RNA, cDNA, PCR - RT Inconsistency - High Quality RNA, Working RT, Varying PCR Results (Nov/16/2008 )

I've done my fair share of RNA extractions, Reverse Transcriptions and PCR's but I can't get around the problem I'm currently facing:

During the recent months I've been extracting RNA of plant ovules, but I can't manage to get a solid expression of my known to be working control gene. I don't want to disclose on the plant's species and deem it irrelevant.

I extract my RNA with a pbiozol kit, which seems to be a some kind of modified trizol extraction - my results are a lot better with it than with trizol or CTAB extraction. RNA concentration is decent and definitly sufficient for atleast 10 RT's per sample and my 260/280 is > 1.8 and mostly around 1.9 but never higher than 2. RNA run on agarose looks great.

I polyA the mRNA with a PAP set from invitrogen (how do I pronounce that company's name correctly btw?). Subsequently I do RT-PCR with 3ug of sample RNA and use AMV for RT. A lot of people in my lab use the same RT set-up and get good results. They however use a different lysate for extraction (as they have different plant material).

Afterwards I do PCR with primers for a polyubiquitine gene and use a Tm of 58. This two step method is not giving me stable results though, sometimes some sample have bands, sometimes the same samples don't. Sometimes if I do RT on let say 10 samples 2 have bands, sometimes 9. Sometimes the bands disappear or appear in subsequent PCR's. I can't really pinpoint what is causing this flux or absence in/of PCR yield, although I certainly have been trying to isolate it.

I've purified my RNA with NACL/EtOH precipitation several times and once it gave me 9 out of 10 band for the initial PCR but the subsequent PCR's had first 3 and then only 2 bands remaining. I run a control every time when I do PCR and RT and there seems to be no contamination. The problem has been lingering around now since September and is driving me crazy. I've tried as much as I could think of to isolate the disruptor; which includes skipping the PolyA, changing RT enzymes, PCR enzymes, Buffers, dNTP's etc. Even if I purify my RNA whith 3 to 4 Na/EtOH ptts & 80% EtOH washes.

I extracted, and now wasted, a lot of my plant material and used up quite the sum of resources. As i've extracted and RT'd and PCR'd for so many times (prob. >150 extraction, you can guess how many RT's, PCR's). There is no DNA contamination in the RNA samples, and the high OD rates suggest no contamination (I still used DNAase for a try). Also even if I store my RNA in the freezer for up to a month (-10, but it's often being opened and closed) the quality is OK and suggests very little RNAase activity.

I'm looking forward to move on but how will I ever get to the quantative RT-PCR part of this experiment... Even more so annoying is the fact that I can't get around this seemingly simple dilemma, which really mocks my aspiration of becoming a scientist and is far more aggravating than anything above-mentioned.

For a start, I would play with increasing Mg2+ concentration and add BSA as a stabilizer. It looks for me that it's the PCR step you have problems with, it almost always is. With poor annealing and scarce matrix, results vary depending on the early cycles, if the primers attach or not. cDNAs aren't stable, too; after a few freezings/thawings they get worse and worse. Some colleagues told me that even three weeks in a freezer can make cDNA stick to the wall of your cDNA tube, so you pipette only water. I have kept cDNAs in -80 for months though and haven't got this kind of a problem. Aliquoting your fresh matrix seems like a good idea.
Get good control primers of an abundant gene (I don't know what it would be for a plant), so you can check your cDNA.
And don't worry, shit happens.

Once before I did several PCRs with a MgCL2 gradient, however without luck - I however did keep the dNTP levels constant, I'm not entirely sure how that relates to increased MgCL2 levels. Today I ran a comparative PCR with some of the samples containing BSA - 1ul of 10mg/ml BSA for a 20ul PCR reaction micture, if I remember well - nothing really changed. The samples that had a working, successful PCR reaction had slightly cleared up bands, but the ones without were still without anything at all.

Thanks for your advice, it was worth the try.