Cellular heterogeneity in the human brain obscures the identification of robust cellular regulatory networks, which is necessary to understand the function of non-coding elements and the impact of non-coding genetic variation. Here we integrate genome-wide chromosome conformation data from purified neurons and glia with transcriptomic and enhancer profiles, to characterize the gene regulatory landscape of two major cell classes in the human brain. We then leverage cell-type-specific regulatory landscapes to gain insight into the cellular etiology of several brain disorders. We find that Alzheimer's disease (AD)-associated epigenetic dysregulation is linked to neurons and oligodendrocytes, whereas genetic risk factors for AD highlighted microglia, suggesting that different cell types may contribute to disease risk, via different mechanisms. Moreover, integration of glutamatergic and GABAergic regulatory maps with genetic risk factors for schizophrenia (SCZ) and bipolar disorder (BD) identifies shared (parvalbumin-expressing interneurons) and distinct cellular etiologies (upper layer neurons for BD, and deeper layer projection neurons for SCZ). Collectively, these findings shed new light on cell-type-specific gene regulatory networks in brain disorders.

Despite progress in defining genetic risk for psychiatric disorders, their molecular mechanisms remain elusive. Addressing this, the PsychENCODE Consortium has generated a comprehensive online resource for the adult brain across 1866 individuals. The PsychENCODE resource contains ~79,000 brain-active enhancers, sets of Hi-C linkages, and topologically associating domains; single-cell expression profiles for many cell types; expression quantitative-trait loci (QTLs); and further QTLs associated with chromatin, splicing, and cell-type proportions. Integration shows that varying cell-type proportions largely account for the cross-population variation in expression (with > 88% reconstruction accuracy). It also allows building of a gene regulatory network, linking genome-wide association study variants to genes (e.g., 321 for schizophrenia). We embed this network into an interpretable deep-learning model, which improves disease prediction by ~6-fold versus polygenic risk scores and identifies key genes and pathways in psychiatric disorders.

Progranulin (GRN) haploinsufficiency is a frequent cause of familial frontotemporal dementia, a currently untreatable progressive neurodegenerative disease. By chemical library screening, we identified suberoylanilide hydroxamic acid (SAHA), a Food and Drug Administration-approved histone deacetylase inhibitor, as an enhancer of GRN expression. SAHA dose-dependently increased GRN mRNA and protein levels in cultured cells and restored near-normal GRN expression in haploinsufficient cells from human subjects. Although elevation of secreted progranulin levels through a post-transcriptional mechanism has recently been reported, this is, to the best of our knowledge, the first report of a small molecule enhancer of progranulin transcription. SAHA has demonstrated therapeutic potential in other neurodegenerative diseases and thus holds promise as a first generation drug for the prevention and treatment of frontotemporal dementia.