Strategies to Optimize Expression: Overcoming the Challenges of Multi-Chain and Bispecific Antibodies
In today’s rapidly evolving biopharmaceutical landscape, the development of multi-chain and bispecific antibodies (bsAbs) presents both significant opportunities and challenges. With advancements in automation, artificial intelligence (AI), and new drug modalities, the need for more flexible, high-yield production platforms has never been greater.
To successfully manufacture these complex biologics, strategies must be adapted at every stage of the process—from early DNA and protein sequence design to optimizing the expression platforms for large-scale production. In this context, Lonza has pioneered several strategies to improve the manufacturability of bispecific antibodies (bsAbs) and other complex protein formats, and is now looking to future innovations to address the growing demands of the industry
Key Strategies for Optimizing Expression of Multi-Chain and Bispecific Antibodies
1. Upfront Design: DNA and Protein Sequence Optimization
Manufacturability challenges are best addressed early in the discovery phase. By designing proteins with a clear understanding of manufacturability requirements—such as using tools like knobs-into-holes technology (for HC-HC heterodimerization) or bYlok® technology (for preventing HC-LC shuffling)—manufacturers can mitigate issues that might arise later in production.
By ensuring that expression challenges are incorporated into the molecular design from the outset, expression bottlenecks can be anticipated and solved in a timely manner, streamlining the overall development pipeline.
2. Optimizing the Expression Platform for Bispecific Antibodies
While traditional expression platforms such as CHO (Chinese Hamster Ovary) cells have been optimized for monoclonal antibodies (mAbs), they must be re-evaluated and tweaked to support the production of more complex molecules, such as bsAbs. New developments in expression vectors, such as engineered promoters, UTR elements, and novel gene orders, can significantly impact expression yields and product quality.
At Lonza, leveraging high-throughput screening (HTP) technologies has enabled better clone selection and faster optimization, helping to achieve high-yield, high-quality protein production. AI-driven predictive models are increasingly becoming an important tool for predicting expression performance, reducing reliance on trial-and-error approaches.
3. Enhancing Clone Development with AI and Automation
The design and optimization of expression vectors are complex, with multiple factors affecting the outcome. AI and automation technologies are reshaping the way clones are developed. By analyzing large datasets and integrating multiple omics approaches (e.g., single-cell multi-omics), AI can predict which clones will perform best at industrial scales—cutting down on the need for extensive empirical testing.
4. Streamlining the Clone Screening Process
Lonza has refined its clone screening process using cutting-edge technologies such as the Beacon™ Optofluidic System, a tool that allows for high-throughput clone isolation. The subsequent use of miniature bioreactors helps assess platform compatibility and scalability at an early stage, enabling faster identification of optimal clones and improving time-to-market.
5. Leveraging Design of Experiments (DoE) for Optimization
By incorporating Design of Experiments (DoE) methodologies into the early stages of vector and clone optimization, expression systems can be systematically tested for various genetic configurations. However, balancing the potential of DoE with the flexibility to explore new variables is key. Through empirical testing, it is possible to refine these systems iteratively and unlock new insights for optimizing titres and product quality.
6. Predicting Clone Performance with Omics Data
One of the exciting advancements in biologics manufacturing is the application of single-cell multi-omics technologies to predict clone performance. By analyzing molecular profiles at the single-cell level, it is possible to identify the most promising clones earlier in the development process. This approach could revolutionize how expression systems are optimized, by predicting long-term stability and product quality before large-scale production begins.
This article is posted at lonza.com
Please fill out the form to access the content
