498. Adeno-Associated Virus Type-1 Vectors Efficiently Transduce Human Polarized Airway Epithelia without Polarity Bias

Abstract

Recombinant adeno-associated (AAV) type-2 viral vectors have demonstrated great promise for gene therapy of cystic fibrosis lung disease. However, since the efficiency of rAAV2 apical transduction of well-differentiated human airway epithelia is low, the use of this vector serotype may require adjunct modulators of the ubiquitin/proteasome pathway to circumvent post-entry nuclear trafficking barriers. An alternative approach is to identify AAV serotypes that potentially lack post-entry barriers in transduction. It was reported that rAAV5 and rAAV6 vectors transduced mouse lungs much more efficiently than rAAV2, and we have also observed that rAAV1 vectors transduce mouse lungs with similar efficiency to rAAV5. In the present study, we sought to compare rAAV1, rAAV2, and rAAV5 transduction in human and mouse polarized airway epithelia using a differentiated air-liquid interface model. Such a comparison allowed for direct analysis of species-specific differences in serotype transduction within the same model. An identical rAAV2 proviral construct with eGFP or firefly luciferase reporters was encapsidated into AAV type 1, 2 or 5 particles in an adenovirus-free packaging system. rAAV2/1, rAAV2/2 and rAAV2/5 viruses were purified by iodixanol discontinuous gradient ultracentrifugation followed by HPLC using the same chromatographic method. Polarized mouse and human airway epithelia were infected with 5|[times]|103 DRP/cell of rAAV2/1, rAAV2/2 or rAAV2/5 from either apical or basolateral membranes, and gene expression was assessed at 3 days to 3 weeks post-infection. Results from these experiments demonstrated that apical transduction of human polarized airway epithelia with rAAV-1 was 60-fold more efficient than rAAV2/2 and rAAV2/5. Similarly, apical transduction of polarized mouse airway epithelia with rAAV2/1 was 88-fold greater than rAAV2/2 and 38-fold greater than rAAV2/5. Interestingly, rAAV2/1 transduced apical and basolateral membranes in mouse or human polarized airway epithelia with similar efficiencies. These findings were in stark contrast to rAAV2/2 and rAAV2/5 vectors that demonstrated much greater transduction efficiencies from the basolateral membranes of mouse and/or human airway epithelia. Proteasome inhibitors LLnL and doxorubicin also augmented rAAV2/1 transduction of human polarized airway epithelia from both apical and basolateral surfaces. However, the induction was 10 to 20-fold lower than that for rAAV2/2 and rAAV2/5 infection. In summary, these results demonstrate several new findings regarding the biology of rAAV1 transduction in airway epithelia. First, rAAV1 lacks a bias for transduction of apical and basolateral membranes of polarized human and mouse airway epithelia, unlike rAAV2 and rAAV5. This finding suggests that the biology of rAAV1 transduction in mouse and human airways may be more similar than that for rAAV2 or rAAV5. Second, the finding that proteasome inhibitors induce rAAV1 transduction to a lesser extent than rAAV2 and rAAV5, suggests that enhanced transduction with rAAV1 might be the result of altered intracellular processing and/or nuclear trafficking as compared to other AAV serotypes.