In article <199702042236.QAA20303 at borcim.wustl.edu>,
brett at BORCIM.WUSTL.EDU (brett) wrote:
> >When subcloning DNA fragments larger than 5 kb into pBluescript, I
> >frequently recover DH5a mini-prep clones with various-sized deletions or
> >no inserts. What reagents (bacterial strain, plasmid vector, restriction
> >sites) should I be using so that I can subclone large (5-15 kb) DNA
> >fragments that are intact?
>> We maintain RNA viral genomes as cDNAs (10-13kB) in low copy plasmids with
> fairly good luck. Try pACYC, or pBR322, or derivatives thereof. Bluescript
> is pretty high copy #, so this might help. Also, I tend to use MC1061 for
> routine manipulations but have resorted to WM1100 or SURE for things that
> like to recombine.
Strange. I've never noticed anything recombining in DH5a. We commonly
subclone entire lambda clones into the NotI site of pBS. I've never had
any trouble with inserts up to 14 kb. Some of the time I get poor yields
on plasmids with greater than 14 kb inserts -- but I've never noticed any
deletion occurring. This includes subclones of lambda clones that are
unstable--I occasionally see such lambda clones undergo recombination between
the LTRs of a retrotransposon they happen to contain. But once subcloned and
in a recA minus strain (such as DH5a) they seem stable.
(The record insert size into pBS in the lab is 35 kb of genomic sorghum DNA.)