A roadmap to establish a comprehensive platform for sustainable manufacturing of natural products in yeast

Abstract

Secondary natural products (NPs) are a rich source for drug discovery. However, the low abundance of NPs makes their extraction from nature inefficient, while chemical synthesis is challenging and unsustainable. Saccharomyces cerevisiae and Pichia pastoris are excellent manufacturing systems for the production of NPs. This Perspective discusses a comprehensive platform for sustainable production of NPs in the two yeasts through system-associated optimization at four levels: genetics, temporal controllers, productivity screening, and scalability. Additionally, it is pointed out critical metabolic building blocks in NP bioengineering can be identified through connecting multilevel data of the optimized system using deep learning.

Fig. 1. Schematic overview of the principle of heterologous chemical production in yeast.

The DNA sequences of genes and regulatory elements, promoters, and terminators (to, respectively, start and terminate the transcription of the individual genes) are engineered into the DNA located inside the nucleus in the cell. Followed by translation using the host translational machinery, the expressed enzymes convert the yeast precursor to the intermediate chemicals and, subsequently, the chemical of interest. The schematic shows a chemical biosynthetic pathway consisting of genes ‘Gene₁’, ‘Gene₂’, and ‘Gene_n’, requiring three promoters ‘P₁’, ‘P₂’, and ‘P_n’, and three terminators ‘T₁’, ‘T_2”, ‘T_n’). Enz enzyme, P promoter, T terminator.

Fig. 2. Schematic overview of the platform for sustainable production of NPs in yeast.

To sustainably produce the highly required pharmaceutical NPs in an environmentally friendly manner, a four-step platform is designed. Step 1, input optimization (orange box): For carbon-negative manufacturing (i), a photosynthetic microorganism or engineered yeast fixes CO₂.The fixed carbon can provide yeast with the required sources for production NP precursors. A light-inducible system is used to dynamically control NP metabolite production in yeast (ii). Step 2, system optimization (blue box): The library of light-inducible synthetic regulators is assembled upstream of genes of enzyme isoforms by using combinatorial assembly approaches (i). Protein engineering strategies are applied to design enzyme derivatives with enhanced activity. Next, yeast cells with rewired endogenous metabolites for the enhanced production of the required precursor in peroxisomes (blue box, ii) are chromosomally engineered with the pathway library (iii). The presence of the NP-responsive biosensor in the yeast cell enables rapid detection of NP production (iv). Step 3, output optimization (green box): Cells with a high level of fluorescence output are subjected to automated process optimization, and the top producers are sorted with FACS flow cytometry. Step 4, DL dataset development (red box): The library variants are subjected to single-cell droplet-based transcriptomics and NGS analyses to feed the DL dataset. The multilayer metadata obtained from dynamic optogenetic stimulation, the productivity of library members and growth fitness in real-time, and the genetic identity of low-to-top producers are used to fit model parameters and extract models to predict optimal conditions for heterologous production of other NPs. To simplify the figure, only three cells, Cell 1, Cell2, and Cell n, from the combinatorial library were shown. CO₂ carbon dioxide, DL deep learning, FACS fluorescence-activated cell sorting, NGS next-generation sequencing, NP natural product.

Fig. 3. Synthetic routes for carbon-negative manufacturing to produce NP in yeast.

a The yeast is engineered for the Calvin cycle to convert CO₂ to G3P, RuBisCo expression, and enhanced production of ATP and NADPH. b The yeast is engineered for the synthetic CETCH cycle to produce glyoxylate from CO₂, the β-hydroxyaspartate cycle to convert glyoxylate to NP, RuBisCo expression, and enhanced production of ATP and NADPH. ATP, adenosine triphosphate, BHAC β-hydroxyaspartate cycle, CETCH crotonyl-coenzyme A (CoA)/ethylmalonyl-CoA/hydroxybutyryl-CoA, CO₂ carbon dioxide, G3P glyceraldehyde-3-phosphate, NADPH nicotinamide adenine dinucleotide phosphate, NP natural product.

Fig. 4. Artificial microbial communities to promote circular acetyl-CoA production in yeast.

a An MSBC is created in polymeric microcapsules by encapsulation of subpopulations of cyanobacteria and yeast. Phototrophic cyanobacteria capture the CO₂ in the presence of light. The produced G3P is converted to the required carbon for production of metabolites, including NPs, in yeast. CO₂ produced by the yeast is reused by the autotrophic partner. b Artificial endosymbiosis. Phototrophic cyanobacteria endosymbionts within yeast cells, in which the cyanobacteria perform bioenergetic functions to provide an energy source for the yeast cells to grow and produce chemicals. CC Calvin cycle, CO₂ carbon dioxide, G3P glyceraldehyde-3-phosphate, MSBC microbial swarmbot consortia, NP natural product.

Fig. 5. Light-dependent expression system.

a Light triggers the dimerization of the light-sensing domain and its partner, thus leading to the joining of DBD and AD to form a functional synthetic activator. Next, the synthetic activator targets its BS within a promoter, thus resulting in the expression of a synTF. b In the absence of light, the synthetic activator is not formed, and thus the downstream synTF is not expressed. AD activation domain, BS binding site, DBD DNA binding domain, P synthetic promoter, synTF synthetic transcription factor, T yeast terminator.