What does a cell culture media development project look like?
Cell culture media can influence cell growth, product quality, and manufacturing performance, but there is no one-size-fits-all: the best-performing formulation is specific to cell type and process. Media development therefore needs to start with a clear understanding of the cells, the process, and the desired outcome. In this Q&A, Soraya Padilla, R&D Lead at Multus Biotechnology, explains how her team approaches media development projects, from establishing the cell culture process and agreeing project targets to high-throughput experimentation and validation.
When a client starts a new media development project with Multus, what do you need to understand first?
We start by defining four things: the needs of their cellular model, their process constraints, what they want to achieve, and the experimental strategy needed to do it. For the cellular model, we assess factors such as the specific cell line or clone, its biological state and culture history, and how it behaves and metabolizes in culture. We then identify which parts of the client's process need to remain representative, and what can be adapted for high-throughput screening without changing the underlying biology. From there, we agree on the endpoints that will define success by combining core measures such as growth and viability with application-specific measures of function or product quality. Those decisions then determine what we screen, which analytics we use, and which measurements are better reserved for validation.
How long does a media development project typically take?
This is one of the most common questions we get asked, and the answer depends mainly on where the client is starting from, and how complex the project is. Some clients already have a medium that performs reasonably well and want to refine certain aspects of it through MediOP™ Boost, while others need something developed from scratch through MediOP™ De Novo, which naturally involves more experimental work. The timeline also depends on factors such as the biological cycle time, maturity of the starting process, analytical requirements, and the depth of development and validation required. Based on the types of projects we work on, development can take approximately 3 to 12 months.
How closely do you need to replicate the client's existing process?
We aim to keep the parts of the client's existing process that are important to the biology and adapt the workflow where needed for high-throughput screening. Depending on the system, the retained parameters could include factors such as inoculation density, feeding strategy, culture duration, pH, or osmolality. We then check that any scale-down or adaptation still gives us a representative model, so the data and formulations we generate remain relevant to the wider bioprocess and can transfer back into the client's own workflow.
How does the development approach change for different cell types?
The overall process stays the same – allowing us to follow the same onboarding, screening, and validation framework across different cell types – but the experimental strategy changes with the biological complexity of the cells. More robust cell systems can allow a wider range of measurements at high throughput, whereas more complex systems require us to be more selective about what we assess during screening. Stem cells are a good example of the latter, because proliferation and viability can be measured across hundreds of formulations, but deeper assessment of phenotype, differentiation potential, or function is better suited to validation.
Which metrics are most important when assessing a media formulation?
We look at two types of metrics: core indicators of cell performance – such as growth, morphology, viability, and metabolic behavior – and application-specific measures of function or product quality. For example, we may look at antibody titers and relevant quality attributes for cells producing monoclonal antibodies, whereas measures such as yield and full-to-empty particle ratio are crucial for viral particle production. With stem cells, we need to understand whether they retain characteristics such as stemness or differentiation potential. That is why growth by itself does not tell you whether a medium is successful. A stem cell might grow very well, but if it loses the characteristics needed for a cell therapy application, then the formulation is not delivering what the client needs. We therefore agree which endpoints matter most for the application before designing our screening strategy.
What happens once the project targets have been agreed?
Once the objectives are clear, we move into high-throughput experimentation. A campaign can involve around 400 to 1,000 formulations at a time, all tested against the metrics we agreed with the client at the beginning of the project. Each campaign generates a large experimental dataset, which we use to train our algorithms and guide the next set of formulations. We test one set, learn from the results, and then use those findings to inform the next round. Robotics make this scale of experimentation practical, allowing us to test hundreds of formulations consistently and generate high-quality datasets. Clients are sometimes surprised by how many strong candidates emerge from this stage, with dozens of formulations potentially performing well against the project objectives. Those candidates then move into validation, where we carry out more detailed testing and narrow the pool to a smaller number of formulations for the client to take forward.
Where does biological expertise fit into a process that also uses artificial intelligence and automation?
Our biological expertise is crucial throughout the project, but it plays a different role at each stage. During onboarding, we use our understanding to establish a representative model, define meaningful metrics, and identify potential sources of variability before generating large datasets. During high-throughput experimentation, automation and our AI models allow us to efficiently test many formulations and guide the next steps, but the results still require human interpretation to understand whether a response is meaningful and consistent with the expected cellular behavior. During validation, we look more deeply at whether candidate formulations maintain the required phenotype, function, and performance over longer-term, more process-relevant conditions.
Why does customizing media make such a difference between cells that seem similar?
Even closely related cells can respond differently to the same medium because their physiological, metabolic, and nutritional requirements can vary. Two clones from the same cell background may produce the same type of product, but differences in genetics, adaptation history, or metabolic state can affect how they respond to a formulation. Depending on the type of media, a formulation might contain around 30 components at the lower end, and closer to 100 in a more complex system, creating a huge number of possible combinations and interactions. When I was working in academia, I might have tested five ingredients at a few concentrations, whereas high-throughput automation and data-driven experimental design allow us to explore a much broader formulation space. A traditional approach might screen those clones against a fixed medium to identify the best performers. Custom media development lets us approach the problem from the other direction too, by asking which formulation best suits the selected cells and enhances the performance we need from them.
What happens during validation, and what does the client receive at the end of the project?
We take the strongest candidate formulations from high-throughput screening into more application-relevant conditions, where we carry out a more comprehensive characterization to confirm that they meet the project targets, including retaining the required properties over longer-term culture. We also assess how well the formulations translate towards manufacturing scale. Having several strong candidates provides flexibility if, for example, an ingredient presents a scale-up challenge and needs to be substituted. By the end of the project, we narrow the pool to a smaller number of formulations for the client to test in their own process. The key question is whether they deliver the performance and quality attributes required for the application, whether that means antibody production, maintaining stemness, or another critical cellular property. That is what ultimately determines whether the media is suitable for the client's process.
To learn more about Multus' approach to cell culture media development, explore our media design services.