PCR enzymes

Chris Boyd chrisb at hgu.mrc.ac.uk
Fri Apr 12 08:33:46 EST 1996


Tori (twilliam at BIOMED.MED.YALE.EDU) wrote:
: Does anyone have any suggestions as to the highest fidelity theromostable 
: enzyme to use for PCR? I am  about to start a mutagenesis project and I 
: want to make sure that the only point mutations created are the ones I 
: design.

Every manufacturer claims that their polymerase is more accurate than
that of their competitors, and of course they are right!  However, it
is commonly agreed that Taq is the worst performer because of its lack
of proofreading activity, so for all but short PCRs you should not use
it.  As for high fidelity variants, I have heard good reports of Pfu
(Stratagene) and NEB's VENT, though I have not used them. Boehringer's
High EXPAND worked for me -- it is a proprietary mixture of Pwo and Taq
and is supposed to combine the processivity of Taq with the
proofreading of Pwo.  I'm sure there are other similar products, but I
suspect there's little to choose between them at the level of fidelity.
The question to ask is, "what available proofreading enzyme or mix
reproducibly gives me enough PCR product to clone?". By thereby being
able to minimize the number of effective cycles you obviously also
minimize the probability of misincorporation.  See the papers about
long PCR (Wayne Barnes etc.) for more info on enzyme mixes.

HOWEVER, don't EVER think you can get away with assuming that there
will be no PCR-induced errors just because the polymerase spec predicts
so. There are other variables affecting fidelity such as template, dNTP
and primer quality. You ALWAYS have to check by sequencing the
mutagenized DNA segment.  For one thing, a common source of error is in
the primer itself.  In all the PCR mutagenesis I've done, the majority
(more than 75%) of unwanted mutations have been in the primer tracts.

DISCLAIMER: I have a research collaboration with Boehringer.

Best wishes,
--
Chris Boyd                       | from, | MRC Human Genetics Unit
chrisb at hgu.mrc.ac.uk             |  not  |  Western General Hospital
http://www.hgu.mrc.ac.uk/~chrisb |   for |   Edinburgh EH4 2XU, SCOTLAND



More information about the Methods mailing list