An investigation of the potential side effects of adeno-associated virus as a vector for gene therapy and methods to improve its gene transfer

Abstract

Gene therapy offers a promising strategy to deliver functional genes that can correct mutations causing genetic diseases. Among viral vectors, Adeno-associated virus (AAV) is commonly used as a safe and efficient vector for gene transfer of genetic material. Even though recombinant AAV (rAAV) has long been considered a non-integrative vector, in recent years it has been found to integrate into the genome, raising concerns about the potential safety of using AAV in humans. Nevertheless, variations exist in AAV integration profile between different animal models suggesting integration site selection in animals may not be predictive of insertion site choice by AAV in humans. Human studies are very limited in providing enough data to predict AAV safety and therefore human-based models are needed to understand risk of adverse effects by AAV before moving to clinical trials. To overcome this limitation, the hIngetox model was developed in our lab as the first induced pluripotent stem cell (iPSCs)-derived liver model for genotoxicity studies. This model has been shown that hepatocyte-like cells (HLCs) derived from iPSCs closely mimic human primary hepatocytes, enhancing the likelihood that results from this model will be translatable to predict clinical outcome following AAV gene transfer. The hIngetox model was used to investigate four main genotoxicity hallmarks associated with gene therapy safety and cancer. These are insertion site (IS) selection, gene splicing between the vector and the host genome, gene expression changes caused by the vector following insertion close to a cancer gene and clonal tracking to identify proliferating cells in bulk infected populations. Both iPSCs and iPSCs-derived HLCs were transduced with AAV vectors carrying either a weak Apolipoprotein (ApoE) promoter or a strong Chicken β-actin (CB7) promoter to drive green fluorescence protein (GFP) reporter gene expression to investigate the association of promoter strength on genotoxicity. Integration events and fusion genes were identified in both iPSCs and HLCs after viral transduction and these events were found to be associated with biological pathways and gene expression changes. Contrary to expectations, the ApoE promoter, appeared with the highest level of integration as measured by q-PCR for vector copy number, suggesting vector integration is not dependent on promoter strength. In this study, to provide a positive control for genotoxicity for the hInGetox model, a nucleic sequence previously identified associated with AAV induced hepatocellular carcinoma was cloned into a rAAV vector. This vector was used to transduce HLCs and compared to rAAV vectors not carrying this sequence in this study. To understand further AAV IS selection and the potential for vector tethering to sites within the host genome, AAV IS were used to investigate the presence of transcription factor binding sites (TFBS) on the vector and human genome. It was identified that all TFBS found in rAAV were also present at IS in host genome, suggesting these sites were used by the vector for tethering to the host transcription factors. Also, in attempt to understand whether viral tethering and integration causes damage to the host DNA, the H2A histone family member (γH2AX) on Ser139 protein, which become phosphorylated in the presence of double strand breaks was measured during and after viral integration. In collaboration with Vidiia Ltd, γH2AX was quantified using a newly developed AI-driven diagnostic device that can detect DNA damage in a fast, affordable and sensitive manner. In line with integration site quantification, AAV carrying the ApoE promoter appeared associated with the highest level of DNA damage and, therefore, higher potential for genotoxicity. Viral vector toxicity has also been associated with the level of vector dose required for successful gene transfer. To determine whether vector dose could be reduced yet achieve high gene transfer efficacy by enhancement of transduction, in collaboration with N4Pharma Ltd, the potential of Nuvec® silica nanoparticles (SiNPs) to encapsulate and deliver viral vectors was investigated. By this study, it was found that Nuvec® SiNPs can enhance viral efficiency at low dose compared to the vector alone. Also, the potential of Nuvec® to enhance viral stability at temperatures other than storage at -80°C was explored. It was demonstrated that Nuvec® can enhance vector stability at room temperature (RT) and 4°C up to 30 days post complexing vectors with Nuvec®. Overall, this research provides an analysis of rAAV genotoxicity in a human derived liver model and identifies potential mechanisms of AAV associated genotoxicity. By exploring strategies to mitigate genotoxicity risks, using a genotoxicity model or ways of improving vector transduction to reduce vector dose vector, this thesis aims to provide information that can be used to identify and improve rAAV risk and safety for gene therapy.