Harvard Medical School, Stanley Center at the Broad Institute, and Institute for Biological Psychiatry, Region H
Title: Using brain cell type-specific protein interactomes to interpret genetic data in schizophrenia
Recent studies have revealed the highly polygenic nature of psychiatric disorders, but the specific underling gene networks (pathways) remain obscure in many cases. This is a key bottleneck towards biological understanding and therapeutic intervention. We developed a production framework of homogeneous cell populations at scale (>10 billion neurons per year) and analyzed neuron-specific protein-protein interactions of risk genes using tandem mass spectrometry. In parallel, we developed a computational platform (Genoppi) to QC the data, to integrate proteomic and genetic datasets, and to prioritize emerging pathway models for functional validation experiments. Many of the identified high-quality protein interactions we identified are unique to human neurons (i.e., not reported in the literature nor seen in non-brain tissues). By integrating data from the SCHEMA exome sequencing consortium and the East Asia Cohort of the Psychiatric Genomics Consortium, we found that the interaction networks of several risk proteins (e.g., CACNA1C, HCN1, and SYNGAP1) are significantly enriched for common and/or rare risk variants, illustrating how common and rare genetic risk can converge onto the same cellular networks in human neurons.Overall, we describe an integrated experimental and computational framework to (1) map interactomes of risk proteins in human neurons, (2) model pathway relationships through integration of proteomic and genetic data, and (3) validate key risk network modules in human neurons, human brain tissue, and animal models. Our approach aims to contribute to gaining functional insights from genetic data in psychiatry.