499. Species-Specific Differences in the Polarity of rAAV5 and rAAV2 Transduction of Mouse and Human Airway Epithelia

Abstract

Top of pageAbstract Species-specific differences in the tropism of viral vectors between mice and humans have made it difficult to anticipate the most appropriate adeno-associated virus (AAV) serotypes for clinical trials. In the context of developing gene therapies for cystic fibrosis lung disease, human air-liquid interface (ALI) models of airway epithelium have been useful as an intermediate surrogate for evaluating the human tropisms of various rAAV serotypes. In this model, we have previously shown that apical infection with rAAV2 and rAAV5 vectors have very similar transduction efficiencies. Additionally, in the presence of proteasome inhibitors, transduction is enhanced |[sim]|1000-fold with both serotypes following apical infection. In contrast to these findings in the human ALI model, in vivo rAAV5 transduction is 50-100-fold greater than that of rAAV2 in mouse lung. These differences could be the result of model system dependencies or species-specific differences in the tropism of AAV2 and AAV5 for airway epithelium in mice and humans. To resolve whether species-specific differences (human vs. mouse) or model system-specific differences (in vitro ALI culture vs. in vivo lung) are the cause of the discrepancies in rAAV5 and rAAV2 transduction in these various studies, we have compared transduction of these rAAV serotypes in mouse and human airway epithelia in ALI culture using pseudotyped luciferase rAAV vectors. Differentiated mouse airway epithelia in ALI culture gave electrophysiologic characteristics similar to those previously reported for murine tracheal epithelium, but distinct from that of human ALI cultures. Following apical infection of mouse airway epithelia in ALI culture, rAAV2/5 gave |[sim]|100-fold greater levels of transduction than infection with rAAV2/2. These results were similar to the differences in transduction between these two serotypes seen in vivo in mouse lung. These findings demonstrate that ALI cultures of mouse airway epithelium retain species-specific differences in comparison to human airway epithelium, which is reflected in a greater apical tropism for rAAV5 as compared to rAAV2 capsid serotypes. Interestingly, basolateral infection of mouse ALI cultures also retained a greater transduction efficiency (10-fold) with rAAV2/5, as compared to rAAV2/2. These differences in basolateral tropisms were reversed for human ALI cultures. Viral transduction for both rAAV2/2 and rAAV2/5 from the apical membrane of both mouse and human ALI cultures was enhanced 2-3 logs by the addition of proteasome modulating agents Doxorubicin and LLnL at the time of infection. Basolateral infection in the absence of proteasome modulating was always more effective than apical infection regardless of the species. However, the addition of proteasome modulating agents appeared to normalize transduction from the apical membrane to maximal levels regardless of the serotype or species. These findings suggest that rAAV2 and rAAV5 vectors have different transduction efficiencies for mouse and human airway epithelium due to serotype-specific differences in post-entry blocks modulated by the proteasome pathway.