Succinylation of a KEAP1 sensor lysine promotes NRF2 activation

Abstract

Cross talk between metabolism and stress-responsive signaling is essential for maintaining cellular homeostasis. This cross talk is often achieved through covalent modification of proteins by endogenous, reactive metabolites that regulate key stress-responsive transcription factors like NRF2. Metabolites including methylglyoxal, glyceraldehyde 3-phosphate, fumarate, and itaconate covalently modify sensor cysteines of the NRF2 repressor KEAP1, resulting in stabilization of NRF2 and activation of its cytoprotective transcriptional program. Here, we employed a shRNA-based screen targeting the enzymes of central carbon metabolism to identify additional regulatory nodes bridging metabolism to NRF2 activation. Succinic anhydride, increased by genetic depletion of the TCA cycle enzyme succinyl-CoA synthetase or by direct administration, results in N-succinylation of lysine 131 of KEAP1 to activate NRF2 signaling. This study identifies KEAP1 as capable of sensing reactive metabolites not only by several cysteine residues but also by a conserved lysine residue, indicating its potential to sense an expanded repertoire of reactive metabolic messengers.

Figure 1.. Loss of succinyl-CoA synthetase activates NRF2 driven transcription.

A) Schematic of the cellular screening platform using ARE-GFP-LUC K562 reporter cells. Shown is a representative plot indicating positivity for GFP after transduction with an shRNA targeting KEAP1. B) Schematic depicting the enzymes of central carbon metabolism targeted by the shRNA screen (boxes). C) Replicate fold increases in percent GFP positive ARE-GFP-LUC K562 cells after a 72 hr exposure to the indicated shRNAs of the screen. D, E) Relative transcript levels of SUCLG2 (D) and NQO1 (E) after 48 hr transient transfection of HEK293T cells with shRNAs targeting SUCLG2 (n=3; mean and s.d.; ****P<0.0001, **P<0.01, one-way ANOVA). F) Relative luminescence values of ARE-LUC reporter activity 72 hr after transduction of lentiviruses encoding shRNAs to SUCLG2 and NRF2 in ARE-GFP-LUC K562 (n=3; mean and s.d.; ****P<0.0001, two-way ANOVA).

Figure 2.. Augmented levels of succinic anhydride activate NRF2 signaling.

A) Relative ARE-LUC reporter activity induced shRNAs targeting SUCLG2 in ARE-GFP-LUC K562 cells in the presence of the indicated antioxidants (10 mM each; n=4; mean and s.d.; ns = not significant P>0.05, ****P<0.0001, two-way ANOVA). B) Schematic depicting the increase of succinyl CoA and subsequent spontaneous formation of SA after knockdown of the SCS subunit SUCLG2. C) Representative Western blotting for NRF2 levels after 4 hrs treatment of HEK293T cells with SA (10 mM) in the absence or presence of GSH (10 mM). D) Representative Western blots of NRF2, NQO1, and HMOX1 levels from IMR32 cells treated with SA (0.1 mM) or bardoxolone (BARD; 250 nM). E) Representative anti-MYC Western blotting of anti-FLAG coimmunoprecipitated material after 1 hr of SA treatment in HEK293T expressing MYC-CUL3 and KEAP1-FLAG transgenes.

Figure 3.. SA accumulation leads to covalent modification of KEAP1 lysines.

A) Structure of succinic anhydride alkyne (SA-alkyne). B) Relative luminescence values of ARE-LUC reporter activity from IMR32 cells treated with SA or SA-alkyne (12 mM) for 4 hrs (n=5, mean and s.d.; ****P<0.0001, **P<0.01, one-way ANOVA). C) Representative fluorescence scan of rhodamine azide labeling of anti-FLAG immunoprecipitated material after treatment of HEK293T cells expressing KEAP1-FLAG with the indicated concentrations of SA-alkyne for 1 hr. D) Representative fluorescence scan of rhodamine azide labeling of anti-FLAG immunoprecipitated material after pretreatment with the indicated concentrations of SA followed by a 1 hr labeling with SA-alkyne (10 mM) in HEK293T cells expressing KEAP1-FLAG. E) Relative ARE-LUC activity from IMR32 cells expressing the indicated KEAP1 mutant and MYC-NRF2 and then treated with 1 mM SA for 4 hrs (n=5; mean and s.d.; ***P<0.001, two-way ANOVA). F) Western blotting for anti-N-succinyl lysine positivity of anti-FLAG immunoprecipitated material from HEK293T cells expressing KEAP1-FLAG after 1 hr treatment with the indicated concentrations of SA. G) Western blotting for anti-N-succinyl lysine positivity of anti-FLAG immunoprecipitated material from HEK293T cells expressing KEAP1-FLAG and shRNAs targeting SUCLG2, stabilized by the desuccinylase inhibitor ET-29 (10 μM).

Figure 4.. N-succinylation of Lysine 131 functionally inhibits KEAP1 activity.

A) MS/MS spectra of the N-succinylated K131 containing tryptic peptide immunoprecipitated from overexpressed KEAP1-FLAG from HEK293T cells treated with SA (10 mM). B) Representative fluorescence scan of rhodamine azide labeling of anti-FLAG immunoprecipitated material after pretreatment with the indicated concentrations of SA followed by a 1 hr labeling with SA-alkyne (5 mM) in HEK293T cells expressing the indicated KEAP1-FLAG transgenes. C) Western blotting for anti-N-succinyl lysine positivity of anti-FLAG immunoprecipitated material from HEK293T cells expressing the indicated KEAP1-FLAG transgene mutants after 1 hr treatment with SA (1 mM). D) Relative ARE-luciferase reporter activity from IMR32 cells expressing the indicated KEAP1-FLAG mutant transgenes and then treated with 1 mM SA for 4 hrs (n=4; mean and s.d.; ns = not significant P>0.05; ****P<0.0001, *P<0.05, two-way ANOVA). E) Schematic depicting metabolites from central carbon metabolism that can covalently modify the indicated residues on KEAP1.