In the last decade, many governments have relaxed their cannabis policies and many more governments are grappling with the idea of making changes. Prof. Dr. Oliver Kayser and myself reside in Dortmund, Germany, where this year cannabis was made fully legal for both medicinal and recreational use. The remaining authors reside in the Netherlands, where cannabis was famously decriminalized in 1976. Some might find this to be exciting, others might find it worrisome. It is certain that there are many conversations around health, policy, revenue, occupational safety, consumer safety, driving, harm reduction, and many more topics need to take place as legal procurement continues to become available. One such discussion is the motivation of our paper “Integrated -Omics of Saccharomyces cerevisiae CENPK2-1C Reveals Pleiotropic Drug Resistance and Lipidomic Adaptations to Cannabidiol”, where we examine broadly the impact that cannabidiol has on yeast cells.
While modern cannabis cultivation is feat of decades of agricultural science, it’s not without its vulnerabilities. Cannabis plants are susceptible to heavy metal1 accumulation and mycotoxins2. Cleaning up the cannabinoids from cannabis plants is also challenging and requires harsh solvents which can leave behind residual solvents3. Cannabis cultivation is also energetically intensive, requires great amounts of resources like water and fertilizer, and requires land4. That’s why many are working to develop less expensive and more efficient processes to produce cannabinoids using microbial cell factories.
Microbial cell factories have been made for many products, from insulin, to biofuels, to vitamins5. It’s been done already for cannabinoids and already boasts a reasonable titer of over half a gram of cannabinoid produced in 120 hours6. But systems biology approaches tell us we can optimize, and we’ll need to because we’ll have to beat the market price of traditional cannabinoid cultivation before commercial interests will capitalize on microbial cell factory derived cannabinoids. Of course, there are limitations to how powerfully a model can predict a biological system, but enzyme constraint modeling has predicted that we should see titers nearly double the highest titer reported to date, 300 mg/L/40 hours7, which we can extrapolate to 900 mg over the same duration as the highest reported titer.
We zeroed in on this apparent gap for our study and focused on the fundamental question of, can cannabinoids induce stress on the yeast cells that we want to use to make them? It’s long been assumed that because cannabinoids have low cytotoxicity to yeast that there would be minimal, if any, off-target effects, but we chose to examine that assumption more closely. So, we went back to fundamentals and used a strain of yeast typically used for metabolic engineering called Saccharomyces cerevisiae CENPK2-1C, and without any kind of genetic recombination, we wanted to see what happens when you expose it to cannabinoids. We chose CBD because it's easily accessible and relatively affordable compared to other cannabinoids like CBGA or Δ9-THC. Our first experiments were very basic, we just added CBD to our liquid cell cultures at 0.5 mM, and the result was very surprising. We observed a noticeable post-diauxic shift in our treated cells, where our control cells were already in stationary phase. A post-diauxic shift is an adaptive shift in metabolism from one substrate to another, which can lead to a bump in a growth curve of a microbial cell culture. It was the opposite of what we were expecting, which was for the cell to waste energy metabolizing and secreting the CBD, and therefore to have a lower growth rate. What we saw was not only that cannabidiol treated cells grow to at least a similar density to the negative controls, but they can exceed the density after this post-diauxic shift occurs.
We got to work characterizing this post-diauxic shift. We repeated the experiment and sampled our cells as this shift was dynamically occurring. We simultaneously sequenced the transcriptome of the cells and analyzed the metabolic profiles using broad-non targeted normal phase chromatography and LC/qTOF. And what we found was very consistent with a post-diauxic shift phenotype- increased expression of rRNA’s and initiation factors, mitochondrial upregulation, pathway enrichment of aminoacyl-tRNA biosynthesis, and increased flux to amino acid pathways, among others. These results confirmed that the cells were shifting their central metabolites and growing. We were surprised to find that there were substantial lipidomic changes to the CBD treated cells, where we found a significant increase in the amount of stearic acid, 2-Hydroxystearic acid, 9-Hydroxy-12-octadecenoic acid, and the most surprising of all- 1-docosanoyl-glycero-3-phosphoserine, a 22 carbon monoacyl glycerophosphatidylserine (PS(22:0/0:0)), which has otherwise not yet been reported in yeast. Yeast membranes fatty acids are typically 16 or 18 carbons long, and phosphoserines are present only as diglycerophospholipids8. What was most surprising was that 1-docosanoyl-glycero-3-phosphoserine was entirely absent from all untreated cell pellets that we extracted, both positive and negative control.
