Supramolecular organization of NMDA receptors and the postsynaptic density

Abstract

The postsynaptic density (PSD) of all vertebrate species share a highly complex proteome with ∼1000 conserved proteins that function as sophisticated molecular computational devices. Here, we review recent studies showing that this complexity can be understood in terms of the supramolecular organization of proteins, which self-assemble within a hierarchy of different length scales, including complexes, supercomplexes and nanodomains. We highlight how genetic and biochemical approaches in mice are being used to uncover the native molecular architecture of the synapse, revealing hitherto unknown molecular structures, including highly selective mechanisms for specifying the assembly of NMDAR-MAGUK supercomplexes. We propose there exists a logical framework that precisely dictates the subunit composition of synaptic complexes, supercomplexes, and nanodomains in vivo.

Figure 1. Hierarchy of supramolecular organization in the postsynaptic proteome.

The genome and transcriptome encode individual proteins and instructs their hierarchical organization at different length scales into complexes, supercomplexes and nanoclusters. Different synapses express different numbers of nanoclusters and these synapses are differentially distributed into different brain regions, as indicated by the colour scheme (pink, regions of brain with predominantly single nanocluster synapses; blue, regions with multiple nanoclusters) of the mouse hippocampus (adapted from Ref. [44•]).

Figure 2. Supramolecular ‘fingerprint’ of 65 forebrain proteins.

Adapted with permission from Ref. [5]. Native assemblies were detected by blue non-native PAGE immunoblot of mouse forebrain extracted with various different detergents. Expected and unexpected/unknown native protein assemblies within each lane are indicated by open and filled arrowheads, respectively. Native molecular mass indicated in mega-Daltons (MDa).

Figure 3. The tripartite rule governs NMDAR-MAGUK supercomplex assembly.

(a) The tripartite rule describes the genetic requirement of three proteins that are essential for the assembly of NMDA-MAGUK synaptic supercomplexes in vivo. Schematic of NMDA receptor subunits (GluN1, GluN2, GluN3) in membrane showing the cytoplasmic tail of GluN2B interacts with PSD95 and PSD93. The assembly of NMDAR-MAGUK supercomplexes does not depend on the ESDV C-terminal PDZ binding site. (b) Schematic showing how subunits of NMDA receptors assemble into three receptor complex subtypes and that only those that contain GluN2B can assemble into supercomplexes because of the tripartite rule.

Figure 4. Organization of a family of ~1.5 MDa synaptic supercomplexes.

PSD95-containing supercomplexes can be subdivided into a population containing NMDA receptors (NMDAR) and those lacking NMDA receptors (Non-NMDAR). Each of these can be further subdivided into subpopulations according to their assembly with Kir2.3, IQsec2, Adam22 or Arc.