A molecular epidemiologic survey of HIV in Uganda

Abstract

Objective: Previous data, based on a small sampling of convenience, reported subtypes A, B, C, D, and G in Uganda, but neither the extent nor the proportion of these subtypes could be evaluated. To establish correctly the prevalence and distribution of HIV-1 subtypes, we analysed viral clades in 739 HIV-1-seropositive specimens from different areas of Uganda. Methods: Blood specimens from 1100 patients were collected in five districts of Uganda. Within this collection, 929 HIV-1-seroreactive samples underwent analysis of viral DNA, and 739 were selected for further subtyping in env or pol regions. Results: Using a combination of subtype A- and D-specific probes to C2–V3 region and DNA sequencing, HIV-1 env subtypes were determined in 594 specimens: 341 were of subtype A (57.4%), 250 of subtype D (42.1%), and three of subtype C (0.5%). Sixty-two samples showed reactivity with both probes, suggesting potential mixed infections, cross-reactivity to probes, or possibly other subtypes. Subsequent sequence analysis of 19 randomly selected specimens revealed subtypes A (n = 4), D (n = 12), and C (n = 3). Sequence analysis of the 27 samples chosen from the remaining 83 samples, which could be amplified only with viral gp41 or protease gene primers, classified them as subtypes A (n = 13) and D (n = 14). No significant clinical, demographic, or geographic differences were found between HIV-1 infections with viruses of subtypes A and D, despite considerable genetic diversity within these clades. Conclusions: This is the first major population-based study of the prevalent HIV-1 strains in an African country selected for vaccine trials. The subtyping methods we describe should be of use to investigators seeking to conduct large-scale screening for HIV variants in other populations.