The extensive experience with natural AAV isolates illustrates significant potential of these vectors, yet also a need for improvements in specificity, potency and safety. Here, we propose an approach that combines rational engineering with high-throughput screening to design AAV capsid technologies that address the current needs.

Our previous work shows that ancestral sequence reconstruction (ASR) can be used to predict the sequence of ancestral AAV (AncAAV) capsids. The ASR method predicts the residue in each position of the AncAAV capsid, but with uncertainty in some positions in which two residues are equally probable. An AncAAV library was created to explore all possible combinations of these uncertain position. The screening of this library revealed a single residue that determines hepatotropism. We have designed eight more AncAAV libraries based on ASR of different branches of the AAV phylogeny and screened them by co-injection in mice. The analysis of capsid enrichment in the heart identified several residues within the AAV capsid that control the transduction of cardiac tissue. This effect is conserved across the libraries. Additionally, we have seen that some of the identified positions also affect liver transduction, which points towards a non-tissue specific higher transduction efficacy of some of the variants. As we demonstrated previously, these findings can be used to manipulate the tropism of other serotypes by mutating the identified residue in the existing capsids.

In conclusion, AncAAV libraries are a powerful platform to identify the residues of AAV capsid that determine or alter the tropism for different tissues.