Epigenetic modifications confer stable transcriptional patterns in the brain, and both normal and abnormal brain function involve specialised brain regions, yet little is known about brain region-specific epigenetic differences. Here, we performed, to our knowledge, the largest and most comprehensive genome-wide epigenetic analysis of sorted nuclei from functionally diverse human brain regions. We performed fluorescence-activated nuclei sorting (FANS) using the nuclear neuronal marker NeuN (RBFOX3) on nuclei from 4 different brain regions: anterior cingulate gyrus, hippocampus, prefrontal cortex, and nucleus accumbens. We generated 45 whole-genome bisulfite-sequencing samples from 6 donors, and 22 ATAC-seq samples and 20 RNA-seq samples from an additional 6 donors. All donors were unaffected with neurological or psychiatric disease at time of death.
In neurons, we identified 13,000 brain region-specific differentially methylated regions (neuronal DMRs; 12 Mb), enriched over putative enhancers and with the vast majority involving nucleus accumbens being distinct from the remaining 3 regions, and 19,000 differentially accessible ATAC-seq peaks (neuronal DAPs; 12 Mb). Surprisingly, these DMRs and DAPs largely did not overlap, although we can detect a weak, consistent change in DNA methylation in DAPs and chromatin accessibility in DMRs. In contrast, we found almost no epigenetic changes between non-neuronal cells from different brain regions. By comparing results from sorted and unsorted cell populations, we reiterate the critical importance of nuclei/cell sorting in neuroepigenetic studies.
Using linkage disequilibrium score regression, we found that neuronal DMRs and DAPs were highly enriched for SNP heritability of various neuropsychiatric diseases, notably schizophrenia and neuroticism, and addictive behaviour, such as cigarette smoking. In contrast, we found no enrichment for traits such as height or LDL cholesterol, as expected. This provides evidence of the functional importance of these regions and is consistent with the hypothesis that common psychiatric illness is mediated through brain region-specific neuronal epigenetic differences.
We provide a genome browser to aid integration of these data into subsequent studies. Our results have broad implications for future neuroepigenetic studies: nuclei/cell sorting is critical, it is important to sample diverse brain regions, and that comparisons of a cell type between brain regions may be more informative than comparisons of cell types within a brain region.