3D-QSAR of SIRT1 Activators Targeting Against Diet-Induced Metabolic Syndrome

Abstract

The prevalence of metabolic syndrome associated with increased risk for cardiovascular disease, type 2 diabetes, or cancer has been increasing over the past decade. While traditional drug discovery efforts have been tackling these diseases by aiming at individual targets, recent studies in humans have suggested the possibility that the collection of the metabolic degenerative processes can be approached as a whole by controlling diet, especially calorie restriction. Studies of calorie restriction in a broad range of animals including primates, mice, worms, and yeast have suggested that a reduction in calorie intake lengthens lifespan and protects against cancer and other age-related diseases. There has therefore been much interest in developing pharmacological agents that mimic the effects of calorie restriction. Resveratrol, a natural product derived from grapes, is the first reported mimetic of calorie restriction. Ever since the mode of action of resveratrol was elucidated to activate SIRT1 [sirtuin (silent mating type information regulation 2 homolog) 1] which, in turn, deacetylates p53 and promotes cell survival in a NAD-dependent manner, numerous efforts have been devoted to discover novel activators of SIRT1. Through high throughput screening of a large selection of small molecules, Milne et al. identified three imidazothiazole derivatives as potent SIRT1 activators, which are structurally unrelated to but 1,000-fold more active than resveratrol. Recently, in order to improve the potency as well as the solubility of the imidazothiazole derivatives, two different libraries of analogues, imidazo[1,2-b]thiazole and oxazolo[4,5-b]pyridine, were constructed and extensively investigated. The aim of the present study is to derive a predictive method for designing novel potent SIRT1 activators through construction of 3D QSAR models. For this purpose, from the literature, SIRT1 activation data of 33 imidazothiazole and oxazolopyridine derivatives were obtained (Fig. 1). At first, the 33 compounds were divided into two groups: 27 compounds as a training set and the other 6 compounds as a test set. The CoMFA (comparative molecular force analysis) and CoMSIA (comparative molecular similarity indices analysis) methods were employed for deriving 3D-QSAR models consisting for a training set of 27 imidazothiazole and oxazolopyridine derivatives, keeping in vitro activity as a dependent variable. The 3D-QSAR models were then validated using a test set of 6 compounds, which were not included in the development of the models. Finally, the contour plots of the 3D-QSAR model were analyzed to