Computational assessment of the relationship between metabolism and histone methylation in cancer cells.

Abstract

Aberrant histone methylation and metabolic alterations are key hallmarks of cancer. Metabolic reprogramming during tumorigenesis could impact the histone methylation pattern by altering the availability of substrates and cofactors required for histone methyltransferases (HMTs) and demethylases (HDMs) activities. Despite advances in understanding this complex interplay, quantitative information about the contributions of specific metabolic shifts and histone methylation dynamics remains poorly understood. Here, we used multi-omics data integrated with machine learning models to discover key metabolites, genes, and pathways predictive of histone methylation levels in cancer cell lines. Our cell line models highlighted the significant role of metabolites associated with one-carbon, nucleotide, redox and lipid metabolism on histone marks. Validation in primary tumors confirmed the cell line models' findings. Overall, this study quantifies the contributions of the metabolic network to histone methylation variation in cancer cells.