A Novel CpG Island Set Identifies Tissue-Specific Methylation At Developmental Gene Loci

The human genome contains about 22,000 genes, each encoding one of the proteins required for human life. A particular cell (e.g. blood, skin, etc.) expresses a specific subset of protein genes and silences the remainder.

This week in the open-access journal PLoS Biology, Robert Illingworth, Adrian Bird, and colleagues shed light on the mechanisms that cause genes to be activated or shut down, by showing how they studied DNA sequences called "CpG islands" (CGIs).

These sequences are found at over half of all human genes and can exist in either the active or silent state depending on the presence or absence of methyl groups on the DNA. They devised a method for purifying all CGIs and showed that, unexpectedly, only half occur at the beginning of genes near the promoter, the rest occur within or between genes.

Notably, methylation of CGIs causes stable gene silencing. They tested 17,000 CGIs in 4 human tissues and found that 6-8% were methylated in each. Genes whose protein products play an essential role during embryonic development (and which need to be switched off at later stages) were biased toward methylation, suggesting that gene expression during development is regulated by CGI methylation.