Phosphorylation-driven assembly of the RIP1-RIP3 complex regulates programmed necrosis and virus-induced inflammation

Abstract

Programmed necrosis is a form of caspase-independent cell death whose molecular regulation is poorly understood. The kinase RIP1 is crucial for programmed necrosis, but also mediates activation of the prosurvival transcription factor NF-kappaB. We postulated that additional molecules are required to specifically activate programmed necrosis. Using a RNA interference screen, we identified the kinase RIP3 as a crucial activator for programmed necrosis induced by TNF and during virus infection. RIP3 regulates necrosis-specific RIP1 phosphorylation. The phosphorylation of RIP1 and RIP3 stabilizes their association within the pronecrotic complex, activates the pronecrotic kinase activity, and triggers downstream reactive oxygen species production. The pronecrotic RIP1-RIP3 complex is induced during vaccinia virus infection. Consequently, RIP3(-/-) mice exhibited severely impaired virus-induced tissue necrosis, inflammation, and control of viral replication. Our findings suggest that RIP3 controls programmed necrosis by initiating the pronecrotic kinase cascade, and that this is necessary for the inflammatory response against virus infections.

Figure 1. RIP3 is specifically required for programmed necrosis, but not apoptosis or NF-κB induction

Wild type TNFR2⁺ Jurkat cells were transfected with siRNAs against RIP1, RIP3 or control siRNA against TRAIL-R4 (TR4). (A) Western blot of RIP1, RIP3 and β-actin expression in cells transfected with the indicated siRNAs. Cells were tested for (B) TNF/zVAD-fmk-induced programmed necrosis (* p = 0.005, ** p < 0.001), (C) TNF-induced apoptosis, or (D) FasL-induced apoptosis 48 hours post-transfection. Data are presented as mean ± SEM of triplicates. (E) Wild type and RIP3^−/− MEFs were treated with TNF for the indicated times and IκBα degradation was analyzed by Western blot.

Figure 2. Requirement of kinase and RHIM domains of RIP3 for programmed necrosis

(A) Domain structures of RIP1 and RIP3. (B-C) The RIP3 kinase and RHIM domains are required for programmed necrosis. Wild type TNFR2⁺ Jurkat cells were transfected with siRNA against human RIP3. After 24 hours, cells were transfected with the indicated GFP-tagged plasmids: (B) kinase-dead (KD, D161N) mouse RIP3, (C) RIP3 RHIM mutant. TNF/zVAD-induced programmed necrosis was determined in the transfected GFP⁺ population. (D) The kinase and RHIM domains of RIP1 are required for programmed necrosis. RIP1-deficient Jurkat cells were transfected with the indicated GFP-tagged RIP1 plasmids. TNF/zVAD-fmk induced programmed necrosis was determined as in (B-C). Data are presented as mean ± SEM of triplicates. (E) RIP3 is recruited to Complex II. Complex II was isolated using caspase-8 specific antibody from wild type TNFR2⁺ Jurkat cells stimulated with TNF ± zVAD-fmk and the recruitment of RIP1 and RIP3 was determined by Western blot. WCE: whole cell extract. C8: caspase-8. (F) RIP3 is not recruited to the TNFR-1 signaling complex. TNFR-1 immune complexes from TNFR2⁺ Jurkat cells stimulated with TNF ± zVAD-fmk for the indicated times were examined for the recruitment of RIP1 and RIP3 by Western blot. RIP1-Ub: polyubiquitinated RIP1.

Figure 3. The RIP1-dependent pro-necrotic kinase activity was induced in Complex II

(A) TNFR2⁺ Jurkat cells were treated with TNF ± zVAD-fmk for the indicated times. Complex II was isolated by FADD immunoprecipitations and subjected to in vitro kinase assays with [³²P]-γATP. The fold induction of kinase activity was quantified by densitometry of the ∼60 kDa band indicated by the arrow. Lower panels: Western blot analyses of the isolated FADD complexes. (B) Induction of Complex II kinase activity can be measured using MBP as artificial substrate. FADD complexes from caspase-8-deficient Jurkat cells were subjected to in vitro kinase assay using [³²P] γ-ATP and MBP as substrate. The arrows indicate the phosphorylated MBP and Complex II components. Lower panels show the Western blot analyses of the isolated Complex II. (C) Complex I did not exhibit TNF-dependent induction of kinase activity. Top panel: IP kinase assays were performed with the isolated TNFR-1 complexes in the presence of [³²P]-γATP. Lower panels: Western blot analyses of Complex I. (D) Complex II kinase activity was not detected in cells that do not undergo programmed necrosis. TNFR2⁻ Jurkat cells were treated with TNF for the indicated times. The isolated FADD complexes were subjected to in vitro kinase assays as in (A). (E) RIP1 is required for induction of Complex II kinase activity. RIP1-deficient Jurkat cells were treated with TNF and the isolated FADD complexes were tested in in vitro kinase assays as in (A).

