Knockdown of RIPK1 Markedly Exacerbates Murine Immune

Knockdown of RIPK1 Markedly Exacerbates Murine Immune-Mediated Liver Injury through Massive Apoptosis of Hepatocytes, Independent of Necroptosis and Inhibition of This information is current as NF-κB of September 25, 2021. Jo Suda, Lily Dara, Luoluo Yang, Mariam Aghajan, Yong Song, Neil Kaplowitz and Zhang-Xu Liu J Immunol 2016; 197:3120-3129; Prepublished online 7 September 2016; Downloaded from doi: 10.4049/jimmunol.1600690 http://www.jimmunol.org/content/197/8/3120 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2016/09/07/jimmunol.160069 Material 0.DCSupplemental References This article cites 48 articles, 12 of which you can access for free at: http://www.jimmunol.org/content/197/8/3120.full#ref-list-1

Why The JI? Submit online. by guest on September 25, 2021

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2016 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Knockdown of RIPK1 Markedly Exacerbates Murine Immune-Mediated Liver Injury through Massive Apoptosis of Hepatocytes, Independent of Necroptosis and Inhibition of NF-kB

Jo Suda,*,1 Lily Dara,*,1 Luoluo Yang,*,† Mariam Aghajan,‡ Yong Song,x Neil Kaplowitz,* and Zhang-Xu Liu*

Receptor-interacting protein kinase (RIPK)1 has an essential role in the signaling pathways triggered by death receptors through activation of NF-kB and regulation of caspase-dependent apoptosis and RIPK3/mixed lineage kinase domain-like protein

(MLKL)-mediated necroptosis. We examined the effect of RIPK1 antisense knockdown on immune-mediated liver injury in Downloaded from C57BL/6 mice caused by a-galactosylceramide (aGalCer), a specific activator for invariant NKT cells. We found that knockdown of RIPK1 markedly exacerbated aGalCer-mediated liver injury and induced lethality. This was associated with increased hepatic inflammation and massive apoptotic death of hepatocytes, as indicated by TUNEL staining and caspase-3 activation. Pretreatment with zVAD.fmk, a pan-caspase inhibitor, or neutralizing Abs against TNF, almost completely protected against the exacerbated liver injury and lethality. Primary hepatocytes isolated from RIPK1-knockdown mice were sensitized to TNF-induced cell death k that was completely inhibited by adding zVAD.fmk. The exacerbated liver injury was not due to impaired hepatic NF- B http://www.jimmunol.org/ activation in terms of IkBa phosphorylation and degradation in in vivo and in vitro studies. Lack of RIPK1 kinase activity by pretreatment with necrostatin-1, a RIPK1 kinase inhibitor, or in the RIPK1 kinase-dead knock-in (RIPK1D138N) mice did not exacerbate aGalCer-mediated liver injury. Furthermore, RIPK3-knockout and MLKL-knockout mice behaved similarly as wild- type control mice in response to aGalCer, with or without knockdown of RIPK1, excluding a switch to RIPK3/MLKL-mediated necroptosis. Our findings reveal a critical kinase-independent platform role for RIPK1 in protecting against TNF/caspase-dependent apoptosis of hepatocytes in immune-mediated liver injury. The Journal of Immunology, 2016, 197: 3120–3129.

eceptor-interacting protein kinase (RIPK)1 is the founding inflammation (5, 6). Cell death in multiple organs, including the member of the serine/threonine kinase family that has an liver, is the most striking histological finding in RIPK12/2 mice, and by guest on September 25, 2021 R essential role in regulating caspase-dependent apoptosis and apoptosis (cleaved caspase-3) is increased in the liver (5, 6). Al- RIPK3/mixed lineage kinase domain-like protein (MLKL)-dependent though the mechanisms by which RIPK1 prevents perinatal lethality necroptosis, as well as in promoting cell survival and inflammation by are complex, recent studies of genetic rescue of perinatal death in serving as a platform for activation of the canonical NF-kBpathway RIPK12/2 mice demonstrated that combined loss of caspase-8 and (1–4). Complete RIPK1 deficiency (RIPK12/2) in mice results in RIPK3 protects against the perinatal lethality and systemic inflam- perinatal death characterized by multiorgan cell death and systemic mation, indicating that RIPK1 can function as an inhibitor to suppress caspase-mediated apoptosis and RIPK3-mediated necroptosis (6–8). In support of this, the conditional deletion of RIPK1 in intestinal *Research Center for Liver Diseases, Keck School of Medicine, University of South- epithelial cells (IECs) was reported to result in apoptosis of IECs and ern California, Los Angeles, CA 90033; †Department of Gastroenterology, Bethune First Hospital of Jilin University, Changchuan 130021, China; ‡Ionis Pharmaceuti- intestinal inflammation (9, 10), suggesting that RIPK1 maintains x cals, Carlsbad, CA 92008; and YSL Bioprocess Development Co., Pomona, CA epithelial homeostasis by inhibiting apoptosis in the intestine (9). 91767 Hepatocyte death (apoptosis or necrosis) was critically impli- 1 J.S. and L.D. are cofirst authors. cated in liver pathology. Necroptosis, a recently defined necrosis ORCIDs: 0000-0002-0121-7186 (L.D.); 0000-0001-8099-9982 (L.Y.); 0000-0002- regulated by the RIPK1/RIPK3/MLKL signaling pathway (1–3, 9424-393X (N.K.); 0000-0001-7290-3506 (Z.-X.L.). 11), was also suggested to account for hepatocyte death in liver Received for publication April 22, 2016. Accepted for publication August 16, 2016. diseases, including alcoholic liver disease (12), drug-induced liver This work was supported by National Institutes of Health Grants R01DK067215 and injury (13, 14), and immune-mediated liver injury (15, 16). In P30DK48522 (to N.K.), and R01DK078586 and U01AA021857 (to Z.-X.L.). studies of perinatal lethality in RIPK12/2 mice, the liver is among Address correspondence and reprint requests to Dr. Zhang-Xu Liu, Research Center for Liver Diseases, Keck School of Medicine, University of Southern California, several organs exhibiting cell death and inflammation (5, 6), sug- 2011 Zonal Avenue, HMR 101, Los Angeles, CA 90033. E-mail address: zxliu@usc. gesting a possible role for RIPK1 in regulating apoptosis and/or edu necroptosis of hepatocytes. However, in more recent studies using The online version of this article contains supplemental material. inducible conditional knockout of RIPK1 in adult mice (17), the Abbreviations used in this article: ALT, alanine aminotransferase; APAP, acute acet- liver appears to be spared from cell death and inflammation, aminophen; ASO, antisense oligonucleotide; c-FLIP, cellular FLIP; cIAP, cellular inhibitor of apoptosis protein; Cont ASO, control ASO; aGalCer, a-galactosylceramide; despite the fact that acute ablation results in intestinal epithelial IEC, intestinal epithelial cell; iNKT, invariant NKT; MLKL, mixed lineage kinase domain- and hematopoietic cell death. Little is known about the physio- like protein; Nec-1, necrostatin 1; qPCR, quantitative PCR; RIPK, receptor-interacting logical role of RIPK1 in regulating hepatocyte death (apoptosis protein kinase; WT, wild-type; zVAD, zVAD.fmk. or necroptosis) triggered by death receptors, such as TNF and Copyright Ó 2016 by The American Association of Immunologists, Inc. 0022-1767/16/$30.00 Fas, in immune-mediated liver injury. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1600690 The Journal of Immunology 3121

