Advancing AAV production

Despite decades of remarkable scientific advancements in the fields of gene and cell therapy precision medicine, the translation of these breakthroughs into regulatory approvals has largely lagged behind in pace. In recent years, however, several FDA approvals, including seven in 2023, have emerged which aim to treat, or even cure, diseases of genetic origin. These approvals mark a significant shift in the field, and set the stage for continued acceleration in the space. In fact, the National Bureau of Economic Research estimates that over 1 million patients will receive gene therapies in the next 15 years, requiring a robust pipeline and a manufacturing infrastructure that is prepared to handle the demand.

Notably, adeno-associated virus (AAV) gene therapy has emerged as a frontrunner in this transformative landscape, due to its favorable safety profile, long-term transgene expression, minimal immunogenicity, and a lifetime single-dose administration for the treatment of a majority of disease indications. As the therapeutic potential of AAV gene therapy becomes increasingly evident, so does the demand for its widespread application; however, this surge in demand poses a formidable challenge in the need for increasingly scalable and efficient AAV production platforms. 

Currently, large-scale AAV-production is predominantly accomplished through transient transfection or through the generation of producer cell lines. Both of these platforms typically use human cell lines, however, insect cells have also been used for the production and manufacturing of AAV. The producer cell line method is advantageous for long term production at lower cost with higher batch-to-batch consistency; however, the generation of these cell lines is labor-intensive, time-consuming, and requires a significant screening effort to determine the best candidate cell lines for a manufacturing process. Leveraging recent advances in technology, we outline here an automated and data-driven future for robust producer cell line development that beckons as a transformative solution. 

Challenges in current producer cell line development

The current stable AAV producer cell line development process is manually intensive and time consuming, demanding meticulous optimization across multiple interdependent stages. Producer cell line generation begins with the introduction of essential components for recombinant AAV production, including AAV genes for replication and packaging, selectable markers, and a gene of interest (GOI) within the framework of AAV inverted terminal repeats (ITRs). Once cell lines are created, an extensive and repetitive cycle of cell line monitoring, screening and selection is performed.

After transfection of parental host cells, large pools of producer cells are generated. These cells are separated and undergo an extended screening process under selective pressure to develop stable cell lines with the desired viral vector expression cassette. Hundreds to thousands of cell populations are seeded into multiwell plates at optimized cell concentrations, and at a scale that balances cell survival and population diversity. Generating and maintaining these cultures is time-consuming, requiring constant monitoring of cell culture health through various methods, such as image analysis and viability assays, to ensure only cells with desired characteristics are carried forward.

The significant workload required for this process becomes evident in the subsequent screening phase, where AAV production titers are assessed in a multi-day process. Infection of cultures with wild-type adenovirus (wtAd), herpes simplex virus (HSV), or standalone helper-specific genes is followed by titer output determinations, typically measured in viral genomes (vg) on a per cell or per volume basis. Typical clinical doses for AAV therapies range from 10¹¹ to 10¹⁴ vg per kg per patient, dependent on the disease indication. To meet such dosage needs on a realistic and manufacturable scale, lead candidate cell lines need to exceed productivity on a level greater than 200,000 vg per cell. Ensuring the generation of high-producing cell lines of this quality necessitates largescale screening efforts involving thousands of individual cultures. This puts a significant strain on the early development pipeline, where producer cell line generation largely relies on physically demanding and repetitive manual pipetting of each individual culture; therefore, this process is an excellent fit for high-throughput automation.

Along with the physical demands of cell line generation are the secondary challenges of sample tracking and data management. As cell lines are generated, it is imperative that the totality of the data associated with each individual culture is correctly linked, to support downstream decision making. In addition, cell line lineage and biosafety data documentation are required for regulatory submissions. At higher throughput, the task of data management and analysis requires specialized software, but errors resulting from manual data entry and potentially subjective decision-making are still prominent hurdles for the field to overcome. Current methods are also limited by the amount of data and types of variables that can be considered in the evaluation of cell lines. While large data sets are generated for each sample throughout the process, candidate cell line selection is still largely based on a small subset of data composed primarily of quantitative values, such as AAV titer, population doubling time and a subset of product quality parameters. More qualitative data, such as cell images collected throughout the process, provide an end-to-end picture of a culture’s growth characteristics and morphology, but are harder to evaluate and weigh against purely quantitative metrics. As throughput requirements in the field continually advance, old-school data management practices no longer support the demand; a great challenge exists in revolutionizing data management and analysis.

The commonality among these challenges lies in the physical demand, resource and time constraints of scientists, and the lack of a developed technology for data management inclusive of both quantitative and qualitative metrics. Addressing these challenges is imperative to advancing AAV cell line development towards a more efficient future.

Emerging automation and data-driven technologies

In 1894, Emil Greiner’s groundbreaking invention of an ‘automated’ pipette marked the inception of a laboratory automation revolution, introducing a device capable of precisely and repetitively pipetting a defined volume.This innovation not only laid the foundation for modern-day automated pipettes and liquid handlers, which have become commonplace in laboratories worldwide, but also initiated a transformative journey that forever impacted the scientific community. While basic automated pipettes and liquid handlers continue to be fundamental tools, remarkable advancements in both process complexity and the requisite technologies needed to support them have emerged. In fact, these developments have given rise to integrated systems resulting in entire ecosystems designed to enhance laboratory functions.

