Our goal is to develop precise nucleases that will permit the targeted inactivation of latent HIV-1 provirus in patients.
Highly active antiretroviral therapy (HAART) has dramatically changed the prognosis for individuals infected with HIV-1. Yet, even when HIV-1 viremia has been well controlled by these drugs for years, termination of HAART results in viral rebound, most likely coming from latent provirus in long-lived memory CD4+ T cells. So long as latent HIV-1 provirus persists – probably for the life of the infected individual - HAART will be required. Most efforts to eradicate latent HIV-1 proviruses have focused on reactivation of proviral transcription to potentiate the elimination of cells bearing HIV-1 provirus. To date, though, such reactivation efforts have largely been unsuccessful. Alternative approaches for the effective elimination of latent HIV-1 provirus are therefore needed.
Recent advances in the development of targeted gene editing tools provide a potential method for direct inactivation or excision of latent HIV-1 provirus. Specifically, the Cas9/CRISPR programmable nuclease system, a versatile platform for the generation of targeted double-strand breaks within the genome, has been shown to excise HIV-1 provirus in cell lines. However, the activity and precision of the Cas9/CRISPR system is suboptimal for clinical application. We have developed a novel nuclease architecture that combines the favorable cleavage activity of Cas9 with the targeting specificity of Transcription Activator-Like Effector (TALE) domains. This Cas9-TALE system dramatically improves the activity and precision of DNA cleavage. Here we will optimize our Cas9-TALE system for the inactivation of HIV-1 provirus in memory CD4+ T cells harvested from humanized mice specially engineered for our experimental purposes and from patients undergoing HAART.
Development of a robust nuclease system for the neutralization of HIV-1 provirus will require consideration not only of vulnerable sequence elements that are prone to conservation, but also the local chromatin landscape of the provirus in quiescent reservoir cells. The local chromatin landscape will be critical not only for defining assessable regions for nuclease targeting in resting cells, but also for identifying the folding of proviral DNA that may facilitate its excision between two appropriately positioned breaks. To obtain sufficient material for characterization, we will employ genome-editing technologies to generate populations of CD34+ cells containing a proviral insertion at a specific site and orientation within the genome. Once engrafted in NSG-BLT mice, these CD34+ cells will generate populations of resting CD4+ cells with uniform proviral integration sites for in-depth characterization. This resource will permit the first high-resolution analysis of the organization of proviral nucleosome architecture and its genomic conformation in primary cells in disease-relevant genomic contexts, which may reveal important and unexpected drivers of latency in reservoir cells.