A biotin switch-based proteomics approach identifies 14-3-3ζ as a target of Sirt1 in the metabolic regulation of caspase-2

Abstract

While lysine acetylation in the nucleus is well characterized, comparatively little is known about its significance in cytoplasmic signaling. Here we show that inhibition of the Sirt1 deacetylase, which is primarily cytoplasmic in cancer cell lines, sensitizes these cells to caspase-2-dependent death. To identify relevant Sirt1 substrates, we developed a proteomics strategy, enabling the identification of a range of putative substrates, including 14-3-3ζ, a known direct regulator of caspase-2. We show here that inhibition of Sirtuin activity accelerates caspase activation and overrides caspase-2 suppression by nutrient abundance. Furthermore, 14-3-3ζ is acetylated prior to caspase activation, and supplementation of Xenopus egg extract with glucose-6-phosphate, which promotes caspase-2/14-3-3ζ binding, enhances 14-3-3ζ-directed Sirtuin activity. Conversely, inhibiting Sirtuin activity promotes14-3-3ζ dissociation from caspase-2 in both egg extract and human cultured cells. These data reveal a role for Sirt1 in modulating apoptotic sensitivity, in response to metabolic changes, by antagonizing 14-3-3ζ acetylation.

Figure 1. Inhibition of cytoplasmic Sirt1 enhances caspase-2-dependent cell death in breast cancer cell lines

(A) Breast cancer cell lines (MCF-7, MDA-MB-435, MDA-MB-468, MDA-MB-231, and MDA-MB-436) were treated with 100 uM ex527 and/or 50 nM taxol. Cell death was measured 36 hours after taxol treatment by propidium idodide staining and FACS analysis. Results of three independent experiments are summarized with error bars representing standard error. (B) Breast cancer cell lines from panel A were fractionated into nuclear (n) and cytoplasmic (c) fractions. Each fraction was immunoblotted for Sirt1, lamin and actin. (C) MDA-MB-231 cells were treated with 180 nM leptomycin B or control, fractionated as in panel C then immunoblotted for Sirt1 (red) and p53 (green). Proteins visualized by Li-Cor imaging. (D) MDA-MB-231 cells were incubated with etoposide (30 h) or staurosporine (4 h), fractionated, and immunoblotted for Sirt1, lamin, and actin.

Figure 2

Summary of steps in acetyl-bst proteomics approach.

Figure 3. Validation of a subset of putative Sirt1 substrates, and C2 involvement in the effect of Sirt1 inhibition on taxol-induced death

(A) GAPDH, SOD-1, 14-3-3ζ, and malate dehydrogenase were expressed in 293T cells, immunoprecipitated and immunoblotted with anti-acetylated lysine antibody. The chromatographs show LC-MS/MS quantitation of the relative abundance of each protein captured by acetyl-bst in the presence or absence of Sirt1. (B) His-14-3-3ζ was acetylated (¹⁴C acetyl-CoA) then retrieved, washed, and incubated with Sirt1 in the presence or absence of nicotinamide. Radiolabled 14-3-3ζ was resolved by SDS-PAGE/phosphorimager. (C) MDA-MB-231 cells were transfected with HA-tagged 14-3-3ζfollowed by transfection with siRNA (20 uM) 24 h later. 48 h after transfection, HA-14-3-3ζ was immunoprecipated and immunoblotted with anti-acetylated lysine antibody. (D) MCF-7 cells were transfected with 20 uM scrambled or C2-targeted smartpool siRNA then treated with ex527/taxol and analyzed for cell death as in panel A.

Figure 4. Metabolic regulation of 14-3-3 and C2 via sirtuin deacetylase activity

(A) Sequence alignment of human and Xenopus 14-3-3ζ showing acetylated lysines (asterisks) identified by mass spectrometry in Xenopus egg extract. (B) Egg extract was incubated with combinations of nicotinamide (20 mM), c646 (500 uM), and G6P (5mM). Caspase activity was determined by a colorimetric measurement of DEVD-pNA cleavage. (C) ³⁵S-labeled C2 was incubated in mock- or nicotinamide-treated egg extract, and samples were resolved by SDS-PAGE/phosphorimager (top panel). Caspase activity was measured (middle panel) as in panel B. In parallel, aliquots of extract were fractionated to produce mitochondria-free cytosol, and immunoblotted for cytochrome c and actin (lower panel). (D) His-tagged recombinant 14-3-3ζon resin was incubated in extract and retrieved for immunoblotting with anti-acetylated lysine antibodies. Lower panel shows caspase activity in parallel. (E) His-tagged recombinant 14-3-3ζ on resin was acetylated with ¹⁴C-labeled acetyl CoA and incubated in mock-, G6P- or nicotinamide-treated egg extract. 14-3-3ζ was retrieved and resolved by SDS-PAGE/phosphorimager. (F) ¹⁴C-acetylated 14-3-3ζ was incubated in fresh, or aged (3 hr) extract supplemented with 2mM NAD or buffer control. Deacetylation was carried out for 60 min and quantified by phosphorimager and normalized to protein levels as determined by coomassie blue staining and Li-Cor imaging. Results of 3 experiments are summarized (error bars represent standard error). (G) Resin-bound GST-C2 was phosphorylated by CaMKII in buffer then allowed to bind His-14-3-3ζ in vitro Bound complex was incubated in mock- or 20 mM nicotinamide-treated extract. GST-C2 was retrieved and immunoblotted for 14-3-3ζ (His). (H) The experiment was performed as in panel F, but in the presence or absence of the acetylase inhibitor c646. (I) 239T and MDA-MB-231 cells were transfected with HA-tagged 14-3-3ζ followed by ex527 (100 uM) treatment for 6 h. HA-14-3-3ζ immunoprecipitates were immunblotted for endogenous C2.