I am studying a genomic locus in some tumoral specimens with the aim of clarifying if there are differences in the methylation status of about 30 CpGs among different samples (two groups of samples differing for the expression pattern of a specific gene).
I designed primers following the most common recommendations (e.g. amplicon not longer than 350 nt, primer with a CpG at the 5' end) as reported in a recent Biotechniques paper (Hernandez et al., 2013). At the moment I started testing my first primer's pair that apparently work well using as template a bisulfite treated DNA (whose methylation status is unknown). To be sure that primers recognize with the same efficiency methylated and non-methylated alleles before starting with my patients' DNA I ordered the Human Methylated & Non-methylated DNA Set from zymo research. I treated the fully methylated, the non-methylated DNA and a combination of both (50:50) using the EZ DNA Methylation-Gold™ Kit again from zymo and try to PCR amplify them.
The result was in my opinion very strange. While the fully methylated DNA amplify well with my primers I can not say the same for the non-mehtylated and for the 50:50 combination. The BS treated non-methylated DNA revealed a very faint band while no product at all was obtained for 50:50 combination. I was worried that something happen in the BS conversion and/or purification so I repeated the treatment but I got the same results in PCR.
Then I tried to use the DAPK1 primers included in the Human Methylated & Non-methylated DNA Set and apparently they work fine (I'm a bit in overcycling, but it doesn't look as differences were present among samples).
At the moment I don't know ho to proceed. Should this data indicate me that my primer combination is more prone in amplifying methylated alleles compared to the non-methylated one? Is it conceivable that the two alleles show such a difference? Do you have any suggestion on how to proceed?
Thanks in advance for any help