Background:  Knowledge of the factors influencing vertical transmission are of particular importance in pediatric AIDS, as the majority of HIV infected children acquire the virus from their mothers. Breast-feeding has been identified as an important route for vertical transmission, predominantly in regions where interventions are not widely available. To understand better the mechanisms of transmission through breast milk, we have performed a cross-sectional analysis of paired breast milk and blood plasma samples to assess the extent of compartmentalization of HIV populations.

Methods:  Paired breast milk and blood plasma samples were provided with informed consent by women attending a postnatal clinic in Lilongwe, Malawi. Virus loads were determined using the Roche Amplicor HIV-1 Monitor (blood plasma) or the NucliSens HIV-1 QT assay (breast milk). Viral RNA (vRNA) was isolated from the blood plasma and the skim milk fraction of breast milk, and the V1/V2 env region was amplified by RT-PCR. Amplicons were analyzed both by heteroduplex tracking assay and by cloning and subsequent nucleotide sequence analysis. Heteroduplex tracking assay populations were quantitated by phosphorimaging, with the percent difference of populations in matched samples compared using the Wilcoxon Rank Sum test.

Results:  Adequate sampling of the virus pool is necessary for sufficient sensitivity to assess whether these samples contain similar virus populations. Comparison of heteroduplex tracking assay patterns from 2 separate RT-PCR reactions of each sample demonstrated that those samples with <1000 vRNA copies/mL often cannot be accurately measured for population content. By this method we identified 8 sample pairs for further analysis. We have examined theV1/V2 env region because its high degree of variability will identify a maximum number of variants, increasing our sensitivity for population differences between compartments. Heteroduplex tracking assay patterns did not demonstrate qualitative or quantitative differences in the presence or abundance of genotypic variants between paired samples. Two V1/V2 env nucleotide sequence data sets were then assessed. These analyses demonstrated that co-migrating heteroduplex tracking assay variants from different compartments were closely related.

Conclusions:  Neither heteroduplex tracking assay nor sequence analysis of the V1/V2 env region of HIV from paired breast milk and blood plasma samples demonstrated significant differences in the virus populations in these compartments. We conclude that compartmentalization of cell-free HIV between the breast milk and blood plasma does not occur at a significant level.

Keywords: compartmentalization; breast milk; transmission