CD1d-restricted invariant NKT (iNKT) cells are a specialized protein (cIAP)1 mAb (1E1-1-10) was from Enzo Life Sciences (Farmingdale, subset of T cells that recognize lipid Ags presented by CD1d mol- NY). Anti-RIPK3 was provided by Genentech. Purified TNF-a mAb was ecules on APCs (18). iNKT cells are enriched in the liver, accounting from BioLegend (San Diego, CA), anti-Fas Ab (Jo2), was from BD Bio- sciences (San Jose, CA), necrostatin-1 (Nec-1) and inactive Nec-1 were for up to 40% of hepatic lymphocytes in mice, and are implicated in from Calbiochem (San Diego, CA), aGalCer was from Funakoshi (Tokyo, the pathophysiology of a variety of liver diseases, including viral and Japan), pan-caspase inhibitor zVAD was from R&D Systems (Minneapolis, autoimmune hepatitis, alcoholic liver disease, and nonalcoholic fatty MN), and LPS and Con A were from Sigma-Aldrich. RIPK1 and scrambled liver disease (19–22). Direct activation of hepatic iNKT cells by ASO were provided by Ionis Pharmaceuticals (Carlsbad, CA). a-galactosylceramide (aGalCer), a specific high-affinity ligand for Animals, treatments, and assays for serum alanine iNKT cells, results in rapid release of large quantities of various aminotransferase cytokines, including TNF and IFN-g from iNKT cells. This promotes Pathogen-free male C57BL/6J mice, 6–8 wk of age, were obtained from the the activation and infiltration of NK cells, macrophages, and neu- Jackson Laboratory (Bar Harbor, ME). MLKL2/2 mice (on a C57BL/6J trophils, all of which lead to a mild immune-mediated liver injury background) were provided by Professor Warren Alexander (Walter and (18, 19, 23, 24). In this study, we used this well-established animal Eliza Hall Institute of Medical Research, Parkville, Australia). RIPK1 D138N 2/2 model of hepatitis to explore the function of RIPK1 in regulating kinase-dead knock-in (RIPK1 ) mice and RIPK3 mice on a immune-mediated liver injury. After RIPK1 antisense oligonucleo- C57BL/6N background were provided by Dr. Vishva Dixit (Genentech), and male C57/BL6N mice (6–8 wk old) were obtained from Harlan Bio- tide (ASO) treatment, we found that knockdown of RIPK1 markedly products for Science (Indianapolis, IN) as controls. RIPK1 and scrambled exacerbates aGalCer-mediated liver injury and induces lethality as- ASO were dissolved in PBS, and animals were injected i.p. with ASO sociated with increased hepatic inflammation. We provide evidence (50 mg/kg) every other day for a total of five injections, as previously that the exacerbation of liver injury is due to TNF- and caspase- described (25). The additional treatments were started the day after the last Downloaded from ASO injection. The high efficiency of hepatic RIPK1 knockdown was dependent massive apoptosis of hepatocytes because neutralization confirmed by Western blot analysis and mRNA levels of liver tissues. We of TNF or inhibition of caspases by pan-caspase inhibitor zVAD.fmk also confirmed that scrambled or RIPK1 ASO treatment alone is not (zVAD) confers almost complete protection. We further show that hepatotoxic, as evidenced by normal serum alanine aminotransferase RIPK3/MLKL-mediated necroptosis does not contribute to the ex- (ALT) and liver histology. aGalCer was stored as a stock solution of 2 2 2 2 m acerbation of liver injury because RIPK3 / and MLKL / mice 200 g/ml in vehicle (0.5% w/v polysorbate 20) and diluted in PBS before i.p. injection (4 mg per mouse in a total volume of 200 ml). For LPS- exhibit exacerbated liver injury and lethality and respond similarly to induced liver injury, a nonlethal dose of LPS (3 mg/kg) was injected i.p. http://www.jimmunol.org/ wild-type (WT) control mice. Unexpectedly, the exacerbation of liver For Fas-mediated liver injury, 4 mg of anti-Fas (Jo2) Ab was injected i.p. injury is not due to impairment of the prosurvival NF-kB signaling For Con A–induced liver injury, a low-toxic dose of Con A (10 mg/kg) was pathway. Moreover, lack of RIPK1 kinase activity in RIPK1D168N injected via the tail vein. For inhibition of caspases, zVAD (10 mg/kg) was injected i.p. 15 min prior to aGalCer or Con A treatment. For neutrali- mice does not exacerbate aGalCer-mediated liver injury, indicating zation of TNF, 0.6 mg per mouse of anti–TNF-a mAb was administered that it is the platform function of RIPK1 protein, and not the kinase i.p. 15 min prior to aGalCer. Serum ALT levels were measured with ALT activity, that regulates TNF/caspase-dependent apoptosis of hepato- reagents from Teco Diagnostics (Anaheim, CA). Animal experiments were cytes in response to iNKT cell activation. carried out in accordance with guidelines from the University of Southern California Institutional Animal Care and Use Committee.

Materials and Methods Preparation of liver extracts and Western blot analyses by guest on September 25, 2021 Reagents Total liver lysates or whole-cell lysates of cultured primary mouse hepa- Abs to RIPK1, caspase-3, IkBa, p-IkBa (Ser32/36), and b-actin were from tocytes were prepared in RIPA buffer with complete protease and phos- Cell Signaling Technology (Danvers, MA). Anti–c-FLIP mAb (Dave-2) phatase inhibitor mixture (Roche), as previously described (26). Aliquots of was from Adipogen (San Diego, CA), and cellular inhibitor of apoptosis protein extracts were subjected to 12% SDS-PAGE. After electrophoresis,

FIGURE 1. Knockdown of RIPK1 markedly exacerbates aGalCer-induced liver injury and causes mortality. Groups of mice were treated with scrambled (Cont ASO) or RIPK1 ASO every other day five times to silence RIPK1 prior to aGalCer (4 mg per mouse i.p.). (A) Serum ALT values at indicated times after aGalCer. Data are representative of three separate experiments. *p , 0.05, versus scrambled controls at 8 h. (B) Mouse survival for 24 h after aGalCer. *p , 0.05, versus scrambled controls. (C) H&E staining of liver sections at indicated times after aGalCer (original magnification 3400). (D) Western blotting of whole-liver lysates for RIPK1 at indicated times after aGalCer. 3122 RIPK1 IN IMMUNE-MEDIATED LIVER INJURY proteins were transferred to a polyvinylidene difluoride membrane and c-FLIP and qPCR array analyses of NF-kB target genes (RT2 Profiler (Amersham Biosciences) and incubated with the desired primary and sec- PCR Arrays; QIAGEN) were performed with the 7900HT Sequence De- ondary Abs. Finally, the proteins were detected using Luminol ECL reagent tection System (Life Technologies). GAPDH was analyzed as an internal (Thermo Scientific). Densitometry was done using ImageJ software. All gels control. Data are expressed as fold changes. and densitometry shown are representative of at least three experiments, un- less otherwise indicated. Bead-based multiplex immunoassays for cytokine analysis Hepatocytes isolation, culture, and cell death assay Concentrations of a panel of cytokines and chemokines in mouse serum Primary mouse hepatocytes were isolated and cultured as described previously samples were determined using the AimPlex Mouse 18-Plex Assay Kit (YSL (27). Three hours after plating of isolated hepatocytes, TNF (20 ng/ml), Bioprocess Development, Pomona, CA), according to the manufacturer’s with or without pan-caspase inhibitor zVAD (10 mM), was added. After instructions. In brief, 15 ml of serum sample was diluted with 30 mlofassay 16 h, cells were double stained with Hoechst 33258 (8 mg/ml) and SYTOX buffer in each well of a 96-well filter plate and incubated with capture Ab- Green (1 mmol/l; both from Invitrogen). Quantitation of cell death was conjugated AimPlex beads for 1 h. After washing the beads three times performed by counting a minimum of 300 cells in different fields. using the AimPlex EZPrep Filter Plate Washer, a mixture of biotinylated detection Abs was applied, followed by the addition of PE-conjugated Histology and TUNEL staining streptavidin. Data were acquired on a FACSCalibur flow cytometer (BD Biosciences) and analyzed with FCAP Array v3.0 (Soft Flow Hungary). For histology, liver tissue was fixed in 10% neutral buffered formalin, and sections were stained with H&E to determine morphologic changes. TUNEL staining was performed using a fluorescent detection kit (Roche Preparation of hepatic leukocytes and flow cytometric analysis Diagnostics), following the manufacturer’s instructions. Hepatic leukocytes were isolated and prepared as described previously 6 RNA isolation and real-time PCR (20, 26). For flow cytometric analysis, 10 cells were incubated for 10 min with Fcg receptor blocker (anti-mouse CD16/32, clone 2.4G2; Downloaded from Liver tissues were homogenized in TRI reagent (Sigma-Aldrich), and RNA Tonbo Biosciences, San Diego, CA) and then stained with mAbs at a was extracted using an RNeasy Plus Mini Kit (QIAGEN). Total RNA (2 mg) concentration of 0.5–1 mg/100 ml PBS containing 0.2% BSA (Boehringer was reverse transcribed in a 20-ml reaction mixture containing 4U Mannheim, Indianapolis, IN) and 0.02% sodium azide for 30 min on ice. Omniscript Reverse Transcriptase (QIAGEN) and oligo dT(12–18mer) primer The following Abs were used: FITC-conjugated anti-CD3 (145-2C11), (Life Technologies). The following sense and antisense primers were used anti-Ly6C (HK1.4); PE-conjugated anti NK1.1 (PK136), anti-F4/80 (BM8.1); in the PCR reaction: cIAP1 sense 59-GGCCACCTAGTGTTCCTGTT-39 PerCP-conjugated anti-Ly6G (1A8), anti-CD19 (1D3); allophycocyanin- and antisense 59-CAACATCTCAAGCCACCATC-39, and cellular FLIP Cy7–conjugated anti-CD45 (BioLegend); and PE-Cy7–conjugated anti- http://www.jimmunol.org/ (c-FLIP) sense 59-TCAGGGTTGTCTCGTCAGAG-39 and antisense CD11b (M1-70) (BD Biosciences). Flow cytometric analysis was performed 59-CAGAGACCACCAGCTACGAA-39. Quantitative PCR (qPCR) for cIAP1 on a FACSVerse (BD Biosciences).

