Ligand binding and dynamics of the GABAB receptor Venus flytrap domain

Abstract

<p>The gamma-amino-butyric-acid (GABA) is the main inhibitory neurotransmitter in the central nervous system (CNS) and exerts its physiological role by binding to the ionotropic GABA_A and GABA_C receptors and the metabotropic GABA_B receptor (GABA_B-R). The GABA_B-R is an obligate heterodimer that belongs to class C of guanine-binding proteins (G-protein) coupled receptors (GPCRs). Each monomer, GABA_B1a/b and GABA_B2, is comprised of an extracellular bi-lobed domain connected by a short loop to a heptahelical transmembrane domain (7TM). The extracellular domain is called the Venus flytrap (VFT) due to the architectural and mechanical resemblance to the carnivorous flower. The GABA_B1a/b VFT contains the orthosteric GABA binding site, while the 7TM domain of GABA_B2 hosts an allosteric binding site and is responsible for binding of G-proteins. <p>Previous studies have shown that the GABA_B-R is associated with numerous neurological and neuropsychiatric disorders including learning and memory deficits, depression and anxiety, addiction and epilepsy. The role of GABA_B-R in pathophysiology makes it an exciting target for drug interventions, especially since there is only one drug on the market targeting the receptor, the agonist baclofen. <p>In the present study, we tested the applicability of both classical structure-based and ligand-based methods in a virtual screening (VS) workflow using databases containing in total 8.2 million compounds, to discover novel orthosteric ligands targeting the GABA_B-R. A total of 34 ligands were purchased and tested in a functional cAMP assay. Currently, two of the compounds have showed antagonistic properties. We also used classical molecular dynamics (MD) simulations in combination with metadynamics to gain insight into the structural movements of the VFT. The results show that the barrier between the open and closed conformation is high, indicating that a ligand might be needed for receptor transition between the open/inactive and closed/active states, but additional simulations are needed to confirm the result.