Yes- I think we have seen this a couple of times.  Once during my PhD with Robert Insall and once recently.
In both situations, we had to screen an unusually high number of clones to get few targeted integrants.  Both the initial southern and PCR were compelling.  The first problem was revealed after we did a few digest combinations for southerns.  The second incident was revealed using a Western blot showing the protein was still there.  Subsequent PCRs suggested an insertion in the genomic region, but 3' to the coding sequence.  Frightening stuff.  In the first instance, the targeting vector was verified by many digestions.  In the second instance, we sequenced the whole thing and it matched the Dictybase sequence, and the same 2kb genomic fragment was able to deliver a point mutation.  Several other labs have tried to KO the first without success.  I was very surprised the second gene targeted without a notable phenotype.  I wonder if one pushes hard to target an essential gene, these sorts of things turn up- often when you do a Southern on a bunch of candidate KOs, there are often the odd one or two lanes with funny bands, even though the wt band has gone- which may be part of the same phenomenon.

Dr. Jonathan Chubb
Division of Cell and Developmental Biology
School of Life Sciences
University of Dundee
Dow Street
Dundee
DD1 5EH
+44 (0) 1382 386336
>>> David Knecht <[log in to unmask]> 01/08/09 5:42 PM >>>
We have been trying to knock out a gene and have a strange result.   
The construct is standard with 5' and 3' non-coding flanking region  
DNA cloned upstream and downstream of Bsr in the LoxP vector.  We cut  
out the insert, purified and transformed cells.  We got what appeared  
to be knockouts based upon PCR screen with internal and external  
primers (we don't have antibody yet).  However further analysis gave  
us some anomalous PCR bands.  The best explanation for what we have is  
a single crossover recombination of a circular version of our linear  
fragment into the 5' homology region of the gene.  Has anyone seen  
this before?  Any suggestions as to how to prevent it so we can get  
the double cross-over gene deletion?  Thanks- Dave

Dr. David Knecht
Department of Molecular and Cell Biology
Co-head Flow Cytometry and Confocal Microscopy Facility
U-3125
91 N. Eagleville Rd.
University of Connecticut
Storrs, CT 06269
860-486-2200
860-486-4331 (fax)



The University of Dundee is a registered Scottish charity, No: SC015096