Problem with amplifying bisulfite treated DNA from FFPE tissue - (Dec/07/2007 )
I've got a bit of a problem that I thought you might be able to help me with. I am trying to develop pyrosequencing assays to be able to look at the methylation status of the promoter regions on six different cancer related genes. To do this I'm using formaline fixed parrafin embedded colon cancer tumor tissue that I have first deparrafinized using the xylen-ethanol method. I used Qiagen’s QIAamp DNA FFPE Tissue to lysate and purify my DNA which was then (500ng) bisulfite treated using the EZ DNA Methylation Direct kit from Zymo research.
The thing is that I'm not at all succesful in optimizing an amplicon in the PCR or at least not the right one..... I can get amplicons up to 300bp on some of the genes, but as mentioned that is unspecific binding of the primers. At least I know that my DNA has not degraded or been cut down to too short fragments for my amplification to occur since I'm trying to amplify amplicons that are no longer than 274bp, with the shortest one being only 122 bp.
I know for a fact that the primers are acctually compatilbe with eachother since I have used the same primers to optimize methods to study the methylated fraction of the same genes, mentioned above, but in DNA from human leukocytes (purified with Qiagen’s EZ DNA 200μl Bloodkit and bisulfite treated with the EZ DNA Methylation kit from Zymo Research).
I made an amplification control (a mix of different house-keeping genes) on my untreated DNA from the tumors and saw that I can get amplicons up to 600bp (but this length differs between samples) so I know that I do get DNA out from the purification, which indicates to me that up until that point everything is probably in its order and that my problems occur later on in the process.
However I know nothing about how degraded my samples might be after the bisulfite treatment nor do I now how well the treatment acctually worked. I might still have a lot of unconverted DNA in my samples. Is there any way to check that??
So my questions to you are:
* Could there still be components, like proteins or chemicals (like formaline) for examples, in my DNA from the tumor samples that inhibit the PCR to do "its job" and if so what would you suggest would be the best way to avoid this problem.
*Is there and modifications that I could try with in the PCR reaction that could help me get an amplification?? I have of course tried with temperature gradients and different MgCl2 concentrations. I have also tried adding a carrier DNA (normal E. coli DNA) both in the PCR to get a more competitative system, as well as in the purification step after the bisulfite treatment to get a better yield of converted DNA in the elution. I thought that might help me if I had too little template, but without much luck. I have tried amplifying my samples for 40 cycles and then taking 2μl of the PCR product as a template in a new PCR (50 cycles) but nothing works… Frustrating!!
* Are there any other suggestions you might have that could help me solve my problems.
I'm really out of new idea's that I could try here so I would be very greatful if you could take the time to help me with my problems and try to find a solution.
Zarah
Hi Zarah,
I feel your pain,
I was wondering with your primers, are they methylation specific primers (MSP) as you mentioned you have tested them out on methylated DNA.
Primer design is the singlemost reason why things don't work for people.
Nick
Hi Nick,
Thank you for trying to help me!!
My primers are not MSP. I have designed them to anneal to specific methylated sequences around the area of interset. They however bind in over C's but no C's involved in any CpG sites. And since they bind in and give me the right amplicon on bisulfite treated DNA from leukocytes, I believe the primers are alright. But please correct me if I'm wrong!!
My personal opinion is that as long as I have pure DNA that I want to amplify, it shouldn't matter if the DNA comes from leukocytes or FFPE tissue... But since this is my first experience with this kind of starting material I might be terribly wrong 
Best regards,
Zarah
Thank you for trying to help me!!
My primers are not MSP. I have designed them to anneal to specific methylated sequences around the area of interset. They however bind in over C's but no C's involved in any CpG sites. And since they bind in and give me the right amplicon on bisulfite treated DNA from leukocytes, I believe the primers are alright. But please correct me if I'm wrong!!
My personal opinion is that as long as I have pure DNA that I want to amplify, it shouldn't matter if the DNA comes from leukocytes or FFPE tissue... But since this is my first experience with this kind of starting material I might be terribly wrong
Best regards,
Zarah
Hi
Having primers that work on good quality (leukocyte) DNA is no guarantee that they will work on DNA from FFPE material - some primer pairs "need" much more template than others. Here are a few ideas for you to try
1. Bisulfite treat for less time - this will lead to less degradation of your DNA, but BEWARE you are in more danger of having incomplete bisulfite conversion - but at least you should be able to tell if this is the case because you are pyrosequencing (unlike many papers that use poor MSP primers that would probably amplify non-converted DNA)
2. Try fully or hemi nested PCR - ie design primers either just outside (or inside) your current amplicon and
perform say 30 cycles with the outside primers then use take 1uL of a 1:10 dilution of this PCR as template for another 20 to 30 cycles with the inside primers
3. Decrease the size of your amplicon - with FFPE we try to keep it as small as possible, under 200bp is good
4. Post your primer sequences and we can give you an idea of how well they are designed
Hope this helps
Good luck!
3.