PCR: mutations in the primer region?

Frank O. Fackelmayer fof1 at chclu.chemie.uni-konstanz.de
Thu Sep 3 05:19:48 EST 1998

> Tim,
> Last month, I had an eeringly similar experience... mutations only in the
> primer region (in my case extra g's) and not in the PCR'ed region. I was
> ready to call the oligo synthesis company when I figured I would try the
> PCR/ligation again (it was Friday and I wouldn't get the new primers until
> at least the middle of the next week). This time I tried a hi-fidelity Taq
> and ligated them into pGEM. When I sequenced the products, there were no
> mutations. So... I guess, I can assume the mutations in the first reaction
> were actually from the PCR and not from the oligo synthesis. I still,
> though,  have no good explanation as to why I found mutations in only the
> primer region and not in the remainder.
> Good luck and let me know what you figure out,
> Zal

Just a thought: Is is possible that Taq makes more errors close to the end of
a linear DNA than in the middle? 
What I think of is the following: Usually a polymerase dissociates from the
template after a misincorporation. Sometimes, the polymerase molecules will
bind again and extend the strand, giving rise to a "fixed" mismatch. In other
cases, it will not bind again but leave the strand unextended. Now, if the
error is "in the middle" of a PCR product, the non-extended strand will not
amplify any more. If, however, the misincorporation is in the primer region,
the strand ending with a misincorporated base might very well serve as a
template for amplification. Over many cycles, you could have more errors in
the primer region than in the internal sequence. 

If you use a high-fidelity polymerase with proofreading, you should not get
the misincorporation, and hence no errors in the primer region, as you
described. A higher annealing temperature may do the same job (although not as
efficient), because the non-extended strand has a shorther overlap with the
primer and should not anneal efficiently.

As I said, it´s just a thought.


More information about the Methods mailing list