Directed evolution of adeno-associated virus 5 capsid enables specific liver tropism

Impressive achievements in clinical trials to treat hemophilia establish a milestone in the development of gene therapy. It highlights the significance of AAV-mediated gene delivery to liver. AAV5 is a unique serotype featured by low neutralizing antibody prevalence. Nevertheless, its liver infectivity is relatively weak. Consequently, it is vital to exploit novel AAV5 capsid mutants with robust liver tropism. To this aim, we performed AAV5-NNK library and barcode screening in mice, from which we identified one capsid variant, called AAVzk2. AAVzk2 displayed a similar yield but divergent post-translational modification sites compared with wild-type serotypes. Mice intravenously injected with AAVzk2 demonstrated a stronger liver transduction than AAV5, roughly comparable with AAV8 and AAV9, with undetectable transduction of other tissues or organs such as heart, lung, spleen, kidney, brain, and skeletal muscle, indicating a liver-specific tropism. Further studies showed a superior human hepatocellular transduction of AAVzk2 to AAV5, AAV8 and AAV9, whereas the seroreactivity of AAVzk2 was as low as AAV5. Overall, we provide a novel AAV serotype that facilitates a robust and specific liver gene delivery to a large population, especially those unable to be treated by AAV8 and AAV9. Impressive achievements in clinical trials to treat hemophilia establish a milestone in the development of gene therapy. It highlights the significance of AAV-mediated gene delivery to liver. AAV5 is a unique serotype featured by low neutralizing antibody prevalence. Nevertheless, its liver infectivity is relatively weak. Consequently, it is vital to exploit novel AAV5 capsid mutants with robust liver tropism. To this aim, we performed AAV5-NNK library and barcode screening in mice, from which we identified one capsid variant, called AAVzk2. AAVzk2 displayed a similar yield but divergent post-translational modification sites compared with wild-type serotypes. Mice intravenously injected with AAVzk2 demonstrated a stronger liver transduction than AAV5, roughly comparable with AAV8 and AAV9, with undetectable transduction of other tissues or organs such as heart, lung, spleen, kidney, brain, and skeletal muscle, indicating a liver-specific tropism. Further studies showed a superior human hepatocellular transduction of AAVzk2 to AAV5, AAV8 and AAV9, whereas the seroreactivity of AAVzk2 was as low as AAV5. Overall, we provide a novel AAV serotype that facilitates a robust and specific liver gene delivery to a large population, especially those unable to be treated by AAV8 and AAV9.