Bisulfite conversion, PCR end of the rope - (Jan/02/2018 )
I don't know how frequent everyone is on this forum but I think this is the last straw for me regarding DNA methylation studies. I'm facing immense amount of problems in this area. I'm studying a gene DMBT1 and it's methylation status in saliva on volunteer samples. I'm not foraging into disease samples yet. Since I have no one to consult, my guide being new to epigenetics as well, I thought I'd post my problem here.
1. Found that I could design bisulfite specific primers and I used MethPrimer for it. Using the promoter sequence of DMBT1 I designed one set of primers.
Is it necessary to design more primer sets just to be on the safe side? My primers are
Left primer TGAATTTTTTAAGATTAAAATTAAGTT
Right primer AAAAAATAAAAACTTTTACTAACAATC
2. I used a Zymo EZDNA Methylation Direct kit which states that samples can be used directly for this kit and there's no need to extract DNA first. So I used DNA first. I used some previous good quality DNA too for conversion. But nothing worked in this kit and I have negligible purity of converted DNA. I ran the samples on a gel too and found nothing. Tried chilling the gel in EtBr buffer also and again 0 results. The 260/280 falls between 0.5 to 0.9 with some of the samples' ratio falling between 1.1 to 1.5. I have used freshly prepared CT conversion reagent yet I'm facing the same problems. The concentration of the samples is between 20-70ng/ul
3. Now, I used some of the samples to run a bisulfite PCR (only the samples with the highest purity i.e. ~1.5 260/280) and again 0 bands. I tried a gradient, again 0 bands, changed a number of cycles from 35-40 cycles. I'm using EpiMark Hot Start taq pol for the reactions since many papers have mentioned using a hotstart taq pol. In the papers I've read, they mention 72 degrees as the extension temperature, but in the booklet of the HotStart taq they mention using 68 degrees. i'm using 72, could this be a problem? I have tried a positive control using GAPDH and unconverted sample but all the reagents were same, I got bands. Now, I'm using two different PCR cyclers for my reactions.
4. I'm now starting with touchdown PCR the first reaction of which, didn't work. I've put another one on a different temperature gradient. I'm not immensely hopeful of the results of this one either.
Can anyone suggest what could be going wrong? Also, what else can I try regarding the PCR because my time is running out. i have only a couple of months to finish complete sequencing of 30 samples.
Take my advice with a grain of salt as I have never done bisulfite conversion PCR before. Having said that:
1) Primer design seems to be key to this sort of analysis. I would attempt a few more primer pairs. I note that your primers are very A/T rich and have long runs of these bases in them. This will cause some secondary structure and primer-dimer problems. Try including some secondary structure inhibitors such as Betaine or DMSO in your PCR. Bisulfite conversion is harsh, so the DNA will be fragmented - make sure that your primers are suitable for amplification of short sequences up to 300 bp.
If you are looking for methylation specific PCR (MSP) then you need to have primers that incorporate the changed site you are looking for as well as primers that do not incorporate this change.
2) No idea.
3) 72 vs 68 could be a problem for an already tricky PCR, follow the manual for the product you have. Did you try amplifying unconverted DNA with the primers you supplied in your post? did it work? if not - you should try to get this to work before you try your amplification of your bisulfite converted samples.
Few years ago I worked with MSP. I used Zymo but don't know if the kit had changed since then.
1st. Working with saliva not always give what we are looking because of the presence of inhibitors and the constant shredding. Make sure that subjects haven't eat or drink at least 30 min before the sample collection, they most wash 1-2 times with plain water before the sample collection. My advice if possible use PBLs instead of saliva. Yes is more cumbersome, but DNA quality is way higher. If saliva is the only way I would suggest to extract DNA first and do a quality control test by acquiring 1 forward primer of any HKG and several reverse primers that give different weights from example 1k bp, 500 bp, 250 bp, 100bp and you decide what is your cutoff for example any thing shorter that 500bp its eliminated.
2nd. Recheck the primers! Also remember that you must have 1 pair for the met and 1 pair for the umet
3rd. Use controls. There are plenty of companies that sells methylated and unmethylated DNA.
We described the UBC integrity assay (both for native and bisulfite-treated DNA). Might be useful to compare the integrity before and after bisulfite treatment (doi: https://doi.org/10.1101/168195) as suggested by merlav. In addition we recently described our bisulfite sequencing method, including some troubleshooting. Might be helpful to find out what might be wrong (doi: https://doi.org/10.1101/239566). Good luck.