Sirtuin 6 (SIRT6) rescues the decline of homologous recombination repair during replicative senescence

Edited by James E. Cleaver, University of California, San Francisco, CA, and approved June 12, 2012 (received for review January 12, 2012)
July 2, 2012
109 (29) 11800-11805

Abstract

Genomic instability is a hallmark of aging tissues. Genomic instability may arise from the inefficient or aberrant function of DNA double-stranded break (DSB) repair. DSBs are repaired by homologous recombination (HR) and nonhomologous DNA end joining (NHEJ). HR is a precise pathway, whereas NHEJ frequently leads to deletions or insertions at the repair site. Here, we used normal human fibroblasts with a chromosomally integrated HR reporter cassette to examine the changes in HR efficiency as cells progress to replicative senescence. We show that HR declines sharply with increasing replicative age, with an up to 38-fold decrease in efficiency in presenescent cells relative to young cells. This decline is not explained by a reduction of the number of cells in S/G2/M stage as presenescent cells are actively dividing. Expression of proteins involved in HR such as Rad51, Rad51C, Rad52, NBS1, and Sirtuin 6 (SIRT6) diminished with cellular senescence. Supplementation of Rad51, Rad51C, Rad52, and NBS1 proteins, either individually or in combination, did not rescue the senescence-related decline of HR. However, overexpression of SIRT6 in “middle-aged” and presenescent cells strongly stimulated HR repair, and this effect was dependent on mono-ADP ribosylation activity of poly(ADP-ribose) polymerase (PARP1). These results suggest that in aging cells, the precise HR pathway becomes repressed giving way to a more error-prone NHEJ pathway. These changes in the processing of DSBs may contribute to age-related genomic instability and a higher incidence of cancer with age. SIRT6 activation provides a potential therapeutic strategy to prevent the decline in genome maintenance.
Aging is associated with an increased mutation rate (1) and the appearance of genomic rearrangements (2). The accumulation of mutations and rearrangements is a contributing cause of aging and leads to a decline of tissue functionality and an increased incidence of tumors. These mutations and genomic rearrangements arise from aberrant repair of DNA double-stranded breaks (DSBs).
DSBs are dangerous DNA lesions. If left unrepaired or repaired incorrectly, DSBs result in a massive loss of genetic information, chromosomal aberrations, or cell death. DSBs are repaired by two major pathways: nonhomologous end joining (NHEJ) and homologous recombination (HR) (3). NHEJ modifies the broken DNA ends and ligates them together with no requirement for homology, often generating deletions or insertions (4). In contrast, HR uses an undamaged DNA template to repair the break, leading to the reconstitution of the original sequence (5). HR repair is responsible for approximately one quarter of DNA repair events and has much slower repair kinetics than NHEJ (6). HR repair begins with the MRE11, NBS1, and Rad50 complex binding to DNA ends and mediating end resection. The RPA protein is then recruited to DNA ends, in a process regulated by CtIP (7). Once the ends are resected, Rad51 forms nucleoprotein filaments and mediates strand invasion of the filament into duplex DNA, usually on the sister chromatid (8). Rad51 is assisted by Rad52 and other members of the Rad52 family (9). Missing genomic information is copied from the donor chromosome, leading to precise reconstitution of the original sequence. Because mitotic recombination preferentially uses the sister chromatid as a template for repair, HR takes place in S/G2/M stages of cell cycle (10). SIRT6 has recently been identified as a component of the HR pathway. SIRT6 participates in HR at two steps, by activating PARP1 at the early stages of DSB processing (11) and, later, by stimulating CtIP (12).
Aging organisms exhibit a diminished capacity to repair DSBs. It was shown that the recruitment of DSB repair proteins to DNA damage sites is delayed in lymphocytes of aged donors (13) and that cells from aged individuals contain a higher number of DSBs (14, 15). We showed previously that NHEJ declines up to fourfold and becomes more error-prone in senescent cells (16). Furthermore, the levels of Ku70 and Ku80, two critical NHEJ proteins, decline in senescent cells (17). However, the changes in HR-mediated repair during replicative senescence and aging have not been directly analyzed.
Here, we examined the efficiency of HR repair of induced DSBs in normal human fibroblasts at increasing replicative age. HR was measured using chromosomally integrated GFP reporter cassettes. We show that HR-mediated repair declines sharply with increasing replicative age. This decline occurs even in “middle-aged” and presenescent cells, before the cells enter senescent growth arrest. The levels of several HR factors decline with replicative age; however, only supplementation with SIRT6 is able to partly restore HR activity. These studies reveal a precipitous decline of HR during replicative aging and suggest that stimulation of SIRT6 may be a potential therapeutic strategy to overcome the age-related decline in DSB repair.

Result

Construction of Primary Human Fibroblast Cell Lines for Measuring HR Efficiency.

