SIRT2 regulates oxidative stress-induced cell death through deacetylation of c-Jun NH2-terminal kinase

Abstract

c-Jun NH₂-terminal kinases (JNKs) are responsive to stress stimuli and their activation regulate key cellular functions, including cell survival, growth, differentiation and aging. Previous studies demonstrate that activation of JNK requires dual phosphorylation by the mitogen-activated protein kinase kinases. However, other post-translational mechanisms involved in regulating the activity of JNK have been poorly understood. In this work, we studied the functional significance of reversible lysine acetylation in regulating the kinase activity of JNK. We found that the acetyl transferase p300 binds to, acetylates and inhibits kinase activity of JNK. Using tandem mass spectrometry, molecular modelling and molecular dynamics simulations, we found that acetylation of JNK at Lys153 would hinder the stable interactions of the negatively charged phosphates and prevent the adenosine binding to JNK. Our screening for the deacetylases found SIRT2 as a deacetylase for JNK. Mechanistically, SIRT2-dependent deacetylation enhances ATP binding and enzymatic activity of JNK towards c-Jun. Furthermore, SIRT2-mediated deacetylation favours the phosphorylation of JNK by MKK4, an upstream kinase. Our results indicate that deacetylation of JNK by SIRT2 promotes oxidative stress-induced cell death. Conversely, SIRT2 inhibition attenuates H₂O₂-mediated cell death in HeLa cells. SIRT2-deficient (SIRT2-KO) mice exhibit increased acetylation of JNK, which is associated with markedly reduced catalytic activity of JNK in the liver. Interestingly, SIRT2-KO mice were resistant to acetaminophen-induced liver toxicity. SIRT2-KO mice show lower cell death, minimal degenerative changes, improved liver function and survival following acetaminophen treatment. Overall, our work identifies SIRT2-mediated deacetylation of JNK as a critical regulator of cell survival during oxidative stress.

Fig. 1

p300 acetyltransferase play key role in regulating JNK acetylation and activity. a Western blotting analysis showing the acetylation of endogenous JNK in HEK 293T cells. Endogenous JNK was immunoprecipitated and acetylation was detected by western blotting using pan-acetyl lysine antibody. Whole-cell lysates (WCLs) of 293T cells were probed for JNK and actin. n = 3 independent experiments. b Western blotting analysis depicting the acetylation status of JNK in 293T cells. Endogenous acetylated proteins were immunoprecipitated with anti-pan acetyl lysine antibody and probed for JNK by western blotting. WCLs of 293T cells were probed for JNK and actin. n = 3 independent experiments. c Western blot analysis showing the acetylation, phosphorylation status and activity of JNK in cells treated with Class I and II HDAC inhibitor, Trichostatin A (TSA) or Class III HDAC inhibitor, nicotinamide (NAM). 293T cells were treated with vehicle or 10 µM TSA for 12 h. 293T cells were treated with vehicle or 50 mM NAM for 6 h. Acetylation was assessed by immunoprecipitating endogenous JNK with a specific JNK antibody. The WCL was tested for the phosphorylation of JNK by western blotting. The activity of endogenous JNK was tested by measuring the phosphorylation of c-Jun, a downstream transcription factor of JNK. n = 4 independent experiments. d Graph showing the relative mRNA expression levels of acetyl transferases, CBP, GCN5 and TIP60 in HeLa cells treated with either scramble (control) or specific pool of shRNAs targeting individual acetyl transferases, as measured by real-time qPCR analysis. n = 3–4 samples per group. Data are presented as mean ± s.d. *p < 0.05. Student’s t-test was used to calculate the p-values. e Western blotting analysis showing the phosphorylation status and activity of JNK in HeLa cells treated with scramble or specific pool of shRNAs targeting individual acetyl transferases. The WCL was used to check the phosphorylation of JNK by western blotting. Depletion of p300 or PCAF was tested by assessing their protein levels by western blotting with specific antibodies. The depletion of acetyl transferases, CBP, GCN5 and Tip60 was assessed by measuring the mRNA levels, as shown in Fig. 1d, due to lack of specific antibodies. f Western blotting analysis showing in vitro acetylation of JNK by p300. Purified recombinant JNK was used as a substrate for acetyltransferases p300 and PCAF in the presence or absence of acetyl CoA. The acetylation of JNK was assessed by western blotting with pan-acetyl lysine antibody. Recombinant p53 was used as a positive control for acetylation assay. Recombinant p300 strongly acetylates JNK, when compared to PCAF, as assessed by western blotting. g Endogenous JNK was immunoprecipitated and probed for its interaction with p300 by western blotting. IgG was used as negative control. WCLs were probed for the presence of JNK and p300 by western blotting. h Western blotting analysis showing acetylation, phosphorylation and activity of JNK in HeLa cells expressing luciferase shRNA (control) or p300 shRNA (p300-KD). WCLs were probed for the depletion of p300 by western blotting. JNK was immunoprecipitated and probed for Ac-Lys antibody. JNK activity was measured by assessing the phosphorylation of c-Jun. Specific antibody was used to detect the phosphorylation of JNK. i Western blotting analysis showing acetylation and phosphorylation of JNK in HeLa cells overexpressing HA-p300. JNK was immunoprecipitated from HeLa cells transiently overexpressing p300 were analysed for the acetylation and phosphorylation levels of JNK by western blotting. WCLs were probed for the overexpression of p300