FIGURE 2. Knockdown of RIPK1 by guest on September 25, 2021 causes massive hepatic apoptosis in response to aGalCer. Animals were treated with Cont ASO or RIPK1 ASO to silence RIPK1 prior to aGalCer (4 mg per mouse i.p.). (A)TUNEL staining of liver sections at 6 h after aGalCer (original magniﬁcation 3200). DAPI was used for nucleus staining. Data are representative of images with similar results. (B) Western blotting for hepatic caspase-3 and cleaved caspase-3 in whole-liver lysates at 6 h after aGalCer. Data are representative of three separate experiments. (C–E) Pretreatment by pan-caspase inhibitor zVAD almost completely protects against the exacerbated liver injury and lethality in RIPK1 ASO+aGalCer- treated mice. RIPK1 ASO mice were treated with vehicle or zVAD (10 mg/kg i.p.) 15 min prior to aGalCer treatment (4 mg per mouse i.p.). (C)SerumALT measured at 6 h after aGalCer (n =5 per group). (D) Representative images of H&E staining of liver sections at 6 h after aGalCer (original magniﬁ- cation 3400). (E) Mouse survival at 8hafteraGalCer (n =6).*p , 0.05. The Journal of Immunology 3123

Statistical analysis remained much higher compared with control mice (Fig. 1A). Data are expressed as mean 6 SEM. Group comparisons were performed RIPK1 knockdown caused massive hepatocyte death and hemor- using the Student t test or ANOVA. The log-rank test was used for mouse rhage in the liver at 6 and 24 h compared with small patchy ne- survival. Differences were considered statistically significant at p , 0.05. crotic foci with inflammatory cell infiltration in the H&E-stained liver sections from Cont ASO–treated mice at 6 and 24 h Results (Fig. 1C). Accompanying the exacerbated liver injury, RIPK1 Knockdown of RIPK1 markedly exacerbates knockdown significantly increased serum levels of proinflammatory a GalCer-mediated liver injury and induces lethality cytokines and chemokines, including TNF and IFN-g,at6hafter First, we examined the effect of RIPK1 knockdown on the liver aGalCer and caused highly increased inflammatory cell infiltration injury induced by i.p. injection of aGalCer. In control mice pre- into the liver compared with control mice (Supplemental Fig. 1). treated with scrambled control ASO (Cont ASO), aGalCer caused However, no differences in the serum levels of cytokines/chemokines mild to moderate liver injury, as evidenced by increased serum wereobservedat3hafteraGalCer (data not shown), suggesting that ALT values starting at 6 h, which peaked around 12–24 h RIPK1 knockdown has no direct effect on the rapid activation of (Fig. 1A), without any mortality (Fig. 1B). In striking contrast, iNKT cells by aGalCer. We confirmed that RIPK1 ASO efficiently when hepatic RIPK1 was knocked down by RIPK1 ASO pre- knocked down hepatic RIPK1 by Western blotting (Fig. 1D) and treatment, aGalCer markedly increased serum ALT at 6 h and qPCR (data not shown). Together, these results indicate that the lack induced ∼80% mortality by 8 h (Fig. 1A, 1B). For the remaining of RIPK1 proteins in the liver markedly exacerbates aGalCer- RIPK1 ASO mice surviving to 12 and 24 h, the serum ALT levels mediated liver injury. Downloaded from http://www.jimmunol.org/

FIGURE 3. Pretreatment by neutralizing Abs against TNF-a almost completely protects against the exacerbated liver injury and lethality in

RIPK1 ASO+aGalCer-treated mice. by guest on September 25, 2021 RIPK1 ASO–treated mice were treated with vehicle or anti-TNFa Ab (600 mg per mouse, i.p.) 15 min prior to aGalCer treatment (4 mgpermouse i.p.). (A)SerumALTvaluesat6h after aGalCer (n = 4–6) *p , 0.05. (B) Representative H&E-stained liver sections at 6 h after aGalCer (original magniﬁcation 3400). (C)Mousesur- vival at 8 h after aGalCer (n =6). (D and E) RIPK1 knockdown sensitizes hepatocytes to TNF- induced apoptosis. Primary mouse hepatocytes isolated from Cont ASO– or RIPK1 ASO– treated mice were treated with TNF-a (20 ng/ml) for 16 h in the absence or presence of pan-caspase inhibitor zVAD (10 mM). (D) Data are representative of microscopic images of primary hepatocytes stained by Hoechst and SYTOX Green (original magniﬁcation 3200). (E) The percentages of cell death at 16 h. *p , 0.05, versus hepatocytes from Cont ASO–treated mouse. 3124 RIPK1 IN IMMUNE-MEDIATED LIVER INJURY

Knockdown of RIPK1 causes massive caspase-dependent ASO mice were sensitized to TNF-mediated cell death and dis- apoptosis of hepatocytes in response to aGalCer played a significantly increased cell death, as indicated by SYTOX Apoptosis of hepatocytes is often associated with immune-mediated Green staining. Cotreatment with zVAD almost completely ab- hepatitis. To determine whether apoptosis contributes to the exacer- rogated the cell death. Thus, RIPK1 knockdown sensitized to TNF bated liver injury in RIPK1 ASO–treated mice in response to induced caspase-dependent apoptosis of hepatocytes in vitro or aGalCer, we performed TUNEL staining of liver sections at 6 h after after aGalCer in vivo. Although increased serum TNF levels due aGalCer. As shown in Fig. 2A, aGalCer treatment drastically in- to RIPK1 knockdown in iNKT cells or other inflammatory cells creased TUNEL+ hepatocytes in RIPK1 ASO mice compared with may contribute to the enhanced apoptosis of hepatocytes, our control mice. Furthermore, liver extracts from RIPK1 ASO–treated in vitro findings indicate that hepatocytes, in the absence of mice exhibited increased protein levels of cleaved caspase-3 at 3 and RIPK1, become intrinsically more susceptible to TNF. In addition 6hafteraGalCer, whereas no cleaved caspase-3 was detected in the to TNF, the Fas/FasL pathway was implicated in aGalCer-mediated liver of control mice (Fig. 2B). These data suggest an involvement of liver injury (20). To examine whether knockdown of RIPK1 caspase-mediated apoptosis. To further validate the role of apoptosis, sensitizes hepatocytes to Fas-mediated apoptosis, animals were pretreatment with zVAD afforded almost complete protection against injected with a sublethal dose of anti-Fas (Jo2) to induce acute the exacerbated aGalCer-mediated liver injury in RIPK1 ASO mice liver injury (28). The liver injury (ALT and histology) was com- in terms of serum ALT levels and histologic liver injury (Fig. 2C, parable between scrambled Cont ASO mice and RIPK1 ASO mice 2D).By8hafteraGalCer, only ∼20% of the RIPK1 ASO mice in response to Jo2 treatment (Supplemental Fig. 2), excluding the without zVAD pretreatment survived, whereas all of the RIPK1 ASO role of Fas-mediated apoptosis in the exacerbated cell death in mice with pretreatment of zVAD were alive (Fig. 2E). Thus, these response to aGalCer. Downloaded from results indicate that the exacerbated aGalCer-mediated liver injury Knockdown of RIPK1 does not impair NF-kB activation in the and the lethality in RIPK1 ASO mice are due to caspase-dependent liver and hepatocytes massive apoptosis of hepatocytes. Ubiquitination of RIPK1 is believed to be required for TNF- Knockdown of RIPK1 sensitizes to TNF-dependent apoptosis mediated prosurvival signaling (3, 5, 29). Therefore, it is con-

