Adeno-Associated Virus Gene Therapy Vectors: Molecular Interactions on Cell Entry

项目摘要/摘要

Project Summary / Abstract Adeno-associated virus (AAV) is a leading delivery vector for gene therapies of a wide range of genetic disease and predilections. The FDA has recently approved AAV-mediated treatments for spinal muscular atrophy (SMA) and RPE-associated retinal dystrophy, while 150 clinical trials are ongoing. However, inefficient transduction requires high doses with risk of immunotoxicity that has been evident in clinical trials. A genome-wide screen for host proteins, needed for AAV cell transduction, changed the understanding of AAV entry. Hitherto-uncharacterized AAVR was identified as a key receptor, while the downstream role of GPR108 is most likely in endosomal escape. AAVR cryo-EM structure will be put into biological context. Studies of GPR108 will start by pinpointing the entry step and cell location where AAV and GPR108 associate, and identification of the domains with which AAV interacts. The approach to structure will be holistic. Overall configuration will come from cryo-electron tomography (ET) of AAV complexed with native-like GPR108 nanodiscs. This will be integrated with high resolution cryo-electron microscopy (EM) of AAV complexed with expressed extracellular domains of GPR108. Triggers will be investigated of an AAV conformational transition that releases a virally-sequestered phospholipase A2 for endosomal escape, as AAV traffics to the nucleus. Antibody neutralization of AAV is widely considered mediated by interference with glycan-binding or of a post-entry step. This will be reevaluated after finding overlap in the binding of AAVR and several neutralizing monoclonal antibodies. Prevalent in vivo modes of binding will first be established through cryo-EM using polyclonal antibodies from pooled human serum. Mutations, directed towards escape, will be used to test whether receptor-binding and entry are, in fact, primary mechanisms of neutralization, and whether escape to rapid neutralization by pre-existing antibodies can be designed without harm to receptor-mediated cell entry. Computer methods, optimizing atomic structure vs. cryo-EM maps, prototyped during AAV studies, will be developed and disseminated. Modules will implement the RSRef algorithm, unique in matching to map values, a sum of 3D atomic profiles of refinable experimental resolution. This will be combined with open source force fields, opening an emerging application of atomic resolution cryo-EM in determining hydrogen bonding networks and protonation states. For resolutions below 3 Å, restraints eliminating unnecessary freedom will be introduced to limit the over-fitting that is an under-appreciated problem in EM refinement. The project’s main goal is a molecular understanding of AAV’s host interactions, providing structural and mechanistic foundations for engineering needed improvements in vector efficiency and specificity in development of safe gene therapies. Broader collateral impact will come from a model system for viral entry that will be put to rigorous practical test, and computer methods for improving the precision of EM structures.