I'm studying gene methylation on some DNA samples. I use the Zymo Research EZ DNA Methylation-Lightning kit to convert my DNA, then perform a PCR using bisulfite-specific primers and analyze the sample on QIAxcel to make sure my DNA amplified. Subsequently, I purify and sequence the DNA through sanger sequencing.
While the sequence has no significant background noise, every time I have a cytosine peak (indicative of a methylated cytosine residue) I also have a thymine peak.
Do you know why this could happen?
Steph Cc
If you look at older papers where cytosine methylation was studied, a common method was to actually clone the bisulfite treated products and then sequence a number of clones to characterize the methylation. This method shows the methylation status of each C residue in individual DNA molecules and it often shows that there is heterogeneity in the pattern, with not every cytosine having the same methylation status in every molecule. This can be represented as in Figure 2 of this paper: Article Hypermethylation of RASSF1A in Human and Rhesus Placentas
So your sequencing result is just showing you the same thing. Even in a "methylated" region, not every C is going to be methylated in every molecule.
You get the same idea from looking at methylation by melt curve analysis. If a "methylated" region had EVERY C methylated and an "unmethylated" region had no C methylated, then the corresponding melt curves would give a single smooth melt temperature transition. But often what you see is a complex transition representing multiple melt temps and that corresponds to molecules which have different proportions of methylated C's. Take a look at Figures 3 and 5 in this paper:
Article Quantitation of DNA methylation by melt curve analysis
I agree with David F. Barker sequencing of individual clones would produce 'clean' reads. Depending on your research outcome, you may have to sequence 10-20 clones to know the percentage of methylation at a given cytosine position in the sequence. I seem to recall that there may be software that could analyse peak heights between the T's and C's that produce the Y to give a percentage methylation. My preference would be to sequence individual clones.
- Badges
- Science method