The retrotransposon Osvaldo from Drosophila buzzatii displays all structural features of a functional retrovirus

The Osvaldo retrotransposon has shown a high transposition rate in some strains of Drosophila buzzaiti and in hybrids between D. buzzatii and its sibling D. koepferae. In order to understand the molecular basis of this phenomenon, we developed a procedure to clone a recently transposed copy with the aim of characterizing an active, full-length Osvaldo element. The complete nucleotide sequence of Osvaldo, obtained from a recent insertion site, was determined. Osvaldo is 9,045 bp long and is composed of a central coding region flanked by identical long terminal repeats (LTRs) of 1,196 bp each. Sequences homologous to the polypurine tract and tRNA-primer-binding site of retroviruses are located adjacent to the 3' and 5' LTRs, respectively. The internal region of Osvaldo contains three long open reading frames (ORFs 1, 2, and 3), comparable in size and location to gag, pal, and env retroviral genes. The conceptual translation of Osvaldo ORF1 exhibits sequence homology to HIV1 and SIV capsid (p24) and nucleocapsid (p7) mature proteins. ORF2 encodes the putative protease (PR), reverse transcriptase/ribonuclease H (RT/RH), integrase (IN), and a significant portion of the surface envelope (ENV) protein that is interrupted by a putative intron. A third ORF encodes the remaining part of the ENV protein. The predicted 62-kDa ENV protein shares several general features with membrane glycoproteins, including a potential signal peptide, a transmembrane domain near the C-terminus that could function as a membrane anchor, four consensus N-linked glycosylation motifs, and, finally, a potential protease cleavage site. The phylogenetic relationships of Osvaldo are explored, and they suggest that Osvaldo may constitute a new family of retroviruses in insects, distantly related to the previously described group of gypsy retroviruses.