Which strand of a plasmid is replicated?

tfitzwater at gilead.com tfitzwater at gilead.com
Thu Sep 14 13:56:14 EST 2000

>From: Wolfgang Schechinger (Wolfgang.Schechinger at med.uni-tuebingen.de)
>Date: Wed 13 Sep 2000 - 22:07:47 BST

>Tim, thanks a big lot for this precise answer. May I conclude that it
>is theoretically sufficient to modify one strand (the lower strand in
>common sequence notations) for mutagenesis (by the quick change
>protocol) and to use a standard primer for the other one (in order to
>generate a fully non-methylated DNA necessary for the DpnI digest
>wich degenerates the template DNA)?

>All the best,


I assume that you are not modifying the Primer RNA sequence.  While the
strong secondary structure did not prevent me from generating mutants in
Primer RNA via the Kunkel protocol 7 years ago, many changes are lethal or
induce compensating changes to maintain basepairs required for proper
function of the Primer RNA.

You must use mutant primers for both strands in the Stratagene QuikChange

The parent plasmid can be cut by Dpn I because it is dam methylated.
PfuTurbo product from first cycle is hemi-methylated and still cut by Dpn

Dpn I digestion will trash 2/4 plasmids from PfuTurbo products from second
cycle, because one strand is methylated.  Note that only the plasmid
template strands coming into the reaction are methylated, but these are
maintained throughout the PCR reaction.  By the 3rd cycle, only 2/8
plasmids are hemi-methylated, 2/16 in the 4th cycle, 2/32 in the 5th cycle,
etc. (assuming complete amplification).
But, none of the progeny are completely mutated.  In the 2nd cycle for
example, 3 plasmids are mutated on only one strand, and this mismatch is
maintained throughout the amplification.  The DNA will not be base-paired
in these positions and following transformation DNA repair mechanisms in
the E. coli will, presumably randomly, repair the mismatch and 50% of your
colonies will be wild type.

The use of two complementary mutant primers solves this problem, since both
strands are mutated, E. coli DNA repair mechanisms never come into play and
83-100% of the colonies generated are mutant (according to Stratagene).

Good luck,

Tim Fitzwater
Principal Research Associate
Gilead Sciences


More information about the Methods mailing list