Problems with qRT-PCR - Lots of unspecific products (Sep/21/2009 )

Hi there,

I'm currently trying to establish primer systems for the detection of tumor-specific marker mRNA. I'm using Roche's Light Cycler 480 and did my reverse transcripition with Qiagen's Omniscript Transcriptase and random hexamers. I designed all my primers with IDT's Assay Design Tool. So much for the background info

I tested my primers on an arbitrary chosen sample using SybrGreen and picked the ones which showed the cleanest melting curve for my further experiments. I also included No-RT-controls and No-Template-Controls, and they showed always some unspecific signals after 35 cycles, but I guess thats fine. So far no problems.
After that first experiment I decided to run the chosen primers on more samples. At this point I got some weird results (well, at least for me):

In some samples the former clean melting curves looked completely different (smaller extra peaks with lower Tm mostly), indicating unspecific products. The same primers on other samples looked fine. What worries me most is that these inconsistencies also happend on the sample that I tested the primers on in the first place (well, not technically the same sample, I did a new RT first).
I'm now a bit unsure what happend. I blasted the primers and they seem to be specific for my target region. I'd rule out RNA degration because other PCR products of these samples of question look fine (as well as RNA integrity checked prior to RT on a gel). But there seems to be a problem with my sample preparation. Might the RT be an issue?

I repeated the run with a annealing temperature of 63°C (Primer where designed with 61° in mind), but it didn't really help.

I don't think it would be wise to try probes with that kind of background?

Any help is really appreciated, I can also provide pictures of melting curves tomorrow (I'm already at home), if this is going to help.

Thanks a lot!
T.

Did you also analyze the expression of a reference gene like GAPDH? I sometimes get unspecific products for my target gene with samples having a low amount of cDNA. You could indentify those samples because of their high Ct-Value for GAPDH.