AAV vector distribution in the mouse respiratory tract following four different methods of administration

Background: Targeted delivery of gene therapy vectors to the mouse respiratory tract is often performed via intranasal or intratracheal administration; however, there can be a great deal of variability between these methods, which could potentially influence experimental results. Improving the accuracy and precision of lung delivery will not only reduce the number of animals required to detect statistically significant differences, but may reduce the variability of studies from different laboratories. Results: Here we evaluated three different methods of adeno-associated virus (AAV) vector administration to the respiratory tract in mice (intranasal, intubation, and intratracheal injection) and discuss the advantages, challenges, and shortcomings of each. We also present a modified-intranasal delivery technique that is superior to passive administration of vector into the nares of anesthetized supine animals. Transgene expression was consistently visible in the nasal cavity, trachea, and proximal to middle aspect of all lung lobes for all four methods, whereas transgene expression was consistently observed in the most distal aspect of lung lobes only with the intubation and intratracheal injection techniques. AAV vector genome copy numbers in the lung were approximately four-fold lower in mice that received vector via intranasal administration in comparison to the other three methods of vector delivery. The modified intranasal, intubation and intratracheal injection methods of vector administration did not yield statistical differences in AAV vector genome copy numbers in the lung. With regard to reproducibility of vector distribution within and between animals, the modified-intranasal technique was superior. Conclusion: Our results show that mode of AAV vector administration to the murine respiratory tract should be selected based on desired target site and skill of the researcher, and that appropriate technique selection may greatly influence experimental outcomes.