56. AAV Vector-Mediated CRISPR Attacks on Proviral HIV-1 DNA for Purging of Cellular Reservoirs

Abstract

HIV, the human immunodeficiency virus, is highly successful as a viral pathogen due to its ability to persist in the infected host as an integrated proviral DNA within cellular reservoirs, particularly in resting CD4+ T cells and macrophages. Current antiretroviral combination therapy can control viral spread and viremia but cannot eradicate latent HIV genomes, raising an urgent need for alternative strategies for long-term remission or cure. We combined two powerful technologies, synthetic AAV vectors and CRISPR-mediated gene engineering, with the ultimate aim to eliminate integrated HIV from latently infected human cells. Therefore, we first screened a library of peptide display mutants based on 12 different AAV serotypes and identified novel capsids that mediate potent transduction of relevant HIV target cells, including primary human T-lymphocytes and macrophages. Our lead candidates – two peptide-modified variants of AAV9 or AAVrh.10 – were capable of transducing >65% of primary humanCD4+ T cells and were superior to all 12 wildtype AAV isolates. Concurrently, we created AAV vector genomes that express – alone or in combination – the two essential CRISPR components, i.e., the Cas9 protein and a g(uide)RNA directed against either the HIV-1 long terminal repeats (LTRs), or against viral structural genes. AAV/CRISPR vector efficiency and specificity were validated using a T7 endonuclease assay, showing successful Cas9-mediated induction of double-stranded DNA breaks at the specific HIV-1 target regions and subsequent error-prone repair in HeLaP4 cells with a stably integrated HIV-1 genome. To further potentiate the approach, we exploited a highly customizable self-complementary AAV vector backbone recently developed in our lab that permits multiplexing up to three gRNAs under control of the U6, H1 and 7SK promoters in a single vector particle. We engineered this template to co-express our best gRNAs against the two HIV-1 LTRs and the viral gag gene, and compared its potency with our different first-generation single gRNA vectors. In reporter cells carrying a GFP-encoding HIV-1 genome, simultaneous cleavage was indeed detected at three different sites (5’LTR, 3’LTR and gag), reaching an impressive 63.5% cutting efficiency (versus 36.6-50% for the single gRNA vectors) and demonstrating the power of these vectors to cleave HIV-1 proviral DNA. Finally, we will present results from our ongoing validation of proviral cleavage in human T-cell lines that are latently infected with HIV-1 as a physiologically relevant model of HIV-1 latency. The described experiments prove the possibility to target, cut and destroy integrated HIV-1 proviral DNA with AAV/CRISPR vectors in human cells, and therefore support the further development of our tools and approach into a novel clinical modality for purging of latent HIV-1 reservoirs.