To examine the effect of replicative senescence on DNA DSB repair by HR, we integrated a HR reporter cassette (Fig. 1A) (6, 10, 18) into the genome of HCA2 cells, a primary foreskin fibroblast strain, at population doubling (PD)26. The challenge of studying HR in the context of replicative senescence is that the reporter construct must be integrated in primary (nonimmortalized cells). A standard protocol of antibiotic selection, followed by isolation and expansion of individual clones, requires extensive passaging that would drive primary cells into replicative senescence. To overcome this difficulty and obtain cells carrying the HR reporter construct at the youngest possible replicative age, we pooled together ∼250 colonies formed after antibiotic (G418) selection and allowed the cells to reach confluence in a 10-cm plate. Because the number of cells at confluence on a 10-cm plate is ∼2,000,000, we estimated the newly established cell line, named as HCA2-HR, to be at approximately PD39 [PD26 + log2(2,000,000/250)]. We electroporated 0.1 μg of linearized HR reporter cassette into HCA2 cells. Our previous analysis of integrated clones indicated that transfection with 0.1 μg of plasmid results in the integration of a single copy of the reporter per genome (18). Three separate transfections were performed, giving rise to three independent pools of integrants. HCA2-HR cell lines were serially passaged until they reached senescence at PD71, consistent with the PD at which its parental cell line, HCA2, becomes senescent, indicating that the calculation of the starting PD for HCA2-HR lines was correct. The same protocol was followed for construction of HRF and HRIF cell lines.
Fig. 1.
HR declines with replicative senescence. (A) Reporter cassette for detection of HR. The construct contains two nonfunctional copies of the GFP gene. GFP gene is engineered to contain Pem1 intron. The GFP activity of the first copy is eliminated by a 22-nt deletion and insertion of two I-SceI recognition sites in the first exon. The second GFP gene lacks ATG start codon and the second exon. After induction of DNA DSBs by expressing I-SceI, gene conversion events reconstitute functional GFP gene. (B) Growth curve of HCA2-HR cell line. Gray bars show replicative ages at which young (Y), middle-aged (MA), and presenescent (Pre) cells were used for HR analysis. (C and D) Efficiency of HR decreases with replicative age. Western blots for I-SceI expression using antibodies against HA tag are shown (C and D, Upper). In C, cells HCA2-HR cells at all ages were cotransfected with 5 μg of I-SceI vector and 0.1 μg of pDsRed2-N1. In D, cells were transfected with increasing amounts of I-SceI vector (2 μg for PD45, 4 μg for PD53, 6 μg for PD62) to compensate for somewhat lower expression of I-SceI in older cells. Four days after transfection, cells were analyzed by FACS, and the GFP+/DsRed+ ratio was used as the measure of HR efficiency. HR efficiency in middle-aged and presenescent cells was significantly different from young cells. *P < 0.001. (E) Cutting efficiency of I-SceI does not decline in aged cells. Cells transfected with I-SceI vector as described in D were analyzed by real-time PCR using primers that amplify across the I-SceI recognition site. I-SceI transfection reduces the number of PCR templates. Cutting efficiency was calculated from a ratio between intact reporter cassettes in control and I-SceI–transfected cells. All experiments were repeated at least three times (nine times, three repeats for each pool of integrants) (C and D), and error bars show SD.

HR Decreases Sharply with Replicative Age.

We then measured the efficiency of HR mediated repair in HCA2-HR cells at PD42–45 (young), PD52–53 (middle-aged), and PD60–62 (presenescent) (Fig. 1B). Importantly, the presenescent cells are still in the logarithmic growth phase and, hence, have a similar fraction of actively dividing cells as the young and middle-aged cells. We did not include terminally senescent PD > 71 cells in this analysis, because these cells are cell cycle–arrested and would not be expected to repair DSBs by the HR pathway. HCA2-HR cells were transfected with 5 μg of I-SceI for inducing DSBs and 0.1 μg of pDsRed2-N1 as an internal control for transfection efficiency. Cells were incubated for 4 d to allow sufficient time for GFP and DsRed expression and analyzed by FACS. The ratio of GFP+ to DsRed+ cells was used as a measure of HR efficiency. HR efficiency decreased dramatically with replicative age (Fig. 1C). HR efficiency in middle-aged cells was reduced 5-fold compared with young cells, whereas the repair in presenescent cells was reduced 38-fold compared with young cells. Although HR efficiency was normalized by internal control pDsRed-N2, the expression level of I-SceI was somewhat lower in older cells (Fig. 1C). To compensate for the lower expression of I-SceI, the amount of pCMV-I–SceI plasmid was increased in older cells (Fig. 1D). Under these conditions, the protein levels of I-SceI were much higher in older cells, and HR efficiency showed a similar reduction with age (Fig. 1D). To rule out the possibility that the cutting efficiency of I-SceI was reduced in older cells because of potential changes in chromatin structure, we measured the reduction in the number of intact I-SceI sites following I-SceI transfection using quantitative real-time PCR. The primers were designed flanking the I-SceI sites, so that the loci that were cut by I-SceI would not yield a PCR product. Upon transfection with I-SceI as described in Fig. 1D, genomic DNA was extracted and the levels of intact I-SceI sites were analyzed (Fig. 1E). The percentage of I-SceI–digested reporter cassettes was higher in older cells, indicating that reduced HR efficiency was not attributable to inefficient induction of DSBs. These results suggest that the decline of HR efficiency is caused by intrinsic changes associated with replicative aging, and not by inefficient induction of DSBs in older cells.

