Does DNA Damage Impact the Transcriptome and Epigenome in a Cell-Specific Manner?

June 19, 2025 By Stuart P. Atkinson

Cortex cell-type UMAP and genomic feature enrichment bars. — UMAP of mouse cerebral-cortex single-cell clusters annotated by cell type and associated enrichment of DNA damage per cell type and genomic functional region. From Bai et al.

Does DNA Damage Impact the Transcriptome and Epigenome in a Cell-Specific Manner?

Parts 1 and 2 of this series of articles from the Epigenome Technologies blog summarized the description and validation of Paired-Damage-seq and described how this exciting technique could define relationships between DNA damage formation and epigenetic alterations, as reported in full in a recent Nature Methods paper from researchers led by Chenxu Zhu (New York Genome Center/Weill Cornell Medicine) (Bai et al.). Paired-Damage-seq hoped to support the study of how DNA damage and epigenomes interact by overcoming multiple obstacles faced by previous related techniques. Part 3 now describes the application of Paired-Damage-seq to cells of the mouse cerebral cortex, the exploration of cell-type-specific genome vulnerabilities to facilitate cell-type predictions, and the identification of dysregulated molecular programs contributing to disease risk. Importantly, the range of products and services that Epigenome Technologies provides can empower your research aims with flexible, high-resolution technologies that turn hidden regulatory layers into actionable discoveries ready to transform our understanding of health and disease.

DNA, RNA, ATAC tracks at marker genes by cell type. — Genome-browser tracks of Paired-Damage-seq DNA (top), Paired-Damage-seq RNA (middle) and snATAC-seq (bottom) at eight marker genes (Slc17a7–Ptgds) across eight brain cell types. From Bai et al.

Identifying Brain Cell-specific DNA Damage

The authors explored whether cell-type defining epigenetic states in complex tissues contribute to cell-type-specific genome vulnerabilities by performing Paired-Damage-seq on cells isolated from the mouse cerebral cortex

Cell clustering and annotation based on RNA profiles revealed all main brain subclasses

Aggregation of DNA damage signals for each cell group according to transcriptome-based clustering results generated mouse brain cell-specific maps of DNA damage hotspots

Higher DNA damage levels occurred at genes highly expressed in non-neuron cells, but DNA damage signals did not tend to occur at open chromatin regions
The relationship disappeared in neurons/non-brain-specific cells, suggesting intra-tissue/-cell heterogeneity regarding the impact of transcription-associated chromatin states on genome vulnerability

Compartment B contained higher DNA damage levels than Compartment A for all mouse brain cells

The results suggest that heterochromatin inhibits DNA repair and that the organization of genome regions closer to nuclei membrane via compartmentalization increases susceptibility to reactive oxygen species

DNA damage hotspots prefer CpG islands, LINEs, low complexity repeats, long terminal repeats, simple repeats, and satellites in brain cells

High DNA damage levels at LINE1 5′ untranslated regions for neuronal/non-neuronal cells (contain a variable number of GC-rich tandem repeats; Li et al.) suggests the conservation of base sequence compositions-associated DNA damage hotspots across cell types

Heatmap of peak expression with GO terms and motif logos. — Heatmap of DNA-damage peak frequency across cell types, with associated term enrichment and TF motifs. From Bai et al.

Neuronal Cis-regulatory Elements Identified as DNA Damage Hotspots

The comparison of cell-type-resolved DNA damage signals with public snATAC-seq and H3K27ac Paired-Tag datasets (Li et al. and Zhang et al.) explored whether DNA repair hotspots display enrichment in regulatory regions to protect the identities and functions of postmitotic cells

DNA damage levels positively correlated with enhancer-associated H3K27ac levels in specific brain cells at 250-kb resolution; however, they did not correlate with heterochromatin-associated H3K9me3 levels, even though heterochromatin suffers from higher average DNA damage levels
At a finer resolution, DNA damage displayed strong enrichment at H3K27ac peak regions in neurons and non-neuronal brain cells, while other cells displayed weaker enrichments
A higher fraction of active enhancers (H3K27ac+/ATAC-seq+ peaks) overlapped with DNA damage peaks compared to weak enhancers (H3K27ac-/ATAC-seq+ peaks) for brain-specific cells only

Disease GWAS enrichment heatmap across cell types. — Heatmap of relative enrichment of disease-associated variants (GWAS) across cell types. From Bai et al.

Exploring Cell-specific Genome Vulnerability

Differential analysis to explore the cell identity-dependence of DNA damage hotspots suggested that cis-regulatory elements associated with DNA damage peaks correlated with specific cell functions
Motif analysis of cell-type-specific damage peaks revealed potential regulators affected by DNA damage

The data indicated that brain cells display diverse patterns of selective genome vulnerability (shaped by conserved and brain-specific mechanisms), which functionally influence different cell identities

The identification of human orthologs of mouse DNA damage hotspots sought to define potential links between selective genome vulnerability and disease risk identified traits associated with specific cell types

Major depressive disorder in excitatory neurons represented the highest enriched trait, with elevated DNA damage known to occur in affected patients compared to healthy individuals (Szebeni et al.)
Additional examples included the associations of DNA damage hotspots with neurodegenerative disorders, including ALS in oligodendrocytes and astrocytes and Alzheimers disease in microglia
The data suggests the non-random and cell-type-specific nature of DNA damage accumulation, which may increase the risk of gene program dysregulation in different disease-associated cell types

Comparing DNA damage and H3K9me3 distribution patterns explored relationships between DNA damage and epigenetic alterations in regions outside cis-regulatory elements employing a dataset describing age-associated heterochromatin loss in the mouse brain (Zhang et al.)

Analysis of the top 1% most damaged regions revealed negative associations for excitatory and inhibitory neurons, oligodendrocytes, and astrocytes but not proliferative oligodendrocyte progenitor cells
These data suggest that the progressive loss of epigenome information in postmitotic cells may be caused by the persistent formation and repair of DNA damage

Damage versus H3K9me3-change scatter with PCC for ExN, ODC, AST. — Scatter of DNA damage levels (RPKM) versus ΔH3K9me3 (18 vs 3 months) in 100 kb bins for excitatory neurons (ExN), oligodendrocytes (ODC) and astrocytes (AST). From Bai et al.

Paired-Damage-seq: Brain Cell-specific DNA Damage and Cell-specific Genome Vulnerabilities

These data reveal how the application of Paired-Damage-seq to the mouse brain can identify hotspots for the age-associated loss of epigenetic information, with this cell-type-specific genome vulnerability perhaps linked to the risk of pathologic gene programs by DNA damage accumulation and epigenome erosion over time. Paired-Tag from Epigenome Technologies, which generates joint epigenetic and gene expression profiles at the single-cell resolution and detects histone modifications and RNA transcripts in individual nuclei with an efficiency comparable to single-nucleus RNA-seq/ChIP-seq assays.

Nature Methods, March 2025.