Seminar – Lucie DELEMOTTE

Common evolutionary mechanism of activation of class A GPCRs

Cells communicate with their environment by integrating signals, often chemical in nature, triggered by specific molecules binding to specific membrane-bound receptors, resulting in a downstream signaling cascade. Arguably, G-protein-coupled receptors (GPCRs) constitute the most pharmacologically important family of such receptors, binding small molecules, peptides, lipids, and hormones with high specificity. However, despite a highly conserved fold and sequence similarity, GPCRs are still mostly studied on a case-by-case basis.

Here, Lucie Delemotte infers a general, evolutionarily conserved mechanism of class A GPCR activation. By leveraging coevolution and machine learning methods applied to all class A GPCRs structures, she derives a mathematical description (a so-called collective variable – CV) of the receptor’s activation state which is independent of its sequence. Then, she biases molecular dynamics simulations along this CV to obtain transitions between activation states of a diverse set of class A GPCR family members. To demonstrate that the model generalizes beyond GPCRs in her training set, she obtains conformational transitions of an orphan receptor, GPR183. Finally, she will show that it is possible to model ligand effect on the receptors by converging Free Energy Surfaces of activation of the β2-adrenergic receptor within this common mechanism framework.

These results for the first time the existence of a mechanism uniting all class A GPCRs. The approach used here thus facilitates direct comparisons between receptors and opens up the possibility of structural and dynamical studies of many orphan and understudied GPCRs. It also serves as a blueprint for inferring family-wide protein mechanisms.