Search

There are multiple subtypes of HIV-1 circulating worldwide, but recently, subtype C has become highly prevalent, particularly in certain geographic regions. It is unclear whether the dominance of subtype C or other subtypes is due to increased fitness of certain subtypes for transmission, or a founder effect in new, rapidly growing epidemics. To examine whether the prevalence of one subtype increases over the course of an expanding epidemic that includes several circulating subtypes, we examined the distribution of HIV-1 subtypes in Kenya from 1986 to 2000. We found no evidence for an increase in the prevalence of subtype C, which remained low throughout this approximately 15-year period. Interestingly, the percentage of subtype D present in the population decreased significantly over that period, with a slight increase in subtype A. Throughout that period, intersubtype recombinant viruses were detected, including at the early stages of the epidemic. This latter finding suggests that reinfection may have occurred in high-risk groups early in the epidemic, leading to intersubtype recombinant viruses that underwent secondary spread.

Diversity among global isolates of HIV-1 presents a formidable challenge for vaccine development. As distinct clades of the virus are recognized, it will be important to monitor their geographic distribution and divergence. In this study, we characterized HIV-1 subtypes from 17 seropositive individuals in Nairobi and Mombasa, Kenya. Seventy-one percent of viruses were clade A and 29% were clade D. The most divergent clade A isolate in our survey, Q45-CxA, grouped closely with two other taxa that were previously reported as having no distinct clade affiliation. Thus, these data may suggest the emergence of an outlier group of clade A variants or a new subtype of HIV-1. Phylogenetic relatedness of the 17 Kenyan isolates was determined separately for C2-V3 and V2 sequences of envelope and subtype designation for these isolates was independent of the region analyzed. However, evaluation of transitions, transversions, and specific character state changes indicated that mutations characterizing V2 differed from those in V3 for clade A and clade D isolates. Comparison of secondary structural characteristics of the V1-V3 region between a clade A and a clade D virus revealed conservation of motifs.

For a HIV vaccine to be effective, it will be essential that it protect against the virus variants to which individuals are most frequently exposed. HIV-1 is predominantly a sexually acquired virus, thus, variants in genital secretions are a potentially important reservoir of viruses that are transmitted. Because there are no data available on variants in the genital mucosa, we analyzed this provirus population and compared it to the proviruses in the blood of individuals chronically infected with HIV-1. A major genetic difference between variants within a patient were insertions, which were apparently created by duplication of adjacent sequences, that resulted in acquisition of new potential glycosylation sites in V1 and V2. Comparisons of mucosal and PBMC variants suggest that these tissues harbor distinct, but related populations of HIV-1 variants. In two of three patients, the mucosal variants were most closely related to a minor variant genotype in blood. In a third individual, viruses in both tissues were surprisingly homogeneous, but the majority of variants in the cervix encoded a V1 sequence with a predicted glycosylation pattern similar to a minor variant in blood. The V3 sequence patterns of the mucosal isolates indicate they may be predominantly macrophage-tropic viruses.