Protocol Online logo
Top : Forum Archives: : Molecular Biology

not reproducible : reverse - transcription PCR - (Jun/29/2007 )

Could anyone please help me with a problem in my semi-quatitative RT-RCR? I attempt to compare the transcript level of a gene in two samplers. I got very good result last week---- sampler A has dramatically higher expression of that gene than sampler B. B-actin was used as loading control and 1:4:16 serial dilutions of both B-action and the interested gene was performed.

But when i tried to reproduce this result this week, everything goes wierd. The dramatic difference between the two samplers disappeared. And the B-actin control looks strange-- not following 1:4:16 pattern and not the same between two samplers. I have been trying repeating the experiments for a whole week but the results were always like that. I am wondering whether anyone else has similiar experience in non-reproducibility of semi-quatitative RT-PCR? Any suggestion for me to fix the problem?

I think the number of cycles in my PCR program might be one problem. I was using 40 cycles because my mRNA was extracted from only 1000 cells. I could reduce it to 35 cycles but probably cannot go any further.


How did you store your template cDNA? If you had gone through a few freeze-thaw cycles, that my be your problem.


oh. I stored it at -20 and have re-frozen it for 6 or 7 times. So, that could be the problem.


the samples can be the problem..but also try with a different kit and the same samples....
If you keep repeting several times the experiments and you always get the sames results....that might mean that the first one you did was not right....


Thanks. But I was keeping getting bizarre results, which were not interpretable. The serial dilution of my actin controls kept yielding wierd results. For example, the 1:16 dilution sometimes yields even higher signal than the 1:4 and 1:1 . The Same were the serial dilutions for my gene of interest.

I tried 35 cycles yestoday with only actin controls , the actin results looked better. I plan to change all the reagents and my samplers with 30 or 35 cycles.


6 to 7 freeze/thaw cycles are very likely killing every reproducible result! I wouldn't trust your experiments. Is there a way to get new samples? You should aliquot also the cDNA if you want to use it several times...