Pharmacological Cellular Sanctuaries in HIV Patients Resistant to the Antiviral Effects of Protease Inhibitors

A. Valentin,1 R.H. Poirier,1  M. Morrow,1  K. Aleman,2 R. Little,2  R. Yarchoan2 and G.N. Pavlakis1. 1Human Retrovirus Section, BRL, CCR, National Cancer Institute, Frederick, MD 21702 and 2HIV and AIDS Malignancy Branch, CCR, National Cancer Institute, Bethesda, MD 20892 (USA).

Background: HIV persistence in patients receiving highly active antiretroviral therapy (HAART) is the result of both viral latency in long-lived cells, and continuous low levels of viral propagation in cells where the pharmacological effects of the antiretroviral drugs are suboptimal. We have shown that, in addition to T cells, Natural Killer (NK) cells are persistently infected by HIV (Valentin, PNAS 99:7015, 2002). NK cells have a highly active P-glycoprotein (P-gp) pump, a member of the ABC family of transporters that regulates the intracellular concentration of protease inhibitors. Material and Methods: GFP-tagged HIV-1 molecular clones were used to study P-gp activity in infected primary cells. Rh123 and TMRE, two fluorescent P-gp pump substrates, were used to study the P-gp activity in different subsets of primary cells from HIV-1 infected individuals (n=63) and healthy controls (n=24) by flow cytometry. CD4+ cells with high and low P-gp activity were sorted, and their susceptibility to the antiviral effect of Ritonavir was studied after infection with HIV. Results: We identified a population of CD4+CCR5+ primary cells with very high P-glycoprotein efflux activity. These cells are phenotypically heterogeneous and were recognized as T lymphocytes and Natural Killer cells. The presence of these cells was found in both healthy HIV-1 seronegative controls and HIV-1 infected individuals. Functional studies demonstrated that higher concentration of protease inhibitors is required to fully inhibit P-gp activity in these cells. Monitoring GFP-tagged infected cells demonstrated that very high P-gp activity is present in cells actively replicating HIV-1. Infection experiments with sorted cells showed that 50 nM of Ritonavir completely inhibited HIV-1 replication in primary CD4(+) P-gp(-) cells. In contrast, under identical conditions, active HIV-1 replication was found in primary CD4+ cells with high P-gp activity. Conclusions: Cellular reservoirs and sanctuaries for HIV include cells other that T lymphocytes. Their identification and study is critical for improving therapies. NK cells remain infected during therapy and are an important part of the reservoir. Our results suggest the presence of pharmacological cellular sanctuaries resistant to the antiviral effects of protease inhibitors. The presence of these cells could be responsible for HIV-1 persistence during HAART. New therapeutic interventions to eliminate these cells are required to successfully treat HIV-1 infected patients.