Pharmacophore-guided Virtual Screening to Identify New β3 -adrenergic Receptor Agonists.

Abstract

The β3‐adrenergic receptor (β3‐AR) is found in several tissues such as adipose tissue and urinary bladder. It is a therapeutic target because it plays a role in thermogenesis, lipolysis, and bladder relaxation. Two β3‐AR agonists are used clinically: mirabegron 1 and vibegron 2, which are indicated for overactive bladder syndrome. However, these drugs show adverse effects, including increased blood pressure in mirabegron patients. Hence, new β3‐AR agonists are needed as starting points for drug development. Previous pharmacophore modeling studies of the β3‐AR did not involve experimental in vitro validation. Therefore, this study aimed to conduct prospective virtual screening and confirm the biological activity of virtual hits. Ligand‐based pharmacophore modeling was performed since no 3D structure of human β3‐AR is yet available. A dataset consisting of β3‐AR agonists was prepared to build and validate the pharmacophore models. The best model was employed for prospective virtual screening, followed by physicochemical property filtering and a docking evaluation. To confirm the activity of the virtual hits, an in vitro assay was conducted, measuring cAMP levels at the cloned β3‐AR. Out of 35 tested compounds, 4 compounds were active in CHO−K1 cells expressing the human β3‐AR, and 8 compounds were active in CHO−K1 cells expressing the mouse β3‐AR.