SIRT6 stabilizes DNA-dependent protein kinase at chromatin for DNA double-strand break repair

Abstract

The Sir2 chromatin regulatory factor links maintenance of genomic stability to life span extension in yeast. The mammalian Sir2 family member SIRT6 has been proposed to have analogous functions, because SIRT6-deficiency leads to shortened life span and an aging-like degenerative phenotype in mice, and SIRT6 knockout cells exhibit genomic instability and DNA damage hypersensitivity. However, the molecular mechanisms underlying these defects are not fully understood. Here, we show that SIRT6 forms a macromolecular complex with the DNA double-strand break (DSB) repair factor DNA-PK (DNA-dependent protein kinase) and promotes DNA DSB repair. In response to DSBs, SIRT6 associates dynamically with chromatin and is necessary for an acute decrease in global cellular acetylation levels on histone H3 Lysine 9. Moreover, SIRT6 is required for mobilization of the DNA-PK catalytic subunit (DNA-PKcs) to chromatin in response to DNA damage and stabilizes DNA-PKcs at chromatin adjacent to an induced site-specific DSB. Abrogation of these SIRT6 activities leads to impaired resolution of DSBs. Together, these findings elucidate a mechanism whereby regulation of dynamic interaction of a DNA repair factor with chromatin impacts on the efficiency of repair, and establish a link between chromatin regulation, DNA repair, and a mammalian Sir2 factor.

Keywords: DNA damage; DNA repair; SIRT6; Sir2; aging; genomic stability.

Figure 1.. SIRT6 interacts with the DNA-PK DSB repair factor.

(A) Endogenous SIRT6 protein associates with large molecular weight complexes. HeLa cell nuclear extract (NE) was separated by gel-filtration and fractions subjected to Western analysis with SIRT6 antibody. Fractions with molecular weight standards are indicated (arrows). (B) Coomassie stain of proteins in Flag-SIRT6 or negative control IPs from 293T cells. The identities of proteins detected by mass spectrometry are indicated. (C) SIRT6, but not SIRT1, associates with DNA-PK subunits. Western analysis of SIRT6, SIRT1, and negative control IPs with the indicated antibodies. WCE: whole cell extract. (D) SIRT6 interaction with DNA-PKcs, but not Ku70 and Ku80, is resistant to ethidium bromide (EtBr). Western analysis as in (C) except EtBr was added as indicated to disrupt DNA-mediated interactions. (E) Endogenous interaction between SIRT6 and DNA-PKcs. Western analysis of SIRT6-bound proteins in co-IPs from 293T cells.

Figure 2.. DNA damage stabilizes SIRT6 interaction with chromatin.

(a) Gel shift showing SIRT6, but not SIRT1, binding to purified native nucleosomes in vitro. Left: coomassie stain of recombinant SIRT6 and SIRT1 proteins. Right: EtBr stain of nucleosomal DNA on a non-denaturing gel; arrow indicates SIRT6-mononucleosome complex. (b) Schematic of chromatin purification protocol (see methods). Dt, detergent extractable fraction; Rn, RNase extractable fraction; Chr, purified chromatin fraction. (c) DNA damage-dependent stimulation of SIRT6 association with chromatin. Western analysis with the indicated antibodies of HeLa cells treated with 45 nM neocarzinostatin (NCS) for 1 hour and fractionated as described in (c). H4 and Tubulin are detected in the expected fractions.

Figure 3.. SIRT6 is required for global deacetylation of H3K9 in response to DNA damage.

(a) Western analysis of SIRT6 expression levels in the indicated cell lines stably expressing two different SIRT6 shRNAs (S6KD1 and S6KD2) or empty vector control (pSR). (b) SIRT6 knock-down leads to hypersensitivity to γ-irradiation (IR). Gy, Gray. (c) Western analysis of H3K9Ac levels in S6KD2 (KD) or control (Co) HT1080 cells in response to DNA DSB agents neocarzinostatin (NCS) or bleomycin (Bleo) treatment (1hr). (d) Western analysis of H3K9Ac levels in 293T cells over-expressing SIRT6 or empty vector control (pcDNA), following exposure to NCS (1hr). In (c) and (d), total H3 or H3K14Ac levels are shown as controls.

