Development of a human-based platform to assess adeno-associated viral vector genotoxicity

Abstract

Gene therapy is an effective treatment for genetic disorders by delivery of therapeutic genes to tissues, such as the liver. Gene transfer can use viral vectors such as adeno‐associated virus, where, once delivered to the nucleus can remain episomally or integrated into the host DNA. Recently, adeno‐associated viruses have shown to been found associated with hepatocellular carcinoma in mice, which is likely due to vector effects on gene(s) local to the site of integration. Models to determine potential gene therapy vector genotoxicity have been developed, however, these often use animals, cells of murine origin or immortal cells that may be oversensitive in predicting human genotoxic risk. To circumvent this, we developed a human based model using induced pluripotent stem cells and their hepatocyte‐like cell derivatives. Cells were profiled for markers of pluripotency and terminal differentiation, respectively, then transcriptomically demonstrating hepatocyte‐like cell expression aligning with primary human hepatocytes. Cells were infected with adeno‐associated vectors using the strong Chicken β‐actin (CB7) or weak Apo‐lipoprotein (ApoE) promoters to drive GFP expression and investigated for integration by EPTS/LM‐PCR that showed twice as many insertions in cancer genes by the ApoE vector than the CB7 vector. RNASeq profiling of infected cells also identified cancer gene fusion transcripts between adeno‐associated virus and host cells and analysis of methylation changes in the infected cells showed differences in global methylation profiles by each vector. We present this model as useful to identify and measure the genotoxic potential of these vectors on the host.