Decline of HR Efficiency with Replicative Age Is Not Attributable to Decreased Number of Cells in S/G2/M.

The repair of DNA DSBs by HR is dependent on the cell cycle because HR requires sister chromatids, which are only available during the S/G2/M phases of the cell cycle (10, 19). To examine whether the reduced HR efficiency is caused by reduced number of cells in S/G2/M phase, we examined cell cycle distribution of HCA2-HR cells by propidium iodide (PI) staining and flow cytometry. The mean percentage of cells in S/G2/M stages was 33 ± 2.2% at PD41, 29 ± 6.1% at PD51, and 31 ± 3.2% at PD60 (Fig. 2). These fluctuations in S/G2/M content cannot account for the drop in HR efficiency and suggest that the changes in cell cycle distribution do not explain the observed age-related decline in HR efficiency.
Fig. 2.
Changes in cell cycle distribution with replicative senescence. HCA2-HR cells were harvested at indicated PDs 2 d postsplitting, stained with PI, and analyzed by Flow cytometry. The experiment was repeated three times, and the numbers next to the peaks show means and SD.

Levels of HR Proteins Decline with Cellular Senescence.

To understand the mechanisms responsible for the senescence-related decline of HR repair, we examined the levels of HR proteins at different passages (Fig. 3). Rad51 is a central protein in HR, and its levels decreased strongly with increasing cellular age. Furthermore, Rad51C, Rad52, and SIRT6 were reduced in older cells. The levels of MRE11 and Rad50 remained unchanged, whereas NBS1, which is involved in end resection, was reduced in senescing cells. CtIP also declined with replicative senescence. In contrast, the levels of PARP1 protein, which is involved in the early response to DNA damage, increased with replicative age, which suggests that the reduction in the levels of HR proteins contributes to the decline of HR repair efficiency.
Fig. 3.
Senescence-related changes in the levels of proteins involved in HR. Cell lysates were prepared from HCA2-HR cells 2 d postsplitting. In each lane, 50 μg of protein was loaded for Western analysis. Actin was used as a loading control. Quantification of the changes in middle-aged and presenescent cells relative to the young cells is shown on the right. Each blot was repeated three times using independent batches of cells, and SDs are shown.

Overexpression of SIRT6 Rescues HR Repair in Aging Cells.

