Chimeric AAV Cap sequences alter gene transduction

Recombinant adeno-associated viruses (rAAV) are vectors for gene delivery. rAAVs occur in several serotypes shown to have different transduction capabilities. Capsid sequences of the serotypes are different suggesting that differences in gene delivery are in large part due to capsid structure. Since the available serotypes are inefficient transducers of cell lines on a particle per cell basis and inefficiently transduce desired target cells in vivo there have been numerous attempts to create better vectors by modifying the capsids. A question of interest is whether natural selection has led, and whether laboratory-based selection methods will lead, to viruses with a sharply defined optimal specificity. Here we created multiple randomly recombined capsid species using the known AAV serotype capsid sequences by STEP and shuffling methods. We then used selection to identify a viral capsid better adapted to Hep G2 cells than the known serotypes. This capsid was then tested with other cells to determine whether the selection produced a virus specifically tailored by the selection methodology or one of wider applicability. The selected virus turned out to be surprisingly limited by its target cell and method of selection.