Deep Learning Dynamic Allostery of G-Protein-Coupled Receptors

Abstract

G-protein-coupled receptors (GPCRs) are the largest superfamily of human membrane proteins and represent primary targets of ~1/3 of currently marketed drugs. Allosteric modulators have emerged as more selective drug candidates compared with orthosteric agonists and antagonists. However, many X-ray and cryo-EM structures of GPCRs resolved so far exhibit negligible differences upon binding of positive and negative allosteric modulators (PAMs and NAMs). Mechanism of dynamic allosteric modulation in GPCRs remains unclear. In this work, we have systematically mapped dynamic changes in free energy landscapes of GPCRs upon binding of allosteric modulators using the Gaussian accelerated molecular dynamics (GaMD), Deep Learning (DL) and free energy prOfiling Workflow (GLOW). A total of 18 available high-resolution experimental structures of allosteric modulator-bound class A and B GPCRs were collected for simulations. A number of 8 computational models were generated to examine selectivity of the modulators by changing their target receptors to different subtypes. All-atom GaMD simulations were performed for a total of 66 microseconds on 44 GPCR systems in the presence/absence of the modulator. DL and free energy calculations revealed significantly reduced conformational space of GPCRs upon modulator binding. While the modulator-free GPCRs often sampled multiple low-energy conformational states, the NAMs and PAMs confined the inactive and active agonist-G protein-bound GPCRs, respectively, to mostly only one specific conformation for signaling. Such cooperative effects were significantly reduced for binding of the selective modulators to non-cognate receptor subtypes in the computational models. Therefore, comprehensive DL of extensive GaMD simulations has revealed a general dynamic mechanism of GPCR allostery, which will greatly facilitate rational design of selective allosteric drugs of GPCRs.

Diverse toxins exhibit a common binding mode to the Nicotinic Acetylcholine Receptors

“…Since the boost potential usually exhibits a near Gaussian distribution, GaMD can maintain the overall shape of the original free energy surface, and cumulant expansion to the second order (“Gaussian approximation”) can be applied to achieve proper energy reweighting 16 . GaMD has been demonstrated to properly sample the biological processes of ligand binding 17 and protein conformational changes upon ligand binding 18 .…”