An investigation to optimise viral vector production and to generate an in vitro model for genotoxicity assessment

Abstract

Viral vectors, including adenoviruses and lentiviruses, have been extensively developed for gene therapy. Gene addition using these delivery mechanisms have proven a successful method in correction of genetic diseases. The rapid expansion in clinical and non-clinical trials in this field indicate a present need to optimise viral vector production to achieve pure, high titre vectors for downstream applications. The current protocols for adenoviral vector purification remain time consuming and present with low recovery yields. We explore aqueous two-phase systems as a cheaper, low maintenance alternative for adenoviral and lentiviral vector purification. These data show high recovery of adenoviral vectors using a PEG 600/ (NH4)2SO4 system. However, lentiviral vectors are not successfully partitioned using these systems. HIV-1 based lentiviruses have been shown to contribute towards genotoxicity in vivo. While models exist, there is need to provide a sensitive, humanised platform of genotoxicity. Here, we present data using induced pluripotent stem cells and their hepatocyte and cardiomyocyte derivatives. Using HIV-1 vectors carrying transgenes flanked by either the native or self-inactivating LTR configurations, we see clear differences in integration site profiling between these vectors in these stem cell and their hepatocyte derivatives. Examination of integration sites over time in replicating stem cells also provides a rapid model for clonal outgrowth. These data suggest the usefulness of using both a naïve and terminally differentiated cell types to sensitively understand lentiviral mediated genotoxicity further. These data also provide evidence for transcription factor binding site tethering of lentiviruses to common sites in the host genome. Comparative analysis to differentiated cardiomyocytes suggest the importance of examining vector mediated adverse events further to understand tissue specific side effects. This work provides for optimised vector production and a model of lentiviral vector genotoxicity which can be used to improve the safety profile of these vectors before clinical applications.