In Vivo Delivery of Precision Genome Editing Agents

Abstract

The capabilities of researchers to modify the mammalian genome has expanded greatly in the last decade, but translation of these editing technologies to therapeutics is limited by the lack of efficient means to deliver the editing agents in vivo. The focus of my doctoral thesis research has been to enhance delivery of base editors and prime editors to therapeutically relevant tissues and to apply these technologies to preclinical animal models of human disease. First, I describe the development of single-AAV delivery for adenine base editors. Base editors had previously been too large to be packaged along with the necessary cis elements for expression within the roughly 4.7kb packaging capacity of adeno-associated virus (AAV). Through minimization of the expression elements on the AAV genome, as well as characterization of novel base editors based on compact Cas9 orthologues with diverse protospacer-adjacent motif (PAM) recognition, we developed a platform for single-AAV delivery of a suite of compact ABEs. Single-AAV ABEs together offer substantial target flexibility, and mediate high efficiency base editing in vivo, especially in tissues where co-transduction is limiting such as heart and muscle. We demonstrate near complete reduction in serum/plasma Pcsk9 at relatively low doses of AAV, demonstrating the robustness of the single-AAV platform. Single-AAV ABEs should facilitate the translation of in vivo base editing by increasing the efficiency of editing in vivo, lowering the total dose of AAV required for a given application, and simplifying production. Second, I describe how we developed methods to perform prime editing in vivo. Prime editing has the potential to facilitate the study or treatment of many genetic disorders caused by DNA substitutions, insertions, or deletions. Realizing this potential requires delivery methods that support efficient prime editing in vivo. By identifying and engineering solutions to bottlenecks limiting AAV-mediated in vivo prime editing efficiencies, we developed optimized dual-AAV prime editor delivery systems (v1em and v3em PE-AAV) that enable therapeutically relevant editing in mouse brain (up to 42% in neocortex), liver (up to 46%), and heart (up to 9.1%). Prime editor expression is a bottleneck of in vivo editing efficiency that is largely overcome by the v3em PE-AAV system. We applied these systems to install putative protective mutations in vivo for Alzheimer’s disease in astrocytes and for coronary artery disease in hepatocytes, cell types of therapeutic relevance to the respective diseases. In vivo prime editing with v3em PE-AAV did not cause detectable off-target editing and did not result in significant liver enzyme or liver histology differences compared to untreated animals. Optimized PE-AAV systems support the highest levels of in vivo prime editing reported to date, facilitating the study and potential treatment of diseases with a genetic component. Finally, I describe the application of in vivo base editing to address a mouse model of prion disease. Prion disease is a fatal condition caused by the templated misfolding of cellular prion protein (PRNP), which can be acquired by exposure to misfolded PRNP or caused by mutations that increase the likelihood of a stochastic misfolding event occurring spontaneously. Lowering PRNP levels via ASO-mediated knockdown improves disease outcomes in a dose-dependent manner in a mouse model of prion disease, but ASOs require repeated dosing and are poorly distributed in the brain. We used an optimized dual-AAV cytosine base editor to install the nonsense allele PRNP R37X to precisely and permanently knock down cellular PRNP in human cells and in humanized PRNP mice using a blood-brain barrier crossing AAV. In vivo base editing resulted in 37% PRNP R37X and 42% knockdown, and base editor-treated PRNP R37X animals after challenge with pathogenic human prion isolates had median lifespans 1.6-fold that of control mice treated with base editor targeting Dnmt1. Furthermore, we assess more potent base editing strategies to achieve comparable levels of editing with lower doses of AAV. These experiments are the first demonstration of lifespan extension in a humanized model of prion disease and provide the basis for further study of PRNP lowering therapies via in vivo genome editing.