Tumor hypoxia reduces TET activity and causes tumor suppressor gene hypermethylation and repression.
Major finding: Tumor hypoxia reduces TET activity and causes tumor suppressor gene hypermethylation and repression.
Mechanism: Decreased TET activity leads to reduced 5hmC and enhanced promoter hypermethylation.
Impact: Tumor hypoxia may promote tumor progression in part by regulating DNA methylation.
Tumors exhibit oncogenic epigenetic alterations including global hypomethylation and local promoter hypermethylation that can repress the transcription of tumor suppressor genes and promote cancer cell growth. Oxygen-dependent tet methylcytosine dioxygenase (TET) family proteins catalyze the oxidation of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), resulting in DNA demethylation. 5hmC is frequently lost in cancer, but in the majority of tumors, the underlying mechanism is unclear. Thienpont and colleagues observed that hypoxia induced 5hmC loss in a majority of tested human and murine cell lines, resulting from reduced activity, but not reduced expression, of TETs. DNA-immunoprecipitation sequencing (DIP-seq) revealed global loss of 5hmC in response to hypoxia, predominantly at gene promoters, many of which had concomitant gain of 5mC, and RNA-seq confirmed that expression of these genes was repressed. Data from The Cancer Genome Atlas showed increased promoter hypermethylation in hypoxic tumors, which was associated with reduced gene expression. TET2 and TET3 expression were inversely correlated with hypermethylation, and TET1 and TET3 mutations were positively correlated, supporting a role for reduced TET activity in limiting 5hmC generation and promoting DNA hypermethylation. Additionally, TET enzymes are completely inactivated in IDH1-mutant glioblastoma, and these tumors exhibited extensive hypermethylation and reduced expression of many of the same genes hypermethylated by hypoxia. However, IDH1-mutant glioblastomas exhibited greater methylation than hypoxic glioblastomas, suggesting that hypoxia only partially inactivates TET. In breast tumors, hypoxia induced hypermethylation of the promoters of tumor suppressor genes, but not of oncogenes, and in a mouse model of spontaneous breast tumors, hypoxic areas of the tumor exhibited 5hmC depletion. Reduction of tumor oxygenation further increased 5mC at tumor suppressor gene promoters, and this effect was reversed by normalizing the tumor vasculature. Taken together, these findings demonstrate that hypoxia suppresses TET activity, leading to reduced expression of tumor suppressor genes, and provide a link between hypoxia and DNA methylation.