Just an additional FYI:
Getting DNA into E.coli works best with supercoiled plasmid DNA, but
getting DNA into B.subtilis works best with linear DNA.
The reason is that B.subtilis uses an active uptake system to absorb DNA
in the environment, which can only absorb DNA if it can find a loose end
to chew on. It's been speculated that most B.subtilis transformants from
plasmid solutions are actually those that found damaged or cut plasmids
and absorbed them, which contributes to B.subtilis' undeserved
reputation for being genetically unstable (that is, the DNA was already
damaged, it's not the bug's fault! :))
The differences in efficiency are pretty dramatic. Especially because
the active uptake system includes chaperone proteins that help trigger
homologous recombination with the chromosome if there are suitably
homologous regions.
So, my suggestion is the build the DNA you want in a shuttle plasmid,
with nice big (500bp+) chromosomal flanking regions, in E.coli. Then use
PCR to generate loads of linear copies of the desired flanking regions
and genes, and transform with PCR product. You'll *probably* get better
results that way.
That's only true if you're trying to get chromosomal integration. For
plasmids in B.subtilis, efficiencies are pretty good with lab strains,
so you can ditch E.coli and just work with B.subtilis if you have the
right plasmids and workflow.. an area I'm working on right now myself.
Protip: if you're designing DNA for B.subtilis 168 derivatives, avoid
the XhoI restriction enzyme site, and the TopoI recognition site for
B.subtilis: WCATWTAWWA (W = A/T, acc. to IUPAC DNA notation). You'll get
considerably higher efficiency and DNA stability.
--
www.indiebiotech.com
twitter.com/onetruecathal
joindiaspora.com/u/cathalgarvey
PGP Public Key:
http://bit.ly/CathalGKey