Role of Viral Reservoirs in the Pathogenesis of HIV Disease

National Institute Of Allergy And Infectious Diseases

Investigators

Linked publications & trials

Abstract

In chronic infections, such as HIV, the persistent exposure to antigen leads to T cell exhaustion and inefficient control of the pathogen. Exhausted CD8+ T cells are characterized by their loss of effector functions, persistent and elevated expression of inhibitory receptors, and abnormal metabolic states. In order to reverse the state of immunologic abnormality and to reinvigorate exhausted T cells, therapeutic strategies involving the blockade of inhibitory receptors have been under active investigation. The recent success of immune checkpoint inhibitor therapies in cancer treatments has led to consideration of similar pathways in HIV infection. Currently, PD-1 and CTLA-4 blockade are the most well-characterized, but T cells in the setting of persistent viral infection may also express other immune exhaustion markers, such as Tim-3, LAG-3, and TIGIT. TIGIT is a member of the poliovirus receptor nectin family that binds to CD155 and is expressed on multiple cell types including CD4+ and CD8+ T cells and natural killer cells. While co-inhibition of both PD-1 and TIGIT may be necessary to fully reverse immune exhaustion, it has been shown that the blockade of TIGIT could lead to the clearance of tumors and the improvement of T-cell functions in animal models, suggesting a role for TIGIT in the restoration of cytolytic CD8+ T cells. However, the mechanism by which TIGIT modulates immune exhaustion and its impact on intrinsic CTL activity in HIV infection remains unclear. To directly assess the intrinsic CTL capacity of cells expressing TIGIT and/or PD-1, enriched CD8+ T cells of HIV-infected and uninfected individuals were cocultured with a chronically infected cell line CEM-IIIB cells with the virus-specific HIVxCD3 dual affinity retargeting protein (DART) molecule or RSVxCD3 DART as a control molecule. Frequencies of TIGIT, but not PD-1 nor TIGIT/PD-1, positively correlated with the degree of DART-mediated CTL activity in HIV-aviremic and healthy individuals but not in HIV-viremic individuals. Of note, the vast majority of HIV Gag-specific CD8+ T cells were found in the TIGIT+ population, further corroborating the above observation. To further delineate the relationship between TIGIT and CTL activity, we conducted transcriptome analyses of enriched TIGIT+ versus TIGIT-CD8+ T cells from HIV-aviremic, HIV-viremic, and healthy individuals. The principal component analysis revealed distinct transcriptional profiles of TIGIT+ versus TIGIT- cells within each study group, and a greater relatedness of such differentially expressed genes between HIV aviremic and healthy individuals. Gene ontology analysis of the 180 differentially expressed genes between the TIGIT+ and TIGIT-CD8+ T cells common for the HIV-aviremic and healthy individuals revealed the enrichment of genes in categories involving defense response, innate immune response, cell activation, positive regulation of cytokine production, and immune effector process. Collectively, our data suggest that TIGIT+CD8+ T cells do not necessarily represent a state of immune exhaustion and maintain an intrinsic cytotoxicity in HIV-infected individuals. It has been shown that a small percentage of HIV-infected individuals in whom ART was initiated during the acute/early phase of infection can control plasma viremia for extended periods following analytical treatment interruption (ATI). A better understanding of the underlying mechanism(s) of virologic control in these individuals could provide insight into the development and application of novel therapies. To this end, frequent and long-term comprehensive analyses of a wide array of immunologic and virologic parameters, as well as routine screening for surreptitious taking of antiretroviral drugs and HIV superinfection, could reveal factors that maintain virologic remission in these individuals and those that lead to plasma viral rebound. We conducted a long-term study of two HIV-infected individuals who had previously participated in the placebo arm of a randomized, controlled therapeutic vaccine trial that included ATI. Two study participants (04 and 30) initiated ART during the acute phase of their infections and had been receiving clinically effective ART for 6.7 and 6.5 years, respectively, prior to initiating ATI. Following ATI, Participant 04 experienced a first substantial plasma viral rebound on day 56 (26,967 copies of HIV RNA/ml), followed by sustained spontaneous virologic control that lasted over 500 days. Thereafter, his plasma viremia rebounded briefly to 778 and 1,784 copies/ml on days 581 and 875, respectively. To gain insight into the genomic nature of his sequential plasma viral rebounds, we conducted phylogenetic analyses using sequences derived from single genome amplification (SGA) of the 3-half of HIV RNA in the plasma. The sequences of rebounding plasma HIV on day 56 were relatively homogeneous. However, the sequences derived from the second plasma viral rebound on day 581 were largely distinct from those on day 56, suggesting continuous viral evolution and/or reactivation of other pre-existing viral reservoirs. There were no antiretroviral drugs detected in Participant 04 prior to day 1,250, including the time points immediately following the three major episodes of plasma viral rebound. However, the drug testing revealed that Participant 04 had initiated undisclosed, suboptimal ART at approximately day 1,250. In contrast, there were no antiretroviral drugs detected in Participant 30 for the entire follow-up period. Despite near complete virologic suppression throughout the entire follow-up period and without any recognizable precipitating event at the time, Participant 30 experienced a significant level of plasma viral rebound on day 1,434. Phylogenetic showed that the HIV env sequences of rebounding plasma HIV on day 1,434 did not resemble those from previous time points and were as divergent as different HIV subtype B strains, strongly suggesting that HIV superinfection had occurred in Participant 30. To investigate the potential role of HIV-specific CD8+ T cells in suppressing HIV in the absence of ART, we performed longitudinal intracellular cytokine staining assays using a pool of overlapping peptides that spanned the entire HIV Gag protein. Frequencies of polyfunctional HIV Gag-specific CD8+ T cells were consistently higher in Participant 04 compared to those of Participant 30. Finally, we investigated whether neutralizing antibodies in the plasma of the study participants potentially contributed to virologic suppression in the absence of ART. For Participant 04, heat-inactivated plasma exhibited low plasma 50% neutralization titer (ID50) (<1:30 dilution), consistently over all time points examined with or without IgG or IgM depletion. In striking contrast, day 117 plasma of Participant 30 showed strong IgG-mediated neutralization activity (ID50 of 1:1,607 dilution) against autologous contemporary infectious isolates, suggesting that one of the potential mechanisms by which Participant 30 achieved near complete virologic suppression in vivo, prior to superinfection, may have involved neutralizing antibodies against HIV. Collectively, our data provide insight into distinct mechanisms of post-treatment interruption control and highlight the importance of frequent monitoring of undisclosed use of ART and superinfection during the ATI phase.