Combined cellomics and proteomics analysis reveals shared neuronal morphology and molecular pathway phenotypes for multiple schizophrenia risk genes

An enigma in studies of neuropsychiatric disorders is how to translate polygenic risk into disease biology. For schizophrenia, where > 145 significant GWAS loci have been identified and only a few genes directly implicated, addressing this issue is a particular challenge. We used a combined cellomics and proteomics approach to show that polygenic risk can be disentangled by searching for shared neuronal morphology and cellular pathway phenotypes of candidate schizophrenia risk genes. We first performed an automated high-content cellular screen to characterize neuronal morphology phenotypes of 41 candidate schizophrenia risk genes. The transcription factors Tcf4 and Tbr1 and the RNA topoisomerase Top3b shared a neuronal phenotype marked by an early and progressive reduction in synapse numbers upon knockdown in mouse primary neuronal cultures. Proteomics analysis subsequently showed that these three genes converge onto the syntaxin-mediated neurotransmitter release pathway, which was previously implicated in schizophrenia, but for which genetic evidence was weak. We show that dysregulation of multiple proteins in this pathway may be due to the combined effects of schizophrenia risk genes Tcf4, Tbr1, and Top3b. Together, our data provide new biological functions for schizophrenia risk genes and support the idea that polygenic risk is the result of multiple small impacts on common neuronal signaling pathways.

Automated summary

A combined cellomics and proteomics approach was used to disentangle polygenic risk in schizophrenia by searching for shared neuronal morphology and cellular pathway phenotypes of candidate schizophrenia risk genes. It was found that certain schizophrenia risk genes shared a neuronal phenotype marked by a reduction in synapse numbers and converged onto the syntaxin-mediated neurotransmitter release pathway. This study provides new biological functions for schizophrenia risk genes and supports the idea that polygenic risk results from multiple small impacts on common neuronal signaling pathways.