A transposition of the reverse transcriptase gene reveals unexpected structural homology to E. coli DNA polymerase I

The rational design of antiviral agents targeting the reverse transcriptase (RT) of the human immunodeficiency virus (HIV) would greatly benefit from a more intimate knowledge of the structure of RT. Until now, the degree of sequence similarity between RT and E. coli DNA polymerase I (Pol I) has been thought to be confined to several small regions, suggesting little basis for homology molecular modeling. However, we have found that a region in the C terminal of the RT polymerase domain is homologous to a central region of Pol I that lies between the universal polymerase motifs A and C (specifically, helices N-O-P of the Pol I crystal structure); a single transposition closely aligns the RT and Pol I genes, revealing a similar domain structure with 20% residue identity, as well as the possible structural correlates of several RNA-dependent polymerase motifs. The RT from Myxococcus xanthus (a bacterium believed to have diverged from other species 2 billion years ago), if similarly transposed, shows homology to both HIV-1 and E. coli, suggesting the possibility of a very ancient divergence between the RT and Pol I polymerase genes. A second even more significant match to this E. coli region was found in the retroviral ribonuclease H (RNase H) domain, and corresponds precisely to a region that has been aligned by previous investigators with the E. coli RNase H, suggesting that Pol I helices O and P are homologous to helices A and D of the RNase H crystal structure, respectively. These results are consistent with a modular theory of molecular evolution.