Mitochondria as a common target of aging-induced NAD+ decline

It is now clear that aging-induced inactivation of SIRT1 has a direct and deleterious effect on mitochondria, as first suggested by the important associations between SIRT1 and PGC-1α 56 and SIRT1 and TFAM 43. A reduction in SIRT1 activity downregulates mitochondrial biogenesis, oxidative metabolism, and associated anti-oxidant defense pathways, leading to damage to complex I of the electron transport chain and a decline in mitochondrial function (Figure 4, right). A similar effect could result from the failure of SIRT1 to deacetylate another of its substrates, FOXO, which would lead to a reduction in mitochondrial anti-oxidant defenses in worms 57 and mammals 58.
Strikingly, other mechanisms have also been recently unveiled, which connect sirtuins to mitochondrial health. Inducing the activity of the worm SIR-2.1 or mammalian SIRT1 triggers the mitochondrial unfolded protein response (UPRmt) pathway, but not other protein quality control pathways, such as those affecting the endoplasmic reticulum 18. Indeed, genetic inactivation of the UPRmt pathway prevents the longevity induced by SIR-2.1 overexpression or by NR supplementation in worms. Recently, it has also been reported that SIRT3 regulates the UPRmt and mitophagy 59. Thus, clearance of damaged mitochondria may also be impaired by NAD+ deficiency.
Finally, a defect in expression of mitochondrial-encoded proteins in skeletal muscle of 24-month old mice (only at older ages was a reduction in nuclear encoded mitochondrial proteins also observed) was shown to lead to metabolic decline 43. Depressed mitochondrial gene expression and metabolic decline were due to a defect in SIRT1 activity and were reversed by supplementation with NMN. Thus, NAD+ deficiency again appears to be the primary trigger, in this case reducing mitochondrial gene expression. Surprisingly, this defect arising from SIRT1 inactivation was not related to PGC-1α or the UPRmt. Rather, SIRT1 deficiency prevented its known downregulation of HIF-1α, leading to an inappropriately high level of HIF-1α. This pseudohypoxic state led to sequestration of cMYC by HIF-1α. Thus, cMYC could no longer activate the promoter of the gene for the mitochondrial transcription factor TFAM. Importantly, knocking out SIRT1 in skeletal muscle of young mice recapitulated many of these effects of normal aging.
The connection between low NAD+ pools in the nucleus and the various mitochondrial quality control mechanisms is noteworthy, because mitochondrial dysfunction is a hallmark of aging 60. Moreover, these findings provide a link by which a nuclear NAD+ defect, for example due to PARP activation, may also affect the mitochondrial pool of NAD+. A decline in SIRT1 activity thus leads to mitochondrial dysfunction and compromises electron transport. A buildup of the substrate of electron transport, NADH, at the expense of mitochondrial NAD+ is a necessary consequence. To further the problem, a mitochondrial NAD+ deficiency will inactivate mitochondrial sirtuins, again leading to an autocatalytic downward spiral in this compartment. The fact that NAD+ intermediate supplementation can affect both the nuclear and mitochondrial NAD+ pools is critical to the efficacy of these compounds in health maintenance.

Prospects for treating neurodegenerative diseases?

Transgenic mice overexpressing SIRT1 throughout the body have been shown to counteract detrimental effects of energy-dense diet and aging and also mimic some physiological phenotypes induced by DR 11. Furthermore, SIRT1 transgenic mice overexpressing this protein in the brain are protected in mouse models of Alzheimer’s disease 61, 62, Parkinson’s disease 63 and Huntington’s disease 64, 65. In another mouse model, Wallerian degeneration slow (WldS) mice owe their heightened protection against peripheral nerve degeneration upon injury to triplication of the NMNAT1 gene 66–68. Thus, SIRT1 and NAD+ may be broadly neuro-protective. However, in most of the above studies, the degree of protection by SIRT1 overexpression or resveratrol is at best partial. It seems likely that NAD+ depletion may occur in at least a subset of the neurodegenerative diseases. This hypothesis follows from the observation that these diseases have been associated with an increase in chronic nuclear DNA damage 69, 70. If NAD+ is depleted, then protection by SIRT1 activation could be limited and could decline altogether as the disease progressed and NAD+ levels fall below the Km for SIRT1.
It is of interest that transgenic mice modeled for Alzheimer’s disease are partially protected against memory loss by NR supplementation 71. NR supplementation was associated with an increase in PGC-1α and a decrease in the β-secretase, which generates the toxic amyloid-β peptide. Although SIRT1 was not monitored, it seems a likely immediate target for the effect of NR. Therefore, it is of interest to determine whether NAD+ declines in one, some, or all of the neurodegenerative diseases and whether supplementation of NAD+ intermediates, such as NMN and NR, for the restoration of NAD+ will be broadly beneficial. If so, it will be essential to revisit the effects of SIRT1 activation, either by transgenes or by compounds, in combination with NAD+ intermediate supplementation. There is currently no effective treatment for any of these neurodegenerative diseases, which continue to arise in an increasingly long-lived population. A broad therapy to treat a number of these diseases would be transformative, and undoubtedly, no stone should be left unturned to find one. The combination of sirtuin activation and NAD+ intermediate supplementation to restore NAD+ may be an intriguing way to start down one such path.

Concluding remarks

Recent studies have indicated that NAD+ decline may drive aging through decreased sirtuin activities in the nucleus and mitochondria. NAD+ decline might be caused by the defect in NAMPT-mediated NAD+ biosynthesis and the PARP-mediated depletion of NAD+, both of which appear to occur during the aging process and perhaps in age-associated diseases, including neurodegenerative diseases. Supplementation of key NAD+ intermediates, such as NMN and NR, can ameliorate a variety of age-associated pathophysiologies generated by NAD+ decline. Further investigations will be necessary to clarify outstanding questions that remain in the field (Outstanding questions box).

Outstanding questions

Highlights

Acknowledgments

We apologize to those whose work is not cited due to space limitations. We thank members in the Imai lab and the Guarente lab for critical discussions and suggestions. S.I. is supported by grants from the National Institute on Aging (AG024150, AG037457). L.G. is supported by the Glenn Foundation for Medical Research and grants from NIH. S. I. had a sponsored research agreement with Oriental Yeast Co., Japan and is a co-founder of Metro Midwest Biotech. L.G. consults for GlaxoSmithKline, Chronos, Segterra, and Elysium Health.

References