1 February 2010. Unseasonably warm days gave way to some notable nights at this year’s Keystone Symposium, Alzheimer’s Disease Beyond Aβ, held 10-15 January at Copper Mountain, Colorado. One evening offering was a short talk from Gizem Donmez, a postdoctoral fellow in Leonard Guarente’s laboratory at MIT. Donmez reported that SIRT1, the histone deacetylase linked to longevity, might protect against AD by boosting ADAM10 (aka α-secretase) and promoting non-amyloidogenic processing of Aβ precursor protein (APP). If true, then you might want to eat more carrots because the effect seems to rely on SIRT1 playing vassal to the retinoic acid receptor.
SIRT1 is activated by caloric restriction, which protects against brain atrophy in primates (see ARF related news story). SIRT1 itself also protects against neurodegeneration in mouse models of AD (see Kim et al., 2007), and previous work from Giulio Pasinetti’s lab at Mount Sinai School of Medicine, New York, suggested that activation of α-secretase may be responsible (see ARF related news story on Qin et al., 2006). Pasinetti and colleagues attributed the increase in α-secretase to SIRT1 inhibition of the Rho kinase ROCK1, previously linked to suppression of the non-amyloidogenic secretase (see ARF related news story). But Donmez’s work suggests that there is more to the tale.
To explore the relationship between SIRT1 and AD, Donmez and colleagues made mice with either the SIRT1 gene knocked out or overexpressed. For knockouts, Donmez used the cre/lox system driven by a nestin promoter, limiting SIRT1 loss to neurons. For overexpression, she knocked the SIRT1 gene into the β actin locus, getting a mild, twofold overexpression. Donmez tested the effects of the SIRT1 mice on Aβ pathology by crossing them with APP/PS1 transgenic animals (APPSwe/PS1ΔE9).
Donmez reported that the APP/PS1/SIRT1 knockouts die earlier than control APP/PS1 animals, and that the knockouts have increased amyloid plaques and gliosis. The increased pathology in these mice was accompanied by a reduction in α-secretase activity. In contrast, APP/PS1 mice overexpressing SIRT1 had reduced levels of Aβ42 compared to controls and increased ADAM10 and ADAM10 mRNA. Levels of Notch intracellular domain, which is produced following α-secretase processing of the transmembrane receptor, were also increased when SIRT1 was overexpressed but not when it was knocked out. The results support the theory that SIRT1 can boost expression of the secretase.
Donmez jousted with the ADAM10 promoter using chromatin immunoprecipitation assays to determine exactly how SIRT1 might exert its influence. She reported that the deacetylase attaches to the promoter very close to a binding site for the retinoic acid receptor (RAR)/retinoid X receptor (RXR) heterodimer. Activation of the ADAM10 gene depended on SIRT1 deacetylase activity (an inactive mutant has no effect) and also the presence of retinoic acid. The evidence suggests that SIRT1 deacetylates RAR leading to increased expression of ADAM10, presumably by allowing RAR to bind more tightly to the promoter. In support of this, Donmez found that RARβ is deacetylated in the presence of SIRT1 and that RARβ acetylation is increased in SIRT1 knockout cells. Coming back full circle, she showed that she was able to reverse the reduced production of Aβ in SIRT1-overexpressing cells by knocking down ADAM10 transcripts with RNA interference.
Donmez concluded that SIRT1 activators might be worth pursuing as potential therapeutics for AD. Resveratrol, a SIRT1 activator found in miniscule quantities in red wine, is widely promoted in the popular press as an elixir of life. It has received serious attention from the scientific community as well, since it has been shown to mimic some of the effects of caloric restriction (see ARF related news story) though other research counters that blocking SIRT1 might actually improve cognition (see ARF related news story). Resveratrol, however, does not cross the blood-brain barrier very efficiently. Amongst all of this, vitamin A, which is metabolized to retinoic acid, might be worth a closer look, too. Recent findings suggest that all-trans retinoic acid can protect APP/PS double transgenic mice against Aβ pathology, reducing levels of the peptide without affecting APP expression (see ARF related news story on Ding et al., 2008), while acitretin, a vitamin A analog, was also shown to upregulate ADAM10 (see Tippmann et al., 2009). Because acitretin crosses the blood-brain barrier and has been approved for treating psoriasis since 1997, it would appear to be a candidate for exploratory clinical or preclinical studies.—Tom Fagan.