Fortifying the link between SIRT1, resveratrol and mitochondrial function

Abstract

The molecular mechanisms behind the health benefits of resveratrol remain enigmatic and controversial. Here, Price et al. establish a clear chemical-genetic connection between SIRT1 and resveratrol, providing strong evidence that SIRT1 is critical for resveratrol to stimulate mitochondrial biogenesis and a switch toward oxidative muscle fibers (Price et al., 2012).

A quick Pubmed search for articles containing the keyword ‘resveratrol’ generates more than 4600 papers. Resveratrol is a naturally produced polyphenol with medicinal properties of interest to a wide range of researchers, health professionals and those seeking an alternative nutriceutical route to healthy aging. Often mimicking some of the anti-aging phenotypes of calorie restriction (Barger et al., 2008), the benefits of resveratrol treatment are varied but most revealing in mice studies where resveratrol improved metabolic function and reduced inflammation (Lagouge et al., 2006) (Pearson et al., 2008). Not surprisingly, there has been tremendous effort to uncover the molecular mechanisms of this compound. In this issue of Cell Metabolism, Sinclair and colleagues report a clever chemical-genetic approach that establishes the protein deacetylase SIRT1 as a key mediator of resveratrol action (Price et al 2012).

In 2003, Howitz et al identified resveratrol in a high-throughput screen for effectors of in vitro activity of the NAD⁺-dependent deacetylase SIRT1, the mammalian ortholog of the reported longevity factor sir2 in yeast and other organisms (Howitz et al., 2003). In this assay, resveratrol stimulated the activity of SIRT1. The potential molecular link between a longevity protein such as sir2 and resveratrol was an intoxicating notion. SIRT1 has been implicated as an anti-aging factor in numerous dysregulated physiology that includes glucose homeostasis, neurodegeneration and mitochondrial integrity (Donmez and Guarente, 2010). Several cell-based studies appeared to support the assertion that resveratrol was a direct in vivo target of SIRT1 and seemed to reasonably explain several physiological benefits associated with the compound. However, it was discovered that the apparent activation of SIRT1 reported by Howitz et al was dependent on a fluorescent tag associated with the acetylated peptide substrate used in the screen (Borra et al., 2005; Kaeberlein et al., 2005). These results added doubt to the conclusion that resveratrol is a direct activator of cellular SIRT1 function. Chung and co-workers added more uncertainty when they reported recently that resveratrol directly inhibits cAMP-degrading phosphodiesterases (Park et al., 2012).

To establish a clear chemical-genetic connection between SIRT1 and resveratrol, it was essential to demonstrate the specific physiological effects of resveratrol are lost when functional SIRT1 is absent. In this issue of Cell Metabolism, Sinclair and co-workers report several important advancements in understanding the link between mammalian SIRT1 and resveratrol (Price et al 2012). Because SIRT1 is critical for mouse development, studying the role of resveratrol in adult mice is complicated by severe abnormalities of SIRT1-deficient mice. Using a tamoxifen inducible SIRT1-knockout (KO), Price et al. were able to avoid complications elicited from germline knockouts and study directly the dependency of resveratrol on functional SIRT1 in adult mice (Price et al 2012). In striking fashion, resveratrol improved skeletal muscle mitochondrial function and induced a shift toward more oxidative muscle fibers in WT mice fed a high fat diet, but no effect was observed on adult inducible SIRT1 KO mice fed the same diet. Also, high overexpression of SIRT1 in a transgenic mouse strain mimicked the effects of resveratrol treatment in skeletal muscle, demonstrating SIRT1 requirement to mediate resveratrol’s effects. As resveratrol is known to improve glucose homeostasis, Price et al. examined whether these effects were lost in the inducible SIRT1 KO mice. Surprisingly, there was little if any difference in glucose homeostasis and liver function between the WT and inducible KO mouse in response to resveratrol treatment. Price et al. suggest that this may be due to the incomplete removal of SIRT1 in the adult inducible KO, and/or because the liver response in SIRT1 KO mice relies on signaling from other tissues. These are reasonable explanations, though the possibility that resveratrol has minimal effects on SIRT1 in the liver remains a possibility.

AMP-activated protein kinase (AMPK) is a well-established sensor of low metabolic fuel and another reported target of resveratrol. AMPK-deficient mice exhibit a blunted response to resveratrol treatments, identifying AMPK as a critical mediator of resveratrol action (Um et al., 2010). Price et al. performed experiments addressing the role of AMPK in resveratrol’s action and support a model in which lower doses of resveratrol stimulate SIRT1 upstream of AMPK, specifically via the deacetylation of LKB1, one of the activating protein kinases of AMPK (Price et al 2012). Recently, Park et al. suggested that resveratrol is not directly targeting SIRT1 but instead stimulates the AMPK pathway through direct inhibition of cAMP-degrading phosphodiesterases (mainly PDE4), resulting in increased cAMP levels and stimulation of a Ca²⁺-dependent pathway involving CamKKβ (Park et al., 2012). Some groups suggest that SIRT1 functions much later in the pathway, after AMPK stimulates NAD⁺ production through increased transcription of NAD⁺ synthetic enzymes (Um et al., 2010). Regardless of proposed mechanism, the transcriptional co-activator PGC1α is the ultimate recipient of the signaling pathway. Using C2C12 cells, Price et al showed that lower doses of resveratrol increased SIRT1-dependent phosphorylation of AMPK, while a higher dose led to SIRT1-independent activation of AMPK. Interestingly, the two doses displayed opposite trends; higher resveratrol decreased NAD⁺ and ATP levels, while the lower dose led to increases in both metabolites, though the NAD⁺ change was not evident until 12 hours. These results convincingly demonstrate that different dosage of resveratrol can elicit varied responses.

Most importantly, the work by Price et al. provides strong evidence that SIRT1 is a critical player in mediating the effects of resveratrol on mitochondrial biogenesis and the switch to more oxidative muscle fibers (Price et al, 2012). Given the pleiotropic effects of resveratrol, the molecular mechanism remains in dispute. When and how are SIRT1 and PDE4 involved in the resveratrol-dependent activation of AMPK? How does resveratrol concentration differentially affect the reported targets and their associated signaling pathways? A careful time-course analysis of all the implicated factors is essential. For example, if SIRT1 is directly activated by resveratrol, then upon treatment, deacetylation of LKB1 should precede or coincide with AMPK phosphorylation. Similarly, does SIRT1-dependent deacetylation of PGC1α occur prior to, coincident with, or after initial AMPK activation? It will be important to dissect the initial signaling events from those of the metabolically-reprogrammed state. Increased NAD⁺ synthesis might to be a long-term adaption to drive sustained sirtuin function or to replenish NAD⁺ levels as a result of an initial surge in sirtuin activity. The results from Price et al strengthen the biological link between resveratrol and SIRT1-dependent processes, and provide a backdrop to further studies that resolve the mechanism.