Internal deletions by PCR

Paul Kowalski kowalski at stanford.edu
Fri May 19 02:56:41 EST 2000


Hi - I haven't done much inverse PCR, just this "PCR-around-an-already-made-plasmid" to make
deletions, but point mutations can be done the same way, as you say, with the mismatch in the
middle of a oligo. Use a proofreading Archeal polymerase. I recall a good treatment of a "2nd
Generation" Inverse PCR protocol was in a recent Nature Genetics paper out of Neal Copeland's
lab, on leukemia disease gene cloning in mouse, a few months back. Sorry - not at work and
don't have the paper here.
Good luck-

--
Paul Kowalski
Postdoctoral Fellow
Genetics - Stanford University School of Medicine
kowalski at stanford.edu

Philip Carl wrote:

> Hi Paul,
>
>         I am just about to try this technique for a class I teach on cloning
> methods so I'm afraid I'm in no position to give you advice.   Actually,
> I was wondering if you could give me some.   I've read a lot of
> techniques papers on Inverse PCR, and think I know what to try first,
> but if you had a detailed protocol of what worked in the past or any
> hints that you found important, I'd sure like to learn of them.  In
> particular I'd like to use the technique for creating a substitution
> mutation rather than a deletion.  I was planning on putting the base to
> be altered about midway in a 26-mer primer.   Does that sound about
> right?  I note that Frank Fackelmayer who replied to your request had a
> brief note on the technique in June of 98 that I had somehow saved.  In
> that note he recommende using 50 ng of template.   Now he says no more
> than 1.   I don't know what caused him to change his advice, but I'll
> try to find out.
>
> Paul Kowalski wrote:
> >
> > Ha! And I thought *I* was the only person to use this technique! ;)
> > It's always worked well for me in the past......except for right now! I've been trying
> > to "pcr-around-the-plasmid" of a Tet-off construct containing an IRES from EMCV, in
> > pRC/CMV backbone, and having nothing but trouble. I initially thought it was the
> > extensive secondary structure in the IRES, but that can be PCRed fine by itself- just
> > not the whole plasmid (or almost the whole plasmid- I'm doing this to get rid of the
> > neo marker, to put in puro/hygro) . I've used LTI HiFi DNAP, Qiagen Hotstartaq,
> > Stratagene Herculase, got long primers with a high Tm (doing 2 step cycling, 68C), have
> > tried using both circular and linearized (within the to-be-deleted neo ORF) template
> > DNAs, 10 minute 96C denaturations prior to DNAP addition, and I'm out of ideas. Any
> > troubleshooting suggestions or controls? Thanks in advance-
> > pk
> > --
> > Paul Kowalski
> > Postdoctoral Fellow
> > Genetics - Stanford University School of Medicine
> > kowalski at stanford.edu
> >
> > "Frank O. Fackelmayer" wrote:
> >
> > > Hi Ulf,
> > > We do that quite routinely, and the method that works best is what I
> > > call "PCR around the plasmid". The idea is to design two primers that
> > > flank the region to be deleted, but point in "the wrong direction" i.e.
> > > the other way round than those that would amplify the region in PCR.
> > > Perform PCR with a good proofreading enzyme (we use Pfu) and long
> > > extension times.  You´ll get a (blunt ended) fragment of "plasmid minus
> > > deleted region"-size. Gel-isolate the fragment, then ligate.
> > > Important details:
> > > * primers must be 5´phosphorylated and good quality
> > > * use no more than 1ng of template
> > > * it is often useful to destroy the original template after ligation by
> > > digestion with DpnI (which won´t harm the PCR product because it is not methylated)
> > >
> > > In most cases, greater than 80% of clones are ok.
> > >
> > > Hope this helps,
> > > Frank
> > >
> > > "Ulf S. Andersson" wrote:
> > > >
> > > > Hi,
> > > > I wonder if anybody could tip me on a good and fast way of making
> > > > internal deletions in a plasmid by PCR. I have previously done it by
> > > > introducing new restriction sites by site directed mutagenesis (and then
> > > > digesting and re-ligating) which takes quite a long time... there must
> > > > be a smarter way.
> > > > /Ulf.









More information about the Methods mailing list