Cloning: Unstable Vector??

Malte Buchholz mbuchho at
Fri Dec 18 07:47:03 EST 1998

wtd3 at wrote:
> I have a cloning problem which is new to me, though hopefully someone
> can help.
> My  vector is a 7.5 kb genomic fragment of the gene I am studying, and
> I'm trying to clone a 7.8kb PGKNEO-ires-TauLacZ cassette into this
> vector to generate a targeting construct.  The problem is that I get
> very many different products on minipreps of the colonies, only about
> 25% of which are identifiable as vector religation.  The others are all
> sorts of strange things which make no sense.  When I made minis from 5
> colonies on the vector control plate, I only had 2 colonies that looked
> like  the vector, while the others were strange (and significantly
> smaller).
> The genomic piece is in pGEM 3Z, and the ligation is a single sticky end
> (Xho1).  I have purified the vector (and insert) very carefully, even
> running them out >20hrs on a long gel to remove any contaminants.  Also,
> I have already tried remaking the maxiprep from a newly picked colony.
> Finally, I've been doing a lot of other cloning without any problem like
> this, so the problem is not, for example, some crap in my water.
> Anyone see this before or have any ideas?   How about moving the genomic
> piece to a new vector (?pBluescript?).
> Thanks,
> Bill Dauer

Hi Bill,

I had a similar problem when cloning an 8kb piece of mouse genomic DNA
into a Bluescript vector; in this case the genomic insert by itself was
stable but addition of a marker cassette resulted in all sorts of
strange products. Mostly the reason  for this is the occurence of DNA
conformations (Z-DNA, inverted repeats etc) that normally do not occur
in E. coli and are being rearranged by the E. coli repair machinery.
What I did was switch to the SURE strain which is defective in all
repair pathways and reclone my insert in pBR322 (low-copy vector).
Worked fine from then.

Hope this helps,


More information about the Methods mailing list