The prevailing model for the variety in drug responses is that different drugs stabilize distinct active states of their G protein-coupled receptor (GPCR) targets, allowing coupling to different effectors. However, whether the same ligand generates different GPCR active states based on the immediate environment of receptors is not known. Here we address this question using spatially resolved imaging of conformational biosensors that read out distinct active conformations of the δ-opioid receptor (DOR), a physiologically relevant GPCR localized to Golgi and the surface in neuronal cells. We have shown that Golgi and surface pools of DOR both inhibit cAMP, but engage distinct conformational biosensors in response to the same ligand in rat neuroendocrine cells. Further, DOR recruits arrestins on the surface but not on the Golgi. Our results suggest that the local environment determines the active states of receptors for any given drug, allowing GPCRs to couple to different effectors at different subcellular locations.
Nature uses structural variations on protein folds to fine-tune the geometries of proteins for diverse functions, yet deep learning-based de novo protein design methods generate highly regular, idealized protein fold geometries that fail to capture natural diversity. Here, using physics-based design methods, we generated and experimentally validated a dataset of 5,996 stable, de novo designed proteins with diverse non-ideal geometries. We show that deep learning-based structure prediction methods applied to this set have a systematic bias towards idealized geometries. To address this problem, we present a fine-tuned version of Alphafold2 that is capable of recapitulating geometric diversity and generalizes to a new dataset of thousands of geometrically diverse de novo proteins from 5 fold families unseen in fine-tuning. Our results suggest that current deep learning-based structure prediction methods do not capture some of the physics that underlie the specific conformational preferences of proteins designed de novo and observed in nature. Ultimately, approaches such as ours and further informative datasets should lead to improved models that reflect more of the physical principles of atomic packing and hydrogen bonding interactions and enable improved generalization to more challenging design problems.
The prevailing model for the variety in drug responses is that they stabilize distinct active states of their G protein-coupled receptor (GPCR) targets, allowing coupling to different effectors. However, whether the same ligand can produce different GPCR active states based on the environment of receptors in cells is a fundamental unanswered question. Here we address this question using live cell imaging of conformational biosensors that read out distinct active conformations of the δ-opioid receptor (DOR), a physiologically relevant GPCR localized to Golgi and the surface in neurons. We show that, although Golgi and surface pools of DOR regulated cAMP, the two pools engaged distinct conformational biosensors in response to the same ligand. Further, DOR recruited arrestin on the plasma membrane but not the Golgi. Our results suggest that the same agonist drives different conformations of a GPCR at different locations, allowing receptor coupling to distinct effectors at different locations.