Tier 3 — preclinical

SIRT6 is a DNA double-strand break sensor

Onn L, Portillo M, Ilic S, Cleitman G, Stein D, Kaluski S, Shirat I, Slobodnik Z, Einav M, Erdel F, Akabayov B, Toiber D
eLife 2020 Volume 9, article e51636

Bibliography

DOI
10.7554/eLife.51636
PubMed
PMID 31995034
PubMed Central
PMC7051178
Funding
European Research Council (ERC2014-STG-638801-TEMPOTOIBER to DT); Israel Science Foundation (individual grant 188/17 to DT); DFG SFB1064 (to FE).
Competing interests
The authors declare that no competing interests exist.

Study snapshot

DesignCell-culture imaging with UV laser-induced DNA damage, in vitro gel retardation and plasmid-binding assays with purified SIRT6, LacO/LacR tethering assay for co-localisation studies, site-directed mutagenesis of predicted DNA-binding residues, and comparative sirtuin family analysis.
ModelHuman U2OS and HeLa cell lines (SIRT6 WT and knockout), purified recombinant human SIRT6-His and SIRT6-Flag proteins, LacO-tagged proteins expressed in U2OS reporter cells, all seven mammalian sirtuins (SIRT1–SIRT7) tested in parallel.
SampleLive imaging n=17–52 cells per condition across three replicate experiments; LacO co-localisation n=40–150 cells per condition across 3–9 replicate experiments; DNA-binding assays n=3–4 biological replicates.
InterventionUV laser-induced DNA damage (405 nm laser); PARP inhibition with olaparib; shRNA knockdown of Ku80 and MRE11; SIRT6 catalytically-dead mutant (H133Y) expression; SIRT6 tunnel-mutant expression (A13W, D63H/Y, W188A, D190W, I219A).
DurationLive imaging from 3 s to 60 s post-damage; fixed imaging at various time points from 15 min to 24 h.
EndpointsTime-resolved SIRT6 recruitment to laser-induced damage sites; SIRT6 DNA-binding affinity (Kd) for ssDNA and sticky-ended dsDNA; Co-localisation of SIRT6 with H2AX, ATM, MRE11, Ku70, Ku80, 53BP1, NBS1, BRCA1, XRCC4 at LacO-tethered sites; DSB signalling activation independent of enzymatic activity

What the study showed, in plain terms

Cells have to detect DNA damage before they can repair it. Three proteins were previously known to act as detectors of double-strand breaks — the most dangerous form of DNA damage — in human cells. This paper adds a fourth. SIRT6, previously known as a longevity enzyme, is shown to arrive at broken DNA within seconds, independently of the three other detectors, and to trigger the cellular damage-response cascade before the cell has even chosen which repair pathway to use.

The Toiber lab at Ben-Gurion University used two experimental strategies. First, they filmed SIRT6 arriving at UV-laser-induced DNA breaks in live human cells. When they blocked the other three detectors, SIRT6 still arrived on schedule. Second, they purified SIRT6 in a test tube and measured how tightly it stuck to different DNA shapes. SIRT6 bound especially well to broken DNA ends with single-strand overhangs, a hallmark of double-strand breaks.

The structural explanation is a narrow tunnel in the SIRT6 protein that fits over the exposed end of one broken DNA strand. When both strands break at once, two SIRT6 molecules cap the two ends and pair up — protecting the ends and signalling for repair. Mutating any of six specific amino acids inside this tunnel weakened DNA binding and DDR activation.

The finding matters for the field for two reasons. It places SIRT6 upstream of the decision between homologous recombination and non-homologous end joining, meaning SIRT6 activity affects both pathways rather than just one. And it shows that this sensor function does not require SIRT6's enzymatic activity — a catalytically dead SIRT6 mutant still binds DNA and activates the damage response. Any small-molecule activator that only tweaks catalytic activity will therefore not enhance this arm of SIRT6's function.

Key findings

  • SIRT6 arrives at UV-laser-induced DNA damage within seconds and reaches maximum accumulation between 15 and 60 seconds, comparable to the recognised early sensors Ku80 and MRE11.
  • SIRT6 recruitment is independent of PARP1 signalling (olaparib does not block it), of Ku80 (shRNA depletion does not block it), and of MRE11 (shRNA depletion does not block it). SIRT6 arrives even when all three canonical DSB sensors are inactivated.
  • Purified SIRT6 binds single-strand DNA with a Kd of 1.48 μM and sticky-ended double-strand DNA with a Kd of 3.59 μM. Binding to circular unbroken DNA is negligible, confirming specificity for broken ends.
  • SIRT6 tethered to a LacO array via LacR fusion is sufficient to recruit ATM, drive H2AX phosphorylation, and bring in NHEJ factors (Ku70, Ku80, 53BP1, XRCC4) and HR factors (MRE11, NBS1, BRCA1) — demonstrating full DDR activation without actual DNA damage.
  • The catalytically inactive SIRT6-H133Y mutant retains full DNA-binding and DDR-activation capacity. Enzymatic activity is not required for the sensor function — this arm of SIRT6 biology is independent of deacetylation and mono-ADP-ribosylation.
  • Six amino acids (A13, D63, H133, W188, D190, I219) line a predicted tunnel-like structure in the SIRT6 core domain. Mutation of any of these residues reduces DNA binding by 30–60% and reduces damage-site recruitment.
  • Comparative analysis of all seven mammalian sirtuins found that SIRT6, SIRT2, and SIRT7 all bind DSBs and initiate H2AX phosphorylation at LacO sites, while SIRT1, SIRT3, and SIRT5 do not — suggesting a broader class-defining property of some sirtuins as DSB sensors.

What this study can and cannot tell us

All experiments are in cell culture (U2OS, HeLa, HEK293T) or on purified proteins in vitro. There is no animal work in this paper, no whole-organism validation, and no attempt to extend the sensor function to primary human cells or tissues from healthy donors. Whether the sensor role is equally important in post-mitotic cells like neurons, or in high-turnover tissues, is not addressed.

The paper does not measure SIRT6 sensor function under physiological damage rates — the assays use acute, high-dose UV laser damage, which is far above the endogenous rate of DNA damage in living tissue. Whether the sensor role dominates repair kinetics under normal conditions is inferred, not shown.

Small-molecule SIRT6 activators (including MDL-800, MDL-811, UBCS039) work by enhancing catalytic activity. This paper shows that the sensor function is independent of catalysis. As a result, activators aimed at enhancing SIRT6 deacetylation or ribosylation activity will not necessarily strengthen the sensor arm — a caveat with direct implications for activator development that the paper flags in the discussion.

The paper does not address whether SIRT6's sensor role changes with age — that is, whether age-related decline in SIRT6 function affects DNA binding and damage sensing, or only the downstream enzymatic activities. This is an important open question for translational relevance.

Related papers in the corpus

Tier 3
Sirtuin 6: linking longevity with genome and epigenome stability
Anatoly Korotkov et al. · Trends in Cell Biology · 2021
Tier 3
SIRT6 is Responsible for More Efficient DNA Double-Strand Break Repair in Long-Lived Species
Tian X et al. · Cell · 2019
Tier 3
SIRT6, a protein with many faces
Gertler AA et al. · Biogerontology · 2013
Tier 3
The sirtuin SIRT6 regulates lifespan in male mice
Kanfi Y et al. · Nature · 2012
Reviewed by , Medical Advisory Board · Last verified against PubMed on 08 August 2026