Figure 4. Phosphorylation of RIP1 and RIP3 critically regulates formation of the pro-necrotic RIP1-RIP3 complex

(A) RIP3 is essential for RIP1 recruitment to FADD-associated Complex II. FADD was immunoprecipitated from RIP3^+/+ and RIP3^−/− MEFs after treatment with TNF, cycloheximide and zVAD-fmk for the indicated times. In lanes 4 and 8, cells were also treated with Nec-1. (B) RIP3 is required for necrosis-specific RIP1 phosphorylation. RIP3^+/+ (lanes 1-4) and RIP3^−/−(lanes 5-6) MEFs were labeled with [³²P]-orthophosphate and treated with TNF alone or TNF in the presence of cycloheximide and zVAD-fmk (C/Z) for 6 hours. RIP1 was immunoprecipitated and assessed for phosphorylation by autoradiography. Bottom panels show that equal amounts of RIP1 were present. (C) Necrosis-dependent RIP1 phosphorylation was not mediated by RIP1 autophosphorylation. FADD-deficient Jurkat cells were labeled with [³²P]-orthophosphate and treated with TNF for 2 hours. In lane 3, cells were pre-treated with Nec-1. The bottom panels show that equal amounts of RIP1 were pulled down. (D) RIP1 is a kinase substrate for RIP3. The indicated GFP-tagged wild type (WT) or kinase defective (KD) RIP1 or RIP3 were expressed in 293T cells, immunoprecipiated with anti-GFP antibody and subjected to in vitro kinase assay with [³²P] γ-ATP. MBP was also included as artificial substrate. Top panel: autoradiograph. Bottom panel: Western blot analysis with anti-GFP antibody. IgG_L: immunoglobulin light chain. The fold-increase in RIP1 phosphorylation was determined by densitometry and by normalizing the amount of RIP1 phosphorylation to the expression level of RIP1 and/or RIP3. For comparison, the background signal in lane 8 was set as 1. (E) Specific association of RIP1 and RIP3 during programmed necrosis. Wild type MEFs were treated with TNF in the presence of cycloheximide (T/C) to induce apoptosis or cycloheximide and zVAD-fmk (T/C/Z) to induce programmed necrosis. Recruitment of RIP3 to RIP1 was examined by Western blot. (F-G) The kinase activity of RIP1 is required for formation of the pro-necrotic RIP1-RIP3 complex. (F) Wild type MEFs were treated as in (E) except that in lanes 4 and 8, cells were pre-treated with of Nec-1. (G) Wild type MEFs were treated with T/C/Z for 6 hours. In lane 3, cells were pre-treated with Nec-1. RIP3 was immunoprecipitated and the association with RIP1 was examined by Western blot. (H) RIP1 kinase activity is required for necrosis-specific RIP3 phosphorylation. Wild type MEFs were labeled with [³²P]-orthophosphate and RIP3 was immunoprecipitated with specific antibody after 6 hours of treatment with TNF or T/C/Z. Top panel: autoradiograph. Bottom panel: RIP3 Western blot.

Figure 5. RIP3 regulates downstream ROS production

(A-C) RIP3 is required for downstream ROS production during programmed necrosis. (A) RIP3^+/+ and RIP3^−/− MEFs were treated with TNF and cycloheximide (T/C) or T/C and zVAD-fmk (T/C/Z) for 6 hours. ROS production was quantified using the fluorescent dye H₂DCFDA on flow cytometry. Data are presented as mean ± SEM of triplicates. (B) Cell death was quantified using propidium iodide uptake. Data are presented as mean ± SEM of triplicates. (C) Flow cytometric analysis of RIP3-dependent ROS production during programmed necrosis. The numbers represent the percentages of cells in each quandrant. (D-E) RIP3 is not required for oxidative stress-induced cell injury. (D) L929 cells transfected with the indicated siRNAs were treated with 0.2 mM H₂O₂. Cell death was assessed by MTS assay 6 hours later. Data are presented as mean ± SEM of triplicates. (E) Dose-dependent response of L929 cells transfected with the RIP3 siRNA to H₂O₂-induced cell death. Cell death was measured by MTS assay 6 hours after stimulation. Data are presented as mean ± SEM of triplicates. (F-G) RIP3 is required for zVAD-fmk-induced, autocrine TNF-dependent necrosis. (F) L929 cells transfected with the indicated siRNAs were treated with 50 μM of zVAD-fmk and cell death was determined by MTS assay 24 hours later. The inset shows the reduction of RIP1 or RIP3 protein expression by the respective siRNAs. Data are presented as mean ± SEM of triplicates. * p = 0.01, ** p = 0.002, *** p < 0.001. (G) Dose-dependent response to zVAD-fmk induced necrosis of L929 cells transfected with the RIP3 siRNA. Cell death was measured by MTS assay 24 hours later. Data are presented as mean ± SEM of triplicates. * p = 0.027, ** p = 0.018, *** p = 0.001.

