Engineering and adaptive laboratory evolution of methanol assimilation in Saccharomyces cerevisiae

Our dependence on fossil fuels for energy, commodity chemicals, and food has led to a rapid rise in global temperature and climate change. Since 2005, focus has shifted to move towards a bioeconomy, where renewable biological resources can be used to derive the products that would normally come from fossil resources, as well as new products that can only be produced biotechnologically. A promising solution to produce food, fuels, chemicals, and pharmaceuticals are microbial fermentations. Microbial fermentations are currently used industrially, for example, for production of bioethanol from Clostridium autoethanogenum or the antimalarial drug artemisinin from Saccharomyces cerevisiae. Microorganisms can be engineered to produce heterologous compounds or optimised to exploit their native qualities through metabolic engineering. Metabolic engineering uses genetic tools to optimise carbon fluxes inside a microorganism, which can also be aided by evolution when non-obvious solutions are needed. Successful metabolic engineering of the model organism S. cerevisiae has resulted in the production of industrially-desirable compounds. However, a problem with microbial fermentations is that they normally use sugars as feedstock, which require arable land that competes with food production. Instead of sugars, research efforts have focused on using renewable C1 compounds such as CO₂, methane and methanol that can be derived from organic waste and natural gas. This research describes metabolic engineering efforts to engineer methanol assimilation in the well-characterised and robust yeast S. cerevisiae.

Firstly, to identify the best host strain for synthetic methylotrophy the two most widely used S. cerevisiae laboratory strains, S288C and CEN.PK, were tested for growth and transcriptional response to methanol. CEN.PK showed better growth than BY4741, a strain derivative of S288C in both solid and liquid methanol media. Interestingly, a methanol-specific transcriptional response was observed in CEN.PK, which shares some similarities to the methylotrophic yeast Pichia pastoris as genes involved in alcohol oxidation, peroxisome proliferation and glutathione metabolism were up-regulated. Moreover, the gene MIG3, coding for a transcription factor, was found to be up-regulated on methanol medium in CEN.PK but not in BY4741. Over-expression of MIG3 in BY4741 resulted in better growth on methanol media, suggesting MIG3 is involved in regulating gene expression in response to non-fermentable carbon sources.

After identifying CEN.PK as the best host strain for synthetic methylotrophy, three pathways were designed: a xylulose monophosphate pathway (XuMP) based on P. pastoris, with enzymes targeted to the peroxisome; a 'hybrid' xylulose monophosphate pathway targeted to the cytosol; and a ribulose monophosphate pathway (RuMP) based on methylotrophic bacteria. Functionality of these pathways in S. cerevisiae was tested by growing the yeast in media with methanol as the sole carbon source. Furthermore, ¹³C-methanol tracer analysis was performed to assess methanol assimilation through central carbon metabolism.

¹³C-methanol tracer analysis revealed for the first time a native capacity for methanol assimilation in S. cerevisiae as ¹³C-ethanol and universally labelled ¹³C-intracellular metabolites were detected. No previous literature existed for native methanol assimilation in S. cerevisiae. To characterise and optimise this previously undiscovered capacity, Adaptive Laboratory Evolution was performed for 230 generations until a biomass increase was observed in liquid minimal medium with 2 % methanol and 0.1 % yeast extract. Whole-genome sequencing of the evolved lineages revealed single-point mutations in YGR067C coding for an uncharacterised transcription factor. The mutations from the three independent lineages all coded for premature stop codons causing truncations to the Y gr067c protein. Reverse engineering one of these mutations in the parental strain recapitulated the improved growth phenotype seen in the evolved strain, with a final biomass increase of 44 % compared to the parental strain.

To further characterise the native capacity for methanol assimilation, a systems biology approach was applied, where the abundance of 13C-labelled metabolites, transcripts and proteins was analysed. ¹³Cmethanol tracer analysis showed the reconstructed evolved strain had up to a 5-fold increase in ¹³Clabelled intracellular metabolites compared to the parental strain. The reconstructed evolved strain also had decreased production of ¹³C-ethanol, which was hypothesised to be a consequence of carbon redirected into biomass constituents. The transcript and protein abundance showed the TCA cycle and glyoxylate cycle were downregulated, suggesting a net glycolytic flux is occurring in the reconstructed EC strain. Finally, high ¹³C-labelling of dihydroxyacetone phosphate alongside higher transcript abundance of DAK2, and higher protein abundance of Tallp and Fbalp suggest an incomplete version of a P. pastoris-Iike methanol assimilation pathway is occurring in S. cerevisiae.

In summary, this work demonstrates for the first time a previously undiscovered capacity for methanol assimilation in S. cerevisiae as well as its optimisation through laboratory evolution and partial characterisation. Together, these results contribute to our understanding of S. cerevisiae and will aid the engineering of either synthetic or native methylotrophy. Furthermore, the discovery of a new metabolic capacity in this model organism will encourage new fundamental and applied research, which will extend the current scientific efforts towards using C1 compounds as feedstock for the production of valuable compounds.