semi-specific PCR

Bryan L. Ford fordb at
Thu Feb 26 04:01:41 EST 1998

Dom Spinella wrote:
> Bryan Ford writes (in response to my earlier posting)
> > > Undoubtedly, your smear contains your gene of interest in addition to a lot of other
> > > products. You could always try cloning the smear and probing the clones
> >
> > Ho Ho, and it is always possible to tow a Mercedes-Benz with a team of
> > horses!
> >
> > If you have a probe then you have, or can easily have, the sequence
> > necessary to design specific PCR primer(s) to keep the "Mercedes" on the
> > road.
> >
> > Sincerely,
> > Bryan L. Ford
> Well Bryan, if you think about this more carefully instead of instantly
> being a smart ass, you might come up with reasons why you might want to
> tow this particular Mercedes with a horse.  For example, suppose our
> friend was looking for a full-length gene but had only partial sequence
> information (lacking the 3' end for example).  He could then use this
> strategy to enrich for a population of longer clones that could be
> subsequently probed (in fact, this is pretty much what is done in the 3'
> RACE protocols.)  The again, perhaps the PCR "smear" could be cloned in
> an expression vector to look for function upon transfection with what is
> now a much smaller, enriched population than the starting cDNA library
> (notice I also alluded to the possibility of a functional assay rather
> than a nucelic acid probe).  Since neither of us knows the particular
> experimental details in question, let's not rule out possibilites a
> priori. And Ho Ho yourself...
> --Dom Spinella

Yes, we are lacking some details. He is clearly not working with a
library of clones but instead a library of cDNA products. It appeared to
me that our friend, I assume to be Dr. Bahn, had 3' ends in abundance
since he used an oligo dT primer together with a "universe primer"
(which, in spite of its name, may be somewhat specific). Before jumping
on but one of your suggestions, I should have pointed out that your
assertions concerning the causes of Dr. Bahn's problems, in my opinion,
were absolutely correct. 

However, your last solution of probing the smear I found amusing in view
of the apparent success, that is a single or discrete PCR product, from
his specific PCR, i.e. using the "universe primer" with the "specific
reverse primer". If by some chance the "smear" from the universe primer
and the oligo dT consists of a vast number of unrelated cDNAs then his
clean PCR product can itself be used as, for example, a "megaprimer"
against the 3' anchored (as in your worthy suggested alteration) oligo
dT primer... or if he simply TA clones and sequences the "clean" product
generated with the specific reverse primer, he can then have newly
synthesized primers made with even higher specificity to improve his
chances of easily reverse transcribing to the 5' end (a bigger problem I
should think, than getting the 3' end, which appears to be close at hand
for Dr. Bahn). On the other hand the "specific reverse primer" may not
be so specific and might instead be giving an undesired product, an
"artifact" or "accident" of some sort. But, if this were the case then
it is likely that TA cloning then sequencing would readily reveal the
congruence or lack thereof with the presumed homologous gene(s).

Of course such a specific PCR product could also be labeled in several
ways to make a very nice probe if one had a phage or plasmid library to
look at-- but I don't see evidence of this in Dr. Bahn's posting.  Of
course, such a probe could 
also be used to do the probing of the "smear" that you suggested. But,
my own prejudice is do the simpler things first, and for me this means
"probing" with the often exquisite selectivity PCR primers to
discriminate desired signals, while using the tremendous amplification
if necessary to bring up such weak signals. (Prejudice born no doubt
from having done thousands of times as many PCR reactions as I have blot

I have no objection to expression libraries or functional assays based
on them, but as I read it Dr. Bahn is asking for quick solutions to
simpler questions than may require such greater efforts. That is, if all
the Mercedes needs is just a little petrol then let's not advise the
owner to buy a new one quite yet.  

Hopefully these posts will be helpful for our friend from Goettingen.


More information about the Methods mailing list