Abstract Body

Background: Control of HIV replication mediated by CD8 T cells is well documented, though some evidence suggest that CD4 T cells with cytolytic properties (CD4 CTL) may also contribute to control of viral replication. Indeed, reports have shown that HIV controllers possess higher frequencies of CD4 CTL than progressors. However, very little is known about the immunobiology of CD4 CTL and how to generate them with an HIV vaccine. Moreover, it is unclear whether they are actively infected and depleted by HIV. Methodology: To further define HIV-specific CD4 CTL, we assessed gene and surface expression patterns of relevant markers in PBMCs, from chronically HIV-infected patients, responding to HIV-Gag peptide pools. We compared HIV-specific CD4 CTL versus HIV-specific CD8 CTL and T-helper cells using Fluidigm Technology and multicolor flow cytometry. We further evaluated susceptibility to HIV infection and by assessing the infectability of sorted HIVspecific CD4 CTL using a panel of four R5 primary isolates. Results: Using index-cell-sorting paired with Fluidigm Technology as well as flow cytometry, we demonstrated that HIV-specific CD4 CTL have a distinct expression profile, which clearly distinguishes them, not only phenotypically, but also functionally from HIV-specific CD8 T cells and Th1 CD4 cells without cytolytic activity. In particular, expression of surface markers CD57 and KLRG1 was significantly higher on HIV-specific CD4 CTL (p=0.0036, p=0.0263 respectively) than on Th1 CD4 cells. Moreover, we found that both t-box transcription factors T-bet and eomesodermin (Eomes) are involved in the cytolytic program of CD4 T cells, similar to cytotoxic CD8 T cell. In contrast to CD4 CTL, HIV-specific CD8 T cells expressed NKG2D with higher level of Eomes, KLRG1 and CD57. Even though HIV-specific Th1 CD4 cells expressed T-bet, Eomes expression was absent in those cells, suggesting a role of Eomes in the cytolytic program of CD4 CTL. Nevertheless, HIV-specific Th1 CD4 cells are characterized by high expression of CD161. Interestingly, the expression levels of CCR5 were the lowest on CD4 Th1 cells and the highest on CD4 CTL. In a viral replication assay, we found that CD4 CTL, defined as expression of CD57 or KLRG1 on CD4 T cells, have a lower infectability than CD57- or KLRG1- CD4 T cells suggesting a distinct HIV-specific CD4 T cell subset that resists HIV infection. Conclusions: Our data demonstrate that HIV-specific CD4 CTL have a distinct functional and phenotypic profile compared to HIV-specific CD8 CTL and CD4 Th1 cells. Furthermore, CD4 CTL express higher levels of CCR5 and are less prone to HIV infection compared to general Th1 CD4 T cells suggesting a CD4 T cell population that is more likely to persist during HIV infection.