Activation of iNKT cells by aGalCer rapidly releases the proin- ceivable that the exacerbated hepatocyte apoptosis in RIPK1 http://www.jimmunol.org/ ﬂammatory cytokine TNF, which plays a critical role in aGalCer- ASO–treated mice may simply be attributed to impairment of mediated liver injury (20, 24). To determine the role of TNF in the NF-kB activation. To this end, we examined the activity of NF-kB exacerbation of liver injury in RIPK1 ASO mice, animals were by Western blot analysis of phosphorylated IkBa and total IkBa pretreated with neutralizing Abs against TNF. As indicated in proteins in the liver and primary mouse hepatocytes. Surprisingly, Fig. 3, neutralization of TNF conferred almost complete protec- RIPK1 ASO mice exhibited levels of hepatic phosphorylation and tion against the exacerbated liver injury caused by aGalCer in degradation of IkBa proteins that were comparable to those ob- RIPK1 ASO mice, as evidenced by a dramatic reduction in serum served in control mice at 1, 3, and 6 h in response to aGalCer ALT and histologic liver injury (Fig. 3A, 3B), and complete (Fig. 4A). Furthermore, primary mouse hepatocytes from RIPK1 protection against lethality (Fig. 3C). Next, we performed in vitro ASO–treated mice also exhibited a similar time course of phos- by guest on September 25, 2021 studies to examine the direct role of TNF in primary mouse he- phorylation and degradation of IkBa proteins compared with that patocytes. As shown in Fig. 3D, hepatocytes isolated from RIPK1 from control mice in response to TNF stimulation in vitro (Fig. 4C).

FIGURE 4. Knockdown of RIPK1 has no effect on hepatic NF-kB activation and causes degradation of c-FLIP and cIAP1. (A) Western blot for detection of p-IkBa, total IkBa, RIPK1, c-FLIP, and cIAP1 in the liver tissues from Cont ASO and RIPK1 ASO mice at the indicated times after aGalCer. (B) Densitometry of Western blotting bands for hepatic c-FLIP and cIAP1 from three separate experiments. (C) Western blot for detection of p-IkBa, total IkBa, and RIPK1 in primary hepatocytes isolated from Cont ASO and RIPK1 ASO mice. Primary hepatocytes were cultured with TNF-a (20 ng/ml) for the indicated times before being subjected to whole-cell lysate protein extracts. Data are representative of three separate experiments. (D)Westernblot for detection of c-FLIP and cIAP1 in primary hepatocytes isolated from Cont ASO and RIPK1 ASO mice and cultured with TNF (20 ng/ml) for the indicated times. Data are representative of two separate experiments with similar results. *p , 0.05. The Journal of Immunology 3125

Moreover, we examined mRNA expression of a group of downstream genes regulated by NF-kB activation in the liver tissues and isolated primary hepatocytes at 3 h after aGalCer treatment (Supplemental Fig. 3). The results showed that these genes, including IkBa,c-FLIP, cIAP1, TNF, IFN-g, iNOS, and SOD2, were signiﬁcantly induced by aGalCer treatment in the liver tissues, as well as the primary hepatocytes, from scrambled Cont ASO– and RIPK1 ASO–treated mice. Together, these results indicate that NF-kB activation is intact when RIPK1 is knocked down in the liver and hepatocytes, suggesting that the exacerbated liver injury and hepatocyte death in RIPK1 ASO– treated mice in response to TNF are not due to the impairment of NF-kB signaling. RIPK1 was shown to promote cell survival in TNF-stimulated cells by stabilizing antiapoptotic proteins, such as cIAP1 and c-FLIP (9, 29, 30). c-FLIP–deﬁcient hepatocytes were demonstrated to be highly susceptible to apoptosis induced by death receptors (TNFR1, Fas, and TRAIL) (31, 32). To this end, we examined the protein levels of hepatic c-FLIP and cIAP1 by Western blot analysis. As shown in Fig. 4A and 4B, we found that the induction of Downloaded from hepatic protein levels of c-FLIP and cIAP1 was decreased after aGalCer treatment in RIPK1 ASO–treated mice compared with control mice. We further examined the protein levels of c-FLIP and cIAP1 in primary hepatocytes in response to TNF stimulation in vitro. As shown in Fig. 4D, the induction of cIAP1 and c-FLIP at

3 and 6 h after TNF was decreased in primary hepatocytes from http://www.jimmunol.org/ RIPK1 ASO–treated mice compared with that of control mice. Coupled with the lack of impairment of NF-kB–induced increases in mRNA expression of c-FLIP and cIAP1 in liver tissues and primary hepatocytes, these results suggest that degradation of c-FLIP and cIAP1 in the hepatocytes of RIPK1 ASO–treated mice in the absence of RIPK1 may contribute to the massive apoptosis of hepatocytes and the exacerbated liver injury.

RIPK1 kinase activity is dispensable for the exacerbation of by guest on September 25, 2021 aGalCer-mediated liver injury Next, we used transgenic RIPK1 kinase-dead knock-in mice D138N FIGURE 5. Loss of RIPK1 kinase activity does not exacerbate aGalCer- (RIPK1 ) (33) to examine whether the exacerbation of D138N aGalCer-mediated liver injury in RIPK1 ASO mice is due to lack mediated liver injury in RIPK1 kinase-dead knock-in (RIPK1 ) mice. (A) Serum ALT values at 0 h and at 8 h after aGalCer treatment (4 mg per of the platform function or the kinase activity of RIPK1. As shown . D138N B D138N mouse, i.p.) (n = 5). p 0.05, RIPK1 versus WT. ( ) Representative in Fig. 5, RIPK1 mice lacking RIPK1 kinase activity ex- H&E-stained liver sections at 0 h and at 8 h after aGalCer (original hibited a similar response to aGalCer as did WT control mice magnification 3400). (Fig. 5). The lack of exacerbation of aGalCer-mediated liver injury in RIPK1D138N mice suggests that the scaffolding function of RIPK1 plays a critical role in regulating hepatocyte death. In showed that loss of RIPK3 or the necroptosis executioner MLKL support of this hypothesis, pretreatment with Nec-1, a specific did not protect against the markedly exacerbated aGalCer- RIPK1 kinase inhibitor, showed no exacerbation of aGalCer- mediated liver injury and lethality after RIPK1 ASO treatment, mediated liver injury in WT mice (Supplemental Fig. 4). excluding the involvement of RIPK3/MLKL-mediated necroptosis of hepatocytes as the mechanism for increased injury. Further- RIPK3/MLKL-dependent necroptosis is not involved in the more, as shown above, inhibition of caspases with zVAD treat- exacerbated liver injury and lethality in RIPK1 ASO mice in ment did not cause a switch to necroptosis (Figs. 2, 3). response to aGalCer RIPK1 is known to regulate caspase-dependent apoptosis and Knockdown of RIPK1 markedly sensitizes to LPS- and RIPK3/MLKL-mediated necroptosis (1–8). To exclude the possi- Con A–mediated liver injury and causes lethality bility that necroptosis is implicated in the exacerbation of Finally, we examined whether the effect of knockdown of RIPK1 in aGalCer-mediated liver injury in RIPK1 ASO–treated mice, we aGalCer toxicity extends to other TNF-related models (i.e., LPS- performed a series of experiments using RIPK32/2 and MLKL2/2 and Con A–mediated liver injury). LPS toxicity is associated with mice. In Fig. 6, RIPK32/2 mice behaved similarly to WT control activation of macrophages/Kupffer cells in the liver and release of mice, exhibiting a mild to moderate liver injury in response to inflammatory cytokines, particularly TNF. In contrast, Con A aGalCer alone. However, when RIPK1 was knocked down in causes T cell–mediated hepatitis in which TNF also plays a criti- RIPK32/2 mice, aGalCer still induced markedly exacerbated liver cal role in hepatocyte death, although necrosis, rather than apo- injury and high mortality by 8 h (Fig. 6). Furthermore, MLKL2/2 ptosis, appears to be the predominant mode of cell death. We mice also behaved similarly to WT mice and RIPK32/2 mice, performed studies on LPS-induced liver injury using a dose that is exhibiting markedly exacerbated aGalCer-mediated liver injury not lethal to normal mice. As shown in Fig. 8A and 8B, a low and lethality after RIPK1 ASO (Fig. 7). Together, these results nonlethal dose of LPS (3 mg/kg, i.p.) caused very mild liver injury 3126 RIPK1 IN IMMUNE-MEDIATED LIVER INJURY Downloaded from

