There's quite a few things that could be going on here. As a starting point...

1. What are the amplicon sizes for your reference gene and gene of interest?

2. Was a melt analysis run with the qPCR? A picture of the melt curves would be really helpful.

3. Did you run the RNA on a gel or the BioAnalyzer or any other analysis method that would show if it's intact or degraded?

4. Have your primers been previously tested?

5. Would either primer set amplify gDNA if it was present?

Also worth noting that Cq values of 28+ indicate very low expression (like, countable numbers of target molecules per well).

Here (though I'm not entirely sure what 36B4 is supposed to be) it looks like you have one very abundant gene that doesn't change much, and one gene that is barely expressed. It might change from "barely expressed" to "slightly more than barely expressed", but it's going to be noisy data no matter what, because it's barely expressed.

More detail would be very helpful.

Dear Laura Leighton,

Thank you very much for your thoughts.

1, Unfortunately, I don't know the amplicon size since these primers were ordered by another person who left the lab. But generally, when we order qRT-PCR primer, the size is between 80bp-250bp.

2, We only did the Standard Curve analysis. We didn't perform a melting curve analysis.

3, I ran the RNA sample on a gel and attached the image. Do you think there is massive RNA degradation in tissue samples? Sonication might be a major issue. The RNA degradation during sample collection is unlikely since we directly drop our tissue into RNAlater.

4, These primers have been tested before. They all are valid.

5, We particularly purchased an RT kit with a gDNA remover reaction. Of course, we won't be able to remove all the gDNA in it. But I kind of doubt that these primers would amplify. I am concerned that the sonication breaks gDNA, which will present in the aqueous phase during Trizol RNA extraction.

Dear John Hildyard,

The 36B4 is the control gene we use in qRT-PCR.

We also believe that a CT value over 30 indicates a very low level.

The mouse cell line RNA samples look relatively intact, although the lowest part of the gel (where very degraded RNA would be visible if present) is cut off. The mouse tissue RNA is extremely degraded, to the point where it's probably incompatible with quantitative PCR.

Why were your samples sonicated? This is the likely reason why the RNA is so degraded and it seems incompatible with your experimental goal?

Do you have the primer sequences? If so, you should be able to use an in silico PCR to determine the the amplicon size from the sequences. I ask because the larger the amplicon is, the worse it will perform if the RNA is degraded. So if we imagine your reference gene amplicon is 100bp but your gene of interest is 250, that would explain both why the amplification of the gene of interest is weak and why it's inconsistent between samples. I should be clear that this is information useful for troubleshooting only - the fundamental issue here is RNA degradation which won't be fixed by using different primers.

Laura Leighton,

Since we don't have a homogenizer in the lab, I have to use a sonicator to break intestinal tissue. I also think that sonication shouldn't shear RNA easily. I used a relatively mild condition: 10s on, 10s off for 6 cycles with 30% amplitude.

I think that the smear is not RNA; it might be DNA. To test this, I took 10ug to RNA and treated them with DNase at 37-degree for 30 mins. I then extract RNA with pheno-chloroform. Surprisingly, the RNA I extracted seemed worse, and I couldn't even get any gene amplified in qRT-PCR.

I ran a gel and found that the smear had disappeared. I am unsure why the newly extracted RNA after DNase treatment failed in qRT-PCR.