In silico Design and Computational Assessment of Small-Molecule GLP-1R Agonist as a Possible Treatment for Obesity and Type 2 Diabetes

Abstract

Obesity is one of the most concerning health hazards in recent years as it associated with a large number of diseases such as diabetes mellitus. Glucagon-like peptide-1 receptor agonists are clinically tested treatment options for obesity along with type 2 diabetes. We present the rational in silico design and computational assessment of ThiaGlip-1, an innovative small-molecule GLP-1R agonist candidate based on a previously unreported indolo[1,2-b]quinoxaline tetracyclic framework. A Tanimoto similarity of 0.41 compared to the closest known analogue confirming that the structure is new. By using Auto Dock Vina 1.2.3 to do molecular docking toward active-state GLP-1R, it was found that ThiaGlip-1 had the best binding affinity of −9.222 kcal/mol, which is similar to the reference agonist Danuglipron (−9.763 kcal/mol). After that, a virtual library of similar compounds was generated via Swiss Similarity. Virtual screening of 380 Enamine_Tang library compounds found 294 candidates (77.4%) with Excellent binding affinity. The best compound had a binding affinity of −12.931 kcal/mol. The SAR analysis found that LogP, the number of aromatic rings, and the number of heavy atoms were the main factors that affected GLP-1R binding affinity. ADMET profiling demonstrated that 270 library compounds (71.1%) have no toxicity flags and a mean drug score of 1.00. The three best leads, Z4594432747, PV-002918833325, and PV-004003938144, were chosen as the best candidates for further experimental follow-up to be framework for a new orally active anti-obesity and antidiabetic drugs in the future.