To test whether HR efficiency in aging cells can be rescued by supplementing DNA repair factors, we transfected HCA2-HR cells with expression vectors encoding the HR factors that exhibited a decline with replicative aging. The ORFs for Rad51, Rad51C, Rad52, NBS1, SIRT6, and CtIP (a kind gift from Junjie Chen, MD Anderson Cancer Center, Houston, TX) were cloned under a CMV promoter and expressed in HCA2-HR cells (Fig. 4A). Surprisingly, when Rad51, Rad51C, Rad52, and NBS1 were coexpressed together, HR was repressed at all PDs compared with control cells that overexpressed a hypoxanthine phosphoribosyltransferase 1 (HPRT) gene (Fig. 4B). We then examined the effect of individual proteins on HR. Supplementing Rad51, Rad51C, Rad52, or NBS1 individually showed that Rad52 suppressed HR, Rad51C and NBS1 had no effect, and Rad51 stimulated HR by 1.6-fold in middle-aged cells but not in presenescent cells (Fig. 4B). When we excluded Rad52, a combination of Rad51, Rad51C, and NBS1 stimulated HR by about 1.8-fold in middle-aged cells and had no effect on presenescent cells, which was similar to the effect of Rad51 alone.
Fig. 4.
SIRT6 rescues the decline of HR efficiency in senescent cells. (A) Western blot showing overexpression of HR-related proteins in HCA2-HR cells. (B) Effect of overexpression of Rad51, Rad51C, Rad52, and NBS1 on HR at different PDs. Mix is a mixture of Rad51, Rad51C, Rad52, and NBS1 expression vectors. For expression of individual HR components, 5 μg of the expression vectors were cotransfected with 0.1 μg of pDsRed2-N1 and I-SceI expression vector into HCA2-HR cells at different PDs. Older cells were transfected with higher amounts of I-SceI expression vector, 2 μg for PD45, 4 μg for PD53, and 6 μg for PD62. For the experiments involving a mixture of HR components, 1.25 μg of each expression vector was cotransfected with DsRed2-N1 and I-SceI expression vectors as described above. Analysis of HR was performed as described in Fig. 1. (C) SIRT6 rescues HR in senescent cells, and this effect is mediated by PARP1. HCA2-HR cells were cotransfected with 5 μg of expression vectors encoding SIRT6, PARP1, or CtIP was cotransfected into HCA2-HR cells together 0.1 μg of DsRed plasmid and increasing amounts of I-SceI vector as described in B. PARP1 inhibitor PJ34 was supplemented at 20 μM starting 1 d before transfection until the cells were harvested for analysis. All of the experiments were repeated at least three times and error bars show SD. Asterisk indicates HR efficiencies significantly different from corresponding controls (P < 0.01); two asterisks indicate HR values significantly different from cells transfected with SIRT6 vector alone (P < 0.01). (D) SIRT6 overexpression reduces the number of γH2AX foci in presenescent cells. (Left) Quantification. (Right) Representative images. The experiment was repeated three times, and error pars show SD. (E) Mono-ADP ribosylation activity but not deacetylation activity of SIRT6 is required to stimulate HR in presenescent cells. HCA2 cells were transfected with plasmids encoding SIRT6 separation of function mutants, and the HR assay was performed as described above. The experiment was repeated three times, and error pars show SD. (F) PJ-34 does not inhibit SIRT6 deacetylation activity. To measure deacetylation activity SIRT6 was overexpressed in HCA2 cells in the presence of 20 μM PJ34 or 5 mM nicotinamide, and the level of acetylated histone H3 was analyzed by immunoblot. (G) PJ-34 does not inhibit SIRT6 mono-ADP ribosylation activity. Purified SIRT6 was incubated with radiolabeled NAD, and the level of labeled SIRT6 protein detected by autoradiography. NAM, nicotinamide, inhibitor of SIRT6 activity. Mono-ADP ribosylation activity of SIRT6 was measured as self-ribosylation reaction in the presence of radiolabeled NAD+. (H) SIRT6 overexpression restores the ability of presenescent cells to recruit Rad51 to DNA breaks. HCA2 cells at PD62 were irradiated with 8 Gy of γ-irradiation and hybridized in situ with Rad51 antibodies. The experiments were repeated three times, and error bars show SDs.
Remarkably, expression of SIRT6 strongly stimulated HR at all PDs tested, including the presenescent cells (Fig. 4C). SIRT6 stimulated HR by 1.4-fold in young cells, 1.7-fold in middle-aged cells, and 3.9-fold in presenescent cells. Thus, SIRT6 was the only factor capable of stimulating HR repair in presenescent cells. To confirm by an independent assay that SIRT6 overexpression rescues the effects of senescence on DSB repair, we examined the levels of spontaneously occurring γH2AX foci in presenescent cells. SIRT6 reduced the number of γH2AX foci (Fig. 4D). SIRT6 expression did not change cell cycle distribution (Fig. S1), ruling out a possibility that SIRT6 stimulates HR in older cells by promoting cell proliferation. Furthermore, SIRT6 overexpression did not alter the levels of HR proteins (Fig. S2), suggesting that the effect of SIRT6 is mediated by protein–protein interactions at the break site.
We then tested which biochemical activity of SIRT6 is required for the stimulation of HR in presenescent cells. To this end, we transfected presenescent HCA2-HR cells with SIRT6 separation of function mutants G60A, R65A, and S56Y. The G60A has deacetylation activity, R65A has mono-ADP ribosylation activity, and S56Y is catalytically inactive (11). Mono-ADP ribosylation was the only activity required for the stimulation of HR in presenescent cells (Fig. 4E).
Because SIRT6 can stimulate HR by two pathways, via interaction with CtIP or PARP1 (11), we next examined which pathway is responsible for the stimulation of HR in aging cells. Overexpression of CtIP had no effect on HR (Fig. 4C). We were unable to assay the effect of PARP1 overexpression on HR because it caused massive cell death. Instead, we used the specific PARP1 inhibitor PJ34. Importantly, PJ34 had no inhibitory effect on either deacetylation or mono-ADP ribosylation activities of SIRT6 (Fig. 4 F and G). In the presence of PJ34, SIRT6 overexpression had no effect on HR (Fig. 4C), indicating that SIRT6-mediated rescue of HR in aging cells is dependent on PARP1.
To gain an insight into the mechanism by which SIRT6 rescues the decline of HR in presenescent cells, we examined the formation of RAD51 foci after γ-irradiation. RAD51 is the key protein in HR and is often a rate-limiting factor. Interestingly, presenescent cells exhibited a drastic decline in the ability to recruit RAD51 to DNA damage sites (Fig. 4H). This decline was fully rescued by SIRT6 overexpression, indicating that SIRT6 promotes recruitment of DNA repair proteins to DNA lesions (Fig. 4H).