Fig. 2

SIRT2 deacetylase regulates the acetylation of JNK. a Western blotting analysis showing acetylation and phosphorylation of JNK in HeLa cells overexpressing the Sirtuin isoforms, SIRT1–SIRT7. Cells were transiently overexpressed with either Flag- or Flag- SIRT1-7, and phosphorylation of JNK was analysed by western blotting. JNK was immunoprecipitated from these overexpressed lysates and tested for its acetylation by western blotting. Whole-cell lysates (WCLs) were probed with Flag-antibody for detecting the expression of Sirtuins. b Western blotting analysis showing the phosphorylation, acetylation and activity of JNK in HeLa cells treated with vehicle or SIRT2 inhibitor, 10 µM AGK2 for 6 h. The activity of AGK2 was tested by probing the acetylation status of tubulin with specific antibody detecting the acetyl-tubulin. JNK was immunoprecipitated from the WCLs and probed for Ac-Lys antibody. The phosphorylation of JNK was tested by western blotting. The activity of JNK was tested by detecting the phosphorylation of c-Jun with specific antibody. c Endogenous JNK or SIRT2 was immunoprecipitated and probed for its interaction with SIRT2 or JNK respectively, by western blotting. IgG was used as negative control. WCLs were probed for the presence of JNK and SIRT2 by western blotting. Actin was used as a loading control. d Co-localization of JNK (green) with SIRT2 (red), as determined by confocal microscopy. Endogenous JNK and SIRT2 were stained using specific antibodies. The merged image shows yellow colour, indicating the co-localization of JNK and SIRT2, both in the cytoplasm and the nucleus. Scale bar = 10 µM. e Western blotting analysis showing acetylation, phosphorylation and activity of JNK in cells expressing scramble shRNA (control) or SIRT2 shRNA (SIRT2-KD). WCLs were probed for the depletion of SIRT2 by western blotting. JNK was immunoprecipitated and probed with Ac-Lys antibody to detect the acetylation. JNK activity was measured by detecting the phosphorylation of c-Jun by western blotting. Phosphorylation of JNK was detected with the use of a specific antibody. f Western blotting images depicting acetylation, phosphorylation and activity of JNK in HeLa cells overexpressing either control vector or SIRT2. WCLs were probed for the overexpression of SIRT2 by western blotting. JNK was immunoprecipitated and probed with Ac-Lys antibody to detect the acetylation. JNK activity was measured by detecting the phosphorylation of c-Jun by western blotting. The phosphorylation of JNK was detected by a specific antibody. g Representative western blotting images showing in vitro kinase assay for recombinant acetylated and deacetylated JNK1. Endogenously acetylated JNK1 was incubated with SIRT2 in the presence and absence of NAD⁺ in a test tube for deacetylation. Then the enzymatic activity of acetylated and deacetylated JNK1 against the c-Jun-fusion protein, which contains the JNK-specific phosphorylation site. n = 3 independent experiments. h Representative western blotting images depicting in vitro kinase assay for endogenous JNK immunoprecipitated from HeLa cells overexpressing either control vector, Flag, SIRT2 or SIRT2-H187Y. Immunoprecipitated JNK was incubated with c-Jun fusion protein to assess the enzymatic activity and the phosphorylation of c-Jun was analysed by western blotting. WCLs were probed for the overexpression of SIRT2 and endogenous JNK levels by western blotting. i Representative western blotting analysis showing acetylation and phosphorylation of JNK immunoprecipitated from wild-type (WT) and SIRT2-deficient (SIRT2-KO) mice. Liver tissue lysate of WT and SIRT2-KO mice was analysed for acetylation and phosphorylation of JNK by western blotting. n = 5 mice per group. j Representative western blotting images showing phosphorylation and activity of JNK in liver lysates of WT and SIRT2-KO mice. Tissue lysates of WT and SIRT2-KO mice was analysed for phosphorylation of JNK and SIRT2 levels. JNK activity was measured by detecting the phosphorylation of c-Jun. Actin was used as a loading control. n = 8 mice per group. k Graph depicting the relative mRNA expression levels of AP-1 target genes, p53, p16 and FasL in WT and SIRT2-KO mice, as measured by real-time qPCR analysis. n = 3–4 mice per group. Data are presented as mean ± s.d. *p < 0.05. Student’s t-test was used to calculate the p-values

