heterologous expression problem in E. coli

Frederik Boernke boernke at nospam.ipk-gatersleben.de
Thu Dec 9 14:35:57 EST 1999

Hi all,

we have cloned this novel gene and all we need to complete the
is a functional proof that the enzyme is really that what we claim.
Therefore we initially planned to complement the respective E. coli 
mutant but things have been much more difficult than we thought.
The first thing we observed was that when ever the respective fragment
was cloned in frame into any E. coli vector, i.e. pGEM-T, pBS or pQE
the ligation yielded only constructs were the fragment was oriented
in antisense. And we did a hell of a lot of minis. Thus, we assumed
a somehow toxic effect of the protein on E. coli. However, when the
ligation reaction was transformed directly into M15 cells harbouring a
repressor plasmid which confers very tight control of the lac promoter,
we were able to obtain plasmids with the proper orientation of the
fragment. From our point of view this finding supports the notion that
expression of the protein has some toxic effect. When those cells (M15)
were induced to express the protein absolutely no expression  could be
detected. We verified the construct via sequencing and everything seems
to be fine. 
After analysing the codon usage of E. coli with respect to our cDNA, we
came aware that our gene of interest contains a hell of a lot of
codon rare in E. coli. At the moment we are trying to express the
using Stratagene's BL21 codon plus as a host in order to see whether we
can overcome a possible codon usage problem. But even if so, that would
not help us with our mutant complementation experiment. So we came up
with the idea to coexpress the respective tRNAs together with our gene
in the mutant. How can this be achieved? Where to get the respective
vectors from? Any other ideas to solve the problem?



Frederik Boernke
Research Group of  Molecular Plant Physiology
Institute for Plant Genetics and Crop Plant Research (IPK)
Corrensstr. 3
06466 Gatersleben
Tel.  039482 -5 321
Fax. 039482 -5 515

More information about the Methods mailing list