SIRT6 Preferentially Stimulates the Precise HR Pathway.

In addition to gene conversion and crossing over, which are precise pathways of HR repair, HR may proceed by a single-stranded annealing (SSA) pathway, leading to deletion of sequences between two direct repeats. The HR reporter we used (Fig. 1A) measured gene conversion, which is the predominant HR repair pathway in mammalian cells (8). To test the effect of SIRT6 overexpression on the more mutagenic SSA pathway, we used two modified reporters: HRF and HRIF (Fig. 5 A and B). HRF detects all three pathways of HR (gene conversion, crossing over, and SSA), whereas HRIF detects gene conversion and crossing over. The constructs were integrated in HCA2 cells as described for HCA2-HR, and the frequency of repair by SSA was measured as HRF minus HRIF. To validate this method of measuring SSA, we tested whether this parameter is stimulated by Rad52, a known mediator of SSA. As expected, Rad52 overexpression stimulated SSA (Fig. 5C). The efficiency of SSA declined with replicative age (Fig. 5D). Interestingly, SIRT6 overexpression did not stimulate SSA (Fig. 5D), indicating that SIRT6 preferentially stimulates the precise pathway of HR repair. In summary, our findings open avenues for preventing age-related genomic instability and decline in DSB repair by activating SIRT6.
Fig. 5.
SIRT6 does not stimulate SSA pathway. (A) Reporter cassette HRF for detection of gene conversion, crossing over, and SSA. The cassette consists of two duplicated copies of GFP-Pem1 as described in Fig. 1A, but the second copy contains full-length GFP-Pem1. (B) Reporter cassette HRIF for detection of gene conversion and crossing over. This cassette differs from HRF in that the second copy of GFP-Pem1 is placed in front of the copy containing I-SceI sites, disabling SSA. (C) Rad52 overexpression stimulates SSA. HRF and HRIF were integrated in HCA2 cells. SSA efficiency was calculated as HRF minus HRIF. To validate this method of measuring SSA, 5 μg of the Rad52 expression vector were cotransfected with 0.1 μg of pDsRed2-N1 and I-SceI expression vector into reporter cells at PD52. (D) SIRT6 overexpression does not stimulate SSA. Cells were cotransfected with 5 μg of I-SceI vector and 0.1 μg of pDsRed2-N1. Four days after transfection cells were analyzed by FACS, and the GFP+/DsRed+ ratio was used as the measure of HR efficiency. All experiments were repeated at least three times and error bars show SDs.

