Back

Design of allosteric modulators that change GPCR G protein subtype selectivity

Moore, M. N.; Person, K. L.; Alwin, A.; Krusemark, C. L.; Foster, N.; Ray, C.; Inoue, A.; Jackson, M. R.; Sheedlo, M. J.; Barak, L. S.; Olson, S. H.; Fernandez de Velasco, E. M.; Slosky, L. M.

2024-11-21 pharmacology and toxicology
10.1101/2024.11.20.624209 bioRxiv
Show abstract

G protein-coupled receptors (GPCRs), the largest family of drug targets, can signal through 16 subtypes of G proteins. Biased compounds that selectively activate therapy-relevant pathways promise to be safer, more effective medications. The determinants of bias are poorly understood, however, and rationally-designed, G protein-subtype-selective compounds are lacking. Here, using the prototypical class A GPCR neurotensin receptor 1 (NTSR1), we find that small molecules binding the intracellular GPCR-transducer interface change G protein coupling by subtype-specific and predictable mechanisms, enabling rational drug design. We demonstrate that the compound SBI-553 switches NTSR1 G protein preference by acting both as a molecular bumper and a molecular glue. Structurally, SBI-553 occludes G protein binding determinants on NTSR1, promoting association with select G protein subtypes for which an alternative, shallow-binding conformation is energetically favorable. Minor modifications to the SBI-553 scaffold produce allosteric modulators with distinct G protein subtype selectivity profiles. Selectivity profiles are probe-independent, conserved across species, and translate to differences in in vivo activity. These studies demonstrate that G protein selectivity can be tailored with small changes to a single chemical scaffold targeting the receptor-transducer interface and, as this pocket is broadly conserved, present a strategy for pathway-selective drug discovery applicable to the diverse GPCR superfamily.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.