upper limit to Inverse PCR products?

Vivian Miao vmiao at unixg.ubc.ca
Sat Jan 20 14:59:47 EST 1996

In article <4dome7$edf at elna.ethz.ch>, Michael Kertesz
<kertesz at micro.biol.ethz.ch> wrote:

> having the correct dilution is crucial to make circles and not polymers,
> but if you go TOO dilute, you run out of DNA.

I've talked with a couple other people since posting and we came to this
point also - how to be sure the ligation involves mostly intramolecular
reactions.  This might be expected to be an increasing problem with circle

The way to trouble shoot this - which I was hoping (am still hoping to
avoid) is to spike the genomic DNA with various amount of the linearized
plasmid that I have been using as a positive control to determine the
efficiency of intramolecular ligation in a complex mixture.  One problem
is that you have to be sure the plasmid is completely linearized.  The
only ways I can see to do this is to isolate the linearized plasmid out of
a gel, and 1) attempt to PCR from that and hope for the negative result,
or 2) transform E. coli and again hope for a negative result or both.  If
the result is negative, THEN do spiking and the ligation for PCR.  

> adding an equivalent aliquot of mock ligation mix to my control PCR
reactions >significantly  reduced the PCR yield. Maybe the ligation needs
to be cleaned up 

I thought about this too - perhaps ligase hangs onto the DNA somehow and
prevents the polymerase from going through (possible?). I did the same
control as you, but in my case found no decrease (by inspection of EtBr
stained band) in the amt. of product from the spiked plasmid I added
compared to the plasmid alone.  I used 1 ul of the ligation in a 40 ul PCR
reaction; maybe this dilution was enough to eliminate the effect of "bad"
stuff in the mix.  

> Do you have a RECENT reference for inverse PCR (or a couple of publications 

No, I haven't seen any yet, but would be glad to learn of some.  I have
seen recent papers using inverse PCR, but they were not dealing with
genomic DNA (e.g. they were making mutant plasmids etc, which is a
different situation).

More information about the Methods mailing list