Discussion

We present a systematic analysis of the changes in HR repair during cellular senescence. We used a sensitive fluorescent reporter assay developed by our laboratory (18), which specifically measures repair events mediated by the HR pathway. A challenge of analyzing HR repair in the context of replicative senescence is that reporter constructs must be integrated in primary human cells, and by the time integrants are selected, cells exhaust their replicative potential. We overcame this problem by pooling integrating clones at low PD numbers. Another advantage of this approach is that clones with multiple integration sites are combined, removing a potential bias attributable to position effect.
HR efficiency declined sharply with replicative age and was almost completely abolished at PDs >60. The reduction in HR repair was not explained by cell cycle arrest in the older cells because the presenescent cells were actively dividing and had a similar ratio of S/G2/M compared with the younger cells. Thus, cellular aging is intrinsically associated with decline in DNA-repair capacity.
Our earlier work demonstrated that DNA repair by NHEJ becomes less efficient and more error-prone with replicative senescence (16). NHEJ in presenescent cells resulted in large deletions. The reduction in NHEJ efficiency was approximately threefold between young and presenescent cells (16), which is a much smaller change compared with 38-fold reduction in the efficiency of HR repair. Taken together, these observations paint the following picture of the DSB repair in aging cells. The precise HR repair is strongly repressed, leaving the DSBs to be repaired by an error-prone NHEJ pathway. The NHEJ pathway itself becomes less efficient and prone to larger deletions, leading to mutations and the loss of genetic information. Decline in DSB repair capacity in senescing cells may be directly relevant to human aging. Aging tissues accumulate senescent and presenescent cells (2026); thus, the landscape of DNA repair may change with age. Inefficient and error-prone repair leads to a higher mutation rate, dysregulation of transcription patterns, and, finally, loss of cell and tissue function. Furthermore, the shift to the error-prone NHEJ pathway of DSB repair is likely to lead to generation of large deletions, joining of inappropriate DNA ends, and chromosomal rearrangements. We hypothesize that these changes in DSB processing are a possible mechanism responsible for age-related genomic instability and higher incidence of cancer in the elderly.
Studies of DSB processing in the context of organismal aging indicate a decline in repair. Sedelnikova et al. reported that the rate of recruitment of DSB repair proteins to γH2AX foci were slower in older donors; furthermore, the frequency of spontaneous γH2AX foci increased with age (13). Earlier studies also observed that cells from older donors contain a higher number of DSBs (27), and analysis of X-ray–induced DSBs in lymphocytes showed that these cells exhibit an age-associated decline in repair efficiency (14, 28). Thus, our studies provide mechanistic insight into the observed decline in DSB repair, showing that HR repair is sharply reduced in older cells.
Interestingly, the frequency of spontaneous mitotic recombination was reported to increase in the pancreas, but not in the skin, of mice carrying a fluorescent reporter cassette (29). Spontaneous mitotic recombination is a rare event likely triggered by a spontaneous DSB. We hypothesize that as the frequency of DSBs increases with age the spontaneous HR events are recorded more frequently despite an overall decline in HR repair. Similarly, in Drosophila, the frequency of ectopic recombination increases with age (30), which may reflect a decrease in accurate intersister gene conversion events, the type of repair measured by our assay.
We found that the levels of several HR proteins decline with replicative age. Rad51 showed the strongest decline with replicative age, and supplementing it rescued repair in middle-aged cells but failed to do so in presenescent cells, indicating that past a certain age, additional Rad51 is no longer capable of helping repair. Rad52 was inhibitory, which likely reflects the role of Rad52 in the error-prone SSA pathway (31). We hypothesize that increased levels of Rad52 channeled repair toward SSA. Supplementing Rad51, Rad51C, and NBS1, either alone or in combination, did not lead to a rescue of the HR defect in presenescent cells. Why did they fail to rescue HR repair? All of these proteins could be classified as enzymes directly involved in the HR reaction. Because these proteins work in a complex, it may be difficult to obtain ideal ratios when overexpressing them. Furthermore, supplementation of additional components of the HR complex may be required. A better solution than supplementing downstream factors would be altering an upstream regulator of HR. Indeed, overexpressing SIRT6 stimulated HR in both middle-aged and presenescent cells. This was the only treatment to work in presenescent cells. Furthermore, we showed that stimulation of DSB repair in aged cells depends on mono-ADP ribosylation activity of SIRT6 and requires PARP1. We propose that SIRT6 serves as an upstream regulator of HR repair, by controlling PARP1 via mono-ADP ribosylation (32), and modulating SIRT6 activity can overcome the senescence-related decline in multiple downstream repair factors. Our results suggest that SIRT6 promotes recruitment of DNA-repair proteins to DNA lesions. We show here that SIRT6 stimulates recruitment of Rad51, and our earlier report (11) showed that SIRT6 accelerates recruitment of NBS1. In addition, SIRT6 may regulate early steps in the HR pathway that involve strand resection because it also stimulates alternative NHEJ (11) and strand resection mediated by CtIP (12). Interestingly, SIRT6 did not stimulate the error-prone SSA pathway, suggesting that it may bias DSB repair toward less mutagenic, precise gene conversion events.
The accumulation of genomic rearrangements is a hallmark of aging and contributes to the loss of tissue functionality and the increased incidence of tumorigenesis that characterize the aging process. Here, we demonstrated that the efficiency of HR repair declines precipitously during cellular aging and that overexpression of the DNA repair factor, SIRT6, was uniquely able to rescue this decline. It was recently reported that SIRT6 overexpression extends lifespan in mice (33); it is possible that protection against genomic instability contributes to this phenotype. Other members of the sirtuin gene family have been shown to be amenable to pharmacological stimulation (34, 35), and recent reports have suggested that SIRT6 may be an intriguing candidate for activation to ameliorate the effect of multiple age-related pathologies. In addition to lifespan extension, a separate study indicated that SIRT6 overexpression protects against diet–induced obesity (36). Additionally, we have observed that SIRT6 overexpression is selectively cytotoxic to multiple cancer cell lines (37), and several studies have suggested that SIRT6 may function as an agonist of NF-κB–induced inflammation (38, 39). Our findings with regard to DNA repair lend credence to the strategy of activating SIRT6 to alleviate age-related disease and suggest that constitutive SIRT6 activation may even delay the onset of age-related pathologies.

Methods

Cell Culture.

HCA2 cells are primary human foreskin fibroblasts. All cells were grown in a 5% CO2, 3% O2 incubator at 37 °C, in Eagle's minimal essential medium supplemented with 15% FBS, 100 units/mL penicillin, and 100 μg/mL streptomycin.

Construction of HCA2-HR Cell Lines.

HCA2 cells at PD24, which is the lowest PD available in our laboratory stock, were transfected with 0.1 μg of HR construct linearized by NheI. Cells were kept in media with 1 mg/mL G418 for 8 d to select colonies with integrated reporter cassettes. After colonies were formed, cells were trypsinized, reseeded, and cultured in one plate until they reached confluence. Three separate transfections were performed, giving rise to three independent pools. These cells lines were named HCA2-HR. HCA2-HR cell lines were then serially passaged, with splits every 3–4 d, until they reached senescence at PD71.
Methods for quantitative PCR, plasmids, antibodies, transfections, FACS analyses, measurements of SIRT6 biochemical activities, and immunofluorescence are provided in SI Methods.