FIGURE 7. Similar to WT and RIPK32/2 mice, knockdown of RIPK1

markedly increases aGalCer-mediated liver injury and induces lethality in http://www.jimmunol.org/ FIGURE 6. Similar to WT control mice, knockdown of RIPK1 mark- 2/2 2/2 edly increases aGalCer-mediated liver injury and induces lethality in MLKL mice. MLKL mice were treated with Cont ASO or RIPK1 A RIPK32/2 mice. RIPK32/2 mice were treated with Cont ASO or RIPK1 ASO prior to aGalCer treatment (4 mg per mouse, i.p.). ( ) Serum ALT B ASO prior to aGalCer treatment (4 mg per mouse, i.p.). (A) Serum ALT values at 6 h after aGalCer (n = 4–6). ( ) Representative H&E-stained C values at 6 h after aGalCer (n = 5–8). (B) Representative H&E-stained liver sections at 6 h after aGalCer. ( ) Mouse survival at 8 h after aGalCer 3 liver sections at 6 h after aGalCer. (C) Mouse survival at 8 h after aGalCer (original magnification 400). (original magnification 3400). (e.g., actinomycin D), which could be blocked by caspase inhibition. Together, these data indicate that RIPK1 in hepatocytes (serum ALT and liver histology) in scrambled ASO–treated con- acts as an inhibitor of apoptosis mediated through TNFR signaling by guest on September 25, 2021 trol mice. However, as in the aGalCer model, knockdown of and has a prosurvival role in hepatocytes in immune-mediated RIPK1 markedly enhanced LPS-induced liver injury, as evidenced liver injury. Interestingly, our findings are consistent with a recent by dramatically increased serum ALT and massive hepatocyte death report on the conditional deletion of RIPK1 in IECs (9). IEC- and hemorrhage in the liver accompanied by increased positive specific RIPK12/2 mice spontaneously develop intestinal inflam- TUNEL staining of hepatocytes (Fig. 8C). Furthermore, all of the mation associated with IEC apoptosis, leading to early death (9). RIPK1 ASO–treated mice (n = 5) succumbed to LPS treatment by The lethality could be rescued by antibiotic treatment and defi- 8 h, whereas all of the Cont ASO–treated mice (n = 5) survived. ciency of TNFR1 or caspase-8, indicating that gut commensal Similarly, a low toxic dose of Con A (10 mg/kg) markedly in- bacteria–mediated TNF/caspase-dependent apoptosis contributes to creased liver injury (ALT and histology) in RIPK1 ASO–treated cell death and lethality (9). In agreement with recent genetic studies mice compared with Cont ASO–treated mice (Fig. 8E, 8F). Con A showing that the platform function of RIPK1 can suppress caspase- ∼ also induced lethality in RIPK1 ASO–treated mice with an 60% 8–dependent apoptosis and RIPK3/MLKL-mediated necroptosis mortality by 6 h (n = 11), whereas all of the Cont ASO–treated (6–8), we found that it is the platform function of RIPK1, not its mice survived (n = 10). Furthermore, the exacerbated Con A– kinase activity, that contributes to the exacerbation of aGalCer- mediated liver injury in RIPK1 ASO–treated mice was almost mediated liver injury in RIPK1 ASO mice. The lack of RIPK1 completely prevented by pretreatment with zVAD (Fig. 8E, 8F), kinase activity by Nec-1 treatment in WT mice or the knock-in of indicating that a massive apoptosis of hepatocytes contributes to dominant-negative RIPK1 (RIPK1D138N mice) does not exacerbate the exacerbation of liver injury in RIPK1 ASO–treated mice. aGalCer-mediated liver injury. This is also supported by the finding in IEC-specific RIPK12/2 mice that the platform function of Discussion RIPK1, instead of its kinase activity, maintains homeostasis of IECs In this study, we found that knockdown of hepatic RIPK1 markedly by suppressing TNF-mediated apoptosis (9). Collectively, the exacerbates aGalCer-mediated liver injury and induces lethality, platform function of RIPK1 in hepatocytes and IECs appears to and we clearly demonstrated that exacerbation of liver injury and protect these epithelial cells against apoptosis in response to the lethality in RIPK1 ASO–treated mice is due to a massive TNF/ TNFR signaling pathway. caspase-dependent apoptosis of hepatocytes. Our conclusions are After TNFR stimulation, it is generally believed that RIPK1 protein, based upon the following findings: increased TUNEL+ staining of not its kinase activity, in the TNFR1 complex serves as a platform for hepatocytes associated with increased caspase-3 activity in the recruitment of kinases, such as NEMO(IKKg)andTGF-b–activated liver; almost complete protection by neutralization of TNF or kinase 1, which are required to activate NF-kB and induce pro- inhibition of caspases by zVAD; and sensitization of primary survival and proinflammatory gene expression (3, 5, 29). Previous mouse hepatocytes from RIPK1 ASO–treated mice to TNF- studies showed that NF-kB plays a critical role in maintaining the mediated cell death in the absence of a transcription inhibitor homeostasis of the liver by preventing death receptor–induced The Journal of Immunology 3127

FIGURE 8. (A–D) RIPK1 ASO markedly exacerbated LPS-mediated liver injury characterized by massive hepatic apoptosis. (A) Serum ALT at 6 h after LPS (3 mg/kg i.p.) (n = 5 per group). (B) Histology of liver sections at 6 h after LPS, showing massive cell death and hemorrhage in the liver of RIPK1 ASO–treated mice (original magnification 3400). (C)TUNEL staining of liver sections at 6 h after LPS, displaying massive TUNEL+ cells in the liver of RIPK1 ASO mice (original magnification 3200). (D) Downloaded from Western blotting for RIPK1 in the liver lysates, confirming the efficient knockdown of RIPK1 proteins in RIPK1 ASO mice. (E and F) Knockdown of RIPK1 markedly exacerbated Con A– mediated liver injury, and zVAD pre- http://www.jimmunol.org/ treatment prevented the exacerbation of liver injury. (E)SerumALTat6hafter Con A (10 mg/kg, i.v.) (n = 6–10). For caspase inhibition, zVAD (10 mg/kg) was administered i.p. 15 min prior to Con A treatment. (F) Representative H&E staining of liver sections at 6 h after Con A treatment, showing massive hepatocyte death and hemorrhage in the liver of RIPK1 ASO–treated mice, by guest on September 25, 2021 which is prevented by pretreatment with zVAD (original magnification 3400). *p , 0.05.

