Tier 3 — preclinical

The sirtuin SIRT6 regulates lifespan in male mice

Kanfi Y, Naiman S, Amir G, Peshti V, Zinman G, Nahum L, Bar-Joseph Z, Cohen HY

Nature 2012 Volume 483, issue 7388, pages 218–221

Bibliography

DOI: 10.1038/nature10815
PubMed: PMID 22367546
Funding: US National Institutes of Health grant 1RO1 GM085022 (to Z.B.-J.). Grants from the Israeli Academy of Sciences, the United States–Israel Binational Science Foundation, the Israel Cancer Association, the Koret Foundation, the Israel Cancer Research Fund, the Israel Health Ministry, the I-CORE program (41/1), the Israel Science Foundation, and the European Research Council (to H.Y.C.).
Competing interests: The authors declared no competing financial interests.

Study snapshot

Design	Interventional lifespan study of two independent Sirt6-transgenic mouse lines with parallel molecular characterisation of the IGF1 signalling pathway in liver, white adipose tissue, and muscle.
Model	Sirt6-transgenic mice on a segregating CB6F1 background (equal contribution from C57BL/6J and BALB/cOlaHsd, both long-lived strains), two independent founder lines (line 55 and line 108), with wild-type littermate controls.
Sample	245 mice total (119 males, 126 females) across both lines. Molecular readouts (Western blots, quantitative PCR, serum ELISA) run in subgroups of 4–7 mice per genotype per sex.
Intervention	Whole-body constitutive Sirt6 transgene overexpression from birth (also referred to elsewhere as MOSES mice).
Duration	Lifelong follow-up until natural death. Molecular and glucose tolerance readouts sampled at 6 months and 19 months of age.
Endpoints	Median and mean lifespan by sex, line, and genotype; maximum lifespan (mean lifespan of the oldest 10%); Cox regression of mortality with genotype, sex, line, and parental identity as covariates; post-mortem tumour spectrum and non-neoplastic pathology; intraperitoneal glucose tolerance area under the curve at 19 months; serum IGF1 at 6 and 19 months; hepatic Igfbp1, Igfbp3, and Als expression; phosphorylation of IGF1R, AKT (Thr308 and Ser473), FOXO1 (Thr24), FOXO3 (Thr32), and S6 (Ser235/236) in perigonadal white adipose tissue

What the study showed, in plain terms

This paper answered a question that had been open for years: does any mammalian sirtuin actually extend lifespan? For sirtuins in yeast, worms, and flies, the answer was disputed. In mammals, SIRT1 overexpression had already been shown not to extend mouse lifespan. So the field was uncertain.

The authors made transgenic mice that carry an extra copy of the mouse Sirt6 gene, expressed throughout the body from birth. They generated two independent founder lines to rule out artefacts from where the transgene randomly landed in the genome, and they used a genetic background made from two long-lived mouse strains combined equally. Then they simply waited and recorded when each mouse died of natural causes.

Male transgenic mice lived roughly 15% longer at the median than their wild-type male littermates, in both lines. Female transgenic mice showed no lifespan difference. The male-only effect was consistent and statistically robust.

Looking at what was different in the transgenic males, the researchers found lower blood levels of insulin-like growth factor 1 (IGF1), higher levels of its binding protein IGFBP1, and reduced activation of the AKT and FOXO signalling pathway in fat tissue. IGF1 signalling is one of the most conserved lifespan-regulating pathways in animals, and reducing it in fat has been shown to extend lifespan on its own. The transgenic females did not show these signalling changes, matching their lack of a lifespan effect.

This was the first paper to show that a mammalian sirtuin can extend lifespan.

Key findings

In line 55, median lifespan of male Sirt6-transgenic mice was 14.5% longer than wild-type males (log-rank test χ² = 10.529, p = 0.001); mean lifespan was 14.8% longer.
In line 108, median lifespan of male transgenic mice was 9.9% longer than wild-type males (χ² = 4.225, p = 0.040); mean lifespan was 16.9% longer.
Female lifespan showed no significant difference between transgenic and wild-type mice in either line (line 55 p = 0.924; line 108 p = 0.319).
Maximum lifespan of transgenic males (mean of the oldest 10%) increased by 15.8% in line 55 and 13.1% in line 108.
Cox regression across all mice showed an additive effect of genotype and line with no line-by-genotype interaction (p = 0.693), indicating the transgene effect was consistent across both founder integration sites.
Serum IGF1 in young (6-month-old) transgenic males was significantly lower than in wild-type males and comparable to female levels; the difference was sustained at 19 months.
Hepatic Igfbp1 expression was upregulated approximately 5-fold in transgenic males, matching baseline female levels; Igfbp3 and Als expression did not change.
In perigonadal white adipose tissue of transgenic males, phosphorylation of IGF1R (Tyr1135), AKT (both Thr308 and Ser473), FOXO1 (Thr24), FOXO3 (Thr32), and S6 (Ser235/236) was reduced compared with wild-type males. No equivalent reduction was seen in females.
Old transgenic mice (19 months) showed a trend towards improved glucose tolerance versus wild-type mice of the same age (ANOVA for genotype p = 0.016 across both sexes).
Whole-genome liver microarray showed 82 significantly differentially expressed genes in transgenic males vs wild-type males, 50% of which overlapped with genes normally differentially expressed between male and female wild-type mice — consistent with a "feminising" transcriptional shift. Thirty per cent of the transgenic-male-differential genes were also known to shift with calorie restriction.

What this study can and cannot tell us

The lifespan effect is male-only. The authors do not resolve whether this reflects a female-specific block on SIRT6's downstream effects, a ceiling effect (females already living longer or with lower baseline IGF1 signalling), or a genuinely male-restricted mechanism. This limits how the finding translates to female biology.

Both transgenic mouse lines were on the same CB6F1 genetic background. Although the two independent founder lines rule out integration-site artefacts, the finding has not been reproduced across genetically diverse mouse populations (for example, the four-way UM-HET3 stock used by the Interventions Testing Program). The magnitude of the effect on any given inbred or outbred background is not established here.

Lung tumours were the most common finding on post-mortem across all groups. Sirt6-transgenic males with lung tumours had a trend toward longer survival than wild-type males with lung tumours, so a partial contribution from reduced lung-cancer lethality cannot be ruled out. The authors argue the tumour-frequency-adjusted effect is still real, but the study is not powered to fully separate cancer-related from non-cancer-related lifespan gains.

This is a genetic overexpression model. It does not test whether pharmacological SIRT6 activation in an already-adult mouse extends lifespan, and it says nothing about safety or efficacy in humans. No human dosing, supplement, or clinical outcome can be inferred from this paper.