New findings on mechanism of activation of sirtuins may vindicate Sirtris founders
The Biopharmconsortium Blog has from time to time been following novel developments in anti-aging medicine, including attempts to develop activators of sirtuins. However, we have not had an article on sirtuins since December 1, 2010. At that time, we reported on the discontinuation by GlaxoSmithKline (GSK) of its lead sirtuin activator, SRT501, a proprietary formulation of the natural product resveratrol (which is found in red wine).
Sirtuins are nicotinamide adenine dinucleotide (NAD+)–dependent protein deacetylases, which have been implicated in control of lifespan in yeast, the nematode Caenorhabditis elegans, and the fruit fly Drosophila. Mammalian sirtuins have been implicated (via animal model studies) in protection against aging-related diseases such as metabolic and cardiovascular diseases, neurodegeneration, and cancer.
As we discussed in our December 1, 2010 article, GSK acquired the sirtuin-pathway specialty company Sirtris (Cambridge, MA) for $720 million in June 2008. This gave GSK ownership of Sirtris’ sirtuin modulator drugs. As stated in that article, although GSK discontinued development of SRT501, it was continuing development of Sirtris’ non-resveratrol synthetic selective sirtuin 1 (SIRT1) activators, which in addition to their greater potency, had more favorably drug-like properties.
Recently, resveratrol and synthetic sirtuin activators such as those developed by Sirtris have come to be known as “sirtuin-activating compounds” (STACs).
Sirtuin-activating compounds (STACs) under a cloud
As we discussed in our February 10, 2010 blog article, researchers at Amgen found evidence that the apparent in vitro activation of SIRT1 by resveratrol depended on the substrate used in the assay. The Amgen group found that the fluorescent SIRT1 peptide substrate used in the Sirtris assay is a substrate for SIRT1, but in the absence of the covalently linked fluorophore, the peptide is not a SIRT1 substrate. Resveratrol did not activate SIRT1 in vitro as determined by assays using two other non-fluorescently-labeled substrates.
Researchers at Pfizer also found that resveratrol and three of Sirtris’ second-generation STACs activated SIRT1 when a fluorophore-bearing peptide substrate was used, but were not SIRT1 activators in in vitro assays using native peptide or protein substrates.The Pfizer researchers also found that the Sirtris compounds interact directly with the fluorophore-conjugated peptide, but not with native peptide substrates.
Moreover, the Pfizer researchers were not able to replicate Sirtris’ in vivo studies of its compounds. Specifically, when the Pfizer researchers tested SRT1720 in a mouse model of obese diabetes, a 30 mg/kg dose of the compound failed to improve blood glucose levels, and the treated mice showed increased food intake and weight gain. A 100 mg/kg dose of SRT1720 was toxic, and resulted in the death of 3 out of 8 mice tested.
The Pfizer researchers also found that the Sirtris compounds interacted with an even greater number of cellular targets (including an assortment of receptors, enzymes, transporters, and ion channels) than resveratrol. For example, SRT1720 showed over 50% inhibition of 38 out of 100 targets tested, while resveratrol only inhibited 7 targets. Only one target, norepinephrine transporter, was inhibited by greater than 50% by all three Sirtris compounds and by resveratrol. Thus the Sirtris compounds have a different target selectivity profile than resveratrol, and all of these compounds exhibit promiscuous targeting.
Finally, as we reported in our December 1, 2010 blog article, NIH researcher Jay H. Chung and his colleagues found evidence that resveratrol works indirectly, via the energy sensor AMP-activated protein kinase (AMPK), to activate sirtuins. Since activation of AMPK increases fatty acid oxidation and upregulates mitochondrial biogenesis, this study suggested that the effect of resveratrol on AMPK may be more important than its more indirect activation of sirtuins in the regulation of insulin sensitivity.
All of these studies left Sirtris/GSK’s STACs under a cloud.
On March 13, 2013, GSK reported that it was shutting down Sirtris and its Cambridge MA facilities, just five years after its $720 million acquisition. GSK also said that it was offering transfers to the Philadelphia area for some of the 60 remaining Sirtris employees. Although GSK was closing Sirtris, it said that it remained confident in Sirtris’ drug candidates. The pharma company said that following Sirtris’ “highly successful” research on the biology of sirtuins, further development of Sirtris’ drug candidates “requires the resource and expertise available from our broader drug discovery organization.” GSK will be “exami[ing] [its] research against a variety of therapeutic conditions, with the aim of moving potential assets into the clinic within the next three to four years.”
New evidence that STACs activate SIRT1 in vitro under certain conditions
On 8 March 2013, the journal Science published a report by Sirtris founder David A. Sinclair, Ph.D. (Harvard Medical School, Boston MA) and his colleagues [from academia and from Sirtris, GSK, and from Biomol (Plymouth Meeting, PA)] that identified conditions under which STACs activate SIRT1 in vitro. This research report was accompanied by a Perspective in the same issue of Science authored by Hua Yuan, Ph.D. and Ronen Marmorstein, Ph.D. (Wistar Institute, Philadelphia, PA).
Dr. Sinclair and his colleagues hypothesized that the fluorophore tags on peptide substrates that were used in the original, successful SIRT1 activation assays might mimic hydrophobic amino acid residues of natural substrates at the same position as the fluorophore (i.e, +1 relative to the acetylated lysine that is engaged by SIRT1). Consistent with this hypothesis, the researchers found that non-fluorophore-tagged natural SIRT1 substrates with a large hydrophobic amino acid residue [i..e, tryotophan (Trp), tyrosine (Tyr), or phenylalanine (Phe)] at positions +1 and +6 or +1 were selectively activated by STACs. Examples of such substrates are peroxisome proliferator-activated receptor γ coactivator 1α acetylated on lysine at position 778 (PGC-1α–K778), and forkhead box protein O3a acetylated on lysine at position 290 (FOXO3a-K290). The PGC-1α–K778 peptide contains Tyr at the +1 position and Phe at the +6 position, and FOXO3a contains Trp at the +1 position. Substitution of these aromatic amino acids on either acetylated peptide with alanine (Ala) resulted in complete abolition of SIRT1 activity.
The researchers identified over 400 nuclear acetylated proteins that are potential SIRT1 targets that support STAC-mediated activation of SIRT1, on the basis of their structure. They tested five of these native sequences and found that three of them supported SIRT1 activation.
Kinetic analysis of SIRT1 activation by STACs in the presence of the above peptide substrates showed that the enhancement in the rate of SIRT1 deacetylation was mediated primarily through an improvement in peptide binding. This is consistent with an allosteric mechanism of activation. In allosteric regulation, an allosteric activator (in this case, a STAC) binds to a regulatory site (also known as an allosteric site) that is distinct from the catalytic site of an enzyme (in this case, SIRT1). Binding of the activator to the allosteric site results in the enhancement of the activity of the enzyme, for example by causing a conformational change in the protein that results in improved biding of the catalytic site to the substrate.
In order to investigate the nature of the hypothesized SIRT1 allosteric site, the researchers screened for SIRT1 mutant proteins that could not be activated by STACs in the presence of an appropriate peptide substrate. As a result of these studies, the researchers identified a critical glutamate (Glu) residue at position 230 of SIRT1, which is immediately N-terminal to the catalytic core of SIRT1. Glu230 of SIRT1 is conserved from flies to humans. Replacement of Glu230 with another amino acid, such as lysine or alanine, resulted in attenuation of SIRT1 activation by STACs, independent of the substrate used. Structural studies identified a rigid N-terminal domain that contains Glu230, and is critical for activation by STACs.
The researchers then studied the effects of STACs on cultured cells (murine myoblasts), expressing either wild-type SIRT1 or mutant SIRT1 in which Glu230 is replaced with lysine (SIRT1-E222K, which is the murine equivalent of human SIRT1-E230K). Cells expressing the mutant SIRT1 did not respond to STACs, but cells expressing wild-type SIRT1 did. Specifically, cells expressing wild-type SIRT1 exhibited STAC-stimulated increases in ATP levels, mitochondrial mass, and mitochondrial DNA copy number, but cells expressing mutant SIRT1 did not. In STAC-treated cells, the researchers found no evidence of SIRT1-independent AMPK phosphorylation. This goes against studies discussed earlier in this article, that indicate that resveratrol works via activating AMPK. They also found no evidence for inhibition of phosphodiesterase isoforms in the STAC-treated cells. This goes against a study, published in Cell in 2012, that indicates that resveratrol ameliorates aging-related metabolic conditions by inhibiting cAMP phosphodiesterases, thus engaging a pathway that activates AMPK.
The researchers conclude that STACs act via a mechanism of direct “assisted allosteric activation” mediated by the Glu230-containing N-terminal activation domain of SIRT1. They further conclude that their findings support the hypothesis that allosteric activation of SIRT1 by STACs constitutes a viable therapeutic intervention strategy for many aging-related diseases. thus apparently vindicating the Sirtris/GSK development program.
However, the authors of the companion Perspective hypothesize that the reason that existing STACs only work with SIRT1 substrates that contain hydrophobic residues at position +1 to the acetylated lysine is because they were identified via screening with a substrate that contained a hydrophobic residue mimetic–i.e., a fluorophore tag. A new screen that is not biased in this way might possibly identify STACs that exhibit selectivity for SIRT1 substrates that contain other sequence signatures. It is possible that such STACs might be better therapeutics for certain aging-related diseases than the current STACs being investigated by Sirtris/GSK. There also remain many unknowns in the biology of SIRT1, and in the biochemistry of STACs –i.e., mechanisms by with certain STACs modulate the activity of biomolecules other than SIRT1 (e.g,, cAMP phosphodiesterases). Such issues might affect the success or failure of any program to develop STACs as therapeutic compounds.
As the producers of this blog, and as consultants to the biotechnology and pharmaceutical industry, Haberman Associates would like to hear from you. If you are in a biotech or pharmaceutical company, and would like a 15-20-minute, no-obligation telephone discussion of issues raised by this or other blog articles, or an initial one-to-one consultation on an issue that is key to your company’s success, please contact us by phone or e-mail. We also welcome your comments on this or any other article on this blog.