In a collaboration with Xiaozhou (David) Luo at SIAT, work led by Will Shaw and my group has yielded a yeast-based biosensor and workflow for screening microbially-produced Δ9-tetrahydrocannbinol, better known as THC. This work has been published now in Nature Communications.
Producing cannabinoids from microbial fermentation is an exciting way to create these promising therapeutic and, let’s be honest, recreational compounds. The number of companies being founded for bioengineering of cannabinoids highlights the lucrative market potential. BUT, metabolic engineering development requires a whole load of screening to identify high producers and this is currently limited to methods such as LC-MS. A cannabinoid biosensor would open the door to quicker, easier, and cheaper development cycles.
We decided to have a go a creating a cannabinoid biosensor by building on our previous work (and the work of many others) for porting G protein-coupled receptors (GPCRs) into yeast and coupling this to a fluorescent protein output. We ported 5 different human GPCRs known to interact with cannabinoids, showing that the cannabinoid type 2 receptor (CB2R) was functional in yeast and had great biosensing characteristics straight out of the box, with a high dynamic and operational range. Having succeeded at that we were interested to see whether our yeast GPCR-based biosensor could be used for reporting relative yields of cannabinoids from microbial production. This turned out to be a challenge!
Cannabinoids have diverse effects on the CB2 receptor. For example, THC is an agonist, while conversely, CBD is an antagonist. Both of which are present in high amounts in plant extract, obscuring precise quantification of either molecule. This picture is simplified in microbial systems, as we can limit cannabinoid diversity by choosing to omit the various synthases that are present in cannabis. However, extracted samples will still be a complex mixture of a few competing by-products. THC was the clear choice to focus on as we expected this very potent compound would “punch through” the effects of other by-products, and so we developed a medium-throughput workflow for screening this molecule from yeast producer strains.
We optimised the yeast biosensor for use on a plate reader, and the team at SIAT, led by the talented Yunfeng Zhang, developed a workflow for preparing samples from yeast and screened > 100 strains they had generated to see whether we could pick out the high producers. While the results of the large screen were quite noisy due to the complexity of the extracted samples, we were able to pick out top producers, showing its usefulness for enriching producer libraries.
We hope this biosensor workflow will be useful to teams trying to produce THC from microbial fermentation and demonstrates yet another example in which GPCRs are an amazing source of biosensing elements.
A big thanks to all involved, including Graham Ladds at Cambridge University and Mo Khalil at Boston University for their support kick-starting and then wrapping up this project, respectively, and to UKRI (EPSRC and BBRSC) for the funding.
Shaw, W. M., Zhang, Y., Lu, X., Khalil, A. S., Ladds, G., Luo, X., & Ellis, T. (2022). Screening microbially produced Δ9-tetrahydrocannabinol using a yeast biosensor workflow. Nature Communications - https://doi.org/10.1038/s41467-022-33207-x
Comments