When creating a manufacturing solution for your product, there are a finite number of alternatives for production strain development:
1.1. In-house expertise:
This resource tends to be more readily available in larger companies. However, the breadth of experience available may be biased towards particular expression technologies traditionally available in the company. Therefore, the depth of knowledge for some systems is likely to be limited compared to others. Consequently, the capabilities needed to resolve challenges adequately may not be available in-house.
1.2. Contract Development Manufacturing Organisations (CDMOs)
Typically, CDMOs have extensive experience in the expression systems they offer. The expression systems used tend to provide “off-the-shelf” solutions to a range of common expression challenges. However, if issues are encountered that require a bespoke solution for your product, resolving them may be impractical, unaffordable, or risky, especially if extensive genome engineering is required for the CDMO’s main expression chassis. Furthermore, suboptimal solutions may transfer problems to other aspects of their end-to-end service offering, e.g. problems that an alternative production strain could have resolved are likely to be passed to downstream processing (DSP) or regulatory support functions.
1.3. Synthetic Biology Companies
They tend to offer platform-specific solutions, taking a collaborative R&D approach to expression systems amenable to their high-throughput strain engineering platforms, which typically leverage automation and AI. While this is often faster than traditional low-throughput methods initially used to develop the established expression systems already used for biomanufacturing, e.g. E. coli, yeasts and Chinese Hamster Ovary (CHO) cells, the approach is essentially the same process of linear research, making stepwise improvements on the original genomic chassis until an adequate solution is achieved. Consequently, it is unlikely to be successful when tackling the more complex engineering challenges required to take these systems to the next level.
1.4. Full genome optimisation using Quantitative Trait Loci (QTL) technology:
This approach applies natural evolution through breeding, which currently focuses on the model eukaryotic host, baker’s yeast (Saccharomyces cerevisiae), to optimise the genome for multiple parameters simultaneously. It combines proven technologies for strain development, such as expression cassette engineering and targeted genome engineering, with a novel technology for powerful evolutionary optimisation of the entire genome. This full genome optimisation technology is uniquely available for Phenotypeca’s proprietary baker’s yeast platform, which was selected due to the advanced genomic knowledge of this microbe, being the model eukaryotic system for many decades, since the first publication of its genome in 1996 (Goffeau et al., “Life with 6000 genes”). However, knowledge of baker’s yeast QTLs is directly applicable to homologous eukaryotic systems used for biomanufacturing, e.g. Pichia pastoris and CHO. Most importantly, full genome optimisation enables “multi-parameter” optimisation for critical strain performance criteria that can improve manufacturing yield, quality, and economics. This is because the manufacturing host profoundly impacts the entire manufacturing process, impacting both upstream and downstream processes. Until recently, many critical bioprocess parameters have been beyond the scope of improvement.
Patent protectable optimised strains
Furthermore, this approach provides multiple genomic solutions for each phenotype being optimised, which is impossible through sequential engineering of a single chassis. Importantly, the availability of multiple strains with different genetic solutions for each bioprocessing improvement allows the most important changes to be identified from the many changes made during each iterative round of improvement. This enables IP to be filed on the improvements because the underlying biological processes causing the improvement can be identified.
Commercial viability requires an optimised solutions for some products
While traditional strain engineering has historically provided adequate solutions for many bioprocesses, this step-by-step approach does not provide extensive genome optimisation solutions compared to QTL technology, and it is ineffective when multiple genes are responsible for the underlying biology. For some products, an optimum solution is essential; without this, the product is unlikely to reach the market or be commercially viable. This may be for commercial reasons related to the nature of the target market, such as a requirement for low cost of goods sold (CoGs) and hence high titre in manufacture. Alternatively, it may be for technical reasons, such as a difficult-to-express protein with challenging folding or other post-translational requirements, or the process may need to be exceptionally robust and sustainable for operation in lower and middle-income countries.
Product specification key to avoiding bias and vested interest
A detailed specification, including critical quality attributes of the final product, is a prerequisite for making an informed selection of the most appropriate expression system. Further consideration must be given to the objectivity of the people consulted for critical decisions. CDMOs may have a vested interest in promoting their proprietary or preferred expression systems. The specific skills and experience of CDMO employees may create a bias towards selecting one particular expression system over another. The ‘not invented here’ viewpoint may also limit the use of alternative expression systems introduced to manufacture your biologic, even if they are demonstrably better for a particular product or process. They may feel their position and/or prospects are challenged by an approach outside their existing skillset, or they can’t provide the required level of end-to-end service for an alternative expression system within their existing facilities.
Genetic engineering suffers from engrained bias
Engineering decisions made during strain development programs are also likely to depend on rational selection of genome targets for improvements, which are typically tainted by past experience with a different product and may be entirely inappropriate for your product. In contrast, full genome optimisation methods using QTL technology remove this human bias by allowing nature to find the regions of the genome giving the most beneficial phenotypic improvements for each different product. These genomic loci responsible for the improvements are then “read” during QTL analysis to identify and prioritise the critical regions for the optimal production of your protein.
Ironically, opting for established tried-and-tested expression platforms and strain improvement techniques will frequently be the “high-risk” choice for future commercialisation in a marketplace where competitor products have undertaken full genome optimisation, rather than the risk-averse choice.