Figure 6. RIP3 regulates cell death during virus infections

(A-B) RIP3 is essential for AICD and TNF-induced death in VV-infected T-cells. Wild type (white bars) or RIP3^−/− T-cells (black bars) were infected with GFP-VV (moi = 10) and stimulated with (A) plate-bound anti-CD3 antibody or (B) mTNF. Cell death in the GFP⁺ infected population and the GFP⁻ uninfected population was determined by PI staining and flow cytometry. Data are presented as mean ± SEM of triplicates. (C) RIP3 is required for T-cell AICD when caspases are inhibited. Wild type or RIP3^−/− T-cells were stimulated with different doses of plate-bound anti-CD3 antibody in the presence or absence of zVAD-fmk. Cell death was assessed by PI uptake on flow cytometry. Data are presented as mean ± SEM of triplicates. (D) VV infection sensitizes MEFs to TNF-induced programmed necrosis in a RIP3-dependent manner. RIP3^+/+ and RIP3^−/− MEFs were infected with GFP-VV (moi = 0.5). Infected cells were treated with TNF and cell death was determined by PI uptake in the infected GFP⁺ populations by flow cytometry. Data are presented as mean ± SEM of triplicates. (E) Electron microscopy shows that VV-infected MEFs underwent necrosis in response to TNF. Panel a: uninfected cell. Panel b: VV-infected cell (no TNF). The electron-dense structures indicated by the arrows are viral particles. Panel c: VV-infected cell treated with TNF. Scale bar: 2 μm. (F) Formation of pro-necrotic RIP1-RIP3 complex in VV-infected MEFs. Uninfected MEFs (lanes 1-4) or VV-infected MEFs (lanes 5-8) were treated with TNF ± zVAD-fmk as indicated. RIP1 was immunoprecipitated and the presence of RIP3 in the immune complex was determined by Western blot.

Figure 7. RIP3-dependent programmed necrosis is required for control of vaccinia virus replication in vivo

(A) Induction of TNF expression in PECs upon VV infection. PECs from uninfected or VV-infected (24 hours post-infection) wild type mice were evaluated for TNF expression by intracellular cytokine staining and flow cytometry. (B) Quantitative PCR analyses of TNF induction upon VV infection. Total RNA was isolated from the indicated tissues 24 hours post-infection. Fold induction of TNF mRNA was determined by comparison with uninfected controls and normalizing the TNF message to that of the 18S RNA. Data are presented as mean ± SEM of triplicates. (C) Visceral fat pads from VV-infected RIP3^+/+ mice (panels a-b), but not RIP3^−/− mice (panels c-d), exhibited massive fatty tissue necrosis and inflammation marked by macrophage and neutrophil infiltration. The red arrows denote the necrotic tissue mass and the inflammatory foci. Panels a, c: 4X objective, Panels b, d: 20X objective of the boxed area in panels a and c. Images shown are representative from 8 RIP3^+/+ and 10 RIP3^−/− mice. (D) H&E staining of liver sections from RIP3^+/+ (panels a-b), TNFR2^−/− (panel c) and RIP3^−/− (panel d) mice. In panels a-b, leukocyte infiltration and necrotic cell masses (arrows) were clearly seen. (E) Formation of the pro-necrotic RIP1-RIP3 complex in VV-infected liver. Liver extracts from uninfected and VV-infected wild type mice were harvested at different times post-infection and subjected to immunoprecipitation with RIP1 antibody. The association of RIP3 was determined by Western blot. (F) Viral titers in the visceral fat pads, livers and spleens of RIP3^+/+ and RIP3^−/−mice were determined on day 3 post-infection. (G) Kaplan-Meier survival plot of VV-infected RIP3^+/+ and RIP3^−/− mice.