Prediction of the 3D structure of FMRF-amide neuropeptides bound to the mouse MrgC11 GPCR and experimental validation

We report the 3D structure predicted for the mouse MrgC11 (mMrgC11) receptor by using the MembStruk computational protocol, and the predicted binding site for the F-M-R-F-NH2 neuro-peptide together with four singly chirally modified ligands. We predicted that the R-F-NH2 part of the tetrapeptide sticks down into the protein between the transmembrane (TM) domains 3, 4, 5, and 6. The Phe (F-NH2) interacted favorably with Tyr110 (TM3) while the Arg makes salt bridges to Asp161 (TM4) and Asp179 (TMS). We predicted that the Met extends from the binding site, but the terminal Phe residue sticks back into an aromatic/hydrophobic site flanked by Tyr237, Leu238, Leu240, and Tyr256 (TM6), and Trp162 (TM4). We carried out subsequent mutagenesis experiments followed by intracellular calcium-release assays that demonstrated the dramatic decrease in activity for the Tyr110Als, Asp161Ala, and Asp179Ala substitutions, which was predicted by our model. These experiments provide strong evidence that our predicted G protein-coupled receptor (GPCR) structure is sufficiently accurate to identify binding sites for selective ligands. Similar studies were made with the mMrg-A1 receptor, which did not bind the R-F-NH2 dipeptide, we explain this to be due to the increased hydrophobic character of the binding pocket in mMrgA1.