Cooperation between cGAS and RIG-I sensing pathways enables improved innate recognition of HIV-1 by myeloid dendritic cells in elite controllers

Spontaneous control of HIV-1 replication in the absence of anti-retroviral therapy (ART) naturally occurs in a small proportion of HIV-1-infected individuals known as elite controllers (EC), likely as a result of improved innate and adaptive immune mechanisms. Previous studies suggest that enhanced cytosolic immune recognition of HIV-1 reverse transcripts in conventional dendritic cells (mDC) from EC enables effective induction of antiviral effector T cell responses. However, the specific molecular circuits responsible for such improved innate recognition of HIV-1 in mDC from these individuals remain unknown. Here, we identified a subpopulation of EC whose mDC displayed higher baseline abilities to respond to intracellular HIV-1 dsDNA stimulation. A computational analysis of transcriptional signatures from such high responder EC, combined with functional studies, suggested cytosolic recognition of HIV-1 dsDNA by cGAS, combined with sensing of viral mRNA by RIG-I after polymerase III-mediated HIV-1 DNA transcription. Together, our work identifies collaborative networks of innate sensing pathways that enhance cell-intrinsic abilities of mDC to induce antiviral innate responses against HIV-1; these observations might be useful for the therapeutic induction of effective antiviral immune responses. Total RNA was extracted from sorted Lin- HLADR+ CD11c+ mDC from the blood of a larger cohort of n=24 EC (85% male, 15% female) characterized by a median plasma viral load of 36 copies/ml (range 5-250 copies/ml) and living with HIV-1 for a median of 17.5 years (range = 3–27 years) and with CD4+ T cell counts median of 821 cells/mL (range: 407-1684 cells/mL) and n=15 HIV-1 positive individuals on antiretroviral therapy (HAART) with undetectable viremia (<20 copies/mL) and diagnosed a median of 11 years (min-max; 1.27 years) and with a median CD4+ T cell counts of 909/mL (min-max; 398-1367) patients using the Qiagen RNeasy Micro Kit. Subsequently, RNA-Seq libraries from mDC were generated as previously described. Briefly, SMART-seq2 (37) was used to prepare whole transcriptome amplification (WTA) and tagmentation-based libraries, and samples were sequenced on a NextSeq 500 Instrument (Illumina). Subsequently, sequences were aligned using the Hg38 human genome database by Bowtie 2 (38), and transcripts per million (TPM) values were obtained for each sample by RNA-Seq using Expectation-Maximization (RSEM) (39). TPM values were then normalized among all samples using the upper quantile normalization method. For a first set of analyses mDC RNA-seq data from all EC and all HAART individuals was compared. In subsequent analyses, RNA-seq data from a total of n=8 EC with known high (n=4) or low (n=4) in vitro response to HIV-1 dsDNA stimulation in mDC were compared. Pathway analyses were performed using Ingenuity Pathway Analysis and DAVID software. Additional gene network images were obtained from selected upstream regulator lists using the NetworkAnalyst software (12).