In order to build platform chassis strains, a greater understanding of cellular metabolism is required alongside improved engineering tools. Three new enzyme-constrained genome-scale metabolic models (ecModels) were generated for the three yeast species (S. cerevisiae, K. marxianus and Y. lipolytica). They were used to predict gene targets for metabolic rewiring, aiming to enhance production of octanoic acid, docosanol and naringenin. Metabolic rewiring strategies to promote enhanced accumulation of metabolic precursors such as L-phenylalanine, 2-phenylethanol and acetyl-CoA was also studied using the ecModels. These models have been made available through the open access repository GitHub (
https://github.com/SysBioChalmers/ecModels(se abrirá en una nueva ventana)) and are already being used by a number of external research groups. In addition, a user-friendly interface for prediction of engineering targets has been developed which opens up new exploitation routes for industry. This is accessible via the Caffeine project (
https://caffeine.dd-decaf.eu/(se abrirá en una nueva ventana)).
To enable the ‘build’ part of the ‘design-build-test-learn’ cycle, genome engineering tools were developed for all three yeasts. Editing tools with multiple CRISPR Cas endonucleases and multiplexing genome editing were implemented. These toolkits are available to the scientific community via Addgene and multiple groups and teams are now using them. Using the metabolic engineering strategies suggested by the ecModels, and the improved expanded synthetic biology toolbox, we were able to optimize the supply of precursors to, and flux through, the shikimate pathway. Therefore, a major outcome of the project is a collection of chassis platform strains that synthesize higher levels of aromatic amino acids. These strains are of interest to the flavour, nutraceutical and personal care sectors. It will now be possible to add further pathways to produce molecules that have commercial potential such as Flavonoids, Stilbenes, Coumarins and Lignans. A similar engineering approach was used to construct chassis platform strains optimized for the production of fatty acids and long chain lipids. By applying the yeast chassis platform strains, several prototypes were developed and tested under industrial-scale conditions. While the results are promising, more optimization is required, e.g. industrial medium, scale-up and preliminary downstream processing, before the process can be moved to full-scale production. Nevertheless, some external collaborations have already been initiated with industry partners to further develop and commercially exploit some of these results with platform strains in both the aromatic and oleochemical space.
Scientific publications formed the backbone of the dissemination strategy for many of these results. CHASSY published a total of 33 peer reviewed scientific articles over the course of the project, including publications in high impact journals such as Nature Energy, Nature Catalysis, Nature Communications, Current Biology and Nucleic Acids Research. These results have also been disseminated at over 60 events over the course of the project. There were also many industry-focused events and extensive dissemination to the SME sector.