apoptosis (34). Interestingly, our results showed that RIPK1 may exist in hepatocytes to activate NF-kB as in IECs. Although knockdown did not impair NF-kB activation in the hepatocytes in RIPK1 knockdown did not impair gene expression of antiapoptotic response to TNF stimulation. Increased sensitivity of mouse embry- protein c-FLIP and cIAP1 in the liver tissues and primary mouse onic fibroblast cells from RIPK12/2 mice to TNF-mediated apoptosis hepatocytes, we found that the induction of hepatic protein levels of was thought to be due to defective NF-kB activation. Thus, the c-FLIP and cIAP1 was decreased after aGalCer treatment in liver perinatal lethality of RIPK12/2 mice was originally believed to be tissue lysates and whole-cell lysates of hepatocytes from RIPK1 caused by defective NF-kB–induced survival genes (5). However, ASO–treated mice. Therefore, degradation of antiapoptotic proteins recent studies argue against the requirement for RIPK1 in the in the absence of RIPK1 may render the hepatocytes more suscep- activation of NF-kB (35). IEC-specific RIPK12/2 mice display no tible to caspase-dependent apoptosis (29–32). impairment of NF-kB activation in IECs, and IEC apoptosis and Our data showed that RIPK32/2 and MLKL2/2 mice behaved intestinal inflammation are not due to defective NF-kB activation similarly as WT control mice and were not sensitized to aGalCer (9). Consistent with these data, we found that TNF-mediated NF- treatment alone. Furthermore, they were not protected from the kB activation is not impaired in the liver of RIPK1 ASO+aGalCer markedly exacerbated liver injury and lethality in response to mice in vivo or in primary hepatocytes from RIPK1 ASO–treated aGalCer treatment when RIPK1 was knocked down. These results mice cultured with TNF in vitro. Therefore, the exacerbation of clearly indicate that a possible RIPK3/MLKL-mediated nec- aGalCer-mediated liver injury and lethality in RIPK1 ASO–treated roptosis in the absence of RIPK1 is not involved in the exacer- mice and the sensitization of hepatocytes to TNF-mediated apoptosis bation of aGalCer-mediated liver injury and lethality in RIPK1 via RIPK1 knockdown are not simply due to an impaired NF-kB ASO–treated mice. In certain cell types, including macrophages, prosurvival pathway, suggesting that RIPK1-independent pathways T cells, and skin keratinocytes, death receptor stimulation can 3128 RIPK1 IN IMMUNE-MEDIATED LIVER INJURY induce necroptosis when there is inhibition of caspase activity or In conclusion, knockdown of RIPK1 markedly exacerbates lack of caspase-8 (1–4, 10, 29, 36). However, our results showed immune-mediated liver injury and induces lethality through massive that the pan-caspase inhibitor zVAD almost completely prevented TNF-a/caspase–dependent apoptosis of hepatocytes, which is not the enhanced liver injury and lethality in vivo and blocked TNF- explained by defective NF-kB activation or a switch to RIPK3/ mediated cell death of hepatocytes from RIPK1 ASO–treated mice MLKL-dependent necroptosis. Therefore, RIPK1 has a unique in vitro. The abrogation of injury by caspase inhibition and lack kinase-independent platform role in protecting against TNF- of necroptotic switch in hepatocytes with caspase inhibition is induced apoptosis of hepatocytes in immune-mediated liver injury. not surprising because hepatocytes do not express RIPK3 under basal conditions (25). Conditional deletion of caspase-8 or cas- Acknowledgments pase inhibition was demonstrated to confer protection against We thank the Analytical/Instrumentation Core, Histology/Imaging Core, acute liver failure due to massive apoptosis of hepatocytes in the and Cell Isolation Core of the University of Southern California Research D (+)-galactosamine/LPS model (37). This is also supported by Center for Liver Diseases (P30DK48522) for the use of various instruments in vitro studies showing no switch from apoptosis to necroptosis to provide histology/imaging, isolation of primary mouse hepatocytes, and when primary mouse hepatocytes were treated with actinomycin flow cytometric analysis. D/TNF and zVAD (38). Together, these results argue against the role of TNF-mediated necroptosis in hepatocytes. Our recent Disclosures studies extensively examined the protein expression of RIPK3, a M.A. is an employee and shareholder of Ionis Pharmaceuticals. Y.S. is a key molecule for MLKL-mediated necroptosis, in the liver (25). shareholder of YSL Bioprocess Development Co. The other authors have

