Atherosclerotic cardiovascular disease (ASCVD) remains an important cause of morbidity in the general population and risk for ASCVD is increased approximately 2-fold in persons living with HIV infection (PLWH). This risk is linked to elevated CD8 T cell counts that are abundant in atherosclerotic plaques and have been implicated in disease pathogenesis yet the mechanisms driving T cell recruitment to and activation within plaques are poorly defined. Here we investigated the role of CD8 T cells in atherosclerosis in a non-human primate model of HIV infection and in the HIV-uninfected elderly; we sought to identify factors that promote the activation, function, and recruitment to endothelium of CX3CR1+ CD8 T cells. We measured elevated expression of CX3CL1 and IL-15, and increased CD8 T cell numbers in the aortas of rhesus macaques infected with SIV or SHIV, and demonstrated similar findings in atherosclerotic vessels of HIV-uninfected humans. We found that recombinant TNF enhanced the production and release of CX3CL1 and bioactive IL-15 from aortic endothelial cells, but not from aortic smooth muscle cells. IL-15 in turn promoted CX3CR1 surface expression on and TNF synthesis by CD8 T cells, and IL-15-treated CD8 T cells exhibited enhanced CX3CL1-dependent chemoattraction toward endothelial cells in vitro. Finally, we show that CD8 T cells in human atherosclerotic plaques have an activated, resident phenotype consistent with in vivo IL-15 and CX3CL1 exposure. In this report, we define a novel model of CD8 T cell involvement in atherosclerosis whereby CX3CL1 and IL-15 operate in tandem within the vascular endothelium to promote infiltration by activated CX3CR1+ memory CD8 T cells that drive further endothelial activation via TNF. We propose that these interactions are prevalent in aging and in PLWH, populations where circulating activated CX3CR1+ CD8 T cell numbers are often expanded. People living with HIV infection and elderly HIV-uninfected persons have increased risk of developing atherosclerotic cardiovascular disease, and have increased numbers and/or proportions of CD8 T cells that express the vascular endothelium-homing receptor CX3CR1. Atherosclerotic plaques contain many activated CD8 T cells, which have been implicated in disease pathogenesis, yet the mechanisms driving T cell recruitment to and activation within plaques are not clear. Here we propose a model in which CX3CR1+ CD8 T cells promote endothelial dysfunction by the combined effects of CX3CL1, IL-15, and TNF. Persistent inflammation triggers endothelial cell activation and dysfunction in people living with HIV infection. Endothelial cell-derived CX3CL1 then directs the migration of CX3CR1+ T cells to the activated endothelium where IL-15 activates T cells to express TNF. TNF drives endothelial expression of CX3CL1 and IL-15, providing a feed-forward loop of activation. We provide evidence that these pathways are active in human atherosclerotic plaques and in the aortic endothelium of SIV/SHIV-infected rhesus macaques. We propose these mechanisms of T cell-induced endothelial damage are operative in traditional risk factor-associated atherosclerosis in the general population and are accelerated in people with HIV infection who live in a state of sustained chronic inflammation.

Cytotoxic CD4 T cells are linked to cardiovascular morbidities and accumulate in both HIV and CMV infections, both of which are associated with increased risk of cardiovascular disease (CVD). In this study, we identify CMV coinfection as a major driver of the cytotoxic phenotype, characterized by elevated CD57 expression and reduced CD28 expression, in circulating CD4 T cells from people living with HIV infection, and investigate potential mechanisms linking this cell population to CVD. We find that human CD57+ CD4 T cells express high levels of the costimulatory receptor CD2 and that CD2/LFA-3 costimulation results in a more robust and polyfunctional effector response to TCR signals, compared with CD28-mediated costimulation. CD57+ CD4 T cells also express the vascular endothelium-homing receptor CX3CR1 and migrate toward CX3CL1-expressing endothelial cells in vitro. IL-15 promotes the cytotoxic phenotype, elevates CX3CR1 expression, and enhances the trafficking of CD57+ CD4 T cells to endothelium and may therefore be important in linking these cells to cardiovascular complications. Finally, we demonstrate the presence of activated CD57+ CD4 T cells and expression of CX3CL1 and LFA-3 in atherosclerotic plaque tissues from HIV-uninfected donors. Our findings are consistent with a model in which cytotoxic CD4 T cells contribute to CVD in HIV/CMV coinfection and in atherosclerosis via CX3CR1-mediated trafficking and CD2/LFA-3-mediated costimulation. This study identifies several targets for therapeutic interventions and may help bridge the gap in understanding how CMV infection and immunity are linked to increased cardiovascular risk in people living with HIV infection.

Cell-to-cell transmission is an efficient mechanism to disseminate human immunodeficiency virus type 1 (HIV-1) and human T cell lymphotropic virus type 1 (HTLV-1). However, it has been challenging to quantify the level of cell-to-cell transmission because the virus-producing cells cannot be easily distinguished from infected target cells. We have previously described replication-dependent vectors that can quantify infection events in cocultured cells. These vectors contain an antisense-oriented promoter and reporter gene interrupted by a sense-oriented intron from the human gamma-globin gene. This strategy prevents expression of the reporter gene in the transfected cells but permits its expression in target cells after infection. However, the gamma-globin intron is not efficiently removed by splicing in the aforementioned vectors, thereby reducing the level of reporter gene expression after transduction into target cells. Here, we used two approaches to improve the replication-dependent vectors. First, we improved the splicing events that remove the gamma-globin intron by optimizing the intron insertion site within the reporter gene. Second, we improved the packaging of the spliced RNA without the gamma-globin intron by targeting the intron-containing RNA via microRNA 30 (miR30)-based short hairpin RNAs. Using two optimized fluorescent reporter vectors and flow cytometry, we determined that multiply HIV-1-infected cells were generated at a higher frequency in coculture than in cell-free infection; furthermore, this increase was dependent upon viruses bearing HIV-1 Env. Compared with previously described vectors, these improved vectors can quantify the infection in lymphocytes and in primary cells with a higher sensitivity and allow the detection and quantitation of multiply infected cells, providing better tools to study retroviral cell-mediated infection. IMPORTANCE The human-pathogenic retroviruses HTLV-1 and HIV-1 can be transmitted more efficiently in vivo via direct contact of infected cells with healthy target cells than through cell-free virion-mediated infection. Despite its importance, cell-to-cell transmission has been difficult to quantify because the previously infected cells and the newly infected cells are mixed together in the same culture. In the current study, we generated vectors that are significantly improved over the previously described replication-dependent vectors. As a result, these improved vectors can efficiently detect and quantify cell-to-cell transmission or new infection events in cells in mixed culture. These luciferase- or fluorescence protein-based reporter vectors can be used to quantify and study HIV-1 or HTLV-1 cell-mediated infection in a simple one-step transfection/infection assay.