Protocol Online logo
Top : New Forum Archives (2009-): : DNA Methylation and Epigenetics

Bisulfite conversion is complete except for the CpG sites in my non-methylated c - (May/26/2014 )

Hi all!


Like many other scientists on this forum I'm having some issues with bisulfite conversion of genomic DNA. 


I am using the EZ DNA methylation kit by ZYMO and I use control 2 control samples, one fully methylated and one fully non-methylated (also by ZYMO). My PCR primers are specific for converted DNA so they bias my PCR towards the converted DNA. Then I use the SNaPshot reaction to interrogate specific CpG sites for their methylation status. 

My PCR works, although not optimal, but I get the amplicons that I expect. When I then perform the SNaPshot on my non-methylated control sample it seems like the conversion is incomplete, some cytosines (in CpG order) remained cytosine and some seem partially converted. So then I decided to sequence my amplicons because I did not understand why my PCR worked on the partially converted DNA. 

The sequence then shows that (for the non-methylated control) all the C's that are not CpG are fully converted, which explains why the PCR works, but the C's in CpG order are not completely converted. 

Could it be that, even though the cytosine is non-methylated, a cytosine in a CpG dinucleotide is still harder to convert by bisulfite?

Or is it possible that the non-methylated control is actually not fully non-methylated?


I am a bit lost at the moment!


Have any of you seen this phenomenon before? 


I hope someone can help me :)





I'd say you have very good evidence that your "unmethylated" control is really methylated. I have no idea why. One thing to note is that plasmid DNA (not genomic) will typically be difficult to convert unless it is first linearlized. This is because the conversion happens on ssDNA, and the plasmids are difficult to pry apart in ssDNA form. So if your control DNA is plasmid, then there might be an issue.


Hi phage434,


thanks for your response :) the standard i use is genomic dna from cell line HTC 116 where methyltransferase is knocked-out which is provided by ZYMO so its not a plasmid. I also don't think the problem lies in the denaturation because the cytosine in non-CpG order are converted, it is just the cytsoine in the CpG sites that are not completely converted. 


I hope someone has had the same experience and knows how to fix it.


Have you talked to Zymo technical services. They were extremely helpful when I worked with them.


Yes I have, they are very helpful! 


They have sent me a new set of standards, the whole-genome amplified version. These should be completely non-methylated so when I receive them I will proceed by testing them. Will post the results on here :)




I realise this post is from a couple of years ago but I wondered if you got to the bottom of the problem?


I am currently having a similar issue.


Kind regards,