Gap19 GPR can signal through not only IP

GPR40 can signal through not only IP3 but also cAMP depending upon the type of agonist ligand used. A full agonist engages both signaling mechanisms in contrast to the endogenous long-chain fatty Gap19 (LCFA) ligands and partial agonists. Thus, stimulation of GPR40 by endogenous LCFAs or by the synthetic partial agonists result in a rather limited incretin response and the consequential insulin secretion, whereas synthetic GPR40 full agonists stimulate robust GLP-1 and GIP responses in addition to the insulin secretion. These incretins, in turn, can also stimulate additional insulin secretion via binding to their own receptors on beta cells. Owing to this dual mechanism of action, leading to greater insulin secretion and potentially superior glycemic control, GPR40 full agonism may be preferable over partial agonism. Representative examples of currently known GPR40 full agonists are shown in . Amgen GPR40 full agonist was the very first GPR40 ligand reported to exhibit greater glucose lowering ability compared to the partial agonists., shows biaromatic GPR40 full agonists which are closely related in structure. The Amgen full agonist differs from the others due to the presence of a central benzenoid core structure. The Takeda and Janssen full agonists contain nitrogenous 6-membered heterocyclic central ring structures. However, when these different central ring structures were identically substituted, was the most potent compound in both human and rat GPR40 Ca assays, (). Our initial optimization work on the lead compound was focused on finding suitable replacements for the carboxylic acid function and the β-cyclopropyl side chain. The traditional medicinal chemistry approach of replacing the acid moiety with acid isosteres did not produce successful results. Additionally, the β -cyclopropyl side chain with the () configuration was also found to be optimal for the GPR40 full agonist activity of (data not shown). The SAR exploration around the 2- fluoro-5-methoxy phenyl ring revealed that it could be exchanged with 4-substituted 5-fluoro-2-methoxypyridine without a loss of potency compared to the parent compound ( vs ). However, the site that accommodates the dimethylcyclopentene moiety of the parent compound was found to be tolerant of a wide range of substituents (). The SAR exploration at this site encompassed the initial introduction of dissected fragments of dimethylcyclopentene followed by further modifications of the active substituents. Several large, lipophilic, structurally diverse substituents were tolerated at this site with improvement in GPR40 potency whereas hydrophilic groups negatively affected the activity of the compound (data not shown). The evaluation of many analogues finally led to the selection of as a compound of interest for further studies based on its favorable biological activity, early ADME and PK profile. Receptor interactions and binding mode of in hGPR40 were determined by docking studies. The compound was docked in the rebuilt co-crystal structure of hGPR40 (PDB ID: ) using MOE for loop modeling and energy minimization and Glide for molecular docking. The docking site used for the docking simulation corresponded to the membrane exposed, full agonist AP8 binding site described recently by Lu et al. The top ranked binding mode of , obtained by docking, is shown in . In this model, the compound fits well in the hydrophobic groove formed by a set of residues from TMD3, 4 and 5 (Val126, Ile130, Ala98, Ala99, Ala102, Leu193 and Ile197) as well as some of the residues from intracellular loop 2 (ICL2), which is folded into an alpha helix due to the presence of the full agonist AP8 in the published X-ray structure. The carboxylate of forms a hydrogen-bond network with Tyr44 and a weak hydrogen bond with Ser123. The F-MeO-pyridyl moiety is stabilized by a CH-π interaction with the Pro194 side chain and the methoxy occupies a sub-pocket formed by Ala92, Val134, and Leu190. The rest of the major interactions, shown in , are mainly hydrophobic in nature.