Development of a Novel rAAV Vector Without Cross-species Barrier to Transduce Human and Ferret Conducting Airways

PROJECT SUMMARY Human and ferret airways share physiologic similarities in the anatomic properties of their upper and lower respiratory tracts and lung physiology. While the susceptibility of ferrets to pandemic influenza has been known for almost a century, with the recent establishment of cystic fibrosis (CF) ferret models and the advance of passive immunization against respiratory infections, ferrets have become an attractive mammalian model in preclinical studies to evaluate the therapeutic and prophylactic approaches for human pulmonary diseases. Recombinant adeno-associated virus (rAAV)-expression of neutralizing antibody in mice and ferret airways has been proven to elicit efficient protection against influenza virus infections. rAAV2.5T was selected by directed evolution of an AAV2 and AAV5 shuffled capsid gene library in polarized human airway epithelium cultured at an air-liquid interface (HAE-ALI) in vitro. It was thought to be a hopeful candidate vector for in vivo gene delivery to human airways from apical lumen. However, studies of its transduction profile in ferret airways in vivo found undesired vector deposition in alveoli, but not in the trachea and lung conducting airways which are the predominant targets for CF gene therapy and also the primary sites where infection of influenza virus and SARS- CoV-2 naturally occurs. Thus, while using ferret models to examine the efficacies of CF gene therapy and influenza and COVID-19 prevention/treatment is favorable in preclinical studies, currently there is a significant lack of an ideal rAAV vector that can transduce both human and ferret epithelial cells on the conducting airways. Both human and ferret conducting airways predominantly express α2-6 N-linked sialic acid (SA), in contrast to the α2-3 N-linked SA that is the primary attachment receptor of the rAAV2.5T vector. The cell surface glycan molecules largely determine the tissue tropism of rAAV vectors, and the directed evolution of the AAV capsid gene has demonstrated its great success in selecting novel rAAV vectors with an altered tropism for favored cell types. We propose to evolve the AAV2.5T capsid from α2-3 N-linked SA tropic to α2-6 N-linked SA tropic through the selections from the AAV2.5T capsid gene libraries. We will employ the evolution in ferret conducting airways in vivo with a productive transduction reporter. Thus, our study will create a novel rAAV vector that can transduce both the conducting airways of ferrets and humans, which will increase the ferret models' applicability in preclinical studies to examine the efficacies of the rAAV-based gene transfer for the expression of neutralizing antibody and the gene therapy of CF lung disease. The outcomes from preclinical studies utilizing the novel rAAV vector and ferret models can then be smoothly translated to developing therapeutics in humans.