Help needed for megaprimer mutagenesis using PCR

Richard Pelletier rpelle1 at
Fri Jan 31 17:49:21 EST 1997

In article <32E7CD75.1671 at>, lbentson at wrote:

> I have been using the megaprimer mutagenesis method to introduce
> EcorI sites in the pR' template in E. coli.  This method uses two Pcr
> reactions to inroduce a mutation.  In the first PCR reaction, a
> universal primer is used (T7) and the mutagenic primer is used to
> amplify the DNA.  The second PCR uses the product form the first PCR
> reaction (which should contain  the mutation) and a second universal
> primer (T3) which should amplify the rest of the DNA.  We have been
> getting strange results from the first PCR reaction which is usually the
> most simple.  At first we were getting PCR products larger than the
> template (from T7 to T3 it is 545 bp).  This does not make any sense or
> seem possible.  We did identical reactions but changed the annealing
> temperature from 50 to 55 degrees C and got products with multiple bands
> one still being larger than the template used and one that is the same
> size as the template (545).  The expected sizes should be around 300
> bp.  The reaction  conditions that are being used have been done before
> in the same manner and have been successful.  I would welcome any
> suggestions.
>                         Lori Bentsen
>                         lbentson at

         You might be using too many cycles, which increases the chances
of amplifying a contaminating DNA template and also getting several
non-specific bands. I would use less than 30 cycles. Also, if you use too
much template, it is more likely that you will get more garbage (non
specific bands). If you are using a plasmid template, I would use less
than 1 ng, maybe 100-200 pg. Also, try doing a hot start, which will
improve the specificity. The products you are getting that are larger that
your template are probably recombination products (as the Taq pol. is able
to switch templates during the reaction). This has been documented in the

Hope this help.

Richard Pelletier
Experimental Medicine
McGill University

More information about the Methods mailing list