Fig. 3

Deacetylation favours MKK4-dependent activation of JNK. a Graph showing the relative mRNA expression levels of JNK isoforms in wild-type (WT) and SIRT2-deficient (SIRT2-KO) mice, as measured by real-time qPCR analysis. n = 3–4 mice per group. Data are presented as mean ± s.d. *p < 0.05. Student’s t-test was used to calculate the p-values. b Graph showing the relative mRNA expression levels of JNK upstream kinases and phosphates in WT and SIRT2-KO mice, as measured by real-time qPCR analysis. n = 3–4 mice per group. Data are presented as mean ± s.d. *p < 0.05. Student’s t-test was used to calculate the p-values. c Western blotting images showing the expression of JNK upstream phosphates in HeLa cells overexpressing either control (Ad-Null) or SIRT2 (Ad-SIRT2). Similarly, the phosphorylation of MKK4, an upstream kinase of JNK, was measured by western blotting. n = 3 independent experiments. d Representative western blotting images showing MKK4 interaction with JNK. MKK4 was immunoprecipitated from cells transiently expressing SIRT2 or SIRT2-N168A mutant and probed for JNK by western blotting. Whole-cell lysates (WCLs) were probed for the overexpression of SIRT2 and endogenous JNK levels by western blotting. n = 3 independent experiments. e Representative western blotting images showing the effect of acetylation on the MKK4-mediated phosphorylation of JNK. Endogenous acetylated JNK was immunoprecipitated from HeLa cells treated with deacetylase inhibitors. The acetyl-JNK was further incubated with SIRT2 in the presence or absence of NAD⁺ in a test tube and then subjected to a kinase assay, where deacetylated JNK was used as substrate for MKK4. Phosphorylation and acetylation of JNK were assessed using western blotting. n = 3 independent experiments. f Representative western blotting images showing the effect of SIRT2-mediated deacetylation on the MKK4-dependent phosphorylation of JNK. MKK4 was overexpressed in HeLa cells transiently expressing either SIRT2 or SIRT2-H187Y. WCLs were probed for the overexpression of SIRT2, SIRT2-H187Y and MKK4. Endogenous JNK levels and its phosphorylation was assessed by western blotting. Lower panel shows the acetylation status of JNK assessed by western blotting

Fig. 4

Modelling and molecular dynamic simulations of JNK. a Representation of the acetylation site on the crystal structure of JNK (PDB ID 4QTD) (I) surface, (II) cartoon and (III) magnified active site representing position of K153 (IV) magnified active site representing position of acetylated K153 (acK153). b Overlay of protein backbone Cα RMSD plots of the wild-type (dark green, light green) and acK153 mutant (dark blue, light blue) of JNK. c Overlay of the wild-type (green) and acK153 mutant (blue) of JNK representing the surface of ATP nucleotide at the (I) start of production run and (II) random snapshot after 10 ns in the MD trajectory. d Overlay of ATP nucleotide RMSD plots of the wild-type (dark green, light green) and acK153 mutant (dark blue, light blue) of JNK. e Overlay of the distance between and γ-phosphate of ATP and nitrogen atom side-chain amine in K153 (dark green) and acK153 (dark blue) as a function of time for the two systems. f Overlay of the structures of JNK wild-type (salmon, K153) and energy-minimized structures of K153R mutant (blue, R153-min). The distance between the oxygen (O3G) atom of γ-phosphate of ATP and side-chain nitrogen atom (NZ) of K153 in wild-type JNK and side-chain NH2 atom of energy-minimized R153 are shown. g Graph showing binding of ATP to recombinant GST-tagged wild-type and mutants of JNK1, JNK1-K153Q and JNK1-K153R. GST-tagged proteins were purified from E. coli BL21 (DE3) and was incubated with either [α-³²P] or [γ-³²P] ATP. The unbound ATP was washed, and the bound ATP was measured in a scintillation counter and reported as counts per minute (cpm). n = 6 independent experiments. Data are presented as mean ± s.d. *p < 0.05. One-way ANOVA was used to calculate the p-values. h Representative western blotting images depicting in vitro kinase assay for wild-type JNK1, JNK1-K153Q and JNK1-K153R. The wild-type and mutants of JNK1 were immunoprecipitated using Flag-antibody-conjugated agarose beads and incubated with c-Jun fusion protein in an in vitro kinase assay. JNK activity was measured by detecting the phosphorylation of c-Jun. The acetylation and phosphorylation of immunoprecipitated JNK was probed by western blotting. Whole-cell lysates (WCLs) were probed for JNK by western blotting. i Graph showing luciferase reporter assay to analyse JNK activity in HeLa cell transiently expressing either JNK1 or JNK1-K153Q or JNK1-K153R. Reporter plasmid containing multiple binding sites for AP-1 was used to track the activity of JNK in HeLa cells. n = 3 independent experiments. Data are presented as mean ± s.d. *p < 0.05. One-way ANOVA was used to calculate the p-values

