Paper discussion: Single-cell DNA methylation and 3D genome architecture in the human brain

This presentation includes slides used when discussing the manuscript by Wei Tian et al., available from https://doi.org/10.1126/science.adf5357, at an LIBD (https://www.libd.org/) team meeting.

Transcript

1. Single-cell DNA methylation and 3D genome architecture in the human brain Wei Tian et al. DOI: 10.1126/science.adf5357 Presented by: Nick Eagles 2024/04/03
2. Background: DNA Methylation and Chromatin Conformation - Chromatin conformation and DNA methylation are key to understanding regulation of gene expression - Chromatin has a 3D arrangement controlling interaction between gene promoters and regulatory elements - Active vs. repressive compartments - Topologically associating domains - Chromatin loops - Methylation is a dynamic chemical modification occurring at cytosines, affecting expression of genes - Manuscript aim: profile chromatin conformation and DNAm at single-cell resolution to produce an epigenetic atlas of the human brain
3. Study Design - 46 brain regions in adult humans: cerebral cortex, basal forebrain, basal nuclei, hippocampus, thalamus, midbrain, pons, cerebellum - 3 biological replicates (adult males) in almost all regions - “Fluorescence-activated nuclei sorting (FANS) was used to isolate 90% NeuN-positive and 10% NeuN-negative cells” - Several single-cell assays! - snmC-seq3 (DNAm at 46 regions) - snm3C-seq (DNAm + chromatin conformation at 17 regions) - snRNA-seq + snATAC-seq for confirmation
4. Fig 1
5. Trait/Disorder Association with DMRs - Used linkage disequilibrium score regression (LDSC) - SCZD + bipolar risk variants enriched in hypo-DMRs in cortical + hippocampal excitatory neurons (Fig 3H) - Alzheimer’s: DMRs in microglia (Fig 3H) Fig 3 Fig S15
6. Fig 4A+B: “Regional Methylation Space” - 4A: low-dimensional space where lower distance between cells indicates similar brain regions and methylation profiles - 4B: cortical neurons placed in this space, colored by dissection location
7. Fig 4B-D: “Regional Methylation Space” - 4C: Like 4B, but flattened along the primary “axis” of variation and discretized by region
8. Fig 6A: snMCodes - Iteratively finding a subset of methylation sites that predict cell type
9. Results I didn’t cover - Assessing conservation of cell types and DMRs across species (in this case human vs mouse) - Chromatin contact distance comparison by cell type (neurons tended to have shorter interaction distances!) - Chromatin features by cell type (compartments, domains, loops) - Chromatin features compared against methlyation
10. Fig 3G Caption: "Gene body mCH, DMR mCG, and 3D chromatin organization around the gene SYT1 in the major types L2/3-IT and MSN-D1." Text reference: "For instance, the gene SYT1, encoding Synaptotagmin-1, a critical synaptic vesicle protein, exhibited lower methylation fractions of both the distal DMRs and the SYT1 gene body in L2/3-IT neurons and stronger interactions between the DMRs and the promoter compared with MSN-D1 neurons (Fig. 3G), leading to a higher expression of SYT1 in L2/3-IT neurons than in MSN-D1 neurons ( Fig. 3G and fig. S14E)"