(A–B) HeLa cells were treated with doxorubicin (1.25µg/mL, 20h for cell viability; 5µg/mL, 8h for caspase activity). (C) HT1080 cells were treated with doxorubicin (1.25µg/mL, 20h). (D) U2OS cells were treated with doxorubicin (1.25µg/mL, 20h). (E) HeLa cells were treated with cisplatin (40µM) or UV (50J/m² or 100J/m²) for 24h. (F) HeLa cells were treated with TNFalpha (10ng/mL, 24h) and cyclohexamide (1µg/ml, 24h) to induce death receptor mediated cell death. Immunoblots were conducted in parallel to show extent of target knockdown. Data are represented as mean +/− s.d. (n=3). (Student’s T-test; *, p<0.05; **, p<0.01; ***, p<0.001)

(A) A schematic of the subtiligase assay. Briefly, cells were gently lysed in NP-40 lysis buffer on ice. Cell debris was pelleted and resulting supernatant was subjected to a subtiligase reaction (1µM subtiligase, 1mM biotin-peptide, 2mM DTT; 1h at room temperature). Biotinylated proteins were affinity purified using Neutravidin beads (Thermo Scientific) and analyzed by SDS-PAGE. Protein N-alpha-acetylation levels were determined by immunoblot against proteins in which the N-terminal residue corresponds to specificity of a Nat complex. Chk2 was included as a negative control, which should not be recognized by subtiligase based on sequence. (B–C) Subtiligase reaction was conducted on HeLa cells transfected with siRNA against NATH (B) or ARD1 (C) as described in (A). Enrichment of biotin labelled NatA substrates was observed in ARD1 or NATH deficient cells compared to that of control. Enrichment in the amount of pulldown suggests a decrease in protein N-alpha-acetylation levels of NatA substrates in NatA deficient cells. Blot quantification for all subtiligase assays was calculated relative to control and normalized to corresponding lysate sample using ImageJ software. (D) A schematic of detecting protein N-terminal peptide modifications by quantitative mass spectrometry. Briefly, cells were treated with RNAi and lysed in NP-40 buffer. FLAG-tagged caspase-2 (active cysteine mutant C320G) is affinity purified using anti-FLAG agarose and eluted with FLAG peptide. Purified protein was digested by Lys-C, which cleaves after a lysine residue. Dimethyl labelling of peptides was conducted as previously described (Hsu et al., 2003; Khidekel et al., 2007). Briefly, peptides were chemically modified using NaCNBH₃ or NaCNBD₃ followed by formaldehyde-H₂ or formaldehyde-D₂ treatment. Heavy or light modified peptides were mixed and analyzed by mass spectrometry. Ratios for levels of N-terminal acetylation were determined by normalizing to 1:1 ratio of shared internal peptides (See Figure S1 for list of peptides).

(A) HeLa cells were transfected with siRNAs as indicated and treated with doxorubicin (5µg/mL, 8h). Cell lysates were analyzed by SDS-PAGE, and caspase cleavage was assessed by immunoblot. Caspase-2 and caspase-3 activation is suppressed in ARD1 and NATH knockdown cells compared to that of control. Arrow indicates cleaved product of caspase-2. (B) Subtiligase reaction was conducted on HEK 293T cells transfected with FLAG-tagged WT, A3P, and A2S caspase-2 (active cysteine mutant C320G of caspase-2) as described in Figure 1. Enrichment of biotin-labeled A3P caspase-2 was observed compared to WT and A2S caspase-2. This suggests that A3P caspase-2 is hypoacetylated. (C) HEK 293T cells were transfected with FLAG-tagged WT, A3P, and A2S caspase-2 as well as with VSV-RAIDD as indicated. Caspase-2 was affinity purified with anti-FLAG agarose and eluted with FLAG peptide. Resulting eluants were subjected to SDS-PAGE for immunblot analysis. RAIDD efficiently co-immunoprecipitates with WT and A2S but not with A3P caspase-2. Blots are representative of at least three independent experiments. (D) Subtiligase reaction was conducted on lysates from HeLa cells transfected with siRNAs against acetyl-CoA synthetase or ATP citrate lyase as indicated. Increased biotin labelling of caspase-2 was observed in knockdown cells compared to control cells.

