Original articleThe first pharmacophore model for potent G protein-coupled receptor 119 agonist
Graphical abstract
Pharmacophore model was generated using HypoGen algorithm in DS v2.1. The well validated Hypo1 perfectly fits GPR119 agonist in trials.
Highlights
► The first pharmacophore model for potent G protein-coupled receptor 119 agonist. ► A highly predictive pharmacophore model was successfully obtained. ► Hypo1 perfectly fits GSK-1292263 as a GPR119 agonist in trials. ► 20 compounds were selected as GPR119 agonists in virtual screening.
Introduction
Type 2 diabetes mellitus (T2DM) is emerging as a disease of staggering proportions in the 21st century, with an estimated 300 million cases worldwide projected by 2020 [1]. Historically, treatment regimens for T2DM have shown significant effectiveness for improving glucose homeostasis; however, increasing failure to maintain glycemic control is observed after about 2 years of therapy [2], [3]. Modulators of G protein-coupled receptors (GPCRs), which represent one of the most successful target classes in drug discovery [4], have been identified as prime candidates for type 2 diabetes and associated disorders for new treatment [5], [6]. GPR119 has been described as a class A (rhodopsin-type) orphan GPCR without close primary sequence relative in the human genome [7]. The initial results with prototypical potent and selective, orally available, synthetic GPR119 agonists indicate that these compounds could be potential therapies for diabetes and related metabolic disorders by i) stimulating glucose-dependent insulin secretion; ii) inducing the release of Glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1); iii) protecting pancreatic β cells through raised cAMP levels; and iv) reducing body weight and food intake [8], [9], [10].
Agonists of GPR119 have emerged from pharmaceutical discovery efforts to identify an improved GLP-1 therapeutic by combining the convenience oral dosage of DPP-IV inhibitors and the pharmacological robustness of GLP-1 receptor agonists. The field of GPR119 agonist research has progressed far enough for some companies to push compounds for the clinical use. For example, GlaxoSmithKline (GSK) and Metabolex have announced that Phase II clinical trials with GSK-1292263 and MBX-2982 have been completed in 2010, respectively [11]. Arena and partner Ortho-McNeil initiated the earliest FTIH studies with APD668 and have recently progressed APD597 into Phase I clinical trials [11]. With the passage of the GPR119 agonist clinical candidates into phase trials and confirmatory reports of clinical proof of concept with respect to glycemic control and incretin release, the spotlight has been set for a new class of therapeutics for type 2 diabetes and associated obesity [12].
In the present study, we have generated pharmacophore models using Discovery Studio V2.1 for a diverse set of molecules as GPR119 agonist with an aim to obtain the pharmacophore model that would provide a hypothetical picture of the chemical features responsible for activity. Further employment of the best pharmacophore model will be used as a 3D query for searching large databases to identify GPR119 agonist and also to utilize this pharmacophore model as a predictive tool for estimating biological activity of GPR119 agonist through virtual screening or molecular designing on the basis of structure–activity analysis.
Section snippets
Selection of molecules
A set of 49 different compounds tested with the same assay (CHO: CRE reporter assay) has been collected from different references [11], [12], [13], [14], [15], [16], [17], [18], [19], [20]. The datasets are divided into a training set and a test set. We selected 24 compounds as training set with the following rules: i) Both training and test sets should cover the widest possible range of molecular bioactivities (EC50); ii) Both the highly active and low active compounds should be included. iii)
Pharmacophore model validation
Validation of a quantitative model was performed in order to determine whether the developed model was able to identify active structures and to forecast their activities precisely. There are several methods to confirm the quality of pharmacophore like preparing test set, Fischer’s randomization method, goodness of fit (GF) etc.
Conclusion
In this study, chemical based pharmacophore modeling of GPR119 agonists has been created by using DS. The best HypoGen model in terms of predictive values consisted of two HBA and three H. Test set, Decoy set and Fischer’s validation methods have been used to validate the pharmacophore model. For predicting activity, the correlation coefficient of Hypo1 with training and test sets were 0.933 and 0.933 respectively. The Fit-Value and Estimate activity of GSK-1292263, which have completed phase
Acknowledgment
The work was supported by the Important National Science and Technology Specific Projects (2009ZX09102-033), the National Natural Science Foundation of China (No. 30772647) and the Fundamental Research Funds for the Central Universities of China (No. 2J10023).
References (29)
- et al.
Annu. Rep. Med. Chem.
(2009) - et al.
Drug. Discov. Today.
(2006) - et al.
FEBS. Lett.
(2003) - et al.
Bioorg. Med. Chem. Lett.
(2010) - et al.
Bioorg. Med. Chem. Lett.
(2007) - et al.
Eur. J. Med. Chem.
(2008) - et al.
Bioorg. Med. Chem. Lett.
(2007) - et al.
Eur. J. Med. Chem.
(2010) Br. J. Nutr.
(2000)Lancet
(1998)
Curr. Opin. Investig. Drugs.
CNS Neurol. Disord. Drug Targets
Expert Opin. Drug Discovery
Nature
Cited by (0)
- 1
These authors contributed equally to this work.