Caspase-8 Cleaves Histone Deacetylase 7 and Abolishes Its Transcription Repressor Function

Abstract

Caspase-8 is the initiator caspase of the extrinsic apoptosis pathway and also has a role in non-apoptotic physiologies. Identifying endogenous substrates for caspase-8 by using integrated bioinformatics and biological approaches is required to delineate the diverse roles of this caspase. We describe a number of novel putative caspase-8 substrates using the Prediction of Protease Specificity (PoPS) program, one of which is histone deacetylase 7 (HDAC7). HDAC7 is cleaved faster than any other caspase-8 substrate described to date. It is also cleaved in primary CD4+CD8+ thymocytes undergoing extrinsic apoptosis. By using naturally occurring caspase inhibitors that have evolved exquisite specificity at concentrations found within the cell, we could unequivocally assign the cleavage activity to caspase-8. Importantly, cleavage of HDAC7 alters its subcellular localization and abrogates its Nur77 repressor function. Thus we demonstrate a direct role for initiator caspase-mediated proteolysis in promoting gene transcription.

FIGURE 1.

HDAC7 is cleaved by caspase-8 in vitro. A, HDAC7 was incubated with caspase-8 for 30 min at 37 °C. Reactions were resolved by SDS-PAGE and immunoblotted with anti-FLAG antibody. B, schematic representation of HDAC7 (splice variant 3). Like all class IIa HDACs, the N terminus interacts with the transcriptional co-repressor CtBP and transcription factors, including MEF2 family members. The conserved phosphorylation sites at Ser¹⁵⁵, Ser³²¹, and Ser⁴⁴⁹ mediate nuclear/cytoplasmic shuttling through interactions with 14-3-3 proteins. The NLS, NES, and the conserved histone deacetylase domain are indicated. The PoPS predicted caspase-8 cleavage site at Asp³⁷⁵ is shown. Cleavage at this position would separate the N-terminal transcription factor binding domain and NLS from the C-terminal deacetylase domain and NES. C, phylogenic tree analysis of HDAC7 proteins with human class IIa HDACs and yeast HDA1. D, multiple sequence alignment of class II HDACs across the PoPS-predicted cleavage site showing conservation in mammalian HDAC7 orthologues. Caspase-8 recognition sequence is in boldface type. wt, wild type.

FIGURE 2.

HDAC7 is cleaved as efficiently as physiological caspase-8 substrates. A, caspase-3 C285A (5 nm, top), GST-Bid (25 nm, middle), or HDAC7-FLAG (bottom) was incubated with 0–1 μm caspase-8 in a ½ dilution series for 30 min at 37 °C. Asterisks indicate nonspecific recognition of caspase-8 large and small subunits by the caspase-3 antibody. Arrowheads indicate caspase-8 concentration at which half the substrate was cleaved (EC₅₀). B, HDAC7-FLAG or 25 nm p35 C2A was incubated with 50 nm of the indicated active site-titrated caspase for 30 min at 37 °C. Asterisk represents HDAC7-FLAG that may be translated from a minor internal initiator methionine. C, conditions are as in B, with (+) or without (-) 10% PEG 6000. D, HDAC7-FLAG was incubated with 0–150 nm caspase in a ½ dilution series for 30 min at 37 °C. Reactions were resolved by SDS-PAGE and immunoblotted with the indicated antibodies. IB, immunoblot.

FIGURE 3.

HDAC7 is cleaved following engagement of the caspase-8-dependent extrinsic apoptosis pathway. HEK293 cells were transfected with HDAC7-FLAG and treated with 200 ng/ml TRAIL or 100 μm etoposide (Etop) for the indicated times (A); wild type (wt) or D375A HDAC7-FLAG and treated with 200 ng/ml TRAIL for 3 h (B); HDAC7-FLAG and treated with 5 μm MG132, 200 μm etoposide, 200 nm staurosporine for 18 h, or co-transfected with 50 ng Bax cDNA during the initial transfection (C); 0.2 μg of HDAC7-FLAG and 0.8 μg of the indicated cDNA and were treated with 10 ng/ml TNFα and 1 μm cycloheximide for 20 h (D). Average and the standard deviation are shown (n = 3). A–D, cell lysates were normalized for protein concentration, resolved by SDS-PAGE, and immunoblotted (IB) with the indicated antibodies. A, B, and C, executioner caspase activity was assessed by hydrolysis of 100 μm Ac-DEVD-afc. A, C, and D, apoptosis was assessed by annexin V binding and flow cytometry. Unless indicated, the data represent one of three independent experiments.

