Thank you for taking the time to reply. Yesterday I did DNase treatment on RNA (from LCM samples) and then measured the RNA and DNA using Qubit flurogenic dyes specific for DNA and RNA, and was surprised that there was DNA even after the DNase treatment, also the the initial amount of RNA was reduced to half. To verify if this I re-checked the DNA and RNA isolated from cell culture using Trizol, which was also DNase treated (37C for 30min) and phenol-chloroform extracted. I still found DNA contamination. does this mean that DNase treatment is not 100%?

If I cannot completely remove the DNA, how can I use this for realtime-PCR? I have ordered Taqman probes spanning introns but for one gene of interest and two internal controls like beta-actin and GAPDH, I could not get probes spanning introns. So in this case will using a no-RT control on the no-template wells in the 96-well plate be useful to detect and subtract any amplification caused by contaminating DNA when doing realtime-PCR? Or prior to doing realtime -PCR should I check for a no-RT control using gene specific primers in a regular reverse transcription PCR and look for amplification on gel?

Thanks a lot for those valuable suggestions and the easy tips. By doing an initial RT-PCR I can now ensure that I'm amplifying RNA, and not DNA by looking at the expected amplicon size, plus I can include a no-RT control. Will try all this soon. Thank you.

Yesterday I did the Real-time PCR analysis for RNA isolated from LCM samples, and as suggested I also loaded a no-RT control sample onto the plate to check for DNA contamination, used GAPDH primer probe. I observed an amplification at Ct 40, in the no-template control, total number of cycles were 45. Does this count as a serious DNA contamination, or can I conclude that it was not high enough to obscure my actual readings in the target wells? the Ct values of the ten other target genes were below 35, and I had used 10ng of template in all wells.

Hi Linda. 5 cycles difference means 2^5 difference in concentration if your reaction is 100% efficient - so at least 30 times less than your measurement. Also, a PCR is supposed to run up to 40 cycles anyway. So I wouldn't worry about it. You said "no-RT" and then "no-template" though, so I hope your no-RT control DOES have the same RNA in it? (it's the enzyme you are supposed to ommit in the no-RT reaction).

Also, two notes on your previous discussion and the very good advice you got. First, including an intron means a decent-sized one (if it's 100bp genomic DNA will still amplify and contaminate your results). Second, DNAse is an enzyme, and as such, it can fail to reach 100% efficiency, so don't be too surprised. Personally, I prefer to design my primes around big introns and skip the DNAse treatment altogether. Also, if you are working with mice as I suspect, a BLAST search using your primers will tell you if there's a chance of amplifying something other than your gene.

Hi Phoebe, Thank you for the suggestions, realized of late that the gene Im working with is a reduntant one, have tried designing specific RT prmers using large introns, when I BLAST, I get at least six different hits including a pseudogene, primers reported in other papers for this gene also were not specific.

and yes, Im working on mice, and no-RT control was all else excluding the RTase, which was included during cDNA synthesis. I used this no-RT control in the Real-time PCR plate to check for genomic DNA and observed an amplification at Ct 40, (used GAPDH primer probe), total number of cycles were 45. let me also correct that there were no aplifications in the no-template control triplicate reactions.

I should have also mentioned that I tried DNase treatment at different time intervals (5 to 20 mins) and temperature (room temp and 37 C). decided to go for 20min @ 37C, added DNase stop solution and kept at room temp for 2 mins, the RNA and DNA was measured with the Qubit assays.

A pity about the mouse, or I could send you some sequences... Ok, then here is a little bit more about how to assess the specificity using blast. Do your search writing both of your primers as ONE sequence - this will allow you to see if they anneal near each other (having one primer anneal alone won't create a product and so isn't a problem). The portion of the primers that anneals is also important: If it's the 10 5'end bases it's not a problem to you, but if it's the 15 3'end bases, then they can be extended just fine. And lastly, don't EVER trust published primers just because they are published... (or anything else for that matter, but that's another sad story). And again, unless there's some specific application I haven't heard of, 45 cycles are too much for PCR - anything above 40 is expected to give rise to non-specific products.

Oh, and a practical "shortcut" of sorts that I use for making primers for RT real-time, is to print your genomic sequence on paper and note down the introns. This way you can pick the general area where each primer will anneal fast, and then use the printout as a map while you work - it helps you keep track of the primers you tried so you won't have to repeat yourself, and move to an entirely different area if the one you were using keeps giving similar hits in blast. Be patient with it, I had primers that took me the best part of a day to design, but it's totally worth it time-wise.

Thank you so much Phoebe for those clever tips. i will try it out. for checking primers specificity i search with both primers by using the e-PCR option in UCSC genome browser, and use the Ensembl database to see the introns of the gene.

You are very, very welcome. I use the standard blast - I only saw that e-PCR when it first came out and didn't really like it, but I bet it's become better. The main thing that put me off is it didn't check the template's melting temperature in its' assessment (had colleagues killing themselves for two months over a PCR whose template Tm was 102 degrees because they trusted it rather blindly). It goes without saying that do those tests on some random RNA and not your precious samples right?

Gooc luck and feel free to send me a msg if you need something else.