Infection with human immunodeficiency virus-1 (HIV-1) results in CD4+ T cell depletion. Antiretroviral therapy (ART), although effective in controlling plasma viremia, does not purge the latent or persistent reservoirs necessary to eliminate the infection. Recently, studies have demonstrated control of viral replication and decreasing viral reservoirs in 50% of rhesus vaccinated with a CMV vaccine vector. They propose that the continuous immuno surveillance of SIV by Tem cells is maintained by the persistent CMV vectors. To mimic the immuno surveillance and increase the HIV-specific CTL activity in vivo, we will genetically modify CMV- specific T cells with a chimeric antigen receptor (CAR) and follow the effects on viral reservoirs in rhesus macaque. The CARs express the CD4 extracellular domain to redirect CTL activity against HIV and intracellular T cell signaling domains to stimulate CTL functions. These "designer T cells" (dTc) target a critical step in the viral life cycle independent of MHC presentation, targeting heterogeneous viruses while avoiding the potential for viral escape. Our preliminary data (and many cancer models) show that CARs with multiple co-stimulatory signals (2nd and/or 3rd generation) have stronger activity than original (1st generation) vectors. Clinical trials with these advanced dTc have been efficacious in reducing metastatic solid tumors and leukemia. Additionally, based our preliminary studies, we will also co-transduce dTc with the potent fusion inhibitor maC46 to protect these dTc from infection. Therefore, we hypothesize that CMV-specific T cells, when co-transduced with CD4-CAR and the fusion inhibitor, will persist in vivo based on their CMV specificity, but will be protected from infection and will targe residual/reactivated HIV+ cells. In the R21 phase, we will determine the feasibility of these experiments by measuring the in vivo anti-HIV activity of CMV-specific dTc in rhesus macaque model. We will follow a clinical scenario and treat animals with ART after SHIV infection. The effects of these CMV-specific dTc on viral load will be measured in lymphoid tissue reservoirs. Therefore, we established these critical milestones: 1) demonstrate efficient co- transduction/gene marking of CMV-specific dTc in vivo, 2) measure the distribution and functional CTL activity in vivo, and 3) measure the effects on plasma and tissue viremia. In the R33 phase, we will expand these studies to measure the effect of CMV-specific dTc expressing CD4- CAR vectors to inhibit viral replication in tissues. We will use an internal control vector as the basis for comparing proliferation and selection of the dTc. The original CD4-CAR vector with only the TCR? the 2nd generation, and the 3rd generation CD4-CAR adding intracellular signaling domains from CD28 and 4-1BB will redirect the CMV-specific T cells towards HIV, as have been accomplished clinically in oncology. We will compare the persistence and homing, CTL activity, the selective expansion, and ultimately, the efficacy of the different vectors. Since rhesus macaque is an important animal model for HIV pathogenesis and gene therapy, the evaluation of dTc in rhesus/SHIV model will translate quickly into the clinic. The ability of dTc t control viremia in the absence of ART, especially in the rhesus/challenge model, would be a significant advancement in HIV treatment and would strongly promote a new clinical trial for dTc in HIV/AIDS.
Currently, HIV positive patients require life-long antiretroviral therapy because these regimens alone are insufficient to cure the infection. To mimic the effector T cell responses against HIV/SIV that CMV vaccine vectors elicit this proposal will genetically modifying CMV-specific T cells with vectors expressing CD4-CAR to redirect CMV-specific T cell towards HIV. Persistence, homing, and activation in vivo will be maintained by the CMV antigens, while killing is mediated through the CD4-CAR. Because the pathological process and disease course is strikingly similar to HIV-1 infection in people, control of viral replication in the rhesus/AIDS model would be highly significant.