Methylation and expression of imprinted genes - (Nov/30/2005 )
I would like to have your opinion on the following:
What's really important for monoallelic expression of an imprinted gene, overall DMR methylation or the methylation status of some CpGs in particular?
Is it possible that transcription/repression factors bind to only some CpGs in particular and in that case only those are important for monoallelic expression or overall methylation is more important and regulates the binding of MBDs leading to repression on that allele?
I've been working on imprinted genes for quite a long time but still couldn't figure this out, any ideas are welcome
Thanks a lot
I think that is one of the big unknowns of methylation; what governs which CpG's become methylated. Single alleles can be methylated in health, as seen in X-inactivation for example. While skewing of the inactivated maternal/paternal X chromosome can be as high as 80:20 can happen(probably even higher sometimes), it isn't quite known what causes this.
In cancer cells, I know that HCT116 cells have a wild type allele of p16 and a mutant allele. For some reason the wild type allele is methylated and the mutant is expressed. There is no sequence difference in the promoters of these two alleles, so something must be controlling it.
I would tend to shy away from the idea that transcription/repression factors bind preferentially to some CpG's over others, with knowledge of the HCT116 promoter situation. I like the overall methylation idea. I believe that something is keeping the silenced allele mostly methylated and the expressed allele mostly un-methylated deliberately. It may be incorrect, so If anyone has a different view I'd be willing to listen, but thats my opinion.
Thank you Dave for your reply
I agree with you, I think it is the overall methylation that controls most imprinted genes expression, although, for example, for H19 there are specific CTCF binding sites that were shown to control IGF2 expression. But my uncertainty is about what is mostly methylated and mostly unmethylated. For instances, if you are studying the repressed allele of one imprinted gene and find that it is not completely methylated, let's say it has 60%-70% of CpGs methylated, is it going to be expressed or not? What is the threshold of methylation to inactivate one gene? Is there any paper elucidating this subject that I should read?
I did a quick google search and found this: http://users.ox.ac.uk/~genemed/pringleasgta.htm
Basically they methylated plasmids with different methylating enzymes and found that the position of the methyl groups was probably more important than the overall level. I couldn't find a published version of this however. Perhaps there is one out there though.
It makes sense that position would be important, and therefore it would probably be different for each genes' promoter, as there would be different sequences and numbers of CpG's. In some genes 60% may be enough to repress the allele while in others 60% may not affect the important CpG's leaving the allele to be expressed? It's something to think about!