We found that RIPK3 is exclusively expressed on hepatic non- no financial conflicts of interest. Downloaded from parenchymal cells (Kupffer cells, sinusoidal endothelial cells, and leukocytes), with little expression of RIPK3 protein in hepatocytes References under basal conditions and in acute acetaminophen (APAP) hep- 1. Silke, J., J. A. Rickard, and M. Gerlic. 2015. The diverse role of RIP kinases in atotoxicity (25). The low expression of RIPK3 at the mRNA and necroptosis and inflammation. Nat. Immunol. 16: 689–697. protein levels is also supported by other studies (38, 39). In sup- 2. Newton, K. 2015. RIPK1 and RIPK3: critical regulators of inflammation and cell death. Trends Cell Biol. 25: 347–353. port of previous findings, we did not detect RIPK3 protein in 3. Vanden Berghe, T., A. Linkermann, S. Jouan-Lanhouet, H. Walczak, and http://www.jimmunol.org/ primary mouse hepatocytes from Cont ASO– and RIPK1 ASO– P. Vandenabeele. 2014. Regulated necrosis: the expanding network of non- treated mice, with or without TNF stimulation (data not shown). apoptotic cell death pathways. Nat. Rev. Mol. Cell Biol. 15: 135–147. 4. Weinlich, R., and D. R. Green. 2014. The two faces of receptor interacting RIPK3 can be activated independently of RIPK1 to cause nec- protein kinase-1. Mol. Cell 56: 469–480. roptosis, particularly when the cells express relatively high levels 5. Kelliher, M. A., S. Grimm, Y. Ishida, F. Kuo, B. Z. Stanger, and P. Leder. 1998. The death domain kinase RIP mediates the TNF-induced NF-kappaB signal. of RIPK3 (40–42). The levels of RIPK3 expression were dem- Immunity 8: 297–303. onstrated to correlate with the sensitivity of cells to necroptosis (2, 6. Rickard, J. A., J. A. O’Donnell, J. M. Evans, N. Lalaoui, A. R. Poh, T. Rogers, 29, 33, 43, 44). Although a role for necroptosis in acute injury J. E. Vince, K. E. Lawlor, R. L. Ninnis, H. Anderton, et al. 2014. RIPK1 regulates RIPK3-MLKL-driven systemic inflammation and emergency hematopoi- models is unlikely, it is conceivable that RIPK3 expression can be esis. Cell 157: 1175–1188. induced in chronic injury and result in necroptosis of hepatocytes. 7. Dillon, C. P., R. Weinlich, D. A. Rodriguez, J. G. Cripps, G. Quarato, P. Gurung, by guest on September 25, 2021 This may play a role in certain chronic inflammatory liver diseases, K. C. Verbist, T. L. Brewer, F. Llambi, Y. N. Gong, et al. 2014. RIPK1 blocks early postnatal lethality mediated by caspase-8 and RIPK3. Cell 157: 1189–1202. such as alcoholic and nonalcoholic steatohepatitis (11, 12, 45). There 8. Kaiser, W. J., L. P. Daley-Bauer, R. J. Thapa, P. Mandal, S. B. Berger, C. Huang, is still much controversy on the role of necroptosis in the liver (12, A. Sundararajan, H. Guo, L. Roback, S. H. Speck, et al. 2014. RIP1 suppresses 13, 15, 16, 25, 38, 46); however, our data indicate that, in this innate immune necrotic as well as apoptotic cell death during mammalian par- turition. Proc. Natl. Acad. Sci. USA 111: 7753–7758. TNF-dependent immune-mediated acute liver injury model, the 9. Takahashi, N., L. Vereecke, M. J. Bertrand, L. Duprez, S. B. Berger, T. Divert, mode of cell death is caspase-dependent apoptosis, and there is no A. Gonçalves, M. Sze, B. Gilbert, S. Kourula, et al. 2014. RIPK1 ensures intestinal homeostasis by protecting the epithelium against apoptosis. Nature 513: contribution by necroptosis. 95–99. In contrast to the protective role of RIPK1 in the current study, in 10. Dannappel, M., K. Vlantis, S. Kumari, A. Polykratis, C. Kim, L. Wachsmuth, APAP-induced hepatotoxicity, we recently showed that knockdown C. Eftychi, J. Lin, T. Corona, N. Hermance, et al. 2014. RIPK1 maintains epithelial homeostasis by inhibiting apoptosis and necroptosis. Nature 513: 90–94. of RIPK1 protects against necrosis of hepatocytes in vivo and 11. Luedde, T., N. Kaplowitz, and R. F. Schwabe. 2014. Cell death and cell death in vitro (25), indicating that RIPK1 participates in the pathways responses in liver disease: mechanisms and clinical relevance. Gastroenterology leading to necrosis of hepatocytes caused by APAP toxicity. 147: 765–783 and. 12. Roychowdhury, S., M. R. McMullen, S. G. Pisano, X. Liu, and L. E. Nagy. 2013. However, mechanistically, APAP hepatotoxicity is different from Absence of receptor interacting protein kinase 3 prevents ethanol-induced liver aGalCer-mediated liver injury triggered by death receptor sig- injury. Hepatology 57: 1773–1783. naling. The mode of cell death in APAP hepatotoxicity is cell 13. Ramachandran, A., M. R. McGill, Y. Xie, H. M. Ni, W. X. Ding, and H. Jaeschke. 2013. Receptor interacting protein kinase 3 is a critical early autonomous, and APAP is directly toxic to hepatocytes (25). Cell mediator of acetaminophen-induced hepatocyte necrosis in mice. Hepatology 58: death in the APAP model predominantly is regulated necrosis and 2099–2108. 14. Wang, H., L. Sun, L. Su, J. Rizo, L. Liu, L. F. Wang, F. S. Wang, and X. Wang. involves the sustained activation of JNK, leading to impairment of 2014. Mixed lineage kinase domain-like protein MLKL causes necrotic mem- mitochondria and resultant oxidative stress (27, 47), and it is not brane disruption upon phosphorylation by RIP3. Mol. Cell 54: 133–146. TNF-a and caspase dependent (48). Interestingly, we demon- 15. Deutsch, M., C. S. Graffeo, R. Rokosh, M. Pansari, A. Ochi, E. M. Levie, E. Van Heerden, D. M. Tippens, S. Greco, R. Barilla, et al. 2015. Divergent strated that, in both models, RIPK3/MLKL-mediated necroptosis effects of RIP1 or RIP3 blockade in murine models of acute liver injury. Cell does not contribute to hepatocyte death, presumably as the result Death Dis. 6: e1759. of little expression of RIPK3 proteins in hepatocytes. The appar- 16. Jouan-Lanhouet, S., M. I. Arshad, C. Piquet-Pellorce, C. Martin-Chouly, G. Le Moigne-Muller, F. Van Herreweghe, N. Takahashi, O. Sergent, ent opposing roles of RIPK1 in APAP- and aGalCer-induced liver D. Lagadic-Gossmann, P. Vandenabeele, et al. 2012. TRAIL induces nec- injury are very interesting, which confirms the critical role of roptosis involving RIPK1/RIPK3-dependent PARP-1 activation. Cell Death Differ. 19: 2003–2014. RIPK1 as a signaling node and a central regulator of cell death and 17. Roderick, J. E., N. Hermance, M. Zelic, M. J. Simmons, A. Polykratis, survival pathways (1–4). It is clear that the function of RIPK1 is M. Pasparakis, and M. A. Kelliher. 2014. Hematopoietic RIPK1 deficiency re- complex in many cell types, including hepatocytes. Detailed sults in bone marrow failure caused by apoptosis and RIPK3-mediated necroptosis. Proc. Natl. Acad. Sci. USA 111: 14436–14441. studies are needed to help understand the various roles of RIPK1 18. Bendelac, A., P. B. Savage, and L. Teyton. 2007. The biology of NKT cells. in different types of liver diseases and experimental conditions. Annu. Rev. Immunol. 25: 297–336. The Journal of Immunology 3129

19. Gao, B., S. Radaeva, and O. Park. 2009. Liver natural killer and natural killer Activity of protein kinase RIPK3 determines whether cells die by necroptosis or T cells: immunobiology and emerging roles in liver diseases. J. Leukoc. Biol. 86: apoptosis. Science 343: 1357–1360. 513–528. 34. Luedde, T., N. Beraza, V. Kotsikoris, G. van Loo, A. Nenci, R. De Vos, 20. Minagawa, M., Q. Deng, Z. X. Liu, H. Tsukamoto, and G. Dennert. 2004. Ac- T. Roskams, C. Trautwein, and M. Pasparakis. 2007. Deletion of NEMO/ tivated natural killer T cells induce liver injury by Fas and tumor necrosis factor- IKKgamma in liver parenchymal cells causes steatohepatitis and hepatocellu- alpha during alcohol consumption. Gastroenterology 126: 1387–1399. lar carcinoma. Cancer Cell 11: 119–132. 21. Syn, W. K., Y. H. Oo, T. A. Pereira, G. F. Karaca, Y. Jung, A. Omenetti, 35. Wong, W. W., I. E. Gentle, U. Nachbur, H. Anderton, D. L. Vaux, and J. Silke. R. P. Witek, S. S. Choi, C. D. Guy, C. M. Fearing, et al. 2010. Accumulation of 2010. RIPK1 is not essential for TNFR1-induced activation of NF-kappaB. Cell natural killer T cells in progressive nonalcoholic fatty liver disease. Hepatology Death Differ. 17: 482–487. 51: 1998–2007. 36. He, S., Y. Liang, F. Shao, and X. Wang. 2011. Toll-like receptors activate pro- 22. Takeda, K., Y. Hayakawa, L. Van Kaer, H. Matsuda, H. Yagita, and K. Okumura. grammed necrosis in macrophages through a receptor-interacting kinase-3- 2000. Critical contribution of liver natural killer T cells to a murine model of mediated pathway. Proc. Natl. Acad. Sci. USA 108: 20054–20059. hepatitis. Proc. Natl. Acad. Sci. USA 97: 5498–5503. 37. Kaufmann, T., P. J. Jost, M. Pellegrini, H. Puthalakath, R. Gugasyan, 23. Osman, Y., T. Kawamura, T. Naito, K. Takeda, L. Van Kaer, K. Okumura, and S. Gerondakis, E. Cretney, M. J. Smyth, J. Silke, R. Hakem, et al. 2009. Fatal T. Abo. 2000. Activation of hepatic NKT cells and subsequent liver injury fol- hepatitis mediated by tumor necrosis factor TNFalpha requires caspase-8 and lowing administration of alpha-galactosylceramide. Eur. J. Immunol. 30: 1919– involves the BH3-only proteins Bid and Bim. Immunity 30: 56–66. 1928. 38. Kang, Y. J., B. R. Bang, K. H. Han, L. Hong, E. J. Shim, J. Ma, R. A. Lerner, and 24. Biburger, M., and G. Tiegs. 2005. Alpha-galactosylceramide-induced liver injury M. Otsuka. 2015. Regulation of NKT cell-mediated immune responses to tu- in mice is mediated by TNF-alpha but independent of Kupffer cells. J. Immunol. mours and liver inﬂammation by mitochondrial PGAM5-Drp1 signalling. Nat. 175: 1540–1550. Commun. 6: 8371. 25. Dara, L., H. Johnson, J. Suda, S. Win, W. Gaarde, D. Han, and N. Kaplowitz. 39. Sun, X., J. Lee, T. Navas, D. T. Baldwin, T. A. Stewart, and V. M. Dixit. 1999. 2015. Receptor interacting protein kinase 1 mediates murine acetaminophen RIP3, a novel apoptosis-inducing kinase. J. Biol. Chem. 274: 16871–16875. toxicity independent of the necrosome and not through necroptosis. Hepatology 40. Moujalled, D. M., W. D. Cook, T. Okamoto, J. Murphy, K. E. Lawlor, 62: 1847–1857. J. E. Vince, and D. L. Vaux. 2013. TNF can activate RIPK3 and cause pro- 26. Liu, Z. X., D. Han, B. Gunawan, and N. Kaplowitz. 2006. Neutrophil depletion grammed necrosis in the absence of RIPK1. Cell Death Dis. 4: e465. protects against murine acetaminophen hepatotoxicity. Hepatology 43: 1220– 41. Kaiser, W. J., H. Sridharan, C. Huang, P. Mandal, J. W. Upton, P. J. Gough, Downloaded from 1230. C. A. Sehon, R. W. Marquis, J. Bertin, and E. S. Mocarski. 2013. Toll-like re- 27. Win, S., T. A. Than, D. Han, L. M. Petrovic, and N. Kaplowitz. 2011. c-Jun ceptor 3-mediated necrosis via TRIF, RIP3, and MLKL. J. Biol. Chem. 288: N-terminal kinase (JNK)-dependent acute liver injury from acetaminophen or 31268–31279. tumor necrosis factor (TNF) requires mitochondrial Sab protein expression in 42. Kearney, C. J., S. P. Cullen, D. Clancy, and S. J. Martin. 2014. RIPK1 can mice. J. Biol. Chem. 286: 35071–35078. function as an inhibitor rather than an initiator of RIPK3-dependent necroptosis. 28. Ogasawara, J., R. Watanabe-Fukunaga, M. Adachi, A. Matsuzawa, T. Kasugai, FEBS J. 281: 4921–4934. Y. Kitamura, N. Itoh, T. Suda, and S. Nagata. 1993. Lethal effect of the anti-Fas 43. Oberst, A., C. P. Dillon, R. Weinlich, L. L. McCormick, P. Fitzgerald, C. Pop,