Acknowledgments

We thank Junjie Chen for kindly providing us with CtIP plasmid. This work was supported by grants from the National Institutes of Health (to V.G.) and the Ellison Medical Foundation (to V.G. and A.S.).

Supporting Information

Supporting Information (PDF)
Supporting Information

References

1
GR Stuart, Y Oda, JG de Boer, BW Glickman, Mutation frequency and specificity with age in liver, bladder and brain of lacI transgenic mice. Genetics 154, 1291–1300 (2000).
2
J Vijg, MET Dollé, Large genome rearrangements as a primary cause of aging. Mech Ageing Dev 123, 907–915 (2002).
3
JE Haber, Partners and pathwaysrepairing a double-strand break. Trends Genet 16, 259–264 (2000).
4
MR Lieber, The mechanism of human nonhomologous DNA end joining. J Biol Chem 283, 1–5 (2008).
5
LH Thompson, D Schild, Homologous recombinational repair of DNA ensures mammalian chromosome stability. Mutat Res 477, 131–153 (2001).
6
Z Mao, M Bozzella, A Seluanov, V Gorbunova, Comparison of nonhomologous end joining and homologous recombination in human cells. DNA Repair (Amst) 7, 1765–1771 (2008).
7
AA Sartori, et al., Human CtIP promotes DNA end resection. Nature 450, 509–514 (2007).
8
RD Johnson, M Jasin, Sister chromatid gene conversion is a prominent double-strand break repair pathway in mammalian cells. EMBO J 19, 3398–3407 (2000).
9
P Sung, H Klein, Mechanism of homologous recombination: Mediators and helicases take on regulatory functions. Nat Rev Mol Cell Biol 7, 739–750 (2006).
10
Z Mao, M Bozzella, A Seluanov, V Gorbunova, DNA repair by nonhomologous end joining and homologous recombination during cell cycle in human cells. Cell Cycle 7, 2902–2906 (2008).
11
Z Mao, et al., SIRT6 promotes DNA repair under stress by activating PARP1. Science 332, 1443–1446 (2011).
12
A Kaidi, BT Weinert, C Choudhary, SP Jackson, Human SIRT6 promotes DNA end resection through CtIP deacetylation. Science 329, 1348–1353 (2010).
13
OA Sedelnikova, et al., Delayed kinetics of DNA double-strand break processing in normal and pathological aging. Aging Cell 7, 89–100 (2008).
14
NP Singh, DB Danner, RR Tice, L Brant, EL Schneider, DNA damage and repair with age in individual human lymphocytes. Mutat Res 237, 123–130 (1990).
15
NP Singh, CH Muller, RE Berger, Effects of age on DNA double-strand breaks and apoptosis in human sperm. Fertil Steril 80, 1420–1430 (2003).
16
A Seluanov, D Mittelman, OM Pereira-Smith, JH Wilson, V Gorbunova, DNA end joining becomes less efficient and more error-prone during cellular senescence. Proc Natl Acad Sci USA 101, 7624–7629 (2004).
17
A Seluanov, J Danek, N Hause, V Gorbunova, Changes in the level and distribution of Ku proteins during cellular senescence. DNA Repair (Amst) 6, 1740–1748 (2007).
18
Z Mao, A Seluanov, Y Jiang, V Gorbunova, TRF2 is required for repair of nontelomeric DNA double-strand breaks by homologous recombination. Proc Natl Acad Sci USA 104, 13068–13073 (2007).
19
N Saleh-Gohari, T Helleday, Conservative homologous recombination preferentially repairs DNA double-strand breaks in the S phase of the cell cycle in human cells. Nucleic Acids Res 32, 3683–3688 (2004).
20
U Herbig, M Ferreira, L Condel, D Carey, JM Sedivy, Cellular senescence in aging primates. Science 311, 1257 (2006).
21
JC Jeyapalan, M Ferreira, JM Sedivy, U Herbig, Accumulation of senescent cells in mitotic tissue of aging primates. Mech Ageing Dev 128, 36–44 (2007).
22
GP Dimri, et al., A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci USA 92, 9363–9367 (1995).
23
WR Pendergrass, et al., Cellular proliferation potential during aging and caloric restriction in rhesus monkeys (Macaca mulatta). J Cell Physiol 180, 123–130 (1999).
24
J Choi, et al., Expression of senescence-associated beta-galactosidase in enlarged prostates from men with benign prostatic hyperplasia. Urology 56, 160–166 (2000).
25
V Paradis, et al., Replicative senescence in normal liver, chronic hepatitis C, and hepatocellular carcinomas. Hum Pathol 32, 327–332 (2001).
26
G Ding, et al., Tubular cell senescence and expression of TGF-beta1 and p21(WAF1/CIP1) in tubulointerstitial fibrosis of aging rats. Exp Mol Pathol 70, 43–53 (2001).
27
NP Singh, CE Ogburn, NS Wolf, G van Belle, GM Martin, DNA double-strand breaks in mouse kidney cells with age. Biogerontology 2, 261–270 (2001).
28
PJ Mayer, CS Lange, MO Bradley, WW Nichols, Age-dependent decline in rejoining of X-ray-induced DNA double-strand breaks in normal human lymphocytes. Mutat Res 219, 95–100 (1989).
29
DM Wiktor-Brown, W Olipitz, CA Hendricks, RE Rugo, BP Engelward, Tissue-specific differences in the accumulation of sequence rearrangements with age. DNA Repair (Amst) 7, 694–703 (2008).
30
CR Preston, C Flores, WR Engels, Age-dependent usage of double-strand-break repair pathways. Curr Biol 16, 2009–2015 (2006).
31
E Van Dyck, AZ Stasiak, A Stasiak, SC West, Visualization of recombination intermediates produced by RAD52-mediated single-strand annealing. EMBO Rep 2, 905–909 (2001).
32
M Van Meter, Z Mao, V Gorbunova, A Seluanov, Repairing split ends: SIRT6, mono-ADP ribosylation and DNA repair. Aging (Albany NY) 3, 829–835 (2011).
33
Y Kanfi, et al., The sirtuin SIRT6 regulates lifespan in male mice. Nature 483, 218–221 (2012).
34
JJ Smith, et al., Small molecule activators of SIRT1 replicate signaling pathways triggered by calorie restriction in vivo. BMC Syst Biol 3, 31 (2009).
35
JC Milne, et al., Small molecule activators of SIRT1 as therapeutics for the treatment of type 2 diabetes. Nature 450, 712–716 (2007).
36
Y Kanfi, et al., SIRT6 protects against pathological damage caused by diet-induced obesity. Aging Cell 9, 162–173 (2010).
37
M Van Meter, Z Mao, V Gorbunova, A Seluanov, SIRT6 overexpression induces massive apoptosis in cancer cells but not in normal cells. Cell Cycle 10, 3153–3158 (2011).
38
TL Kawahara, et al., SIRT6 links histone H3 lysine 9 deacetylation to NF-kappaB-dependent gene expression and organismal life span. Cell 136, 62–74 (2009).
39
TL Kawahara, et al., Dynamic chromatin localization of Sirt6 shapes stress- and aging-related transcriptional networks. PLoS Genet 7, e1002153 (2011).