As the legacy of Greiner’s invention continues to shape the landscape of lab automation, its influence has begun to extend into specialized applications, of which cell line development is a prime example. At its core, an automated cell line development ecosystem is designed to seamlessly execute essential tasks (e.g. cell seeding, media exchanges and other culture manipulations) as well an analytical data generation (e.g. cell viability and confluence monitoring) in and end-to-end workflow with minimal user intervention. Early technology adopters designed specialized systems to address these individualized tasks, leading to a proliferation of standalone automated solutions. These include advanced automated cell counting technologies like LUNA, Vi-CELL, Countess, and TC20, cell imaging systems such as the IncuCyte and the Celigo S Imaging Cytometer, as well as a suite of robotic incubators, centrifuges, and plate manipulators. Evolving and customizable liquid handling systems from companies such as Hamilton, Tecan, and Beckman Coulter can carry out basic liquid transfers via automated pipette drives. They also have the capacity to integrate various stand-alone pieces of equipment through the addition of plate grippers, arms, and tracks which shuttle plates between instruments. This integration allows for a more comprehensive approach, covering the entire end-to-end workflow.

The benefits of these implementations are substantial. Automation, even for single-task ecosystems, immediately alleviates the physical demands of high-throughput and manually intensive processes. Automation ensures consistent sample processing conditions that are otherwise unachievable manually, where multiple users interact with a sample over extended periods of time spanning across different workflows. This is particularly crucial for companies aiming to increase throughput and conserve resources through miniaturized assays in small- volume microtiter plates. While these smaller volume assays offer many benefits, they require high precision when pipetting to avoid variability between assays and samples.

Furthermore, automated instruments are transformative on a business scale as a result of improved process integrity, but also their capacity to function independently, without user intervention. With remote monitoring, work can extend well beyond normal business hours with minimal risk. Ultimately, these factors increase laboratory operating time, efficiency, scale, and overall output, enabling scientists to reallocate time to other aspects of process improvements and innovation.

Beyond the benefits of physically automating cell handling and manipulation, the integration of analytical tools within automated workflows enables informed decision making to be incorporated into these processes at various stages. One notable application is the utilization of cell culture images to determine cell density for every culture at regular intervals. When a preset cell density value is reached, well-designed systems can intelligently flag cultures for additional processing based on these predefined criteria, ensuring optimal growth conditions for each culture throughout the development life cycle. This process, when done manually is time consuming, tedious, and inherently subjective.

As such, the emergence of AI presents opportunities to automate these nuanced steps. For instance, AI algorithms can analyze cell culture images to not only determine cell density, but also identify morphological characteristics associated with optimal growth or high productivity. Cells exhibiting favorable traits based on historical data can be automatically flagged for further processing, streamlining the selection of high-potential candidates. In practice, this means that as new cell lines are developed, an AI system can evaluate them against learned criteria and provide insights into their potential value. This data-driven approach strategically directs resources towards cell cultures with the highest potential to become top candidates, creating a more efficient and effective cell line development process.

As throughput expands with the application of automated processes, so does the need for a sophisticated and automated digital data management platform. Data is the key driver in the decision-making process and guides each subsequent step; therefore, it is paramount that data integrity and traceability are accurate. Maintaining samples in an end-to-end automated workflow facilitates this record keeping and provides clear, traceable sample tracking that is invaluable for internal documentation and decision making.

Automation: challenges and future perspective

Budget remains a primary barrier to adopting automation for many labs, but with evolving technology, and competition in a growing market, automated solutions are becoming more affordable and cost-effective, especially when weighed against the longer-term cost saving benefits of implementing automation within the lab.

Another barrier comes in the need for trained personnel. Automation is a powerful technology but will always need skilled operators for its implementation. As technology improves, more scientists will need to incorporate these skills for the field to fully realize the potential benefits automation has to off er. Improving technical literacy in Big Data, AI, machine learning and automation will be essential skills for the future.

As we look to improve not only the throughput, but the quality of our products, automation and AI promise to be powerful drivers and vital staples in the labs of the future. As more labs are adopting automated systems and workflows, the next wave of innovation will be in the area of data management. Data are being generating at unprecedented levels, and with emerging advancements in AI and machine learning, there are new opportunities to leverage this data in meaningful ways, which has the potential to be truly transformative for the field of genomic medicine.

References

Current FDA Approved Cell & Gene Therapies. Mirus Bio. Accessed: December 27, 2023.

Wong, C. H., et. al. Estimating the Financial Impact of Gene Therapy in the U.S. National Bureau of Economic Research. April 2021.

Maurya, S., et. al. Safety of Adeno-associated virus-based vector-mediated gene therapy—impact of vector dose. Cancer Gene Therapy. Jan. 2022.

Greiner, E. A New Automatic Pipette. J. American Chemical Society. Sept. 1894.