antibody in mice. Nature 364: 806–809. R. Hakem, G. S. Salvesen, and D. R. Green. 2011. Catalytic activity of the http://www.jimmunol.org/ 29. Chan, F. K., N. F. Luz, and K. Moriwaki. 2015. Programmed necrosis in the caspase-8-FLIP(L) complex inhibits RIPK3-dependent necrosis. Nature 471: cross talk of cell death and inﬂammation. Annu. Rev. Immunol. 33: 79–106. 363–367. 30. Gentle, I. E., W. W. Wong, J. M. Evans, A. Bankovacki, W. D. Cook, N. R. Khan, 44. He, S., L. Wang, L. Miao, T. Wang, F. Du, L. Zhao, and X. Wang. 2009. Re- U. Nachbur, J. Rickard, H. Anderton, M. Moulin, et al. 2011. In TNF-stimulated ceptor interacting protein kinase-3 determines cellular necrotic response to TNF- cells, RIPK1 promotes cell survival by stabilizing TRAF2 and cIAP1, which alpha. Cell 137: 1100–1111. limits induction of non-canonical NF-kappaB and activation of caspase-8. 45. Gautheron, J., M. Vucur, F. Reisinger, D. V. Cardenas, C. Roderburg, C. Koppe, [Published erratum appears in 1994 J. Biol. Chem. 291: 2547.] J. Biol. Chem. K. Kreggenwinkel, A. T. Schneider, M. Bartneck, U. P. Neumann, et al. 2014. A 286: 13282–13291. positive feedback loop between RIP3 and JNK controls non-alcoholic steato- 31. Schattenberg, J. M., T. Zimmermann, M. Wo¨rns, M. F. Sprinzl, A. Kreft, T. Kohl, hepatitis. EMBO Mol. Med. 6: 1062–1074. M. Nagel, J. Siebler, H. Schulze Bergkamen, Y. W. He, et al. 2011. Ablation of 46. Yang, X., X. Chao, Z. T. Wang, and W. X. Ding. 2016. The end of RIPK1- c-FLIP in hepatocytes enhances death-receptor mediated apoptosis and toxic RIPK3-MLKL-mediated necroptosis in acetaminophen-induced hepatotoxicity?

liver injury in vivo. J. Hepatol. 55: 1272–1280. Hepatology 64: 311–312. by guest on September 25, 2021 32. Piao, X., S. Komazawa-Sakon, T. Nishina, M. Koike, J. H. Piao, H. Ehlken, 47. Han, D., L. Dara, S. Win, T. A. Than, L. Yuan, S. Q. Abbasi, Z. X. Liu, and H. Kurihara, M. Hara, N. Van Rooijen, G. Schutz,€ et al. 2012. c-FLIP maintains N. Kaplowitz. 2013. Regulation of drug-induced liver injury by signal trans- tissue homeostasis by preventing apoptosis and programmed necrosis. Sci. duction pathways: critical role of mitochondria. Trends Pharmacol. Sci. 34: 243– Signal. 5: ra93. 253. 33. Newton, K., D. L. Dugger, K. E. Wickliffe, N. Kapoor, M. C. de Almagro, 48. Jaeschke, H., C. D. Williams, and A. Farhood. 2011. No evidence for caspase- D. Vucic, L. Komuves, R. E. Ferrando, D. M. French, J. Webster, et al. 2014. dependent apoptosis in acetaminophen hepatotoxicity. Hepatology 53: 718–719. Suppl. Figure 1

Suppl. Figure 1. Knockdown of RIPK1signicantly increases the serum levels of cytokines/chemokines (A) and hepatic leukocytes (B) in response to GalCer treatment. Groups of mice were pretreated with Cont ASO or RIPK1 ASO to silence RIPK1 prior to GalCer (4µg/mouse, i.p.). (A) Serum cytokine/chemokine levels at 6hr after GalCer (n=4‐6). *p<0.05 vs. Cont ASO; (B) Total numbers of hepatic leukocytes per liver tissue at 6hr after GalCer. *p<0.05 vs. Cont ASO. **Reduced NKT numbers in both groups reflected an activation‐induced cell death of NKT cells in response to GalCer as expected. Suppl. Figure 2

Suppl. Figure 2. RIPK1 knockdown does not sensitize to Fas‐mediated liver injury. Groups of Cont ASO and RIPK1 ASO pretreated mice (n=5 per group) were administered i.p with anti‐Fas (Jo2, 4μg per mouse). (A) Serum ALT at 6 and 24hr after Jo2; (B) H&E staining of liver sections at 24hr after Jo2. Suppl. Figure 3

Suppl. Figure 3. Knockdown of RIPK1 does not impair the induction of NF‐B downstream genes. (A) qPCR analysis of mRNA levels of c‐FLIP and cIAP1 in the liver from Cont ASO and RIPK1 ASO treated mice at indicated times after GalCer treatment. (B) mRNA levels in the liver tissues from Cont ASO and RIPK1 ASO treated mice at 3hr after GalCer treatment were analyzed by mouse NF‐κB Signaling Pathway RT2 ProfilerTM PCR array kit (QIAGEN). Fold changes at 3hr over time zero were expressed. (C)mRNA levels in the primary hepatocytes isolated from Cont ASO and RIPK1 ASO treated mice at 3hr after GalCer treatment were analyzed by mouse NF‐κB Signaling Pathway RT2 ProfilerTM PCR array kit (QIAGEN). Fold changes at 3hr over time zero were expressed. Suppl. Figure 4

Suppl. Figure 4. Necostatin‐1 (Nec‐1) does not prevent αGalCer‐mediated liver injury. Groups of mice were pretreated with inactive Nec‐1 or Nec‐1 (1.65 mg/kg, i.p.) 15min prior to αGalCer (4μg per mouse, i.p.). (A)Serum ALT at 24hr after αGalCer (n=5 per group), p>0.05; (B) H&E staining of liver sections at 24hr after αGalCer.