In vitro NIH3T3 mouse embryonic fibroblast cell model does not predict AAV2 or AAVdj-mediated cell transformation

One of the potential risk factors of recombinant adeno-associated virus (rAAV)-based gene therapy is insertional mutagenesis, which has been associated with the development of hepatocellular carcinoma (HCC) in rAAV-treated neonatal mice. The objective of this study was to investigate if well-established in vitro cell transformation assays (CTA) in mouse cell lines can detect AAV2 or AAVdj-mediated cell transformation. Since AAV integration at the Rian locus in neonatal mice has been implicated in AAV-mediated HCC, an rAAV vector specifically targeting the mouse Rian locus and an additional rAAV vector previously shown to cause HCC in neonatal mice were both tested for the induction of cell transformation in NIH3T3 cells. To increase the frequency of AAV DNA integration at the Rian locus in the genome of NIH3T3 cells, double-strand breaks in Rian locus of NIH3T3 cells were created by CRISPR-Cas9 to increase the homologous crossover between viral DNA and the cell genome. When transduced cells were assayed in CTA, the transformation frequency observed in AAV-transduced NIH3T3 cells was not significantly different from that of untreated vehicle cells. The finding that rAAV is unable to transform the NIH3T3 in vitro indicates that either the transformation rate is less than the spontaneous rate of NIH3T3 cellular transformation, or in vitro CTA are not predictive of rAAV-induced HCC in mice.