CA methylation and epigenetic memory
Epigenetics is the study of how information can be transmitted without altering DNA sequence.This can occur through changes in DNA methylation, histone modification or chromatin conformation. DNA methylation on CG residues is considered a long term method of gene repression. In this paper, Stroud et al. characterise newly-described methylation on CA motifs in DNA and its role in the maintenance of transcriptional memory. They suggest that CA
methylation may allow neurons in the developing mouse brain to carry an imprint of early exposures and transcriptional states, recruiting methyl binding proteins to poorly transcribed areas of the genome and maintaining repression throughout development. Original Article Here.
DNA methylation as a repressive epigenetic modification
The methylation of CG motifs in DNA has long been described as a relatively stable repressive epigenetic modification, inhibiting the transcription of nearby genes. These methyl marks can recruit proteins including MBD1, MBD2, MBD4, and MECP2, leading to histone deacetylation, cementing gene repression. Recently, methylation at non-CG sites has been described. This occurs at CA sites, which occur more frequently in the genome than CG sites. In this paper, Stroud et al. describe the emergence of CA methylation in mouse neurons during early life and postulate that CA methylation may enable neurons to carry a transcriptional imprint of early exposures and experiences, patterning them for later life.
CA methylation, DNMT3A and the genome
Like CG methylation, CA modification is catalysed by DNA methyltransferase 3A (DNMT3A). The authors begin by showing that DNMT3A is widely distributed over the genome during early life, but excluded from active enhancers and promoters, as well as from gene bodies which are being actively transcribed. Interestingly, it was also absent from highly repressed areas of heterochromatin. The authors then analyse the stability of CA methylation and find that it is relatively stable over the first 12 weeks of murine development, making it perfect for the long term retention of silencing information.
CA methylation bookmarks repression
The authors then analyse whether gene expression changes alter CA methylation. Ezh2 forms part of the PCR2 complex which adds the repressive H3K27me3 mark to histones. Removing this enzyme causes significant perturbations in gene expression. Using an Ezh2 cKO mouse, the authors analyse the reaction of CA methylation to gene expression change. They focus on genes whose expression increase as a result of Ezh2 deletion and find a reduction in CA methylation at these genes. Interestingly, the disruption of DNMT3A has no further impact on the transcript levels of these genes, indicating that CA methylation is a placemarker of active gene expression, without directly affecting transcription itself.
CA methylation appears to be a late event, happening only after birth. Comparing two different types of neurons during early development, they find that CA methylation is deposited on poorly transcribed genes in each type, further underlining their earlier findings. MeCP2 is a protein which binds to methylated residues and cements gene repression. By ChIP analysis, MeCP2 binding correlates with methylated CA density in both cell types. Furthermore, in MeCP2 KO neurons, CA methylation-marked genes are preferentially upregulated. This indicates that CA methylation could provide a foothold for MeCP2, enabling the fine-tuning of gene repression. CA methylation and then MeCP2 binding would occur at poorly transcribed genes, allowing the cells to maintain a memory of early developmental gene repression.
This paper contributes to our concept of how long-term gene repression is effected by CA methylation. Epigenetics studies how changes outside DNA sequence can affect gene expression and CA methylation does just this, providing a trigger for long-term repression and carrying a memory of gene repression earlier on in development. CA methylation is one method by which early experiences could shape later gene expression.