Dear scientists,
At the moment I am analysing the RNA-seq, ATAC-seq and Methylation datasets of the same mouse-derived cell-type. Methylation was done not as bisulfite sequencing, but as an enzymatic enrichment for DNA fragments (after sonication), containing methylated CpGs with subsequent DNA library prep. It results in ATAC-like peaky view on IGV, so I focus not on the particular cytosines, but on the overall hyper/hypomethylations of promoters, etc.
I have found, that in some cases the Methylation peaks follow the exon structure, and quite often the regions, which were methylated, were not accompanied with Tn5 accessibility on ATAC-seq.
Do you have an idea why introns are not methylated, in contrast to exons? Is it correct to say that the regions of chromatin, which are closed, tend (not?) to be methylated? Does DNA methylation decrease the Tn5 cutting efficensy?
Thank you!
Best,
Michail
For the differential methylation profile between exons and introns, I would be very careful for the interpretation and double-check that is does not come from some kind of contamination (RNA...), but apparently it's a known phenomenon and it relates to splicing mechanisms, see: Article DNA-methylation effect on co-transcriptional splicing is dep...
As for other intergenic regions, it's quite well known that DNA methylation recruits the PRC2 Polycomb complex that adds the repressive H3K27me3 histone mark and induces a compaction of the chromatin. It's not so surprising, therefore, that your methylated regions are not detected by ATAC-seq, as they are probably in a less open conformation and less accessible to Tn5 (I refer to your first statement in the 2nd paragraph of your question, but in your second question you seem to have made a confusion between closed or open). I would say that methylated regions tend to be closed, and open regions tend to be unmethylated.
Dear Fabrice, thank you for your comment! Indeed, H3K27me3 is known to make the chromatin more dense and less accessible. But is it the same for the methylation of DNA itself, CpGs in particular? I just expected that if the chromatin is closed, it is also not exposed to the methylating enzymes, which results in less methylation of DNA... What do you think?
It's a bit of a hen / egg problem here! I would say that DNA methylation is anterior to chromatin compaction and stays stable, so that methylated regions are less accessible. It could also be the case that DNA methylation interferes with Tn5 action at the DNA level and therefore methylated regions appear to be less accessible whereas it's not the case. I don't remember reading anything about it, although I performed ATAC-seq several times, but I know that there is a bias for GC rich regions in ATAC-seq... Could be worth investigating in the literature!
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics