Abstract
Research towards improved adeno-associated virus (AAV) delivery vectors for use in treating human liver diseases continues to be dominated by efforts to characterize potential candidate vectors in mice which often can't recapitulate either the human liver tropism/transduction profiles, or the specifics of the human immune response to these vectors. Research from our lab has shown that despite the impressive hepatocyte transduction levels by AAV-8 in mouse liver, AAV-8 does not functionally transduce human hepatocytes in xenografted humanized liver mouse models to levels of therapeutic relevance, possibly explaining the lower than expected FIX expression levels from the recent AAV-8 hemophilia clinical trial. Efforts to evolve novel capsid variants with improved human hepatocyte tropism were successful with the generation of AAV-LK03, yet limitations to clinical translation with many newly evolved serotypes remains due to high levels of pre-existing neutralizing antibodies that may cross-react and neutralize transduction. For this reason we sought to evolve improved hepatotropic variants–again in humanized liver mice–that not only retained robust human liver tropism, but also had been evolved for resistance to pre-existing neutralizing antibodies in the human population to create the ideal liver vector for use in the clinic. We utilized wild-type replicating AAV libraries of 10e5 variants via DNA shuffling of eleven different parental AAV capsids. Our libraries selectively replicate in human cells when co-administered with wild-type adenovirus, making chimeric humanized liver mice an excellent tool to allow for selection of capsids with tropism for human liver. Screens were carried out for five rounds of selection and rather than waiting for the screen to go to completion, we ended the screen with some library diversity remaining for use in a subsequent screen for neutralizing antibody evasion. All variants from round five of the human liver screen were carried forward and screened for two additional rounds for their ability to resist binding to pooled human immunoglobulins in IgG immunocapture assays. The top 100 highly selected variants were sequenced and vectorized into CAG-driven GFP preparations in seven pools based on capsid relatedness at the amino acid level. Each pool was then resubjected to additional pooled human immunoglobulin screening and the best pool was chosen. The top seven candidates from this pool were tested alongside control serotypes that represent the extremes for humoral neutralization: highest neutralization (AAV-2), lowest neutralization (AAV-DJ), as well as LK03 that falls in between. One variant from this highly selected pool has neutralization profiles on par with AAV-DJ and further characterizations are underway in humanized liver mice in vivo and various human hepatoma lines in vitro to determine if both desired characteristics were robustly retained—human liver tropism and human immune evasion.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call