Figure 4.. SIRT6 stabilizes DNA-PKcs at chromatin in response to DNA damage.

(a) SIRT6 is required for efficient mobilization of DNA-PKcs to chromatin in response to the DNA DSB agent NCS. Western analysis with the indicated antibodies of S6KD2 and control cells fractionated as in (Fig 2c). (b) Fractionation experiments performed as in (c) utilizing S6KD1 cells (second independent SIRT6 shRNA). WCE, whole cell extract. Chr, chromatin. (b) Western analysis showing reconstitution of S6KD2 cells with recombinant wild-type (WT) or catalytically mutant (HY) SIRT6 protein. (d) DNA-PKcs mobilization to chromatin upon NCS treatment was determined in S6KD2 cell lines reconstituted with WT SIRT6, catalytically inactive SIRT6 HY protein, or control vector (Co) as indicated.

Figure 5.. Dynamic association of SIRT6 and DNA-PKcs with chromatin flanking site-specific DNA DSBs.

(a) SIRT6 occupancy at chromatin flanking DSBs induced by I-PpoI. Quantitative real-time PCR amplification of DNA sequences flanking the I-PpoI site from ChIPs performed with SIRT6 antibodies. Data are normalized to no I-PpoI (control) samples. (b) DNA-PKcs occupancy at chromatin flanking DSBs induced by I-PpoI, determined as for SIRT6 in (a). (c) DNA-PKcs occupancy at the indicated distances from an I-SceI DSB site in SIRT6 KD and controls cells. Data are normalized to no I-SceI controls. (d) DNA-PKcs occupancy at chromatin adjacent to (+60 to +223 bp) an I-SceI DSB site, following retroviral over-expression of Flag-tagged wild-type SIRT6 (S6WT), catalytically inactive SIRT6 (S6HY), or empty vector control (pBabe). In all panels, SIRT6 KD cells were generated with S6KD2 shRNA, and the data represent the mean +/- S.E.

Figure 6.. SIRT6 promotes resolution of DNA double strand breaks (DSBs).

(a-c) Impaired resolution of DSBs in SIRT6 knock-down (S6KD2) cells. Control (pSR) and knock-down (S6KD) cells were isolated 1 hour after NCS treatment and tail moment was determined in comet assays. (a) Histogram of tail moments of >100 cells. (b) Representative comet images from S6KD or control cells following DNA damage (+NCS) or mock (-NCS) treatment. (c) Mean tail moment of comet assays shown in (a). Error bars indicate the S.E.M. (d) Wild-type SIRT6 (S6WT), but not the mutant SIRT6 (S6HY), protein rescues the DSB repair defect of SIRT6 knock-down cells. Mean tail moment of comet assays for the indicated cells, following treatment with NCS and quantified as in (c). (e) Resolution of site-specific DNA DSBs in SIRT6 KD (S6KD2) and control (pSR) cells assayed using the I-SceI and I-PpoI systems. Quantitative real-time PCR amplification of DNA using primers flanking the DSB sites is shown. *, p=0.009; **, p=0.02. The data represent the average of triplicate experiments, and error bars indicate the S.E.M.

Figure 7.. SIRT6 is not required for DNA double-strand break rejoining in a cell-free system.

(a) Western analysis showing immunodepletion of SIRT6 from HeLa cell nuclear extracts (NE) with anti-SIRT6 antibodies. Mock-depleted control nuclear extracts were generated using the protein A/G-sepharose beads alone. (b) Ethidium bromide stain (inverted image) of DNA products of cell-free DSB rejoining assay. Reactions contained linearized pUC19 plasmid DNA fragments and SIRT6- or mock- depleted nuclear extracts. No difference in ligated species is observed between SIRT6-depleted and mock-depleted reactions.