DNA methylation in cancer: Gene regulation and biomarkers

Changes in DNA methylation can alter gene expression in cancer cells, affecting cellular processes such as growth and division. Measuring methylation across multiple sites can also reveal patterns that distinguish one tumor type from another. Those patterns can be useful for identifying a tumor even when the changes themselves did not cause the cancer.

On this page

• DNA methylation and CpG islands
• Maintaining and removing DNA methylation
• How cancer changes DNA methylation
• DNA methylation as a cancer biomarker
• Choosing which methylation sites and patterns to measure

DNA methylation and CpG islands

DNA methylation is the addition of a methyl group to DNA. In mammals, it commonly occurs on cytosine where a C is followed by a G along the same DNA strand, a site called a CpG. The modified base is 5-methylcytosine, or 5mC. This epigenetic modification contributes to the regulation of gene activity while leaving the underlying DNA sequence intact. (1, 2)

CpG sites are unevenly distributed across the genome. Regions with a high concentration of CpG sites are called CpG islands and average about 1,000 base pairs in length. They are identified using a combination of sequence length, G+C content and CpG frequency, rather than a fixed number of CpG sites. Many gene promoters overlap CpG islands. Most CpG islands normally remain unmethylated, even though much of the surrounding genome is methylated. An unmethylated promoter island does not, by itself, mean that its gene is active, because other regulatory processes also control transcription. (2, 3)

Methylation of a promoter CpG island can help keep the associated gene silent. Within the transcribed region of a gene, known as the gene body, methylation can be present even when the gene is actively expressed. (2, 4)

Maintaining and removing DNA methylation

Methylation patterns can persist as cells divide. The enzymes DNMT3A and DNMT3B establish new methylation, while DNMT1 has a central role in maintaining it after DNA replication. Replication produces DNA in which an existing methylated strand is paired with a newly synthesized strand that initially lacks the corresponding methylation marks. Maintenance methylation restores the marks on the new strand, helping daughter cells retain the pattern. (1)

If the methylation pattern is not copied onto newly synthesized strands, successive rounds of replication dilute the methylation marks. This is passive demethylation. Cells can also remove methylation through TET-mediated active DNA demethylation. TET enzymes oxidize 5mC through several modified forms; subsequent DNA repair can replace the modified base with unmodified cytosine. TET oxidation products can also be lost through replication-dependent dilution. (1)

A change in a tumor's methylation pattern may reflect altered addition, maintenance or removal of methylation. The pattern alone does not reveal which process was responsible.

How cancer changes DNA methylation

Within a tumor, methylation can increase in some genomic regions and decrease in others. In a colorectal cancer study, Berman and colleagues found focal gains in methylation, mainly at CpG islands, within much larger regions that had lost methylation. They confirmed this coexistence in additional colorectal tumors and matched adjacent tissues. (5)

Hypomethylation across broad genomic regions

Hypomethylation means a loss of methylation relative to a comparison sample or state. Much of the genome-wide loss observed in cancers affects repetitive DNA, including transposable elements such as LINE-1. Methylation normally helps suppress their activity. Loss of methylation at LINE-1 promoters may allow more LINE-1 transcription and, in some cases, insertion of new copies at other genomic locations. (6, 7)

Broad methylation loss in human cancers is also associated with somatic copy number variations (CNVs), meaning gains or losses of DNA segments in tumor cells. Finding hypomethylation alongside these changes does not establish that the methylation loss caused them. It may contribute to the changes or accompany another defect. (6)

Promoter hypermethylation and gene silencing

Hypermethylation means an increase in methylation relative to a comparison sample or state. At a promoter CpG island, it can contribute to persistent gene silencing. One way this happens is through methyl-CpG-binding proteins such as MeCP2, which recruits complexes that repress transcription. Methylation can also limit DNA binding by some transcription factors, including NRF1. (2, 8)

The tumor suppressor gene CDKN2A (cyclin dependent kinase inhibitor 2A) is hypermethylated in several tumor types. It encodes proteins that help restrain cell growth and division. Epigenetic silencing of this gene can therefore weaken that control without changing its DNA sequence. (4, 9)

Finding promoter methylation together with low gene expression does not reveal which came first: methylation may have initiated silencing or helped maintain an already silent state. (2)

DNA methylation as a cancer biomarker

Different tumor types can have distinct patterns of methylation across multiple CpG sites. A methylation pattern can serve as a biomarker for tumor type if it reliably distinguishes one tumor group from another. (10)

