Single Cell RNA-Sequencing-based Analysis of CD4+ T-Cell Subset-Specific Susceptibility to Transcriptional Modulation by HIV-1 Latency-Reversing Agents

Shock-and-kill is one of the conceptually most advanced strategy towards establishment of an HIV-1 cure. Treatment with latency-reversing agents (LRAs), including histone deacetylase inhibitors with chromatin-remodeling capabilities, combined with anti-retroviral therapy, reactivates HIV-1 transcription in vivo. However, LRA treatment fails to significantly reduce the HIV-1 reservoir in HIV-1-positive individuals, indicating that it is probably insufficient to eliminate latently infected cells. The global and T-cell subset-specific impact of individual LRAs on the transcriptome of CD4+ T-cells, the main HIV-1 reservoir containing cell type in vivo, remains understudied. Here, using single cell RNA-sequencing, we characterize LRA treatment-induced alterations of CD4+ T-cell subset composition and of subpopulation-specific transcriptomes, using Vorinostat and Panobinostat as two prototypic HDAC inhibitors. Ex vivo exposure of CD4+ T-cells from an aviremic HIV-1-positive individual to Panobinostat markedly reduced the percentage of TREG cells. Furthermore, it altered expression of a multitude of interferon-regulated genes, resulting in suppression of several well-characterized antiviral genes, and in enhancement of selected interferon-regulated genes with proviral activities. These changes were most pronounced in TN, TCM, TTM and TEM, and less pronounced in TREG. Exposure to Vorinostat resulted in a comparably mild change of cellular transcriptomic profile, regarding both the number of deregulated genes and their fold change of expression. Nevertheless, selected interferon-regulated genes exhibited a subset-specific expression profile upon Vorinostat treatment. Finally, some genes were deregulated by both treatments in a subset-specific manner. We conclude that treatment by both individual HDAC inhibitors induces an overall proviral milieu in CD4+ T-cells subsets. While this proviral state might be favorable for efficient HIV-1 reactivation, we hypothesize that it may impede the instruction of activation of cellular and adaptive immunity required for effective killing of reactivated cells.