We can’t yet be sure what substrate the cells are shifting towards metabolizing in this post-diauxic shift, but the pieces we have found so far are that 1) they certainly are shifting and 2) there are several lipidomic perturbations in the cells. At the level of primary metabolism, our current hypothesis is that lipases are breaking down a very long chain diacylphosphoserine until only the monoacyl 1-docosanoyl-glycero-3-phosphoserine remains, but we don’t have enough evidence to make that conclusion yet, and not enough is known yet about yeast lipid metabolism to incorporate our findings into a pathway model. But there was another important finding of our study, PDR5 expression is a causal factor in the presence of this post-diauxic shift. We confirmed that not only through its overrepresentation in our transcriptomic data, but also through a knockout and recovery assay. And we’ve shown that this phenotype is not seen in the knockouts, and it comes back when we put PDR5 back in with a constitutive promoter.
More systems modeling and metabolic engineering is needed at this point to show how these results will impact cannabinoid titers in yeast cell factories. It needs to be determined through experimentation how PDR5 deletion and overexpression affect cannabinoid productivity in a cannabinoid biosynthetic YCF strain. It’s also an interesting question to explore what the impact of these very long chain lipids are on the system in a cannabinoid-YCF, which would, in theory, be a sponge for necessary cannabinoid precursors like acetyl-coA and malonyl-coA, not to mention ATP. The question of dynamic lipid membrane remodeling has been reviewed before9 and it seems like compounds that we thought were inert can impact the metabolomes of a cell quite dramatically. We hope that this paper adds to that body of evidence and shows that it’s worth it to take a deeper look into how xenobiotics and yeast cells interact.
Read the Open Access article here.
- Golia, E. E. et al. Investigating the potential of heavy metal accumulation from hemp. The use of industrial hemp (Cannabis Sativa L.) for phytoremediation of heavily and moderated polluted soils. Sustainable Chemistry and Pharmacy 31, 100961 (2023).
- Buchicchio, L. et al. Investigation of aflatoxin and ochratoxin A contamination of seized cannabis and cannabis resin samples. Mycotoxin Res 38, 71–78 (2022).
- Clifford, B. et al. Utilizing GC-MS and GC Instrumentation for Residual Solvents in Cannabis and Hemp. in Cannabis Laboratory Fundamentals (ed. Opie, S. R.) 191–222 (Springer International Publishing, Cham, 2021). doi:10.1007/978-3-030-62716-4_10.
- Zheng, Z., Fiddes, K. & Yang, L. A narrative review on environmental impacts of cannabis cultivation. J Cannabis Res 3, 35 (2021).
- Gohil, N., Bhattacharjee, G. & Singh, V. An introduction to microbial cell factories for production of biomolecules. in Microbial Cell Factories Engineering for Production of Biomolecules 1–19 (Elsevier, 2021). doi:10.1016/B978-0-12-821477-0.00021-0.
- Zhang, Y. et al. Development of an efficient yeast platform for cannabigerolic acid biosynthesis. Metabolic Engineering 80, 232–240 (2023).
- Thomas, F., Schmidt, C. & Kayser, O. Bioengineering studies and pathway modeling of the heterologous biosynthesis of tetrahydrocannabinolic acid in yeast. Appl Microbiol Biotechnol 104, 9551–9563 (2020).
- Klose, C. et al. Flexibility of a Eukaryotic Lipidome – Insights from Yeast Lipidomics. PLoS ONE 7, e35063 (2012).
- Ferraz, L., Sauer, M., Maria João Sousa, Sousa, M. J. & Branduardi, P. The Plasma Membrane at the Cornerstone Between Flexibility and Adaptability: Implications for Saccharomyces cerevisiae as a Cell Factory. Frontiers in Microbiology 12, 715891 (2021).