HELP: Coli is rearranging recombinant plasmid

P.J. v. Santbrink santbrin at CHEM.LEIDENUNIV.NL
Tue Apr 21 08:56:14 EST 1998

As many fellow researchers I'am having problems cloning several
(eucaryotic) inserts in an expression vector.

I've to ligate the inserts in the Xho I - Spe I site of my vector.
Problem is that the sites are very close to each other, so after
digesting the vector with the enzymes and gel purification you only know
for certain that you exclude the presence of uncut vector.
When I perform  ligations of the presumably double cut vector with and
without the insert I get in both cases many colonies, but the background
(ligation in the absence of insert) is almost as high as the ligations
in the presence of insert (did check several colonies: all where
reannealed vector).
To get rid of the background I'm currently trying the so called three
fragment ligation : cut vector with Xho I and other unique site in the
vector; gel purify right fragment. In other reaction cut vector with
SpeI and the same unique internal site; gel purify: perform ligation
reaction with two gel purified fragments and the Xho I - Spe I insert
(different molar ratios). in theory, the ligation can only be
successfull in one way.
I did get colonies and I performed a colony check by means of PCR with
primers covering the eucaryotic insert. most of them where positive, so
I was very happy.
Next I made some overnight cultures, did mini preps and digested a
fraction with Xho I and Spe I to confirm the correct arrangement of the
to my surprise the insert was NOT present in the Xho I- speI site: the
plasmid was not cut, and supercoiled vector of all mini preps ran much
lower then the control vector (empty vector). The inserts are between
0,3 - 1 kb in size so there should be a difference in running patern on
a agarose gel. But even if the insert is not present, the vector should
run at least on  the same level as the control vector (though with this
3 fragment method ligation of the vector without the incorporation of
the insert is not possible )
So what's happening ??

I learned on this internet site that Coli sometimes deletes/reaarranges
recombinant plasmids when they don't like what they get (toxic product,
palingdromes etc.). This could explain the low position of the uncut
recombinant vector in the agarose gel.
I use DH5 alpha's to transform. the vector is pUC18 like, so high copy
What can I do to prevent this deleting/ rearranging ?
I read something about using different bugs for this deleting
problem.or  maybe grow at a lower temperature, but after transformation
and plating I checked several  colonies directly (so, no minipreps) and
there was the same pattern on gel.

So has anybody experienced the same problems and if so did you find a
solution ?

all help is welcome !!

Peter van santbrink
Div. of Biopharmaceutics
Leiden University,
the Netherlands
E-mail: santbrin at

More information about the Methods mailing list