Summary Neuroligins and neurexins promote synapse development and validation by forming trans-synaptic bridges spanning the synaptic cleft. Select pairs promote excitatory and inhibitory synapses, respectively, with neuroligin 2 (NLGN2) limited to inhibitory synapses and neuroligin 1 (NLGN1) dominating at excitatory synapses. The cell surface molecules, MAM domain-containing glycosylphosphatidylinositol anchor 1 (MDGA1) and 2 (MDGA2), regulate trans-synaptic adhesion between neurexins and neuroligins, impacting NLGN2 and NLGN1, respectively. We have determined the molecular mechanism of MDGA action. MDGA1 Ig1–Ig2 is sufficient to bind NLGN2 with nanomolar affinity; its crystal structure reveals an unusual locked rod-shaped array. In the crystal structure of the complex, two MDGA1 Ig1–Ig2 molecules each span the entire NLGN2 dimer. Site-directed mutagenesis confirms the observed interaction interface. Strikingly, Ig1 from MDGA1 binds to the same region on NLGN2 as neurexins do. Thus, MDGAs regulate the formation of neuroligin-neurexin trans-synaptic bridges by sterically blocking access of neurexins to neuroligins.

No abstract

No abstract

No abstract

While the pathogenesis of schizophrenia (SCZ) is proposed to involve alterations of neural circuits via synaptic dysfunction, the underlying molecular mechanisms remain poorly understood. Recent exome sequencing studies of SCZ have uncovered numerous single nucleotide variants (SNVs); however, the majority of these SNVs have unknown functional consequences, leaving their disease relevance uncertain. Filling this knowledge gap requires systematic application of quantitative and scalable assays to assess known and novel biological functions of genes. Here we demonstrate loss-of-function effects of multiple rare coding SNVs found in SCZ subjects in the GIT1 (G protein-coupled receptor kinase interacting ArfGAP 1) gene using functional cell-based assays involving co-expression of GIT1 and PAK3 (p21 protein (Cdc42/Rac)-activated kinase 3). Most notably, a GIT1-R283W variant reported in four independent SCZ cases was defective in activating PAK3 as well as MAPK (Mitogen-activated protein kinase). Similar functional deficits were found for a de novo SCZ variant GIT1-S601N. Additional assays revealed deficits in GIT1-R283W’s capacity to stimulate PAK phosphorylation in cultured hippocampal neurons. Also, GIT1-R283W showed deficits in the induction of GAD1 (Glutamate decarboxylase 1) protein expression. Extending these functional assays to ten additional rare GIT1 variants revealed the existence of an allelic series with the majority of the SCZ case variants exhibiting loss-of-function towards MAPK activation in a manner correlated with loss of PAK3 activation. Taken together, we propose that rare variants in GIT1, along with other genetic and environmental factors, cause dysregulation of PAK3 leading to synaptic deficits in SCZ.