Background:  The uterine cervix has high concentrations of T lymphocytes, and may serve as a compartment for evolution of unique HIV-1 variants. Understanding the biology of HIV-1 in the female genital tract, because of its role in heterosexual transmission, has important implications for vaccine development. In previous studies, HIV-1 DNA sequences from cervical biopsies have had low genetic diversity, suggesting that a single punch biopsy may sample only related viruses that spread between adjacent cells. To examine HIV-1 diversity within the cervix and whether viruses clusters in cervical tissue, we compared HIV-1 sequences from multiple locations within the cervix to one another and to sequences from the peripheral blood mononuclear cells (PBMC).

Methods:  Multiple single genomes of HIV-1 DNA from PBMC and 3 cervical biopsies per woman were analyzed in a cross-sectional study. DNA was amplified by polymerase chain reaction (PCR) from each subject’s specimens on separate days to minimize the chance of cross-contamination between specimens. As many as 40 single genome sequences of the C2-V5 gp120 region of env were derived per specimen and subjected to phylogenetic and distance analyses of compartmentalization. The distribution of sequence diversity between cervical biopsies from each woman was compared using a permutation analysis.

Results:  Viruses from 6 of 8 women (median of 65 sequences/woman) clustered into unique clades with significant compartmentalization between the blood and cervical sequences (p = 0.0335 to p <0.0001). Closely related HIV-1 sequences were found within and between all 3 biopsies from each woman. The frequency of identical sequences within a biopsy affected whether compartmentalization was observed. Of note, in 3 of the 8 women, a single biopsy had significantly (p <0.05) less diversity than the other biopsies. Each woman’s phylogenetic tree also had clades with both PBMC and cervical sequences.

Conclusions:  Compartmentalization between cervical and peripheral blood HIV-1, and between individual biopsies, appears to result from bursts of viral replication or clonal expansion of infected cells within each woman. A single biopsy may result in sampling viruses with low genetic diversity that skews the analysis toward compartmentalization. Sampling a larger number of sequences from the blood and cervix consistently demonstrated trafficking of HIV-1 between these tissues, suggesting that the female genital tract does not constitute an isolated compartment of viruses.