Information & Authors

Information

Published in

The cover image for PNAS Vol.109; No.29
Proceedings of the National Academy of Sciences
Vol. 109 | No. 29
July 17, 2012
PubMed: 22753495

Classifications

Submission history

Published online: July 2, 2012
Published in issue: July 17, 2012

Acknowledgments

We thank Junjie Chen for kindly providing us with CtIP plasmid. This work was supported by grants from the National Institutes of Health (to V.G.) and the Ellison Medical Foundation (to V.G. and A.S.).

Notes

This article is a PNAS Direct Submission.

Authors

Affiliations

Zhiyong Mao
Department of Biology, University of Rochester, Rochester, NY 14627
Xiao Tian
Department of Biology, University of Rochester, Rochester, NY 14627
Michael Van Meter
Department of Biology, University of Rochester, Rochester, NY 14627
Zhonghe Ke
Department of Biology, University of Rochester, Rochester, NY 14627
Vera Gorbunova1 [email protected]
Department of Biology, University of Rochester, Rochester, NY 14627
Andrei Seluanov1 [email protected]
Department of Biology, University of Rochester, Rochester, NY 14627

Notes

1
To whom correspondence may be addressed. E-mail: [email protected] or [email protected].
Author contributions: Z.M., X.T., V.G., and A.S. designed research; Z.M., X.T., M.V.M., and Z.K. performed research; Z.M., X.T., M.V.M., V.G., and A.S. analyzed data; and Z.M., M.V.M., V.G., and A.S. wrote the paper.

Competing Interests

The authors declare no conflict of interest.

Metrics & Citations

Metrics

Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.


Citation statements




Altmetrics

Citations

Export the article citation data by selecting a format from the list below and clicking Export.

Cited by

    Loading...

    View Options

    View options

    PDF format

    Download this article as a PDF file

    DOWNLOAD PDF

    Login options

    Check if you have access through your login credentials or your institution to get full access on this article.

    Personal login Institutional Login

    Recommend to a librarian

    Recommend PNAS to a Librarian

    Purchase options

    Purchase this article to access the full text.

    Single Article Purchase

    Sirtuin 6 (SIRT6) rescues the decline of homologous recombination repair during replicative senescence
    Proceedings of the National Academy of Sciences
    • Vol. 109
    • No. 29
    • pp. 11467-11890

    Figures

    Tables

    Media

    Share

    Share

    Share article link

    Share on social media