Problems with GOI Ct's - How to do PCR efficiency test if there is no Ct in GOI? (Dec/10/2009 )

Hey!

I have one BIG problem with my GOI and their dilutions. I'm trying to do a validation experiment for DDCt method (The comparative Ct method) with SYBR Green. I've used 1 ug of total RNA for reverse transcription and after that I did five points serial dilution (1:5, 1:10, 1:50, 1:100 and 1:500) with my cDNA. Housekeeping gene is working well (efficiency=95,7%, RSq=0,999) but my GOI doesn't work at all. Well if I am precise I've got three Ct's, two within 1:5 dilution (32.95 and 33.73) and one in 1:50 dilution (33.79) . It is very strange to have so little differences between Ct's over ten-fold dilution, is it not? What am I doing wrong? Dissociation curve is ok, just one peak for reference gene and GOI. I've put all results in an attachment.
Final volume of qPCR reaction was 20 ul. For GOI I've used 300 nM primers and 100 nM for housekeeping gene.
1 ul of sample was used per reaction.
I don't know what went wrong . Could you help me out? What are my solutions?
Thanks.

E.

P.S. One stupid question for the end. Can someone do primer optimization on genomic DNA and then use it on cDNA samples or not? And if not, why?

Eon's mommy on Dec 10 2009, 08:44 AM said:

When I've done validation curves, I've always done them over 4-6 LOG dilutions to get a good range. However, based on your Ct values for a 1:5, my guess is that you just don't have that much transcription occurring in whatever sample you've got. Most likely, the similarity in Ct between the 1:5 and 1:50 is because there's just too little too measure. Are you running NTCs along with these, or -RT controls to make sure it's not low level genomic DNA giving you the signals? Alternatively, check the melt curve analysis of your products. I had a set of primers once that amplified well and gave a Tm of about 80C, but in -RT samples there were primer dimers that game a peak at about 75C, and when I did -RT LOG dilutions I got a flat line, because every sample had the same Ct due to primer dimers, and it was about 30-35 at each point. The thing is, the melt curve for -RT and +RT both gave only one peak, but Tm of the peaks was different.

If you have a reaction where you know the GOI amplified (maybe from gDNA), make sure the Tm of that product matches the Tm of the product you're getting in your validation curve. If that's not the problem, then perhaps it's just too little mRNA present to give a sufficient amount of cDNA for a good curve. A colleague of mine is having that trouble right now with some bacterial GOIs. He has to really concentrate the mRNA to get sufficient cDNA in the RT reaction to be able to have a good enough dynamic range for the validation curve. The problem that arises with concentrating the samples that much is that you likely may have a lot of residual gDNA being added, which causes problems with -RT controls and all that.

As for the P.S., my understanding for primer optimization and validation curve is that it's a measure of the primers' ability to amplify DNA. Whether that's gDNA or cDNA I don't think matters. If someone knows of why that would be incorrect, I would like to know as well.

I'm always running NTC and it is negative (No Ct). I didn't run -RT controls, because I treated RNA samples with DNase prior reverse transcription.
Before qRT-PCR I did PCR with genomic DNA (using same set of primers) producing clear, one bend PCR product. I'm using the same Tm.
If there is really too little mRNA in my samples will it help if I do reverse transcription with more than 1 ug of total RNA?

As for primer optimization...I asked this question because I have limited amounts of cDNA and I was thinking if I can optimize primers on genomic DNA, I can then use optimized primers on cDNA, without loosing to much cDNA.

Thank you for reply.

E.

Hi, I would say that your GOI is simply not expressed or at the detection level in your samples. The question if gDNA is suitable to check the primers depends on the nature of your primers. If they are intron spanning primers then it will not work. If the primers are located on one single exon then it should be possible. In general, every sample which contains your correct target sequence can be used, no matter if it is gDNA, cDNA, plasmid DNA, PCR-product... Most often it is the easiest way to use cDNA from a sample where you know that your GOI is highly expressed.

And as fishdoc said already correctly, it is important to measure your primer efficiency over several log dilutions like 4-6. 2 logs are not sufficient

I hope this was helpful.

Eon's mommy on Dec 10 2009, 11:09 AM said:

In my experience, DNase treatment will NOT eliminate all gDNA. It will eliminate a lot of it, but not all of it. Take your DNase-treated RNA sample and use 1 ul of it directly as template in a conventional PCR reaction and see if you get a product. If you don't, you've successfully reduced the DNA levels to undetectable. If you get a band, you need to determine how much to dilute your RNA samples before gDNA contamination is undetectable.

Running NTC will tell you if there's contamination in any of your reagents. Running -RT will tell you if there's DNA contamination in your RNA. Never assume DNase takes care of it all.

I don't exactly understand what both of you mean by making serial 4-6 LOG dilutions. Does that mean, for example, that you do approximately 7 by 4-fold dilutions? If I can't do validation experiment, what then? Any suggestions would be appreciated!

E.

Eon's mommy on Dec 10 2009, 04:40 PM said:

6 log dilutions is taking a sample and diluting it 1:10 5 times, so you have 10^0, 10^-1, 10^-2,...,10^-5, or 6 logs of template. It's obviously hard to do that, however, when your 10^0 gives you Ct values of low to mid 30s. I'm not sure what the answer is for that.

You can, for example use a PCR product from a previous initial amplification to get enough template for a broad dilution range. A procedure for this is explained here: