TL;DR
This study maps the whole-genome DNA methylation landscape of the HL-60/S4 human myeloid leukemia cell line as it differentiates into macrophage-like and neutrophil-like states. The research reveals that while the genome remains stable, differentiation primarily drives changes in methylation at enhancers and CTCF binding sites, suggesting a crucial role in large-scale transcriptional reprogramming and chromatin organization.
What Problem Does This Paper Address?
Cell differentiation is a fundamental biological process governed by complex regulatory mechanisms, with epigenetics playing a central role. DNA methylation, a key epigenetic mark, is known to influence gene expression and cellular identity. However, the precise genome-wide methylation changes that occur during chemically induced differentiation of myeloid leukemia cell lines, and how these changes orchestrate the transition to specific cell fates, remain incompletely understood. This paper aims to fill this gap by providing a comprehensive analysis of DNA methylation dynamics in the HL-60/S4 cell line during its differentiation into granulocyte-like (RA-treated) and macrophage-like (TPA-treated) states, integrating this with existing transcriptomic and chromatin data.
Methods
The core technique employed was whole-genome bisulfite sequencing (WGBS) to precisely map DNA methylation across the entire genome in three distinct states: undifferentiated (UN), RA-treated (granulocyte-like), and TPA-treated (macrophage-like) HL-60/S4 cells.
Key analytical steps included:
- Karyotyping: To confirm genomic stability throughout differentiation.
- Differential Methylation Analysis: Identifying differentially methylated positions (DMPs) and regions (DMRs) between cell states using Fisher's exact test.
- Genomic Feature Enrichment Analysis: Determining which genomic elements (e.g., enhancers, CpG islands, CTCF binding sites, genes) were disproportionately represented in DMPs.
- Principal Component Analysis (PCA): To visualize global methylation patterns and assess similarities/differences between cell states.
- Module Analysis: Unsupervised clustering of DMPs to identify distinct methylation patterns associated with differentiation.
- Integration with Existing Data: Correlating methylation changes with previously published transcriptomic data (from Mark Welch et al., 2017) and nucleosome occupancy data (from Teif et al., 2017).
- Chromatin Interaction Analysis: Utilizing data from IM-PET and CHIA-PET experiments to link distal regulatory elements to gene expression.
Datasets
The primary dataset generated for this study was whole-genome bisulfite sequencing data from three states of the HL-60/S4 cell line:
- Undifferentiated (UN)
- RA-treated (granulocyte-like)
- TPA-treated (macrophage-like)
The study also integrated previously published datasets:
- Transcriptome data (RNA-seq) from the same HL-60/S4 cell line differentiation states (Mark Welch et al., 2017).
- Nucleosome occupancy data from the same HL-60/S4 cell line differentiation states (Teif et al., 2017).
- Chromatin interaction data (IM-PET and CHIA-PET) from relevant cell types (e.g., K562, primary monocytes) from the ENCODE and ROADMAP epigenome projects.
No new benchmarks were introduced; rather, the study focused on integrating existing data types to provide a multi-omics perspective.
Key Insights and Ideas
- Genomic Stability During Differentiation: Despite significant cellular changes, the HL-60/S4 cell line maintains a remarkably stable genome throughout chemically induced differentiation, as confirmed by karyotyping.
- Macrophage Differentiation Drives Major Methylation Changes: While RA-induced granulocyte differentiation shows subtle methylation shifts, TPA-induced macrophage differentiation is associated with more pronounced and widespread changes in DNA methylation.
- Enhancers and CTCF Sites are Key Targets: The most significant enrichment of methylation changes was observed in enhancers and CTCF binding sites, suggesting these regions are critical for orchestrating the large-scale transcriptional reprogramming and chromatin reorganisation required for differentiation.
- Immune-Related Functions Dominate Promoter Methylation: Differentially methylated promoters were strongly associated with immune response-related biological functions, highlighting the role of methylation in regulating immune gene expression during myeloid differentiation.
- Complex Correlation Between Methylation and Expression: The relationship between DNA methylation and gene expression is not always a simple negative correlation. While some genes show expected inverse relationships (e.g., promoter methylation correlating with reduced expression), others, particularly those involving transcriptional repressors like EZH2, can exhibit positive correlations.
- Long-Range Regulatory Elements Play a Role: Chromatin interaction data revealed that distal regulatory elements (enhancers) can influence gene expression through long-range interactions, and methylation changes in these regions can have significant impacts.
Key Results
- Genome-wide CpG Methylation Rates: All three cell states exhibited very similar genome-wide CpG methylation rates, with most CpGs being highly methylated (>0.75).
- Differential Methylated Positions (DMPs): A total of 41,306 unique DMPs were identified. The TPA vs. UN comparison yielded more DMPs than RA vs. UN.
- Enhancer and CTCF Enrichment: Enhancers and CTCF binding sites were significantly enriched in hypermethylated DMPs in both RA and TPA comparisons to UN cells. CTCF enrichment was specifically noted in TPA hypermethylated DMPs.
- DMP Modules: Unsupervised clustering identified 12 distinct DMP "modules." Module M6 showed a strong enrichment for enhancers and hypomethylation in TPA cells, suggesting a role in macrophage differentiation.
- Immune Gene Association: Functional annotation of genes with differentially methylated promoters revealed a strong enrichment for immune response-related terms in both RA and TPA comparisons. Key immune genes like CEBPA, GFI1, MAFB, and GATA1 showed differential expression and methylation.
- Correlation Analysis: While some genes showed negative correlation between promoter methylation and expression, others, particularly those involving transcriptional repressors, showed positive correlations. Long-range regulatory regions showed a stronger tendency for negative correlation.
Limitations and Future Work
- Cell Line Model: While HL-60/S4 is a valuable model, it is a leukemic cell line and may not perfectly recapitulate all aspects of primary myeloid differentiation.
- Chemical Induction: The differentiation is chemically induced, which might introduce specific pathways or epigenetic changes not entirely representative of physiological differentiation.
- Limited Resolution of Some Data: Chromatin interaction data was derived from different cell types than HL-60/S4, which could limit direct interpretation.
- Future Work: The authors propose a model of HL-60/S4 differentiation based on their findings, highlighting the roles of key transcription factors and epigenetic modifications. Further research could explore the precise mechanisms by which enhancers and CTCF sites mediate transcriptional reprogramming and investigate the functional consequences of hypomethylation in simple repeat elements. The study also suggests that understanding the context of regulatory regions is crucial for interpreting DNA methylation's effects on transcription.
Summary updated