Background:  Rapid and error-prone replication of HIV-1 yields a genetically diverse virus population from which drug resistant mutants are selected. The emergence of RT and PI mutations during combination antiretroviral therapy remains unpredictable, but may be related to baseline sequence variation. To characterize the diversity of pre-therapy HIV-1 populations and to determine whether pro and pol vary independently over time, we compared sequence variation and phylogenetic relationships of pro and pol in chronically infected drug-naïve individuals.

Methods:  Individual HIV-1 pro-pol sequences (PR and nt 1-1200 of RT) were obtained by single genome sequencing. A total of 950 single genome sequences from 14 drug naïve individuals infected with HIV-1 for at least 1.5 years were obtained from samples spanning a study period of 1 to 14 years. Sequences were aligned using Clustal W and subjected to phylogenetic analysis; recombination was evaluated using the 4 gamete model of Hudson.

Results:  Sequence analysis revealed comparable levels of genetic diversity of pro and pol within individuals. However, a position-specific analysis revealed that nucleotide variation was not uniformly distributed in HIV-1 pro or pol. In pro, synonymous and nonsynonymous changes were concentrated in 4 regions (nt 34-79, 101-130, 175-240, 260-290); nonsynonymous changes largely localized to solvent-exposed regions. Pol had few clusters of nonsynonymous polymorphisms in fingers and palm domains, but highly polymorphic synonymous changes were present throughout. Longitudinal analyses (n = 9 patients) did not identify consistent changes in pro or pol diversity over time. Phylogenetic analyses revealed relatively homogenous HIV-1 pro and pol sequences within individuals; bootstrap analysis did detect distinct pro and pol phylogenetic relationships (bootstrap values > 85%) in 6 of 9 individuals, suggesting independent sequence evolution of the two genes. Recombination analyses revealed evidence of frequent recombination with minimum recombination intervals as short as 7-17 nt.

Conclusions:  In contrast to HIV-1 samples taken from patients early in infection, sequence diversity in pro and pol is comparable and stable over time in chronically infected individuals. Position-specific analysis reveals that pro diversity is localized to specific regions. The presence of frequent recombination provides a potent mechanism for independent pro and pol variation.