No amplification after bisulfite treatment - (Aug/03/2010 )

Hi everyone,

I am planning an experiment of next generation sequencing (454) of PCR amplicons obtained after treating genomic DNA with sodium bisulfite. Unfortunately, the technique doesn’t seem to be really easy to perform or, at least, so far I haven’t been able to get it to work. Maybe you might have some good suggestions.

Very briefly, I sonicate the gDNA and use the EZ DNA Methylation-Gold Kit (ZymoResearch), then I do a PCR using different primer pairs designed with the Kismeth software. Actually, the primers usually work just fine on control non-treated genomic DNA, while I never get any product from the treated samples. I also tried a reamplification using as templates the PCR mix from both the treated and non treated DNA but, again, I obtained a nice band only from the reamplification of the control DNA. So I am beginning to feel a little disheartened.

I am under the impression that the quality of the DNA that I recover after the treatment is really poor but I don’t know what I can do to improve it, as I follow the protocol step by step.

I will be absolutely grateful for any advice that you will be able to give me.

Could you post your primer sequences? Amplifying only non-converted DNA (which should actually create no pcr product) suggests something´s wrong with your designed primers .

Hi!
Sure, below you'll find the reference sequence and the primers sequences.
I designed my primers using the Kismeth software.

>RefSeq
aaagaaaaaaaaagatgagatcgatatattcccgcgcgcgttcctttctttatttgctgttttttttttgtttttggcaaagtgtcactgtacctttggccgccgcggaaggcactgtcgtttgactggtagtacatctgatactactacctaccggtccggacagaataatgacacacgcactgccgcactggcactgc
acttgcgcgggaaagggagggggccccacgtacctgcgtgctttccatttctagcagtagcagaggccgtccgatccattggcggcggcgttggtcctgtgggctgctgcctgcaccggcacggcggctggccagtggccaccataccagtataccaccgccccccgccccgtttgtgtggcgggaaa
aacgtacagtgcgaccacgtcgccgtcttctcttcctgttggacacggcagctagcaacgtcgtcatggccgcaagcccacaacactgttgcctcttggccgtctcaagggacaagttgtaccagctccctctgctcacagcccggataataataataatacttgttgtaaacaaagctaataataatgaaa
aataacatatatatataaatcttgatgccctttccagaaaaagcgggtgcaggtatgaaaggtcgtcaaaaggagcaggagaggagatgcgttggcgatgagcttcctgcctttcccgttacgtttgcctaccttcacacattctacatttgtagaagatgcctgaagtggaaggtgcctaggcctaggttca
catgtttatttatggaaagcgacatttgacaaagaacgcgattccgatagtgctctcacagtctcacttcacatccattccatccatgtatcatgcgttgtagtgacgataagccgccggtcgtaggagtgatctacatagatttatatgacatacatgtatgtcacggtacaatgcatgtcttacatgcacgtacgta
cgtagttattgccggcggtaacagtgtacctttgctactttgcctttgtcttgcttcgtctcgacgcaagcttcacaatgatgccatgtgtcatttctgcttcttcttgctccaacatcaaccagattcaaaggtccaaactcgaaacccgtccgcttatggttttgatgcccaaaccacttgacgtctctttctgtgcgagag
tagtgtatctgatatattctttccatgatgtttgccaaataaaactagttctattgagggatggggaattcccactagtttaggtggtcacgaggttaaatacagcttcctcaattttgctttaacaatatagtctttgtagattaatctttttacaccaagtaaggacaaatttaagagcgatgttacgtgtacgtagatgca
cgagaactatatataatataagtaatttttggacacaaagcgatataatatagtttctggttcgtctaacaaactaattaaacgctagctcggtatcgtcgttttatgacgatgatggatttgacgttaattgcttctttgtggggaattcatcggatcgtacattcgtattcatgtcgaggagaggtgcaagtggactcg
atggatgggaatgcgcacagtctgcctcagctagaaaaatcgtagatttctactccctccgctccaatttataaattcgtttaatttttacaccaaatttgacatacttgtcttattcaaaattttgcaaaaatatataaaaaaatcaaagctacacttaaagtatattatatgctaaacggtatcacactaaaaaaatt
aaataaattgtaaattgaaatggagggactagcatttaaaatctcgaactggaactgatgatctaaatagactagctttcgttatcctgtctactacaggaatgatatttagtaccggttgaaggtaaacactattattttttagtatcttactttttaaatgtctagatatgtctaattcgggtctatttg

N28_bis_610_for GAAAAAGYGGGTGYAGGTATGAAAG
N28_bis_610_rev TTTAACCTCRTRACCACCTAAACTA

Sary, do you need to use degenerate primers? Because you always co-amplify non-converted DNA. One problem could be the length of your amplicon as bisulfite-treatment fragments DNA, although it sometimes works. I also succeeded in the past by simply using another polymerase (no proofreading-enzyme!!).

Hi ElHo,

I see your point but I think that amplification of non converted DNA is necessary, at least in my case, but I may be totally wrong.
In fact, the plan would be to generate these amplicons and then do 454 sequencing, in order to be able to evaluate the frequency of methylation for each single cytosine, so I also need to amplify the molecules in which Cs haven't been modified. Unless there's a way to discriminate between Cs that have not reacted because of incomplete conversion and Cs that have not reacted because they were protected by 5-Me. Have I made myself clear, or not that much?
Also, I have tried shorter amplicons (3-400 bp) but the result is pretty much the same.

Another (maybe silly) question: I've read that most people do a nested PCR, with internal primers. Why can't one just use exactly the same primers as in the first round of amplification?

Thank you for your replies!