inconsistent and low reproducibility of bisulfite sequencing - (Jan/11/2016 )

Hi, 

I am trying to optimize a protocol to investigate the differential methylation in gene promoter.

I am using published primers for the OPRM1 promoter for nested PCR on bisulfite converted DNA (extracted from blood). I am able to obtain single band from my PCR and proceed with BigDye Sequecing of the PCR product. 

However, I am having problems getting consistent seqeuncing results.

1)I find that the % methylation from the same PCR product when sequenced with forward and reverse primers to vary significantly~ up to 20-30% difference. I believe that my DNA is fully converted as I did not observe any non CpG C's in the sequences (no partial C>T conversion observed in these areas). 

2) When I repeat the BigDye sequencing on the same PCR product, I am not able to reproduce the % methylation observed. However, I do notice that the reverse primer tend to give a more consistent result. The sequencing with forward primer shows really big variation. I do note that some papers prefer to use reverse primers for bisulfite sequencing. does anyone know why is it so? is there a reason why the reverse primer is more reliable?

Also, as the first~80bp of sequeing results are usually not very reliable, how can one rely on only sequencing with either one of the primers ? espcecially when longer PCR product size is not recommended for bisulfite sequencing.

3) I believe that my DNA is fully converted as I did not observe any non CpG's C in all my sequencing results (no C minor peaks observed). However, when I repeated the bisulfite conversion, followed by nested PCR and BigDye sequencing on the same DNA sample, I observed huge variation in the % methylation observed as well. 

Does anyone have similar problems and could suggest to me what could I have possibly done wrong? Please help!!

Thank you in advance.

Mary

Are you looking at the electropherogram, or at just the called bases in your sequencing reactions? I would strongly recommend looking at the electropherograms, which provide much more information about what is really happening. The first 30-100 bases of most sequencing reads are less reliable than regions 100-600 bp, which will be obvious if you look at the electropherogram.

Hi,

Thanks for your reply. Yes. I am looking at the electrophetogram. And I calculate the % methylation by measuring peak heights C/(C+T) on chromaslite.

And I agree that the first 30-100base are unreliable and for regular sequencing, I would usually design my primers to start -100bp from my region of interest to overcome this problem.

however with the bisulfite sequencing, I am interested to look at the CpG methylation on the OPRM1 gene promoter which spans ~300bp. As I understand that it is not recommended to work with PCR product size >300bp for bisulfite converted DNA, my sequencing primers are directly next to my region of interest. Initially, I had thought that sequencing  from both forward and reverse would solve this problem. However, I find that in the regions where I am able to get good signals on both forward and reverse sequencing reads, the % methylation observed is vary significantly.

Do you have any idea why is it so?I would appreciate your suggestion.

Mary

Mary,

the 300bp limit is to do with the potential fragmentation of gDNA during bisulfite conversion, it is possible to amplify larger regions, keeping in mind the methof of bisulfite conversion can degrade your sample such that fewer amplifiable templates are available the longer the template you want to target. 

N