Problems seuencing Bisulphite treated DNA - (Oct/01/2004 )
The Story so far!
Treated some DNA with the Zymo Research EZ DNA methylation kit (Good kit, cheap + easy to use)
Got great yield of good sized DNA
Did a PCR on the treated DNA, got a good band
Cloned it into p-GEM-T Easy
All my clones are in the same orientation in the plasmid, which is not unheard of.
Tried sequencing. This is where the fun starts!
I’ve tried sequencing with both Universal/Reverse primers, and the two PCR primers used to create the construct. In both cases, I got poor/failed sequence on one strand (ATG rich, few remaining methylated C) and passable (~400-500bp) sequence on the other strand (ATC rich + a few odd G).
I also tried sequencing out from the middle with internal primers. Again, only the one strand worked OK, and then only for around 220bp before an abrupt stop.
I’m using Beckman DTCS chemistry.
Any thoughts? Is the tribasic nature of the template mucking the chemistry up? Could there be bigger than usual problems with secondary structure? Should I just go for a beer?
All help appreciated!
I have done lots of direct sequencing and sequencing after TA cloning (over 400 bp) and rarely seen sequencing reaction stop in the middle especially with plasmid. I only sequence one strand on plasmid and am able to read the whole insert. Probably extremely long stretches of consecutive nucleotides is the cause, who knows.
I got the similar problem before. Try to add 1M betaine or 5%DMSO in your sequencing reaction. It may work.
Thanks for the replies.
Tried it with ABI chemistry/hardware, same result.
A colleague suggested Betaine, I'll give that a go tonight.
Played with the conditions on the sequencing reaction + got a much improved result.
Today's puzzle: In a 620 bp CpG island, there were only a couple of unconverted C's in my bisulphite converted DNA. I was under the impression that methylation levels were expected to be rather higher than that!
You are right. Bisulfite genomic PCR has bias over unmethylated DNA, which means that from a mixture of methylated and unmethylated DNA strands, the PCR amplify preferentially unmethylated DNA. So methylated DNA will get under-represented in the subsequent sequencing results. In this case, amplifying and sequencing both sense and antisense strand may yield unbiased results
The oligos I use (designed with MethPrimer) are outside the CpG region I'm looking at (I amplify a 750bp fragment with the 620bp island in it).
Are you saying that my DNA template will contain a whole load of different template copies of the island, and that the primers will preferentially amplify those which originated from low methylation copies, even though the primers bind outside the CpG region?
I cloned amplicons from 5 different breeds, and sequenced 4-8 clones of each breed. All showed the same low methylation pattern. Should I have expected to see at least a few clones with some more methylated residues?
Or have I got the wrong end of the stick?
Yes, according to a paper by SJ Clark's group, methylated DNA gives rise to a bisulfite-treated derivative with higher GC content than unmethyated DNA, and it is possible that this higher GC content may raise the melting temperature of the DNA and increase the likelihood of secondary structure formation for some sequences, resulting in a low PCR efficiency when compared to unmethylated sequences.
It is worthwhile to read this paper:
Detection and measurement of PCR bias in quantitative methylation analysis of bisulphite-treated DNA. Nucleic Acids Res. 1997 Nov 1;25(21):4422-6.
Thanks again. By a coincidence I found that paper yesterday!