Since 1989, human immunodeficiency virus type 1 (HIV-1) has spread explosively through the heterosexual population in Thailand. This epidemic is caused primarily by viruses classified as "subtype E", which, on the basis of limited sequence comparisons, appear to represent hybrids of subtypes A (gag) and E (env). However, the true evolutionary origins of "subtype E" viruses are still obscure since no complete genomes have been analyzed, and only one full-length subtype A sequence has been available for phylogenetic comparison. In this study, we determined full-length proviral sequences for "subtype E" viruses from Thailand (93TH253) and the Central African Republic (90CR402) and for a subtype A virus from Uganda (92UG037). We also sequenced the long terminal repeat (LTR) regions from 16 virus strains representing clades A, C, E, F, and G. Detailed phylogenetic analyses of these sequences indicated that "subtype E" viruses do indeed represent A/E recombinants with multiple points of crossover along their genomes. The extracellular portion of env, parts of vif and vpr, as well as most of the LTR are of subtype E origin, whereas the remainder of the genome is of subtype A origin. The possibility that the discordant phylogenetic positions of "subtype E" viruses in gag- and env-derived trees are the result of unusual rates or patterns of evolution was also considered but was ruled out on the basis of two lines of evidence: (i) phylogenetic trees constructed for synonymous and nonsynonymous substitutions yielded the same discordant branching orders for "subtype E" gag and env gene sequences, thus excluding selection-driven evolution, and (ii) multiple crossovers in the viral genome are most consistent with the copy choice model of recombination and have been observed in other documented examples of HIV-1 intersubtype recombination. Thai and CAR "subtype E" viruses exhibited the same pattern of A/E mosaicism, indicating that the recombination event occurred in Africa prior to the spread of virus to Asia. Finally, all "subtype E" viruses were found to contain a distinctive two-nucleotide bulge in their transactivation response (TAR) elements. This feature was present only in viruses which also contained a subtype A 5' pol region (i.e., subtype A viruses or A/D and A/E recombinants), raising the possibility of a functional linkage between the TAR region and the polymerase. The implications of epidemic spread of a recombinant HIV-1 strain to viral natural history and vaccine development are discussed.
PIP: The subtype E of HIV-1 is primarily responsible for the heterosexual HIV-1 epidemic in Thailand. Based on limited sequence comparisons, subtype E viruses seem to be hybrids of subtypes A (gag) and E (env). No complete genomes of subtype E HIV-1 have been analyzed, and there is only 1 full-length subtype A sequence for phylogenetic comparison. Thus, virologists have performed full-length proviral sequences for subtype E viruses from Thailand (93TH253) and from the Central African Republic (CAR) (90CR402) and for a subtype A virus from Uganda (92UG037). They also sequenced the long terminal repeat (LTR) regions from 16 virus strains (clades A, C, E, F, and G). The detailed phylogenetic analyses found that subtype E HIV-1 viruses are subtype A/E recombinants with many crossover points along their genomes. The parts of the genome of subtype E origin include the extracellular portion of env, parts of vif and vpr, and most of the LTR. The remaining parts of the genome are of subtype A origin. The Thai and CPR subtype E viruses had the same pattern of A/E mosaicism, suggesting that recombination took place in Africa before the Thai subtype E HIV-1 (93TH253) spread to Asia. All subtype E viruses that also had a subtype A 5' pol region (subtype A viruses or A/D and A/E recombinants) had a unique 2-nucleotide bulge in their transaction response (TAR) elements. This suggests a possible functional linkage between the TAR region and the polymerase. Intersubtype recombination appears to be a relatively recent phenomenon. The spread of the HIV-1 epidemic and the mixing of clades have created the opportunity for coinfection and recombination. The widespread dissemination and virulence of subtype E viruses reveal that intersubtype recombination can generate potent pathogens. It is very important to study the effects of viral recombination on virus-host interaction in terms of immune protection from natural infection and vaccines.