Stratifying TAD boundaries pinpoints focal genomic regions of regulation, damage, and repair

Bijia Chen, Chao Ren, Zhangyi Ouyang, Jingxuan Xu, Kang Xu, Yaru Li, Hejiang Guo, Xuemei Bai, Mengge Tian, Xiang Xu, Yuyang Wang, Hao Li, Xiaochen Bo, Hebing ChenBriefings in Bioinformatics 2024 · DOI

TL;DR

This study introduces a novel method to stratify topologically associating domain (TAD) boundaries based on how many TADs they separate. This stratification reveals that "high-level" boundaries are not just structural elements but are crucial hubs for gene regulation, DNA damage, and repair, offering new insights into genomic stability and disease.

What Problem Does This Paper Address?

The 3D organization of the genome, particularly into structures like TADs, is known to be critical for gene regulation and cellular function. However, the precise functional implications of TAD boundaries, especially those that are shared across multiple TADs, remain unclear. While TAD boundaries are generally associated with active gene expression and are implicated in various diseases when disrupted, there's a need to understand if different types of boundaries have distinct roles. This paper aims to dissect the functional significance of these shared TAD boundaries by stratifying them and investigating their relationship with gene characteristics, epigenetic features, and DNA damage/repair mechanisms.

Methods

The researchers employed a multi-omics approach, integrating high-resolution Hi-C data with chromatin accessibility (DNase-seq) and DNA double-strand break (DSB) data from multiple cell lines.

  • TAD Boundary Stratification: They used the OnTAD algorithm to identify hierarchical TAD structures. A key innovation here is the stratification of TAD boundaries based on the "level" of sharing – how many TADs flank a given boundary. Boundaries shared by more TADs are classified as higher-level.
  • Gene Characterization: Genes were analyzed for expression levels, essentiality (using CRISPR scores), length, and evolutionary age.
  • Epigenetic Features: Chromatin accessibility (DNase-seq) and transcription factor binding site (TFBS) clustered regions (TFCRs) were investigated. TFCR complexity was assessed.
  • DNA Damage and Repair: DSB data was used to identify regions prone to breaks. Gene Ontology (GO) analysis was performed on genes at high-level boundaries to identify enriched functions, particularly DNA repair. Motif analysis was used to identify DSB-associated motifs at boundaries.
  • Validation: The identified TAD boundaries were cross-validated against five other established TAD calling methods to assess consistency and robustness.

Datasets

The study utilized publicly available datasets:

  • Hi-C data: From Rao et al. [4] and Johnstone et al. [28], covering both normal and tumor cell lines.
  • DNase-seq data: From the ENCODE consortium.
  • DNA DSB data: From GEO database (GSE129529 and GSE78172).
  • RNA-seq data: From the ENCODE consortium.
  • Gene annotations: From GENCODE (V15).
  • CRISPR scores: From Wang et al. [29].
  • Repair-related genes: From Hussmann et al. [61].

Key Insights and Ideas

  • "Hub-Boundaries" as Functional Centers: High-level TAD boundaries (those shared by many TADs) are not just passive dividers but active "hub-boundaries" that integrate multiple regulatory processes.
  • Vulnerability and Resilience: These hub-boundaries are paradoxically enriched for both DNA double-strand breaks (DSBs) and DNA repair-related genes, suggesting a sophisticated mechanism for maintaining genomic stability despite inherent fragility.
  • Complexity Drives Regulation and Repair: Higher complexity of transcription factor binding regions (TFCRs) at high-level boundaries correlates with increased gene expression, essentiality, and a stronger association with DSB motifs, hinting at a complex regulatory machinery.
  • Evolutionary Conservation: High-level boundaries are more conserved across cell lines, underscoring their fundamental importance for genomic integrity.
  • "Sharing the Danger, Sharing the Fix": The concept that these boundaries are sites of both damage and coordinated repair mechanisms is a central theme.

Key Results

  • Gene Characteristics: Genes at higher-level boundaries are more actively expressed, have higher essentiality, are shorter, and are evolutionarily older (more conserved).
  • TFCRs: Higher-level boundaries show increased TFCR complexity, with the most complex TFCRs (TC9) being significantly enriched at these locations.
  • DSB Enrichment: DSBs are significantly enriched at TAD boundaries, with a notable increase at higher-level boundaries, especially those co-localizing with high-complexity TFCRs.
  • DNA Repair Gene Enrichment: Genes located at high-level boundaries are significantly enriched for DNA repair functions, including key players like TOPBP1 and BRCA2.
  • Boundary Consistency: High-level TAD boundaries identified by the OnTAD algorithm show good consistency with those identified by multiple other TAD calling methods.

Limitations and Future Work

The authors acknowledge several limitations:

  • Lack of Experimental Validation: The study relies heavily on computational analysis, and experimental validation is needed to confirm the findings.
  • Tissue Type Diversity: While multiple cell lines were used, exploring a wider range of tissue types could improve generalizability.
  • Single-Cell Analysis: Bulk analyses might overlook cell-to-cell variability. Single-cell resolution could reveal finer nuances in chromatin organization and repair.

Future work could focus on experimentally validating these findings, expanding the analysis to diverse tissues, and employing single-cell technologies to gain a more granular understanding of TAD boundary function. The study also opens avenues for therapeutic interventions targeting these crucial genomic regions.

Summary updated

BibTeX
@article{chen2024stratifying,
  author = {Bijia Chen and Chao Ren and Zhangyi Ouyang and Jingxuan Xu and Kang Xu and Yaru Li and Hejiang Guo and Xuemei Bai and Mengge Tian and Xiang Xu and Yuyang Wang and Hao Li and Xiaochen Bo and Hebing Chen},
  title = {{Stratifying TAD boundaries pinpoints focal genomic regions of regulation, damage, and repair}},
  journal = {Briefings in Bioinformatics},
  year = {2024},
  doi = {10.1093/bib/bbae306},
}
Figure