Conserved endothelial cell-specific DNA methylation alterations in colorectal and breast tumor microenvironments

DNA methylation alterations are well-established contributors to carcinogenesis, yet, within the tumor microenvironment (TME), lineage- and cell-specific patterns remain poorly characterized. We applied a bulk DNA methylation deconvolution approach (HiTIMED) combined with a cell-type-specific differential methylation framework (CellDMC) to identify and validate tumor endothelial cell (TEC) specific DNA methylation alterations in colorectal cancer and nontumor normal tissue. We previously identified TEC-specific DNA methylation alterations in breast cancer. Here, we evaluate the shared endothelial-specific epigenetic programs between breast and colorectal tumors. Endothelial cells, key regulators of angiogenesis and vascular homeostasis, often acquire molecular and structural abnormalities that promote tumor progression. Using genome-scale DNA methylation data from a discovery and validation framework (tumor n = 725; nontumor n = 989) of colorectal tissues, we identified 38,647 CpGs (FDR ≤ 0.05) with altered DNA methylation specific to TECs in both discovery and validation sets, corresponding to 13,249 genes. The majority of CpGs were hypermethylated in TEC compared to endothelial cells in normal tissue. Clusters of CpGs mapped to genes known to regulate angiogenesis, endothelial cell biology, genes in the VEGF signaling pathway, and members of the hallmark angiogenesis gene set. Comparative analysis of TEC-specific methylation alterations in the colorectal and breast TME showed a 24% overlap at the CpG level, and an 81% overlap at the gene level, suggesting conserved endothelial-specific methylation dysregulation programs between tumor types. Nearly all genes whose expression correlated with endothelial cell proportion in our previous breast cancer analysis were also identified in the colorectal datasets. These findings indicate a shared epigenetic architecture underlying endothelial cell reprogramming in distinct tumor contexts. Ongoing expression analyses will extend these comparisons in colorectal tumors to further link DNA methylation with transcriptional regulation. Together, our results demonstrate that high-resolution, lineage-specific DNA methylation landscapes can be reconstructed from bulk tumor DNA methylation profiles and establish conserved TEC-specific epigenetic signatures with implications for tumor vascular biology and precision therapeutic targeting.