Fig. 5

Reversible acetylation of K153 regulates ATP binding and kinase activity of JNK. a Western blotting images showing the phosphorylation of JNK, c-Jun and total protein levels of SIRT2 and K-40 tubulin acetylation in HeLa cells treated with 200 µM H₂O₂ for 30 min. n = 4 independent experiments. b Western blotting images showing the time-dependent phosphorylation, acetylation and activity of JNK upon H₂O₂ treatment. HeLa cells were treated with 500 µM H₂O₂ for different time points and the JNK immunoprecipitated from these cells were analysed for their acetylation and phosphorylation levels by western blotting. The whole-cell lysate (WCL) was used to observe the concomitant changes in the phosphorylation of c-Jun, a downstream target of JNK. c Western blotting images showing the UV-induced changes in phosphorylation, acetylation and activity of JNK. JNK was immunoprecipitated from HeLa cells treated with UV (1 KJ/m²) for 30 min and assessed for acetylation. WCLs were probed for phosphorylation of JNK, phosphorylation of c-Jun and total protein levels of SIRT2 by western blotting. n = 3 independent experiments. d Western blotting images depicting the effect of p300 overexpression on H₂O₂-mediated phosphorylation and activity of JNK. 293T cells transiently overexpressing p300 were treated with 500 µM H₂O₂ for 30 min, and the phosphorylation of JNK and c-Jun were analysed by western blotting. e Western blotting images showing the effect of SIRT2 depletion on H₂O₂-mediated phosphorylation and acetylation of JNK. SIRT2-depleted HeLa cells were treated with 500 µM H₂O₂ for 30 min for different time points. JNK was immunoprecipitated and assessed for the acetylation of JNK. WCLs were probed for phosphorylation of JNK and total protein levels of SIRT2 by western blotting. f Western blotting images depicting the effect of SIRT2 inhibition on H₂O₂-mediated phosphorylation, acetylation and activity of JNK. HeLa cells were treated with 10 µM AGK2 for 2 h and then treated with 500 µM H₂O₂ for 30 min. The phosphorylation of c-Jun was analysed by western blotting in WCLs. JNK was immunoprecipitated from HeLa cells and assessed for acetylation and phosphorylation. g Representative western blotting images showing H₂O₂-mediated phosphorylation of wild-type and mutants of JNK. HeLa cells transiently expressing JNK1 or JNK1-K153Q or JNK1-K153R were treated with 500 µM H₂O₂ for 30 min. JNK was immunoprecipitated from these lysates and assessed for its phosphorylation by western blotting

