Can bacteria transfer genes to eukaryotes? Many people may remember the rather rash assertion of evidence that they can that was made in the initial human genome paper (and the subsequent debunking of those claims by people I used to work with at TIGR).
But just because one study was flawed doesn't mean that such horizontal transfers don't happen, and today in the advance publication section of Science a new study shows solid evidence that the endosymbiotic bacterium Wolbachia has integrated parts (in some cases quite large portions) of its genome into that of numerous strains of Drosophila, the wasp Nasonia, and the worm Brugia malayi. These aren't mere cases of "BLASTology" -- they were confirmed by PCR. Even more stunningly, RT-PCR suggests that in some cases the integrated genes are actually expressed.
Again, there is a TIGR/JCVI connection -- one of the co-first authors, Julie Dunning Hotopp, has a cube only several feet from mine -- and all today I could hear her talking on the phone with science reporters -- so congratulations to her (and best wishes that the reporters don't screw things up). Also, congratulations to all the other authors, particularly the other co-first author, Michael Clark, whom I haven't met.
The results of the study have many implications both theoretical and practical. From an evolutionary perspective, it is interesting to ponder if the ancestors of many current eukaryotic genes came from such bacterial integrations. And from a practical perspective, it really makes one wonder if discarding bacterial sequences during the assembly of eukaryotic genomic data as "obvious contamination" (as is commonly done) is really the right thing to do.