Nonpromoter DNA methylation by Dnmt3a allows transcription

Dnmt3a-Dependent Nonpromoter DNA Methylation Facilitates Transcription of Neurogenic Genes

This paper showed that Dnmt3a can methylate both promoter and non-promoter DNA regions, with the latter being transcriptionally permissive. Dnmt3a antagonizes Polycomb complex binding and thus the silencing impact by H3K27me3 on transcription. Further, the authors suggested Polycomb may help recruit Dnmt via their direct interaction, and eventually Dnmt activity would be sufficient to counteract the repression by Polycomb. It should be very helpful to pursue how the regulation is coordinated on different genes and during different developmental processes.

Keystone Meeting

A short review on some of the research discussed at this years nuclear receptors keystone symposium:

http://www.nature.com/embor/journal/v11/n8/full/embor2010111.html

ATP-citrate lyase links cellular metabolism to histone acetylation

Sorry for the delay.

What is cool about this paper is that the authors show that global histone aceylation is increased in human colon carcinoma cells grown in the presence of glucose and that this is dependent on the enzyme ATP-citrate lyase (ACL). ACL generates acetyl-CoA from citrate (remember the TCA cycle…how many times did we have to memorize that?) and acetyl-CoA has been shown previously in yeast to serve as the acetyl donor for histone acetyltransferases (HATs). In this paper, the authors also show that adipocyte differentiation, and the gene expression changes that occur, are dependent on ACL. This finding, in addition to previous knowledge that histone deacetylation (via NAD+-dependent Sirtuins) and demethylation (via 2-alpha ketoglutarate-dependent prolyl isomerases) are metabolically regulated, argue that our natural metabolism as well as what we eat influences gene expression through dynamic alterations in chromatin modifications.

If you recall, in p23 overexpressing cells there are increases in metabolically regulated chromatin modifying enzymes (not ACL, but an enzyme that generates acetyl-CoA from long chain fatty acids) and I also see increases histone acetylation at p23 sensitive genes.

GR and MeCP2

So, Freddy sent me a thesis written in 2006 from a German University. It is about a methyl CpG binding protein, called MeCP2, that suppresses GR-target genes. MeCP2 has been shown to be required for the normal function of neurons and is usually thought to act as a repressor of genes, including CRH (my favorite gene). Mutations in the MECP2 gene has been strongly linked to a neurodevelopmental disorder known as Rett Syndrome.

In this paper, they used a microarray analysis to compare gene expression from WT and MeCP2-null mice. They discovered that at least 5 of the 11 differentially regulated genes are GR target genes. They show by ChIP (your favorite technique!) that MeCP2 binds to the promoter region of FKBP5 and SGK1, suggesting a relationship between MeCP2 occupancy and gene expression. Its not very clear to me in the paper, but what is really novel in the thesis is that Dex prevents Mecp2 from binding to a GRE within the FKBP5 gene. This is prevented by RU486. Its in the last figure of the RESULTS Section found HERE.

The bigger picture here, I think, is how epigenetics, in particular, DNA methylation, is regulating the expression of these GR target genes in neurons. This has many implications for stress response since one of the identified genes, POMC, is a key player of the HPA axis and CRH also has been shown to be regulated by MeCP2.