Background
HIV-1 latently infected cells that persist in tissues during antiretroviral therapy (ART) are a barrier to HIV-1 cure, but these reservoirs are hard to reach and therefore poorly characterized. In this study we performed a comparative cross-sectional analysis of HIV-1 reservoir characteristics in paired blood and excised lymph node (LN) tissue.
Methods
We studied 12 individuals from South Africa identified with acute infection by virtue of being HIV-RNA positive and antibody negative with subsequent seroconversion. At the time of sampling 5 participants had initiated ART during chronic infection (CT) (median days after diagnosis =424, IQR 141-719; median days on ART =899, IQR 405-1567), 5 during acute infection (AT) (median days after diagnosis =1, IQR 0.5-1; median days on ART =428, IQR 64-843) and 2 remained untreated (UT). One AT was not virologically suppressed at the time of sampling (days after treatment initiation =8). We measured total proviral DNA by ddPCR (n=12) and sequenced viral genomes by full length individual proviral sequencing (FLIP-seq) (n=8) in paired PBMC and excised LN samples.
Results
In CT (n=5) and AT (n=5), LNs had 1.5 and 1.1-fold higher median total HIV DNA load than PBMCs (p=0.04). By FLIP-seq we obtained 65 single genome sequences from PBMCs and 159 from LN samples. Around 50% of participants well-suppressed on ART for more than 2 years had detectable intact HIV-1 DNA in LN but not in PBMC. Pooled analysis revealed that relative to PBMC, LN had higher proportions of intact viral genomes in both CT and UT individuals (CT: 5% vs 0% intact, sampling depth 66 and 20 genomes from 258 888 and 231 501 cells; UT: 54% vs 29% intact, sampling depth 22 and 21 genomes from 11 585 and 12 686 cells, in LN and PBMC respectively). In contrast, in AT, proportions of intact HIV-DNA were similar between compartments: LN and PBMC was 20% vs 25% intact respectively (sampling depth 64 and 24 genomes from 630 293 and 214 561 cells).
Conclusions
Genetically intact proviruses are found at higher frequencies in lymphoid tissues compared to PBMC in both chronic-treated and untreated infections but were similar in acute treated infections. These data suggest that depending on treatment status/stage, HIV DNA genome profiles in peripheral blood may not reflect those in tissue and that viral compartmentalization persists during suppressive ART. Design of HIV curative strategies should target all relevant organ systems.
Table or Figure