Fig. 6

SIRT2-dependent deacetylation of JNK regulates H₂O₂-induced cell death in HeLa cells. a Western blotting analysis showing the SIRT2 levels in scramble (control) or SIRT2-depleted (SIRT2-KD) HeLa cells. b Representative fluorescent microscopic images depicting the H₂O₂-induced cell death in control (scramble) or SIRT2-depleted (SIRT2-KD) HeLa cells. Live (green) and dead (red) were stained after 200 µM H₂O₂ treatment for 12 h. c Graph showing the percentage of H₂O₂-induced cell death in control (scramble) or SIRT2-depleted (SIRT2-KD) HeLa cells as measured by the staining for live (Green) and dead (red) cells after 200 µM H₂O₂ treatment for 12 h. n = 3 independent experiments. Data are presented as mean ± s.d. *p < 0.05. Two-way ANOVA was used to calculate the p-values. d Graph showing luciferase reporter assay to analyse JNK activity in control (scramble) or SIRT2-depleted (SIRT2-KD) HeLa cells treated with 200 µM H₂O₂. Reporter plasmid containing multiple binding sites for AP-1 was used to track the activity of JNK in HeLa cells. n = 3 independent experiments. Data are presented as mean ± s.d. *p < 0.05. Two-way ANOVA was used to calculate the p-values. e Western blotting analysis showing the SIRT2 levels in HeLa cells overexpressing either Flag, wild-type and mutant versions of SIRT2. f Representative fluorescent microscopic images depicting the H₂O₂-induced cell death in HeLa cells overexpressing either Flag, wild-type and mutant versions of SIRT2. Live (green) and dead (red) were stained after 200 µM H₂O₂ treatment for 12 h. g Graph showing the percentage of H₂O₂-induced cell death in HeLa cells overexpressing either Flag, wild-type and mutant versions of SIRT2, as measured by the staining for live (Green) and dead (red) cells after 200 µM H₂O₂ treatment for 12 h. n = 3 independent experiments. Data are presented as mean ± s.d. *p < 0.05. Two-way ANOVA was used to calculate the p-values. h Western blotting analysis showing the SIRT2 levels in Flag or wild-type SIRT2 overexpressing HeLa cells treated with JNK inhibitor (JNK-I). i Graph showing luciferase reporter assay to analyse JNK activity in Flag or wild-type SIRT2 overexpressing HeLa cells treated with JNK inhibitor (JNK-I). Reporter plasmid containing multiple binding sites for AP-1 was used to assess the activity of JNK in HeLa cells. n = 3 independent experiments. Data are presented as mean ± s.d. *p < 0.05. One-way ANOVA was used to calculate the p-values. j Representative fluorescent microscopic images depicting the H₂O₂-induced cell death in Flag or wild-type SIRT2 overexpressing HeLa cells treated with JNK inhibitor (JNK-I). Live (green) and dead (red) were stained after 200 µM H₂O₂ treatment for 12 h. k Graph showing the percentage of H₂O₂-induced cell death in Flag or wild-type SIRT2 overexpressing HeLa cells treated with JNK inhibitor (JNK-I), as measured by the staining for live (Green) and dead (red) cells after 200 µM H₂O₂ treatment for 12 h. n = 3 independent experiments. Data are presented as mean ± s.d. *p < 0.05. Two-way ANOVA was used to calculate the p-values

Fig. 7

SIRT2 deficiency prevents acetaminophen (APAP)-induced liver cell death. a Representative H&E staining images of wild-type (WT) and SIRT2-deficient (SIRT2-KO) mice liver sections showing degenerative changes (arrows) when injected with 100 mg/kg APAP intraperitoneally. SIRT2-KO mice show reduced degenerative changes. n = 5 mice per group/time point. Scale bar = 20 µM. b Representative TUNEL staining images of WT and SIRT2-KO mice liver sections showing TUNEL-positive cells (brown nuclei) 24 h post injection with 100 mg/kg APAP intraperitoneally. n = 5 mice per group. c Graph showing the percentage of TUNEL-positive cells in WT and SIRT2-KO mice liver sections. Mice were sacrificed 24 h post 100 mg/kg APAP intraperitoneal injection. n = 5 mice per group. Scale bar = 20 µM Data are presented as mean ± s.d. *p < 0.05. Student’s t-test was used to calculate the p-values. d Graph depicting the levels of serum alanine aminotransferase (ALT) in WT and SIRT2-KO mice. Mice were sacrificed at different time points after 100 mg/kg acetaminophen (APAP) intraperitoneal injection. n = 5 mice per group. Data are presented as mean ± s.d. *p < 0.05. Two-way ANOVA was used to calculate the p-values. e Graph depicting the levels of serum aspartate aminotransferase (AST) in WT and SIRT2-KO mice. Mice were sacrificed at different time points after 100 mg/kg APAP intraperitoneal injection. n = 5 mice per group. Data are presented as mean ± s.d. *p < 0.05. Two-way ANOVA was used to calculate the p-values. f Western blotting images depicting the acetylation of endogenous JNK in the liver lysates of WT and SIRT2-KO mice injected with vehicle or 100 mg/kg APAP. JNK was immunoprecipitated and assessed for the acetylation of JNK. Whole cell lysates (WCLs) were probed for the protein levels of SIRT2 by western blotting. g Western blotting images depicting the phosphorylation of endogenous JNK in the liver lysates of WT and SIRT2-KO mice injected with vehicle or 100 mg/kg APAP. WCLs were probed for phosphorylation of JNK and total protein levels of SIRT2 by western blotting. h Graph showing the survival of WT and SIRT2-KO mice injected with high dose of APAP (600 mg/kg; APAP). n = 20 mice per group