Pre-clinical development of adeno-associated virus (AAV)-based therapies gives important insights into safety and efficacy before experimental testing in humans. However, translating promising pre-clinical results into clinical success is a struggle. AAV-based therapies have faced numerous safety concerns with several adverse events linked to manufacturing impurities, such as empty capsids or encapsidated non-therapeutic (e.g., partial vector or non-vector) DNA.1
How to effectively characterize these impurities is an evolving regulatory question, and drug developers are increasingly prioritizing manufacturing quality assessments early in pre-clinical development. In the following blog, we discuss some of the critical quality attributes (CQAs) – characteristics that need to be within an appropriate range to ensure quality – of recombinant AAVs (rAAVs), how they’re measured, and how in silico methods can eliminate bottlenecks in this process.
Emerging Trends in rAAV Product Characterization
Bioreactors are cell culture vessels that are used for the production of rAAV-based therapeutics. They have sensors for monitoring key parameters, such as temperature, pH, dissolved oxygen, and nutrient levels, essential for optimizing production yield and quality. A bioreactor is used to assess the manufacturability of an rAAV vector design for preclinical and clinical studies.
Several types of capsids can be produced during the manufacturing of rAAVs in a bioreactor, and accurately characterizing and quantifying them is essential for determining product-related CQAs. They include:2
- Full capsid: Contains the complete viral vector and can deliver the intended therapeutic gene to the target cell type.
- Partially-filled capsid: Contains a partial viral vector or non-vector DNA (e.g., DNA derived from host cell) that may reduce therapeutic efficacy by decreasing overall transduction efficiency and increasing immunotoxicity.
- Empty capsid: Contains no DNA and, like partially-filled capsids, may decrease overall transduction efficiency and increase immunotoxicity.
Quantifying the presence of these different types of capsids is vital for determining the overall quality of rAAVs, and two CQAs in particular, virus titer and content ratio, have emerged as necessary for fulfilling regulatory requirements for efficacy and safety.
Virus titer can refer to infectious, genome, or capsid titers. Assays that measure infectivity can be time-consuming for process development, and methods that quantify genome and capsid titers are higher throughput and more rapid, making them more practical for the early stages of pre-clinical development.
For instance, viral genomes can be measured with quantitative PCR (qPCR), digital droplet PCR (ddPCR), or next-generation sequencing (NGS). Capsid titers can be measured using ELISAs. While these methods are tried-and-true, they can face issues with reproducibility and lengthy turnaround times.
Content ratio refers to the ratio of full capsids to partially filled or empty capsids. This can be quantified using ultracentrifugation, anion-exchange chromatography, or electron microscopy. These methods are considered the gold standard, yet they are low throughput and take significant amounts of time to complete.
Enhancing Gene Therapy Manufacturing Product Quality with Bioreactor Simulation
The time and money associated with producing rAAVs and evaluating the manufacturability of different rAAV vectors creates a major bottleneck for rAAV-based drug developers.
Predictive, AI-powered in silico tools can change this, uncovering production impurities without the capital investment required for biological confirmation.
Bioreactor simulation, a method to predict which constructs will work best in an actual bioreactor, can streamline this process by enabling rAAV developers to efficiently assess numerous construct combinations in a virtual environment, examining the quality, production efficiency, and expression of multiple candidates featuring diverse regulatory elements. It allows for the optimization of vector construct design (e.g., promoter, terminator, ITR, etc. selection) and transfection conditions. In addition, certain models can help identify problematic CpG islands and vector truncation events so vectors can be derisked and improved for enhanced manufacturability.
Using in silico predictive tools and strategic biological validation can help gene therapy developers identify and overcome vector design problems early in pre-clinical development phase of product development.3 Ultimately, these tools help circumvent time-consuming and costly trial-and-error experimentation.
AI Disclosure: Some of this content was generated with assistance from AI tools for copywriting. Feature image was generated by an AI image tool MidJourney.