Capper and colleagues investigated whether these patterns could help distinguish central nervous system tumors that are difficult to classify. They trained a computer model (known as a classifier) to recognize tumor groups from their methylation profiles, then tested it on a separate set of samples. The model’s assignments were compared with diagnoses made by pathologists who had not seen the methylation results. When the model assigned a sample to a tumor group with high confidence but that assignment disagreed with the initial diagnosis, the tissue and molecular findings were reviewed together. That review supported revised diagnoses in some cases, showing how methylation profiling could help pathologists resolve difficult cases. (10)

DNA from non-tumor cells also contributes to a tissue sample’s methylation profile. When tumor cells make up only a small part of the sample, their methylation pattern can be harder to recognize. (10) In blood, cell-free DNA (cfDNA) methylation also reflects a mixture of DNA released by different cell types. (11)

Choosing which methylation sites and patterns to measure

The methylation sites that help explain changes in gene activity may differ from those that are most useful for identifying a tumor. For a study of gene silencing, methylation around the gene's promoter needs to be considered alongside gene expression data. If the aim is to distinguish tumor groups, the priority is a set of sites whose combined pattern separates those groups in additional samples. A site can be relevant to both questions; its role depends on what the study is trying to establish. (2, 10)

Whether to profile methylation broadly or focus on selected regions depends on whether the relevant sites are already known. If those sites are still uncertain, broad profiling can help identify candidates. For example, Capper and colleagues used methylation arrays to measure methylation at CpG sites across the genome, then selected those most useful for distinguishing tumor groups. (10) Once regions have been selected, focused assays or targeted sequencing panels can measure their methylation across further samples.

Many PCR and sequencing approaches to analyze DNA methylation use bisulfite treatment to distinguish methylated from unmethylated cytosines. The treatment converts unmethylated cytosines to uracil, which is read as thymine after amplification, while 5mC remains cytosine. (1)

QIAGEN products support studies of individual CpG sites and selected regions, with options for bisulfite conversion, PCR, Pyrosequencing and targeted sequencing.

References

1. Wu X, Zhang Y. TET-mediated active DNA demethylation: mechanism, function and beyond. Nat Rev Genet. 2017;18:517–534. doi.org/10.1038/nrg.2017.33
2. Deaton AM, Bird A. CpG islands and the regulation of transcription. Genes Dev. 2011;25:1010–1022. doi.org/10.1101/gad.2037511
3. Takai D, Jones PA. Comprehensive analysis of CpG islands in human chromosomes 21 and 22. Proc Natl Acad Sci U S A. 2002;99(6):3740–3745. doi.org/10.1073/pnas.052410099
4. Baylin SB, Jones PA. A decade of exploring the cancer epigenome: biological and translational implications. Nat Rev Cancer. 2011;11(10):726–734. doi.org/10.1038/nrc3130
5. Berman BP, et al. Regions of focal DNA hypermethylation and long-range hypomethylation in colorectal cancer coincide with nuclear lamina-associated domains. Nat Genet. 2012;44(1):40–46. doi.org/10.1038/ng.969
6. González B, et al. Somatic hypomethylation of pericentromeric SST1 repeats and tetraploidization in human colorectal cancer cells. Cancers (Basel). 2021;13(21):5353. doi.org/10.3390/cancers13215353
7. Choi SH, et al. Changes in DNA methylation of tandem DNA repeats are different from interspersed repeats in cancer. Int J Cancer. 2009;125(3):723–729. doi.org/10.1002/ijc.24384
8. Domcke S, Bardet AF, Adrian Ginno P, Hartl D, Burger L, Schübeler D. Competition between DNA methylation and transcription factors determines binding of NRF1. Nature. 2015;528(7583):575–579. doi.org/10.1038/nature16462
9. National Library of Medicine. CDKN2A gene. MedlinePlus Genetics. Accessed October 15, 2026. https://medlineplus.gov/genetics/gene/cdkn2a/
10. Capper D, et al. DNA methylation-based classification of central nervous system tumours. Nature. 2018;555:469–474. doi.org/10.1038/nature26000
11. Moss J, et al. Comprehensive human cell-type methylation atlas reveals origins of circulating cell-free DNA in health and disease. Nat Commun. 2018;9:5068. doi.org/10.1038/s41467-018-07466-6