Abstract
Homology Medicines was a company focused on using proprietary strains of rAAV to introduce payloads that could integrate into a targeted location in the human genome using the body’s own DNA repair pathways to correct and treat rare genetic diseases. Research and development into designing and validating homology arms that are safe and efficient is very time-consuming and costly, making it difficult to get these drugs to patients. To reduce the cost and effort of R&D, we sought to better understand what features of homology arms and the targeted genomic locus leads to greater integration efficiencies, as well as design an integration assay capable of detecting all on and off-target integration events to ensure safety. Utilizing expertise in both wet lab assay development and novel bioinformatics pipeline creation to analyze long-read sequencing data, we were able to determine that combining the single stranded nature of the rAAV genome to target locations in the human genome that were prone to double-strand breaks likely increases the rate of integration. Additionally, we adapted a proviral integration recovery assay to overcome challenges in designing assays for AAV to develop an assay that maps all targeted integration events, ensuring our constructs only integrate at the intended location. While Homology Medicines is no longer in business, other companies utilizing AAV and other genome editing tools could use these tools to progress their programs more quickly and cost-effectively while ensuring the safety of their vectors.