Coming from a TB background, one of the outstanding concerns in the field is whether antibacterial drugs can access and thus clear bacteria residing within granulomas. It is thought that low drug diffusion might partly explain why anti-TB therapy takes months rather than days to work.
In a paper published last week in Nature, Lorenzo-Redondo et al. find that this question of drug access may also apply to HIV infection. Just like the granuloma, they find that HIV can replicate in certain reservoirs that do not take up antiretroviral drugs well. But in this case, it's not granulomas but rather lymphoid tissue.
Using deep sequencing, the authors assessed the evolution of HIV genomes from human blood and lymph node samples taken before and during antiretroviral therapy. Though viral load was undetectable in the plasma during treatment, virus isolated from lymph nodes showed a continued accrual of stochastic mutations, suggesting the virus underwent uninterrupted replication in these sites. These lymphoid lineages could also be found in the blood, implying that low level lymphoid viral replication could serve as a reservoir for reseeding the body.
Interestingly, the new viruses emerging from the lymphoid compartment were not drug resistant, which suggests that there is not enough drug pressure in these sites to select for resistant genotypes. Thus, the authors constructed a computational and spatial model, in which they propose that some tissues with low drug penetrance can enable HIV to replicate uncontrolled and establish the viral reservoir despite the vast majority of niches being effectively silenced through antiretroviral therapy. Together, the work reframes the thinking behind what constitutes HIV's persistent viral reservoir and has important implications for future drug development to find improved lymphoid accessible compounds.