Binding pose depth modulates photoswitchable ligands' efficacy at the 5-HT_2A receptor

Photoswitchable ligands enable reversible control of receptor signaling through light-induced cis-trans isomerization, yet predicting how subtle structural modifications affect efficacy remains challenging. Here, we use molecular dynamics simulations to investigate two azobenzene-based human 5-HT_2A receptor ligands differing only by a methoxy substituent position (para- vs meta-methoxy). Compound 1 (para-methoxy) switches from acting as a weak antagonist (trans) to a moderate agonist (cis), whereas compound 2 (meta-methoxy) maintains agonist activity in both forms, with cis-2 exhibiting the highest efficacy. Our simulations reveal that the key determinant of these efficacy differences lies in the vertical depth of ligand insertion into the orthosteric binding pocket. The para-methoxy moiety of trans-1 forms hydrogen bonds with Asp231^5.35 and Thr160^3.37, anchoring the ligand deeper than typical tryptamine agonists and preventing engagement with activation-critical residues, thereby stabilizing the inactive receptor. Conversely, trans-2 lacks these anchoring interactions and adopts a shallower, agonist-compatible pose. In the active receptor, cis-2 forms a persistent Thr160^3.37 hydrogen bond that allows deeper penetration between TM4 and TM5, whereas cis-1's para-methoxy causes steric hindrances limiting this interaction. Based on these findings, we suggest that ligand insertion depth is a critical determinant of efficacy. This provides a framework for designing light-sensitive GPCR ligands with tunable signaling properties.