Engineering cells has been the focus of immense effort over recent years, withbiomanufacturers using increasingly diverse cell lines to produce more complexbiologics. As with any manufacturing industry, growth of the sector has been accompanied by a drive to improve yields, reducecosts, accelerate development and strengthen supply chains. As a result, many manufacturershave developed better bioreactors and more automated workflows, yet there hasbeen little progress on one critical component of bioproduction – cell culturemedia – which has quietly become one of the biggest workflow bottlenecks. Historically,the industry has struggled to optimise media with the tools available, but thissituation is set to change with the arrival of high throughput, AI-drivenapproaches that allow rapid refinement of culture media targeted to the exactcell type, process and target product.

‍

Customizing media could address many of these limitations, but conventional development methods are resource intensive, expensive to apply routinely, and prohibitively slow, typically taking from two to four years. For many companies, this delay is not commercially viable, forcing them to buy generic off-the-shelf formulations that are perceived as ‘good enough’ but are not best suited to their cell lines or aims. At the heart of the challenge is the fact that the effects of changing an enormous number of possible ingredients, concentrations, and interactions are exceptionally difficult to predict. Metabolic models provide valuable insights but cannot capture every interaction, making media optimization as much a data science issue as a biological one

Changing the focus in this way acknowledges that it is not always necessary to understand every biological mechanism before making a change to a formulation. Instead, developers should take an empirical approach, systematically mapping the relationships between media inputs and measurable cellular outputs. However, the number of potential permutations of a formulation containing, for example, 60 ingredients is immense.

‍

Why media has historically received less attention

Traditional culture media – such as DMEM-F12, FBS or HPL – work reasonably well across many applications, enabling the growth of a variety of cell types. Many of these media are based on undefined or animal-derived components, which include a complex mixture of around 50 to 80 ingredients to drive cell growth, and can introduce batch-to-batch variability. Conventional media development separates these ingredients into functional groups, and treats their roles as relatively independent, which is often not the case in reality. Unfortunately, without an in-depth mechanistic understanding of how each ingredient interacts with other media components for a given cell type, media design has relied on trial and error experimentation, which is both slow and expensive. Progress has therefore been incremental, refining existing media compositions to achieve small gains.

As biomanufacturing has diversified, using progressively more novel cell lines to produce more complex biologics – including HEK293-derived products, viral vectors, bi- and tri-specific antibodies, antibody-drug conjugates and advanced therapies – generic media formulations have become much less effective, leading to inconsistent results, batch-to-batch differences and unpredictable process performance. Engineering principles are therefore increasingly being applied to what has traditionally been viewed as a biological challenge, looking at how cell growth, behaviour, health and stress affect product quality and titres.

‍

Entering a new era of media design

The ‘design, build, test, learn’ approach often favoured by engineers is increasingly being adopted for cell culture media development. Instead of making changes in isolation, this system puts the emphasis on using high throughput automation to evaluate many formulations in parallel, generating large, high quality experimental datasets. These datasets can then be analysed by AI and machine learning algorithms to identify patterns and inform the next round of experiments. Using Bayesian optimisation principles, each round of experiments gives machine learning models an ever-expanding knowledge base to work with, enabling more accurate decisions about the best way forward for testing potential new formulations.

This controlled, automation- and AI-driven approach allows the rapid development of custom, chemically defined cell culture media formulations developed specifically for the target bioprocesses. It provides more control over cell growth, yields and product quality, which is reflected in higher titres and faster cell growth, leading to shorter production cycles and lowering the cost of goods. Because the exact formulation is known – unlike most off-the-shelf media – batch-to-batch consistency is also improved, making troubleshooting easier and reducing the regulatory and supply chain risks associated with undefined or animal-derived components. Customised media can even reduce the number of by products, which can make purification easier and enhance product quality, offering more robust bioprocessing and further reducing downstream costs.

‍

The commercial benefits of better media

Media influences almost every aspect of bioprocess development and manufacturing performance, not just cell growth, and can no longer be viewed simply as a consumable as biologics become more complex and manufacturing economics become increasingly important. Better media can improve the commercial viability of biomanufacturing by increasing productivity, shortening production cycles, reducing manufacturing costs and accelerating development. Decisions made during media selection can influence performance throughout a product’s lifetime, helping manufacturers reach the market faster, run processes more reliably and consistently, reduce costs and build a stronger competitive position.

Culture media selection can be seen as a source of competitive differentiation, and choosing the right ingredients can strengthen manufacturing resilience. A proprietary formulation designed for a specific cell line and process gives manufacturers greater ownership and flexibility over a critical part of production, including the ability to select ingredients and potentially reduce dependence on a single media supplier. For CDMOs or companies with proprietary cell lines, a platform medium can be designed to perform across a defined family of cells or products, creating a more integrated and differentiated offering, rather than competing on the cell line or process technology alone. This makes media a strategic asset, not just a consumable.

Ultimately, media hasn't historically been neglected because it isn’t important to bioprocessing; it has been neglected because it has been extraordinarily difficult to optimise with the tools available. The advent of high throughput, AI-driven approaches changes the paradigm, allowing rapid optimisation of culture media to match the exact cell type, process and target product. This eliminates the hidden bottleneck in biomanufacturing, with huge potential to increase yields and reduce costs for manufacturers.

Talk to one of our scientists about media designed for the cell types you're working with.