Chimeric antigen receptors enable superior control of HIV replication by rapidly killing infected cells

Engineered T cells hold great promise to become part of an effective HIV cure strategy, but it is currently unclear how best to redirect T cells to target HIV. To gain insight, we generated engineered T cells using lentiviral vectors encoding one of three distinct HIV-specific T cell receptors (TCRs) or a previously optimized HIV-targeting chimeric antigen receptor (CAR) and compared their functional capabilities. All engineered T cells had robust, antigen-specific polyfunctional cytokine profiles when mixed with artificial antigen-presenting cells. However, only the CAR T cells could potently control HIV replication. TCR affinity enhancement did not augment HIV control but did allow TCR T cells to recognize common HIV escape variants. Interestingly, either altering Nef activity or adding additional target epitopes into the HIV genome bolstered TCR T cell anti-HIV activity, but CAR T cells remained superior in their ability to control HIV replication. To better understand why CAR T cells control HIV replication better than TCR T cells, we performed a time course to determine when HIV-specific T cells were first able to activate Caspase 3 in HIV-infected targets. We demonstrated that CAR T cells recognized and killed HIV-infected targets more rapidly than TCR T cells, which correlates with their ability to control HIV replication. These studies suggest that the speed of target recognition and killing is a key determinant of whether engineered T cell therapies will be effective against infectious diseases. Infusion of HIV-specific engineered T cells is a promising approach to eliminate HIV infected cells that reverse latency and begin to replicate HIV after ART withdrawal. However, it is unclear whether chimeric antigen receptors (CAR) or TCRs are the best approach to generate HIV-specific T cells for adoptive T cell therapy. We generated CD8 T cells expressing one of three HIV-specific TCRs (TCR-T) or a previously optimized HIV-specific CAR and interrogated which population controlled HIV replication more effectively. HIV-specific CAR T cells were far superior to any of the TCR-T cells. Increasing TCR affinity or interfering with the ability of Nef to downregulate MHC class I had modest effects on the ability TCR-T cells to control HIV. Inserting additional cognate epitopes within the HIV genome did improve the ability of TCR-T cells to control HIV but HIV CAR T cells were still superior. The ability to control HIV replication best correlated with the ability of engineered T cell to rapidly induce cell death in HIV-infected T cells, suggesting that the speed of recognition and initiation of T cell mediated killing is a key determinant on whether engineered T cells can control viral infections.