Prevention of HIV-1 acquisition and determinants of disease progression.

Abstract

Introduction HIV-1 infection can be managed using multiple strategies, including preventative approaches and therapeutic approaches. Current preventative and treatment strategies are suboptimal and there is a need to develop an effective prophylactic or therapeutic vaccine and to improve the public health approaches against the virus. This requires more detailed understanding of the infection, from prevention to natural disease progression. We performed several studies that cover a range of infection attributes, from understanding the mechanism of action of pre-exposure prophylaxis (PrEP) and determining the effectiveness of different compounds in blocking initial infection, to gaining further insight into potential mechanisms of natural control of HIV-1 disease progression in viraemic controllers (VC) with (VC+) and without (VC-) protective class I human leukocyte antigen (HLA-I) alleles. In order to cover this range of infection attributes we investigated two hypotheses: (i) initial low dose infection can be cleared with suboptimal drug inhibition, which allows ongoing viral replication, as long as the drug mechanism acts before the first cell is infected; and (ii) individuals without protective HLA-I alleles have CD8+ T cell-independent mechanisms of control. Methods To understand the mechanism of action of PrEP, the probability of extinction of new infections in the presence of two drug mechanisms (interference of initial infection with tenofovir (TFV), or reduction of burst size with atazanavir (ATV)), or with no drug, was modelled as a function of initial infected cells and viral replication ratio. The fraction of extinguished infections was experimentally determined with low viral input in the presence of either drug, or with no drug, in an in vitro model of PrEP. To gain insight into potential mechanisms of control, we studied immune cells in 12 VC+ and 9 VC- and, compared these 21 controllers with 5 rapid progressors (RP). Measurements included the magnitude and breadth of CTL responses using the ELISpot assay, as well as flow cytometry-based characterization of NK cell and T cell populations, which included the measurement of surface markers for activation, maturation, and exhaustion on these populations. Further, NK cell function was measured by intracellular cytokine staining following stimulation of these cells. Results Our study showed that TFV dramatically increased clearance while ATV did not, both for our mathematical model and our experimental study.

We observed that both VC, in particular VC-, had a higher contribution of Gag CTL responses to the total CTL response than RP (p=0.04), however there was no significant difference in the magnitude and breadth of CTL responses between VC+ and VC-. In addition, VC- NK cells had higher levels of the activation markers HLA-DR (p=0.007) and co-expression of CD38 and HLA-DR (p=0.03) when compared to VC+ and uninfected individuals (UI), and lower cytokine expression (MIP-1β and TNF-α) than VC+ NK cells (p=0.05 and p=0.04, respectively). We found a negative correlation between the expression of MIP-1β and the co-expression of CD38 and HLA-DR (r =-0.45, p=0.05). Furthermore, VC- T cells had higher levels of CD38 and HLA-DR co-expression (p=0.05), and a trend of higher HLA-DR (p=0.07) as well as CD57 expression (p=0.09) when compared to VC+. Conclusions The ability of drugs to clear initial but not established infection depends only on the ability to target initial infection. This implies that in diseases which involve transmission of low pathogen numbers upon exposure, but have robust replication when established, such as HIV-1, a possibility to clear infection should exist even with relatively weak inhibition as long as the drug has the mechanism of targeting the initial infection. This finding is particularly relevant in scenarios of variable adherence that result in sub-optimal drug levels or possible future PrEP strategies with drugs that have long half-lives yet do not completely suppress viral replication. VC have a more Gag focused CTL response than RP, however this feature did not distinguish VC+ from VC-. NK and T cell profiles differ between VC+ and VC-. VC- have a more activated NK cell profile with lower cytokine expression, and a more active and terminally differentiated T cell profile than VC+. A possible explanation for our results is that the increased CD38+HLA-DR+ NK cells in VC- may represent NK cells acting as antigen presenting cells (APCs), which may then directly interact with a more activated and terminally differentiated population of T cells observed in VC-. Further work to test this hypothesis is necessary to better understand the mechanisms underlying control in these two groups of VC patients. It is also suggested that transcriptomic studies may contribute further to understanding the distinct NK and T cell profiles observed between VC+ and VC- and how these may result in differing mechanisms of control.