Posted 09 February 2010 - 06:57 PM
A colleague of mine had a similar problem with a not-too-commonly studied bacterium. There were 3 or 4 stop codons (for E. coli) in the ORF that actually encode an amino acid in the particular organism rather than causing a stop. They didn't realize that cloning it into E. coli would result in a truncation at first. Once they did, they ended up having the gene made synthetically with each of the stop codons modified at the wobble position so that the correct amino acid is added and the stop is removed. Thus far, the synthetic hasn't expressed, but I have a feeling that's more of user error than a sequence problem.
Of course, having an entire gene made synthetically can be expensive. But your other option is to make a point mutation at each stop codon to change the wobble position so that you can clone into E. coli and get the correct amino acid.