Problems with semi-quatitative PCR - (Feb/10/2011 )

Hi all,

I have been working on this project for past 4 months. I am treating my cells in increasing dose of a drug and trying to look for a dose response by quantifying 4 different factors made by my cells.
I have run into many weird problems since the beginning. I isolated RNA using Trizol reagent from invitrogen. I did RT-PCR to get the cDNA (using same amount of RNA to get them consistent). I am using cyclophillin as my housekeeping gene to verify uniformity of my samples. Unfortunately it gave me random bands in the PCR, I repeated it many times, used different primes but the problem didn't go away. I thought something was wrong with my samples, also my 260/230 and 260/280 ratios weren't the best. So I read through the forum and decided on using the qiagen minelute kit, to get more pure RNA, but I don;t know why that didn't work either. Meanwhile I got this Versagene RNA extraction kit, that worked perfectly. So I repeated the whole experiment and got the RNA.
This time cyclophillin although uniform had a second smaller sized band in it. My -RTs and -ve controls do not have any contaminating band. I have seen this problem only with these experiments, and none else. I next did PCR for my factor of interest, that looked fine, no contaminating band, however i have had trouble getting consistent results. It shows one trend one time and something random the other times. So to get past this issue, my boss suggested i do the PCR for the housekeeping and experimental (factor of interest) at the same time. I can see two separate bands when i do them together (bands of expected size), but when I do the PCR with all three conditions together (experimental and housekeeping together, housekeeping alone, and experimental alone), to look for any kind of competition between primers, I see a very small sized band on the housekeeping gene lane. I have done this twice and with same result.
I have no idea how to explain this.

I am in desperate need of suggestions for getting consistency and ways of explaining these observations. All help is greatly appreciated.

Thanks.

You maybe don't want to hear this, but if you want to quantify anything you need really a quantitative method (currently real-time RT-PCR, or before that semi-quantitative gel-based quantitative-competitive PCR) classic PCR is not. It can only tell you something if you titrate your samples low enough to be caught between background and plateau phase (and still be visible).

Nowhere in your description you mention DNAase treatment (Trizol doesn't have it and it's only optional in Qiagen RNeasy kits), but you said your -RT control was negative (for all genes?) so that shouldn't have an effect on the PCR. DNA contamination could however screw your concentration measurement and 260/280 ratio. You can check it by running RNA (denatured, with formamide and buffer) on gel, DNA band would be visible on the top and from the brightness of rRNA bands you can tell quality of your RNA. Being sure you have a good RNA is the first step.