposure. Indeed, increased ALT/AST in response to binge or chronic ethanol feeding was prevented in RIP3-deficient mice. Ethanol feeding also induces lipid accumulation in the liver.19, 34 Absence of RIP3 also reduced ethanol-induced steatosis in the liver. However, the mechanism of RIP3-mediated lipid accumulation is still not understood. MCP-1 is implicated as a key regulator of ethanol-induced steatosis in mouse livers.35 Mice deficient in MCP-1 are protected from ethanol-induced hepatic lipid accumulation.35 Reduction in MCP-1 expression in the livers of RIP3-deficient mice is associated with reduced steatosis, suggesting that ethanol-mediated necroptosis induces MCP-1, which in turn activates steatosis.

In addition to RIP3 protein expression, chronic ethanol feeding also enhanced RIP3-FADD interactions, assessed using the Duolink PLA assay, in liver from WT mice. Interestingly, the number of cells showing RIP3-FADD interactions was much lower than the number of cells expressing RIP3 in the liver. These results suggest that while hepatocytes with higher RIP3 expression are likely at a greater risk for necroptotic cell death, only a subset of these RIP3-positive hepatocytes are actually undergoing necroptosis, as indicated by an increased RIP3-FADD interaction, during chronic ethanol feeding.

Accumulating evidence indicates that RIP3-driven cell death is RIP1 kinase-dependent. Necrostatin-1, a specific inhibitor of RIP1 kinase, has been shown to attenuate necroptotic cell death following ischemia/reperfusion injury in the brain7 and photoreceptor damage-associated retinal cell death.36 However, treatment with necrostatin-1 did not attenuate hepatocyte injury following binge ethanol exposure, indicating that ethanol-induced hepatocyte injury is RIP1-independent. These results corroborate a recent report by Linkermann et al.20 demonstrating that cell death following TNFα-mediated shock is RIP3-dependent but RIP1 kinase–independent. However, we cannot fully exclude that effects of necrostatin-1 were underestimated in our model due to the short half-life of necrostatin-1.

Increased RIP3 expression is implicated in a variety of pathological conditions including pancreatitis, ileitis, and retinal detachment-related tissue injury.11 The current data suggest that ethanol-induced liver injury should be added to the growing list of pathological conditions associated with RIP3 induction and necroptosis. Although a handful of reports indicate that RIP1-RIP3 complex formation leads to ROS overproduction in some cell types,6, 37, 38 other studies indicate that ROS acts as an upstream signal for initiation of necroptosis.39 Ethanol feeding induces ROS overproduction in the liver via multiple pathways, including CYP2E1-dependent ethanol metabolism, TNFα-mediated signaling, and JNK activation.40 CYP2E1-mediated ethanol metabolism is one of the major sources of hepatic ROS production during chronic ethanol feeding and deficiency of CYP2E1 reduces ethanol-induced oxidative stress marker, 4-HNE adducts, steatosis, and liver injury in mouse livers.22 If CYP2E1-mediated oxidative stress is an upstream activator of RIP3, absence of CYP2E1 would prevent ethanol-induced RIP3 expression, as well as liver injury. Indeed, ethanol-induced RIP3 expression and hepatocyte injury were blunted in CYP2E1-deficient mice. Thus, activation of necroptosis during ethanol exposure depends on CYP2E1-mediated ethanol metabolism. Moreover, RIP3 also appears to contribute to ROS production during ethanol feeding, as RIP3-deficient mice accumulate less 4-HNE adducts. Taken together, these data suggest a complex interplay between ROS and RIP3 in the liver.

Prolonged JNK activation is implicated in a variety of hepatic pathologies.41, 42 Interestingly, Yang et al.31 have shown that ethanol-induced oxidative stress in liver is JNK-dependent. Activation of JNK is also known to act as a downstream mediator of RIP3-driven necroptosis.12 Consistent with this data, RIP3 deficiency reduced the number of pJNK-positive cells in the liver following ethanol feeding, indicating that RIP3 contributes to JNK activation during chronic ethanol feeding, likely due to its role in ROS production.

While necroptosis shares the same initiation route with apoptosis, morphologically it resembles necrosis, associated with cell rupture and leakage of proinflammatory debris in the extracellular space.5 RIP3 deficiency serves to genetically suppress necroptosis and prevents inflammation in mouse models of cerulein-induced pancreatitis.6 Therefore, activation of necroptosis should aggravate inflammatory responses during ethanol exposure. Mice lacking RIP3 showed reduction in ethanol-induced inflammatory foci, expression of mRNA for inflammatory mediators and TNFα protein expression. Although, Deaciuc et al15 have reported that lipopolysaccharide-stimulated inflammation in the liver after chronic ethanol feeding is apoptosis-dependent, our previous work demonstrates that inhibition of apoptosis is not sufficient to reverse ethanol-induced expression of the proinflammatory mediators or increased hepatic infiltration of the immune cells.16 Consistent with the current data, we show that ethanol-mediated hepatic inflammation is regulated by RIP3-driven necroptosis, rather than apoptosis. ALD is one of the major health problems in the United States resulting in 80,000 deaths each year. However, there is currently a dearth of effective therapeutic strategies to prevent or treat ALD. Here, for the first time, we provide evidence demonstrating that RIP3-driven necroptosis is a central mediator of ethanol-induced hepatocyte injury, steatosis, and hepatic inflammation. Detection of this alternative cell death mechanism during ethanol-induced liver injury thus identifies a new therapeutic target for treatment of patients with ALD.

Ancillary