Site directed mutations

bugg at bugg at
Thu May 6 19:22:39 EST 1993

In article <1993Apr20.121953.15040 at>, jeffs at (Jeffrey Alan Silverman) writes:
> Has anyone out there any experience with Enzymatic Inverse PCR?  I am trying it
> with little success.  After making the primers - exact match for 20 bases containg a restriction site for a class !! enzyme- I get NO product in the pcr 
> reaction!! (despite trying the obvious manipulations in temperature, Mg, various
> times of anneal and extensions etc).... So does anyone actually use this method?
> how did you get it to work? 
> Related: if you don't use this method but rather favor another for site
> directed mutagenesis... I'd be interested in hearing about that too...
> Thanks  ;-)
> Jeff

I use unique site elimination on double stranded plasmids very effectively.  I
have mutagenized plasmids for E.coli and mammalian expression by this method. 
The protocol is from Deng and Nickoloff.  When I first started doing it someone
in the lab had a Transformer kit from Clontech, which works just fine. 
However, when the reagents ran out, I didn't see any sense in spending that
much again, so I just made them according to the Deng and Nickoloff paper and
bought T4 DNA polymerase and T4 DNA ligase from Promega.  
The basic idea is that you simultaneously anneal your heat-denatured double-
stranded plasmid with both a mutagenic primer and another primer (both PO4) 
which is designed to alter a unique restriction site in the vector (it 
must be in a non-essential area).  Then, you use T4 DNA pol and T4
DNA ligase to extend around the circle and ligate to close up.  You transform a
mutS strain of E.coli with the in vitro reaction and grow up a "pool" with
antibiotic selection overnight.  The next day, you make a miniprep of the
plasmid pool and digest to completion with the restriction enzyme whose site
you are hoping to eliminate.  A second transformation into a "good" E.coli
strain like DH5 alpha is then plated out with antibiotic selection.  This time
you make up minipreps from individual clones and check for the loss of the site
by restriction digest.  Then, you confirm the mutation you want by sequencing. 
I have had anywhere from 1/6 to 6/6 of my minipreps lose the unique site.  When
sequenced, I have found that 100% of those that lost the unique site gained the
mutation I wanted from the other primer.  I think the variation in unique site
elimination is due to specific annealing differences between the two primers
that you design.
This also is relevant to the question about transformation with linear DNA into
E.coli.  I always use electroporation for these USE experiments and I have not
seen any evidence that linear DNA is effective for transformation.  If it were,
I would get huge numbers of transformants and they would all have the site I am
trying to eliminate.  I think the ones that still have the site have just
escaped digestion in the linearization step.

Barbara Bugg
St. Jude Children's Research Hospital
Memphis, TN  USA 

More information about the Methods mailing list