Rare arginine codon

Chen Ho An chen at bsm.bioc.ucl.ac.uk
Fri Sep 11 08:17:46 EST 1998

Bryan L. Ford (bryan.ford at orst.edu) wrote:
: Tomer Tishgarten wrote:

: > Our lab is interested in overexpressing a gene in E. coli, but the
: > protein contains several arginines that translated from a rare codon.  I
: > was wondering if anyone can suggest a person/company that can
: > provide/sell a vector which will compensate for this codon by
: > overproduction of the appropriate tRNA.

: This approach seems difficult since the imbalance of tRNA will likewise
: influence translation efficiency of the native E. coli genes. 

A few papers have been published which showed that this can be done.  For
example Hua et al Biochem Mol Biol Int (94) 32, 537-543 and there
are others. E.mail me if you want more reference and I see if I'll can dig
them up.  I think you can get the argU gene from CGSC at Yale


You'll probably need to clone it into a plasmid with a single copy 
number.  From the experience of other people it would seem that it does
improve the expression but the cells grow much slower, may be as a result
of the disturbance to the change in tRNA composition, but more likely due
to the antibiotic (since there is only 1 copy of plasmid and therefore
possible lower resistence to the antibiotic).

: Let me ask if
: there is any reason that you cannot modify the rare codon into a
: preferred E. coli codon, using the quite straightforward (that is
: *easy*) methods of site-specific or site-directed mutagenesis?  For
: example (using data for E. coli taken from the Codon Usage Database
: found online at http://www.dna.affrc.go.jp/~nakamura/codon.html ) we see
: that AGG is the rarest at 1.5 per thousand, and CGA is relatively rare
: at 3.6/k. Both can be changed into frequently used codons with single
: base changes, that is to AGC with a usage of 15.7/k and to CGC, which
: happens to have the highest usage of all six arginyl codons at 21.5/k.

This approach is useful if there is only a few rare codons close together 
(actually you'll find that in most cases only AGA or AGG in tandem or in
cluster can have any serious effect on the level of expression), otherwise
you might need to do quite a few mutagenesis reaction to eliminate all the
rare codons.  Other people approach this by synthesizing the whole gene,
but this is only practicable if the gene is relatively small and there are
far too many of these rare codons.  

However there is one serious error in your post, AGG (arg) cannot be
change to AGC because AGC codes for serine.  You'll also find that
different codon usage table give different result, depending on how the
sampling is done.  I got one which give CGT as the commonest.

: Let me know if you need more information about about site-specific
: mutagenesis protocols.

: -Bryan

More information about the Methods mailing list