Search

We have investigated the genetic diversity and potential mosaic genomes of HIV-1 during the early part of the HIV-1 epidemic among commercial sex workers (CSWs) in Kinshasa, Democratic Republic of Congo (formerly Zaire). Serologic analysis revealed that 27 (28.7%) of the 94 specimens were seropositive by both peptide and whole-virus lysate EIAs and that 24 were positive by molecular screening assays, using generic primers that can detect all known groups of HIV-1. Phylogenetic analyses of the gag(p24), C2V3, and gp41 regions of these 24 specimens showed that all were group M; none of them had any evidence of group O, N, or SIVcpz-like sequences. On the basis of env sequence analysis, the 24 group M specimens were classified as subtypes G (37.5%), A (21%), F1 (12.5%), CRF01_AE (8%), D (4%), and H (4%); 3 (12.5%) were unclassifiable (U). Similar analysis of the gag(p24) region revealed that the majority of infections were subtype A; however, one-third of the specimens were subtype G. Parallel analysis of gag(p24) and env regions revealed discordant subtypes in many specimens that may reflect possible dual and/or recombinant viruses. These data suggest a predominance of subtype G (both pure G and recombinant CRF02_AG) during the early part of the epidemic in Kinshasa. Infections with group N or SIVcpz-like viruses were not present among these CSWs in Kinshasa.

Up to now, all known env subtype E viruses (CRF01-AE) have had the same mosaic structure with subtype A, and no other env subtype E HIV-1 viruses with non-A subtypes in their genomes have been described. In this report we describe the full-length genome sequence of an env subtype E isolate with a recombinant genome different from the prototype CRF01-AE strains. The 97CD-KTB49 strain, obtained from a tuberculosis patient in Kinshasa, has a complex mosaic genome involving subtypes A, E, G, H, J, K, and several unknown fragments. The U sequences formed well-separated clusters together with previously described unknown fragments from CRF04-cpx (subtype I), and from Z321, the oldest intersubtype recombinant isolated in 1976 in the Democratic Republic of Congo. The complex recombinant virus from our study is not an isolated strain; partial sequencing of a second strain, 97CD-KFE45, confirmed the breakpoints observed in the 97CD-KTB49 strain in the regions sequenced. The complexity of these recombinant strains suggests a longstanding presence of subtype E in Central Africa.

We recently reported a high divergence among African subtype F strains. Three well-separated groups (F1, F2, and F3) have been shown based on the phylogenetic analysis of the p24 gag and envelope sequences with genetic distances similar to those observed for known subtypes. In this study, we characterized the near-full-length genomes of two strains from epidemiological unlinked individual belonging to each of the subgroups: F1 (96FR-MP411), F2 (95CM-MP255 and 95CM-MP257), and F3 (96CM-MP535 and 97ZR-EQTB11). Phylogenetic analysis of the near-full-length sequences and for each of the genes separately showed the same three groups, supported by high bootstrap values. Diversity plotting, BLAST subtyping, and bootstrap plotting confirmed that the divergent F strains correspond to nonrecombinant viruses. The divergence between F1 and F2 is consistently lower than that seen in any other intersubtype comparison, with the exception of subtypes B and D. Based on all the different analyses, we propose to divide subtype F into two subclades, with F1 gathering the known subtype F strains from Brazil and Finland, and our African strain (96FR-MP411), and F2 containing the 95CM-MP255 and 95CM-MP257 strains from Cameroon. The F3 strains, 97ZR-EQTB11 from the Democratic Republic of Congo and 96CM-MP535 from Cameroon, meet the criteria of a new subtype designated as K. The equidistance of subtype K to the other subtypes of HIV-1 suggests that this subtype existed as long as the others, the lower distance between B and D, and between F1 and F2 suggest a more recent subdivision for these latter strains.