Identification of a Critical Amino Acid Position Conferring G‐protein Bias to Delta Opioid Receptor Peptides

The delta opioid receptor (δOR) is a G‐protein coupled receptor that presents as a promising target in the potential treatment of mood, anxiety and alcohol use disorders. Thus far, attempts at the clinical development of δOR agonists have not met desired end‐points, which may be related to the risk of seizures associated with some δOR agonists. Previous studies have associated δOR agonist induced seizures to an agonist's efficacy in recruiting β‐arrestin proteins. Thus, it has been hypothesized that δOR agonists that display minimal β‐arrestin recruitment may show greater promise in developing into successful clinical therapies. A big current limitation in readily producing such signal biased drugs is a lack of conformational insight into such biased signaling. Recently, we determined that the naturally derived rubiscolin peptides YPLDL and YPLDLF possess a preference or signaling bias towards G i protein over β‐arrestin recruitment. Here we hypothesize that the conformationally restricted nature of proline may be contributing to the G‐protein bias of these peptides. Methods To interrogate our hypothesis we systematically exchanged proline with residues of varying flexibility and lipophilicity. Peptides were synthesized using standard Fmoc‐based solid‐phase synthesis with HBTU. Sequences were derived from rubiscolins YPLDLF, YPLDLV and YPMDLV, representing 3 different series. Each series consisted of 5 peptides with substitution at the second position by either proline, alanine, glycine, valine or serine. In vitro characterization of G‐protein (cAMP Glosensor), β‐arrestin 2 recruitment (PathHunter) and radioligand binding ( 3 H‐DPDPE competition) was performed for each peptide at the δOR. Results Within glycine, alanine and proline substituted rubiscolin derivatives, β‐arrestin 2 potency and efficacy showed a consistent trend of Gly > Ala > Pro. Further, these same derivatives all showed full agonism through G i with potency values following the same rank‐order, as did radioligand affinity. Substitution of valine ablated both β‐arrestin 2 and G i activity and dramatically reduced binding affinity relative to proline, while serine substitution reduced G i potency and efficacy relative to proline with more modest effects on β‐arrestin 2 and binding. Discussion Our data suggests an important role for proline in the apparent bias of the rubiscolin peptides. Specifically, at this second position it appears that increasing residue flexibility correlates with increased β‐arrestin 2 recruitment. Our finding that substitution with hydrophilic serine did not strongly increase β‐arrestin 2 preference, nor that valine decreased it, suggests that the G‐protein bias observed for the proline containing peptides is unlikely to be due to the greater lipophilicity of proline relative to glycine and alanine. These findings should provide opportunities to use computational analysis to model the δOR in conformations that are either G protein biased or promote arrestin recruitment. Support or Funding Information This project was funded by a grant from the National Institutes on Drug Abuse (R03DA045897 to RMvR). This abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal .