Metabolic control of methylation landscapes in cancer growth and plasticity: targeting the epigenetic–metabolic axis with natural compounds and emerging therapies

Cancer cell plasticity enables tumor cells to dynamically transition between diverse phenotypic states, including epithelial–mesenchymal plasticity stemness, dormancy, lineage switching, immune evasion, and therapy-resistant phenotypes. Although metabolic reprogramming and epigenetic regulation have each been recognized as hallmarks of cancer, the mechanistic framework linking metabolic adaptation to dynamic cell-state transitions remains incompletely understood. In this Review, we propose that methylation serves as a major mechanistic interface through which metabolic rewiring can influence adaptive cancer cell plasticity. We discuss how one-carbon metabolism regulates the availability of the universal methyl donor S-adenosylmethionine (SAM), thereby influencing DNA methylation, histone methylation, and multiple forms of RNA methylation, with N6-methyladenosine (m 6 A) representing the principal RNA modification discussed in this review. Rather than responding uniformly to metabolic change, these methylation systems represent distinct but metabolically connected regulatory layers that share dependence on one-carbon metabolism while differing in their biochemical sensitivities and regulatory outputs. Net methylation states, however, are also shaped by metabolically constrained demethylation, including α-ketoglutarate-, oxygen-, Fe 2+ -, and flavin adenine dinucleotide (FAD)-dependent eraser pathways. Together, these metabolically responsive writing and erasing processes influence transcriptional stability, chromatin accessibility, and post-transcriptional regulation during reversible cell-state transitions. We further examine how environmental and therapeutic stresses, including hypoxia, nutrient limitation, immune surveillance, and anticancer therapy, reshape methylation programs through metabolic adaptation, and how such remodeling may support phenotypic flexibility and survival under selective pressure. Building upon these advances, we propose the concept of the adaptive methylation landscape, a perturbation-responsive, state-associated, and potentially reversible distribution of methylation states that contributes to shaping the spectrum of cellular states accessible during tumor evolution. Finally, we discuss emerging therapeutic strategies targeting the metabolic-methylation axis and highlight how single-cell multi-omics, spatial epigenomics, functional epigenome editing, and AI may enable dynamic mapping of adaptive methylation landscapes. We propose that cancer cell plasticity is not merely a genetic phenomenon but can also be shaped by metabolically responsive methylation programs, providing a testable conceptual framework for future precision epigenetic therapies.