Background:  The mutagenic capability of HIV-1 due to high replication rate, lack of proof-reading activity of RT, and ability to switch RNA template during transcription, has led to the development of 12 different subtypes and sub-subtypes, 17 circulating recombinant forms (CRF), and a growing number of unique recombinant forms (URF). Recombination seems to be a random process, but some hot-spots have been identified in the HIV-1.

Methods:  We evaluated 32 URF sequences, encompassing 850 bases starting from position 1 of protease gene, obtained from 6 different epidemiological studies and recorded into our own database. Sequences were obtained with home-made procedures and were analyzed with Simplot software v.3.5.1 using a set of reference strains comprising all published subtypes, sub-subtypes, and CRF, and 20 URF sequences selected from the GenBank DataBase (DB). To assess distribution significance we used a Poisson’s heterogeneity test.

Results:  We narrowed down the sequences into portions corresponding to coding protein domains, and evaluated the number of recombination points in each region. After normalizing according to their different lengths, we observed that almost no recombination points were found in the protease N-terminal region (amino acids 1 to 61) (p <0.0001), as most of them were scattered in all other regions. In particular, recombination points seem to cluster in 2 regions, corresponding to the protease C-terminal region (amino acids 62 to 99) and to a region coding for part of the RT finger domain (amono acids 118 to 152). Similar results were obtained when analyzing the URF from GenBank DB, which partly supports our findings.

Conclusions:  Since all samples were obtained from patients with an active infection, our data indicate that recombination involving at least the first 150 amino acids of RT seems not to interfere with viral replication capacity. No assumptions can be drawn regarding viral fitness or infectiveness because the URF studied could have been either generated in the individual host or acquired through infection. Break-points randomly involving the first ~61 amino acids of protease apparently produce viruses unable to replicate, not found in the viral species of the host. It is noteworthy that this region encompasses the entire PR flap region and the active site. Finally, the numerous break-points found in the relatively narrow RT finger region could indicate the presence of a recombination hot-spot caused by underlying RNA sequence or structural features.