RT-PCR on 3kb fragment of viral RNA - (Oct/31/2006 )

I am trying to amplify a 3kb fragment at the 3' end of a single stranded viral RNA (no polyA tail). On the Reverse primer I have engineered a BAMHI restriction site (for cloning later) and the primer is 35 bases long, with a Tm of 68. The forward primers (have tried 2) have been between 30-33 bases with Tms 68-69. I am primarily using the Takara LA PCR kit, but have also tried Platinum taq (invitrogen), Clontech BD Advantage 2 system, and Novagen Hot-start taq. After cDNA synthesis, I have done an internal PCR usign known primers and have amplified a segment from the cDNA so I am assuming the cDNA synthesis is successful. The problem that I am having, is that when the pcr is run on a 1% TAE gel, I get 2 VERY distinct bands at <1kb (though band sizes are differnet with the 2 different F primers) and no band at 3 kb. Several times I've gotten one large smear from about 3 kb to <500bp. I've tried adding 5% DMSO, adding BSA to 1x concentration, done a gradient PCR (with differing temps making no differnce in the banding pattern), lessened amount of template, decreased primer/enzyme concentration, and I always continue to get the same non-specific bands. The thermalcycler has been consistent (Gene Amp 9700) which has worked just fine for other PCR with the same size amplicon. Does anyone have any recommendations? I'm getting frazzled trying to figure this one out. I'm almost out of Takara LA taq and was thinking about trying a new kit, but 3 kb shouldn't be that hard for Most taq. Also, if anyone knows a good program to try to determine non-specific binding of primers to template, that would be great. Thanks a ton!

you've probably already thought of this, but I wanted to cover all the bases...

did you take into account that the melting temp for the initial rounds of PCR is likely much lower, for the primer that has a restriction site?

also, has anything been done to determine integrity of your template? for example, if your DNA is degraded or sheared, then it would not amplify well

as far as nonspecific binding between your template and primer...you did use a sequence alignment software to check that your primer's sequence is unique in your organism? if it is, there shouldn't be any other binding to get concerned about

The internal fragment you amplified (to know that your cDNA was OK), where is it situated? If it is near the primer site used for the cDNA synthesis, you can not be sure that the entire cDNA has been synthesized (and hence you won't be able to know IF you can amplify your 3 kb product starting from the cDNA you have).

Is it a retrovirus? If so, what I'd do is (as we have done in our lab for optimizin RT-PCR): try amplifying your 3kb starting from genomic DNA, and if you get that one specific and sensitive, try it on your cDNA. Sensitivity can be an issue, knowing that cDNA synthesis isn't too efficient (RT's aren't too processive, longer pieces, as 3 kb can be considered longer are quite hard to get).

What RT did you use? Is there heavy secondary structure in your RNA in the region you amplify?