Questions regarding RT-PCR optimization - (Feb/13/2011 )

Hi, all. This is my first time here.

I am currently analyzing the expression pattern of a gene in different organs (brain, liver, lung, testis, heart) of mice.
I have isolated RNA from these organs, converted the RNAs into cDNA, ran PCR.
Very faint band were observed for all of the samples.

My professor asked me to do optimization for the reactions.

My questions are:
1) why do i need to do optimization? The presence of very faint band indicates that the gene is weakly expressed, isn't it?
2) if there is a need for me to do optimization, which sample (brain/liver/lung/testis/heart) should i use for optimization, and why? If, for example, I optimize the reaction using brain cDNA, but the fact is that the gene is not expressed in the brain, I would be wasting the brain sample, wouldn't i? All the samples are very precious...

thanks

To properly answer this question you need to provide a little more information. The fact that you got very faint bands in all tissues suggests that your assay is not working very well since it would be very unusual that your gene would be expressed at such a low level in all tissues. Actually I would guess that the band you are seeing may not be gene-specific. A more expected result would have been to have seen at least one tissue showing a little more robust amplification than the rest.

As for which tissue to use for optimization, that is kind of tricky. Typically I would go with liver since that tissue tends to show robust amplification for a lot of genes. Then again your problem could be your RNA itself. If you are not sure that your RNA is of high quality then the low amplification could be due to your RNA and not your PCR assay. If that is the case, you may want to consider purifying more RNA.

Finally running a control PCR, for example to measure the expression of a housekeeping gene like GAPDH, would be a good idea. This would tell you if your RNA is of good quality and also determine if every step of your experiment is working as it should.

My two cents.

I agree with the above.

Also, using gel analysis to determine whether something is lowly or highly expressed isn't very accurate. For one, its not very sensitive and minor differences in loading each sample could adversely affect your results. If you want to determine levels of expression I suggest using qPCR and not gel analysis. A good PCR reaction should give discrete bright bands everytime, regardless of levels of expression.

Best ways to start optimizing are to alter annealing temperature, try Tm-y touchdown PCR, re-design primers (do you have primer dimer formation?).
Some PCR reactions are just not very efficient due to GC content. Try to use freshly thawed mastermix, and have your reactions on ice until you are ready to start the PCR machine.

If you are seeing a really faint band, did you let your gel run long enough? If you don't run your gel long enough a smear would appear as a faint band. Smears indicate non-specific amplification and you probably want to re-design your primers or increase annealing temperature.

Hope this helps.