FIGURE 4.

Endogenous HDAC7 is cleaved in CD4⁺CD8⁺ thymocytes following engagement of the extrinsic apoptosis pathway. Primary CD4⁺CD8⁺ thymocytes were treated with 100 ng/ml FasL (A) or 5 μg/ml anti-CD3 and 5 μg/ml anti-CD28 antibody, ±1 h preincubation with 100 μm Z-VAD-fmk (B). C, DPK cells were treated with 5 μg/ml anti-CD3 and 5 μg/ml anti-CD28 antibody for 0, 2, 6, and 24 h, ±100 μm Z-VAD-fmk. Cell lysates were normalized for protein concentration, resolved by SDS-PAGE, and immunoblotted (IB) with the indicated antibodies. Executioner caspase activity was assessed by hydrolysis of 100 μm Ac-DEVD-afc. Apoptosis as measured by DNA content was assessed by propidium iodide staining of permeabilized cells. Data represent one of three independent experiments.

FIGURE 5.

The apoptosis-generated N-terminal fragment of HDAC7 is unstable. HEK293 cells transfected with FLAG-HDAC7, HDAC7-FLAG, or FLAG-HDAC7-FLAG were treated with 200 ng/ml TRAIL for the indicated times. Cell lysates were normalized for protein concentration, resolved by SDS-PAGE, and immunoblotted (IB) with the indicated antibody. Asterisks represent nonspecific bands and serve as loading controls. Although the polyclonal HDAC7 antibody is specific and recognizes the N terminus (Santa Cruz Biotechnology, N-18) the position of the epitope(s) is unknown.

FIGURE 6.

Cleavage of HDAC7 alters intracellular localization. A, COS7 cells were transfected with HDAC7 mutants containing C-terminal (C-term) tagged GFP. B, cells were counted and scored for subcellular GFP localization, n >100 cells. C, COS7 cells were co-transfected with HDAC7-FLAG (wild type (wt) or D375A) together with caspase-8 and XIAP. Cells were fixed and stained with mouse anti-FLAG antibody, FITC-conjugated anti-mouse IgG, and 4′,6-diamidino-2-phenylindole (DAPI). D, cells were scored blind for subcellular FLAG staining, n > 100 cells. A representative of two independent experiments is shown. Caspase-8 processing of HDAC7 causes relocalization of C-terminal fragment to the cytoplasm. Scale bar represents 10 μm.

FIGURE 7.

Cleavage of HDAC7 abolishes its transcription repressor function. A, HEK293 cells were co-transfected with 50 ng of β-galactosidase, 200 ng of Nur77-luciferase reporter construct, 200 ng of MEF2D, and 20 ng of the indicated HDAC7-FLAG variant (n = 4). B, to control for expression levels, HEK293 cells were transfected 1 μg of the indicated HDAC7-FLAG variant. C, HEK293 cells were transfected with HDAC7 variants for 24 h and then treated with 100 ng/ml TRAIL for 4 h (n = 3). D, HEK293 cells were co-transfected with 50 ng of β-galactosidase, 2.5 μg of XIAP, 200 ng of Nur77-luciferase reporter construct, and 200 ng of MEF2D with 20 ng of HDAC7-FLAG wild type (wt) or D375A, ±8 ng of caspase-8 plasmid (n = 3). E, primary thymocytes were treated with 5 μg/ml anti-CD3 and 5 μg/ml anti-CD28 antibody ±100 μm Z-VAD-fmk for 0, 1, 6, and 24 h. Nur77 mRNA expression was assessed by quantitative RT-PCR (n = 3 or 4). Luciferase activity was normalized to β-galactosidase activity. Apoptosis was assessed by annexin V binding and flow cytometry. Average and mean ± S.E. is shown. Paired Student's t test: ^**, p < 0.01; ^*, p < 0.05; n.s., p > 0.05. IB, immunoblot.