(A–C) HEK 293T, HeLa, or Jurkat cells were generated to express empty GFP vector or Bcl-xL. Cells were treated with acetate (50mM, 24h) or citrate (10mM, 24h) as indicated. Subtiligase reaction was conducted as described in Figure 1. Enrichment of biotin-labelled NatA and NatB substrates is observed in Bcl-xL cells compared to control cells. Biotin labelling is reduced to control levels by acetate or citrate treatment in Bcl-xL cells. (D) Lysates generated from Bcl-xL +/+ or Bcl-xL −/− MEFs were subjected to subtiligase reaction as described in Figure 1. Bax is hypoacetylated in Bcl-xL +/+ MEFs compared to Bcl-xL −/− MEFs. (E) Lysates generated from wild type, Bax −/−, Bak −/−, or Bax/Bak −/− (DKO) MEFs were subjected to subtiligase assay as described in Figure 1. Blots are representative of at least three independent experiments.

(A) Identification of metabolites using an AB/Sciex 5500 QTRAP mass spectrometer separated with normal phase HILIC NH2 chromatography using positive/negative ion switching in Jurkat cells stably expressing GFP empty vector or Bcl-xL. (B) Acetyl-CoA levels were measured in Jurkat cells stably expressing GFP empty vector or Bcl-xL by mass spectrometry. Peak areas of total ion current (TIC) were processed using MultiQuant 1.1 software. Average TIC for acetyl-CoA analyte is shown. Acetyl-CoA levels are lower in Bcl-xL cells relative to GFP cells. (C) Acetyl-CoA levels were measured in Bcl-xL +/+ or BclxL −/− MEFs using an enzyme-based assay (Abcam). Metabolite concentration was determined by generating a standard curve. Bcl-xL deficiency results in higher acetyl-CoA levels compared to control. (D) 293T cells were transiently transfected with empty vector, WT Bcl-xL, or specific Bcl-xL mutants (F131V/D133A or G148E). Acetyl-CoA levels were measured by mass spectrometry. WT or mutant Bcl-xL cells show a decrease in acetyl-CoA levels compared to empty vector cells. (E) Acetyl-CoA levels were measured in 293T cells transiently expressing WT or mutant Bcl-xL as described in (C). Data are represented as mean +/− s.d. (n=3). (Student’s T-test; *, p<0.05; **, p<0.01; ***, p<0.001).

(A) Jurkat cells stably expressing GFP empty vector or Bcl-xL were fed uniformly labelled 13C-glucose and metabolites were isolated by methanol extraction. Lysates were subjected to mass spectrometry as described in Figure 3A. (B) Percent change in ¹³C-glucose derived metabolites of the pentose phosphate pathway (PPP), glycolysis, and TCA cycle. Fructose-1,6 bisphosphate (FBP), dihydroxyacetone-phosphate (DHAP), 1,3-bis-phosphoglycerate (1,3PG), phosphoenolpyruvate (PEP), α–ketoglutarate (a-KG). Data are represented as mean +/− s.d. (n=3). (Student’s T-test; *, p<0.05; **, p<0.01; ***, p<0.001).

(A–D) HeLa cells stably expressing empty GFP vector or Bcl-xL were treated with doxorubicin as indicated. Cell viability was determined by measuring cellular ATP levels using a luminescence-based assay (Promega). Caspase activity was determined by measuring cleavage of a luciferin substrate containing the caspase cleavage site, DEVD (Promega). (A–B) Doxorubicin treatment efficiently kills GFP cells but not Bcl-xL cells. Pretreatment with acetate (50mM) or citrate (10mM) for 24h sensitizes Bcl-xL cells to doxorubicin-induced cell death. (C–D) Doxorubicin treatment induces substantial caspase activity in GFP cells but is completely suppressed in Bcl-xL cells. Pre-treatment of Bcl-xL cells with acetate or citrate enhances caspase activity induced by doxorubicin treatment. (E) Bcl-xL cells were transiently transfected with a pool of siRNAs (50nM, Dharmacon) as indicated followed by pre-treatment with acetate or citrate prior to doxorubicin treatment. Metabolite enhancement of doxorubicin-induced caspase activity in Bcl-xL cells is suppressed to control levels by RNAi against acetyl-CoA synthetase (Acs) or ATP citrate lyase (ACLY), respectively. Immunoblot shows extent of knockdown. (F) HeLa cells were transiently transfected with a pool of siRNAs as indicated. Double knockdown of ARD1 and Bax provides additive protection against doxorubicin-induced cell death compared to single knockdown of these genes. Graphs are representative of three independent experiments. Data are represented as mean +/− s.d. (n=3). (Student’s T-test; *, p<0.05; **, p<0.01; ***, p<0.001).