Altered Molecular Composition of a Specific Subset of Prefrontal Cortical Excitatory Synapses in Schizophrenia

Abnormal excitatory synaptic transmission in the human prefrontal cortex has been implicated in the pathophysiology of schizophrenia based primarily on genetic evidence. However, changes in synaptic function cannot be predicted from altered gene expressions, but determining the amount, density, and subsynaptic distribution of synaptic proteins is the only reliable indirect readout of function. Detecting proteins in individual synapses of human postmortem tissues has been severely constrained by technical limitations. Here we overcome this limitation by optimizing a high-resolution, quantitative localization method to facilitate antigen recognition at excitatory synapses in postmortem brains of both sexes. Using PSD-95 immunoreactivity as molecular marker of excitatory synapses, we demonstrate the lack of significant differences in synapse density and size in upper cortical layers of control and schizophrenia subjects. The synaptic densities of postsynaptic AMPA and NMDA receptor subunits and presynaptic molecules Bassoon and Munc13-1 are also indistinguishable between control and schizophrenia subjects. The number of Munc13-1 nanoclusters, marking presynaptic neurotransmitter release sites, does not differ either. Excitatory synapses on parvalbumin expressing interneurons contain similar AMPA but significantly lower NMDA receptor densities in schizophrenia compared with control subjects. Our study provides the first comprehensive comparison of key functionally relevant synaptic proteins in individual human excitatory synapses and demonstrates that changes in the molecular composition of only a specific subset of excitatory synapses may contribute to the pathophysiology of schizophrenia.