Impact of AAV capsid redesigns on titres, processing and stability

Project ID: 2228cd1264 (You will need this ID for your application)

UCL Lead department: Biochemical Engineering

Lead Supervisor: Paul Dalby

Project Summary:

AAV-based gene therapies are often mutated and selected for improved clinical efficacy and safety. This has delivered clinical breakthroughs for some diseases, and yet poor manufacturing efficiency and stability brings unsustainably high costs (eg. Zolgensma, £1.79m per dose). AAV protein surfaces are mutated to improve their cell targeting and to evade neutralising antibodies. However, the same protein surfaces are critical for manufacturing and stability (shelf-life), eg. surface mutations can disrupt affinity chromatography, or lead to non-specific binding to membranes used in AAV purification. Others lead to self-association and aggregation during storage. Therefore, translation of clinical promise is frustrated by low yields and poor quality, which limits their access by patients. We currently lack understanding of the highly complex inter-relationships between AAV structure and functional properties. Such knowledge would enable design rules and predictive tools that accelerate the development of new vectors simultaneously for improved clinical outcomes, manufacturability and stability.
This project has two key aims: 1) establish an experimental platform of scale-down process tools and analytics, to rapidly probe the impact of mutations and their combinations on critical manufacturing and stability properties. An already significant library of viable AAV capsid designs at UCL Biochemical Engineering, with a wide range of in vivo, manufacturing, and stability properties, will be expanded to establish a comprehensive dataset and mapping of structural features to properties. 2) Statistical (PLS, MVA and ML) analysis will map the tolerability of all locations to mutations, and their impact on manufacturing and stability, and enable us to build predictive models for the effects of mutations based on location, mutation type and sequence context. These models will then be used to guide the design of new AAVs with altered tropism, and yet also improve their manufacturability and shelf life.