This session summaries new findings on HIV persistence of HIV reservoirs. Steven Yukl (USA) presented very interesting results on CD4+ T cell reconstitution, activation and memory subsets composition in different compartments of the gut and blood during HIV infection. CD4 depletion is incomplete in ileum of HIVinfected patients under ART but occurs in the rectum. There were no significant difference between ileum, rectum and blood regarding T cell activation which remained high in chronically infected patients on antiretroviral therapy. Compared to the blood, the ileum has a lower % of T lymphocytes with a central memory (CM) phenotype, higher % of Effector Memory (EM) cells and a comparable % of Transitional memory (TM) cells. Preliminary data suggests that HIV DNA levels (reflecting the number of infected cells) in ileum TM cells could be much higher than in peripheral TM cells. The ileum and rectum differs in HIV levels and responses to intensification suggesting that mechanisms of persistence may vary between sites. in ART suppressed HIV-1+ individials, HIV persistence in rectum and ileum could be impacted by T-cell activation, absolute CD4 T cell level and composition of memory T-cells. The abnormal immune activation and lack of CD4 reconstitution in the ileum could reflect ongoing replication or chronic virus production.

The second speaker of the session John Zaunders (Australia) said that earlier studies have reported that most CD4 T cell depletion occurs in Gut Associated Lymphoid Tissue (GALT), inferring that most viral replication occurs in these tissues. Memory CD4 T lymphocytes in peripheral blood comprises of 2 main subsets , those with integrins a4b7+ that circulate through GALT and those with a4b1, that do not access GALT. a4b7+ regulates the access of leukocytes to the gut and their return to the blood stream via thoracic duct; b1 integrin define non-gut homing T cells (mutually exclusive). In his study (conducted on frozen samples from 8 individuals) he tested whether a4b7+ CD4+ T cells were preferentially carrying HIV DNA. His results demonstrated that majority of HIV reservoir in PBMC is present in non gut homing in memory CD4+ T cells with a resting CD127 high CD8-CD27+ phenotype. These cells recirculate through secondary lymphoid tissue but not through the GALT. In addition, he found that Tregs (characterized by low IL-7R, CD127, and high IL-2Ra, CD25); CD127high have much more HIV DNA than negative ones. CD38+ memory cells had much less HIV DNA than CD38 neg cells. Therefore, Zaunders concluded, the majority of HIV DNA+ CD4+ T cells were unlikely to have been activated and infected in GALT.

Alexander Pasternak (The Netherlands) discussed the importance of HIV-1 unspliced RNA (usRNA) levels as a marker of HIV-1 progression in ART-naïve individuals and after virological rebound in patients on ART with undetectable plasma viremia. Indeed, no reliable longitudinal virological markers are known for the progression of HIV infection in the asymptomatic phase and also no markers predictive of virological rebound in initial responders to combination anti-retroviral therapy (cART) during the period of undetectable plasma viremia. His approach was to use a semi-nested real time PCR increasing of 1.5 log the sensitivity of the assay without compromising its specificity. Patients with higher levels of usRNA in PBMC were more prone to failing cART, suggesting that HIV-1 usRNA is a strong predictive marker for the outcome of therapy. Moreover, during the asymptomatic phase of infection in the absence of ART, HIV-1 usRNA levels in PBMC steadily increases and is a sensitive and direct longitudinal virological marker of infection progression. The confounding variable of virions attached to cells (in ART-naïve individuals) was estimated by previous studies not to exceed 10% of the total usRNA.

Maria Jose Buzon (Spain) discussed the role of viral reservoirs on adaptive HIV-1 evolution throughout viral recombination under ART and as a consequence of immune pressures. She analyzed the rebound virus after partial treatment interruption. Interestingly, this rebound virus was found to be frequently a recombinant virus of ancestral viral sequences before treatment combined with sequences that have escape mutations on the current drug regime. The recombinant virus showed increased replicative capacity in vitro compared to the viruses under therapy and fitness similar to the ancestral virus. Furthermore, the patients were still neutralizing the ancestral virus implying that despite immune pressure the virus was able to recombine. This provocative study underscores the potential relevance of HIV-1 reservoirs to adaptive viral evolution throughout the recombination in response to selective pressures.

The last speaker, Alessandro Marcello (Italy) was not present.