NMDAR dysfunction in autism spectrum disorders: Lessons learned from 10 years of study

Over the past decade or so, mouse models of autism spectrum disorders (ASD) have been extensively studied in the search for key mechanisms underlying the disorder. Numerous intriguing mechanisms have been proposed, spanning various levels of the neural system, including molecular, synaptic, neuronal, circuit, and systems-level processes. However, no single mechanism has emerged as universally applicable, highlighting the heterogeneous nature of the genetic and neurobiological underpinnings of ASD. Among these, the NMDA receptor (NMDAR) dysfunction hypothesis has garnered significant attention. Many mouse models exhibit NMDAR dysfunction, with NMDAR hypofunction appearing more prevalent than hyperfunction. Nevertheless, not all mouse models display this dysfunction, suggesting that NMDAR abnormalities may not be ubiquitous across models, or that we have yet to fully explore the spectrum of NMDAR-related dysfunction in ASD. These findings underscore the need to consider multiple factors when studying ASD mouse models, including different mutations within the same gene, gene deletion dosage, genetic background, sex, age, brain regions, cell types, and neural circuits.