Universal PCR

[Dick Beeman]

In article <4060220020 at uamercury.uark.edu>, DRHOADS at MERCURY.UARK.EDU ("Doug
Rhoads") wrote, in response to my post about universal PCR:
> When does one decide to add extra of a degenerate primer.  The `universal' 
> primer Dr. Beeman describes would have an approximate million-fold   
> (4^10=3D1,048,576)=degeneracy.  We have worked on amplifying homologous gene 
>  regions based on 
> degenerate primers developed from reverse-translation of conserved 
> protein domains.  Usually when we have 30 or 40,000 fold degeneracies we 
> up the primer 2-3x (from around 0.8 uM in the reaction).  In other tests 
> with specific, non-degenerate primers we have obtained much more 
> specificity in PCRs when we drop the primer concentrations to around 0.1 
> uM.  But with degenerate primers such as the `universal' primer, after the
> first cycle, don't you generate a million-fold excess of potential 
> competitive inhibitors for the priming site, plus reduce the concentration  
> of effective primers a few thousand fold??...(stuff deleted)

I agree with Doug, that, like PCR in general, it seems that universal PCR
should not work.  It's a good thing people roll up their sleeves and try
things that are theoretically impossible!  All of Doug's arguments seem
reasonable to me.  All I can offer are some observations on the conditions
that do work.  The universal primer (1 million-fold degenerate!) and
specific primer 1 are used at 1 pmol per ul and 0.1 pmol per ul,
respectively, in the first PCR reaction. In the second reaction, specific
primer 1 and T7 are both at 1 pmol per ul.  It may be that the first 9 or
10 bases on the 3' end of the universal primer are the only critical ones,
and that the effective degeneracy is only about 250-500-fold (the first 4
or so bases at the 3' end are specific).  Another point is that only a few
copies of the target sequence may be generated in the first round, but this
may be sufficient.  There are never any visible bands after the first
round, and EthidBr-stained gels of round 1 product would suggest a complete
failure of the PCR.  Remember that the 5' linker (T7 promoter sequence) is
effectively (partly) converted to a gene-specific primer for round 2, since
it can only anneal to three types of templates:(A) amplicons generated in
round 1 that were primed by "forward" specific primer 1 and "reverse"
universal primer, (B) extension products of these amplicons generated by
specific primer 2 in cycle 1 of round 2, and (C) amplicons generated in
round 1 that were primed on both ends by the universal primer.  Category
(C) is apparently not a major problem, since most universal PCR clones
turned out to be the right stuff.  
-- 
Dick Beeman
beeman at crunch.usgmrl.ksu.edu
"One of the advantages of being disorderly is that one is constantly making
exciting discoveries".  A. A. Milne