Optimizing Cell Line Development for Next Generation Biologics

Transparent cell with biotechnology and cosmetic concept, 3d rendering. | Image Credit: © Vink Fan - stock.adobe.com

To rapidly develop the next generation of biological therapeutics, cell line developers need to build efficient platforms so that cell line development (CLD) does not become a bottleneck in the development pathway. To make allowance for the increasing array of novel antibody–based and non-antibody modalities, these platforms need to be flexible in terms of both molecule build and production method. Therefore, CLD unit operations need to be considered as modules, with predeveloped alternative modules that can be quickly and easily swapped in and out, plus defined application and predictable outcomes. Additionally, cell line developers need to work effectively across teams to leverage knowledge accumulated during lead candidate design and selection, while working with analytical scientists to apply a well-tuned analytical strategy.

When contracting out steps such as CLD, it is important to choose a contractor who is prepared to work collaboratively with the drug developer and apply a streamlined analytical approach. Working with an integrated contract development and manufacturing organization means that more complex programs can be tackled without significantly increasing CLD timelines or risk of failure.

Improved productivity feeds a growing sector

The rise of biological therapeutics looks set to continue, with recent forecasts predicting approximately a 7% compound annual growth rate over the next five years, a rate which would value the field at more than $600 billion by 2029 (1). The build-out of manufacturing volume has not kept pace with the increasing demand for biologics, but the long-touted “capacity crunch” has not yet occurred. Instead, there has been a maturing of the bioprocessing sector so that increases in productivity have been the focus, rather than building out volume (2).

These productivity improvements have come across the board, including the rise of single-use equipment (3), process intensification (4), and the adoption of platform processes. Mammalian cell line productivity in particular has seen some significant increases in titers. Average commercial monoclonal antibody (mAb) expression titers were 1.95 g/L in 2008; by 2020, this had increased to 3.5 g/L (2). However, these average figures don’t accurately reflect the true increases, with cell line developers for new products now routinely achieving mAbs titers above 5 g/L.

Trend toward more complex biologics

In parallel to these leaps in productivity, there has also been a trend toward more complex products. The production of bispecific antibodies, multiactive fusion proteins, nanoparticles, and other complex biologics is now possible because of advanced protein engineering technology and bioprocessing advances. These advances have made it feasible to express complex modalities, hitherto considered difficult to express, in economically viable quantities; however, challenges remain.

Cell line developers are required to generate manufacturing cell lines for an ever-widening pipeline of new modalities, with shortened timelines and without compromising on quality. Many of these molecules may only be required in small quantities with higher potency, to serve smaller patient populations for orphan and rare diseases or as ancillary materials for cell and gene therapy products. Therefore, there is a requirement for efficient, productive processes to support the economics of smaller batch sizes.

Additionally, many of these new modalities will come with unique critical quality attributes (CQAs) and more complex impurity profiles, challenging traditional mAb upstream and downstream platforms. These challenges can be alleviated by generating and selecting high-quality production clones. The right clone will generate more of the desired product, with CQAs that conform to the quality target product profile, and fewer of the impurities that might complicate downstream removal. Cell line developers need to balance the tension between the efficiency of a platform that delivers highly productive clones against the nuances of more complex proteins. Building flexible CLD platforms, where platform unit operation can be modified or swapped for more bespoke versions, while maintaining the backbone of the platform process is one way in which the competing needs of speed, quality, and cost can be balanced.

It starts with vectors

Having a robust expression vector, adaptable to a wide range of modalities, is critical to having a flexible CLD platform. Most vectors used in CLD have been optimized around mAbs, employing strong constitutive promoters like cytomegalovirus or elongation factor 1 alpha, with enhancer regions to maximize rates of gene transcription. Other technologies include elements to modify chromatin and gene accessibility to transcription complexes (such as ubiquitous chromatin-opening elements and matrix attachment regions) or utilize transposase mediated integration (such as Atum Bio’s Leap-in, Addgene’s PiggyBac, and Sleeping Beauty) to semitarget efficient gene integration. Premium vectors, such as ProteoNic’s 2G-UNic and Abzena’s AbZelectPRO , combine a powerful promoter to maximize transcription with optimized untranslated regions to generate more protein from each message copy, making them well-suited for mAb production as well as proteins that are more difficult to express.

All of these technologies are focused on maximizing gene expression, but when looking at more complex multichain proteins, it is important to consider flexible vectors and transfection strategies that can account for potential differential chain expression. To optimize productivity, expression of each chain needs to be balanced. Balancing the expression of each chain ensures they are produced in the correct ratios, maximizing the amount of desirable product and minimizing product-related impurities (PRI).

Balancing chain expression can be done using promoters of varying strength or through altering ratios of gene copy number. In either case, for any given pair (or more) of subunits, the optimal ratios or promoter combinations will need to be determined experimentally. To optimize quickly and without extending timelines, a vector system is required that allows rapid cloning of genes and a flexible process for rapidly generating multiple stable pools in parallel. This, aligned with a strong analytical strategy, can provide empirical data for analysis, which allows quick decisions without elongating timelines.

Flexible analytical platforms

Throughout a flexible CLD process, cell line developers will need the support of experienced and capable analytical scientists. Complex biological modalities will require a wide range of analytical techniques to distinguish between the desired product and mispaired variants or other PRI. Standard mAb approaches may or may not be applicable, as many of these impurities may be difficult to resolve with a single analytical technique. It is possible to build a comprehensive panel of assays capable of resolving the impurities through many techniques, for example, through mass, charge, or hydrophobicity differences, but employing all these assays routinely adds to costs and timelines unnecessarily.

Instead, it is better to have the support of an analytical team that has the capability to build and use these methods, and the know-how and experience to understand which are the best techniques for any given molecule. Employing an analytical platform, such as LabZient from Abzena, is beneficial in streamlining the assessment of complex molecules by combining well-established laboratory methods with predictive in silico tools. The right suite of predeveloped assays can then be employed, depending on the characteristics of the product. Having analytical scientists working alongside CLD minimizes time in sample transfer, data analysis, and the generation of high-quality data for better decisions.

The power of pools

In a typical CLD platform, the first step after vector construction is the generation of stable pools. This is often the first chance to get a look at how well the product is expressed in the host cell line. For platform mAbs this is fairly predictable, but for complex products this is an opportunity to get a good gauge of expression levels which, with a well-understood platform, can be extrapolated to likely clone productivities. This also gives a good idea of whether the product is forming properly, is active, and what major impurities may need to be addressed.

As stable pools are relatively straightforward to generate, multiple candidates can be run in parallel and assessed for expression levels, CQAs, and platform fit so that CLD can start before the lead candidate is nominated, creating overlap and accelerating the path to the clinic. Furthermore, material produced from stable pools can be used for immunogenicity evaluation, non-GMP in-vivo efficacy, and pharmacokinetics (PK) studies, as well as to drive formulation development, support analytical method development and, if needed, begin unique selling point and direct support professional development. Taking these activities off the critical path for novel therapeutic modalities is key to keeping timelines short.

Rapid cloning strategies

From the stable pools, the next step is to isolate and screen monoclonal populations. International Council for Harmonisation guidance requires cell lines used for manufacturing to be derived from a single cell (5). To streamline the cloning process and keep timelines down, many cell line developers have opted for advanced cloning methods such as fluorescence-activated cell sorting, cell printing, optofluidics, and microfluidics. Allied with modern imaging techniques and capable of capturing images of single cells and documenting monoclonality, these methods can provide a very high statistical probability of monoclonality in a single step.

Screening for upstream platform fit

When considering the clone selection system, it is important to choose one relevant to the production system of choice. As Porter et al. established, the ranking of clones for productivity can change depending on the vessel and culture conditions applied (6). Most cell line developers and upstream scientists understand that the culture conditions also significantly influence product quality. Therefore, it is important to have a good scale-down model for the intended production method, such as the Ambr15 and Ambr 250 microbioreactor systems. These allow for the ranking of clones based on productivity, growth profile and, if connected to a suitable bioanalyzer, the metabolite profiles. They also provide sufficient material for purification and analysis of the product.

Being able to assess product quality for each candidate clone is vital. Ensuring that clones are selected on the basis of both productivity and product quality, especially for bispecific and more complex molecules, streamlines and derisks process development activities. This also includes having a good measure on the protein impurities in the sample. Selecting clones where PRIs are minimized significantly reduces the need to develop new downstream steps. Having good characterization data on clone performance, growth, productivity, metabolite profile, product quality, and impurity profile, in a relevant scale-down model, is critical to a successful development program.

Intensified processes

Most biologics are produced in a fed-batch process, and this has significant advantages in terms of applying and realizing the benefits of a platform approach. For most products this is still the most appropriate production method to rapidly generate investigational new drug enabling material and data.

However, there is a drive toward process intensification (PI) either to improve production efficiency or as a result of product stability in cell culture. There is also a strong environmental argument for PI, with the basic principle of developing smaller, cleaner, and more energy-efficient processes. PI, therefore, not only has benefits in reducing costs through efficiency gains, but also aligns with environmental goals and corporate environmental, social, and governance policies.

Ideally, a flexible platform will be able to select clones that not only perform well in fed-batch mode for early development, but are also able to cope with the demands of an intensified fed-batch or a continuous perfusion process. To allow for this, additional consideration needs to be made for the clone’s growth attributes, such as viable cell density, cell viability in extended culture, metabolite profile, and clone stability. A robust clone that grows to higher cell density in a platform medium and feed, produces little in the way of toxic metabolites such as ammonium, and has good stability will be easier to adapt to intensified processes.

Building flexibility into each of the steps of a CLD platform allows cell line developers to accommodate a wide range of different biotherapeutic modalities from mAbs and bispecific antibodies to vaccines and protein nanoparticles.

References

1. Mordor Intelligence, Biologics Market Size & Share Analysis—Growth Trends and Forecasts (2024-2029); Mordor Intelligence, April 2023.
2. Langer, E.S. and Rader, R.A. Total Global Capacity Finally Shows Improved Productivity. BioProcess Int. online, May 21, 2021.
3. Morrow Jr., K.J. and Langer, E.S. Rise of Single-Use Bioprocessing Technologies: Dominating Most R&D and Clinical Manufacture. Am. Pharm. Rev. online, Feb. 27, 2020.
4. Boodhoo, K.V.K.; Flickinger, M.C.; Woodley, J.M; and Emanuelsson, E.A.C. Bioprocess Intensification: A Route to Efficient and Sustainable Biocatalytic Transformations for the Future. Chem. Eng. Process. 2022, 172, 109793. DOI: 10.1016/j.cep.2022.108793.
5. ICH, Q5D Derivation and Characterisation of Cell Substrates Used for Production of Biotechnological/Biological Products, Step 5 version (1998).
6. Porter, A.J.; Dickson, A.J.; and Racher, A.J. Strategies for Selecting Recombinant CHO Cell Lines for cGMP Manufacturing: Realizing the Potential in Bioreactors. Biotechnol. Prog. 2010, 26 (5) 1446–1454. DOI: 10.1002/btpr.442.