This paper is a review of how DNA methylation works and why new, gene-selective tools are needed to study and treat disease. DNA methylation means adding a small chemical group (a methyl) to DNA. That change can be used by some chemotherapy drugs to kill cancer cells. But in normal cells it is a basic way to control which genes are turned on or off.

The authors summarize decades of work on the enzymes and drugs that change DNA methylation. In mammals, three enzymes called DNA methyltransferases (DNMT1, DNMT3A and DNMT3B) add methyl groups at sites called CpG dinucleotides. The process uses a universal methyl donor called S-adenosyl-L-methionine (SAM) and involves a catalytic cysteine in the enzyme. Cells can also remove methyl groups. The ten-eleven translocation enzymes (TET1, TET2, TET3) oxidize the methylated base and then normal DNA repair restores an unmethylated base.

The review contrasts two uses of DNA methylation. One is toxic alkylation by chemotherapy drugs, such as temozolomide and older alkylating agents, which damage DNA and can kill cancer cells. The other is the normal, reversible role of methylation in gene control. CpG-rich regions called CpG islands often sit near gene promoters; about 70% of annotated promoters overlap such islands. The authors note other methylation patterns too, for example non-CpG methylation (mCH) is rare overall (about 1–3%) but seen in oocytes, neurons and embryonic stem cells.

Clinically, drugs that block DNMTs—Azacitidine and Decitabine—are approved for blood cancers such as myelodysplastic syndromes, chronic myelomonocytic leukemia, and acute myeloid leukemia. But these drugs act broadly across the genome. The paper stresses that global DNMT inhibitors lack gene selectivity, which limits benefit and raises off-target toxicity. It also makes it hard for researchers to study the specific roles of methylation at individual genes, a problem as high-resolution sequencing now maps methylation in fine detail.