Back

Single-cell dissection of schizophrenia reveals neurodevelopmental-synaptic axis and transcriptional resilience

Ruzicka, W. B.; Mohammadi, S.; Davila-Velderrain, J.; Subburaju, S.; Tso, D. R.; Hourihan, M.; Kellis, M.

2020-11-09 psychiatry and clinical psychology
10.1101/2020.11.06.20225342 medRxiv
Show abstract

Schizophrenia is a devastating mental disorder with a high societal burden, complex pathophysiology, and diverse genetic and environmental risk factors. Its complexity, polygenicity, and small-effect-size and cell-type-specific contributors have hindered mechanistic elucidation and the search for new therapeutics. Here, we present the first single-cell dissection of schizophrenia, across 500,000+ cells from 48 postmortem human prefrontal cortex samples, including 24 schizophrenia cases and 24 controls. We annotate 20 cell types/states, providing a high-resolution atlas of schizophrenia-altered genes and pathways in each. We find neurons are the most affected cell type, with deep-layer cortico-cortical projection neurons and parvalbumin-expressing inhibitory neurons showing significant transcriptional changes converging on genetically-implicated regions. We discover a novel excitatory-neuron cell-state indicative of transcriptional resilience and enriched in schizophrenia subjects with less-perturbed transcriptional signatures. We identify key trans-acting factors as candidate drivers of observed transcriptional perturbations, including MEF2C, TCF4, SOX5, and SATB2, and map their binding patterns in postmortem human neurons. These factors regulate distinct gene sets underlying fetal neurodevelopment and adult synaptic function, bridging two leading models of schizophrenia pathogenesis. Our results provide the most detailed map to date for mechanistic understanding and therapeutic development in neuropsychiatric disorders.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.