Principal Component Analysis of Glutamate Receptor Ligand Binding Domains

Principal component analysis (PCA) provides a method for analyzing the characteristic large scale motions of a protein. Here, PCA was applied to ensembles of configurations taken from umbrella sampling simulations performed on the ligand binding domains of a variety of ionotropic glutamate receptors (iGluRs). iGluRs are key mediators of neuronal communication. Two iGluR families are AMPA and NMDA receptors. AMPARs require only glutamate for activation, while NMDARs are obligate heterotetramers that require both glycine and glutamate. We analyzed GluA2, a glutamate binding AMPA domain, GluN2A, a glutamate binding NMDA domain, GluN1 and GluN3A, glycine binding NMDA domains, and a bacterial iGluR homologue, GluR0. The goal of the study was to determine the impact that binding has on the principal components of the LBDs by comparing the apo and holo forms. We found that the LBDs have four common modes that can be described as "hinge bending", "rocking", "twisting", and "sweeping", with hinge bending being the dominant mode in all LBDs in both the apo and holo forms. The other modes, however, vary in their contributions to total motion for the different iGluR families. This variation could have functional impact in these receptors.