gene hunting advice - (Jan/09/2008 )

Hello,

I would like some advice on what would be the best approach to use to find the sequence of a gene on a new species. This gene has been found and sequenced in a very distantly related species (the honeybee), but so far in no other species. I have used the sequence from the honeybee to design primers, but had no luck with my target species. The gene in question is highly variable even withing hoenybees, so its very likely that it is very divergent in my target species. What approach would be the best one to id and sequence this gene? One idea is southern blot, but in case this didnt work, what other approach could I take?

Any suggestions on what approaches to take, or literature I should check, would be greatly appreciated.

Thanks,

hans

Trying using a BLAST search at the NCBI website.

I would also try first by PCR to get at least a small fragment and use it as probe or extend it by RACE. I don't favor the use of heterologous probes and low stringency screenings, for I had many troubles with that strategy. First align as much as protein sequences of the gene as you can find, and look for conserved amino acids. If there's only one sequence, you could try to guess if there are residues that are likely to be conserved (for the function, signal peptides, etc). These are the places where for the primers.
Then, you have to guess a nucleotidic sequence from the proteic one. You could find a codon usage for your organism, or use one from a related organism. While remaining in the conserved (hopefully) regions, try to design the primers in order to have their 3' half on amino acids encoded by a limited number of triplets. Best are M (only AUG) and W (only UGG), so put them at the 3' end if you can. Otherwise, put at the 3' end a partial codon that can ancode many aa.
Use primer up to 1000 fold degenerated or more and use inosine for the aminoacids that can be encoded by many triplets. A good idea is also to use the primers at higher conc than usual.
I had found some information on the internet about this, but I don't have to bookmarks any more. You look yourself on google searching for degenerate PCR or something similar.