Industrial systems biology of Saccharomyces cerevisiae enables novel succinic acid cell factory

Abstract

Saccharomyces cerevisiae is the most well characterized eukaryote, the preferred microbial cell factory for the largest industrial biotechnology product (bioethanol), and a robust commerically compatible scaffold to be exploitted for diverse chemical production. Succinic acid is a highly sought after added-value chemical for which there is no native pre-disposition for production and accmulation in S. cerevisiae. The genome-scale metabolic network reconstruction of S. cerevisiae enabled in silico gene deletion predictions using an evolutionary programming method to couple biomass and succinate production. Glycine and serine, both essential amino acids required for biomass formation, are formed from both glycolytic and TCA cycle intermediates. Succinate formation results from the isocitrate lyase catalyzed conversion of isocitrate, and from the α-keto-glutarate dehydrogenase catalyzed conversion of α-keto-glutarate. Succinate is subsequently depleted by the succinate dehydrogenase complex. The metabolic engineering strategy identified included deletion of the primary succinate consuming reaction, Sdh3p, and interruption of glycolysis derived serine by deletion of 3-phosphoglycerate dehydrogenase, Ser3p/Ser33p. Pursuing these targets, a multi-gene deletion strain was constructed, and directed evolution with selection used to identify a succinate producing mutant. Physiological characterization coupled with integrated data analysis of transcriptome data in the metabolically engineered strain were used to identify 2(nd)-round metabolic engineering targets. The resulting strain represents a 30-fold improvement in succinate titer, and a 43-fold improvement in succinate yield on biomass, with only a 2.8-fold decrease in the specific growth rate compared to the reference strain. Intuitive genetic targets for either over-expression or interruption of succinate producing or consuming pathways, respectively, do not lead to increased succinate. Rather, we demonstrate how systems biology tools coupled with directed evolution and selection allows non-intuitive, rapid and substantial re-direction of carbon fluxes in S. cerevisiae, and hence show proof of concept that this is a potentially attractive cell factory for over-producing different platform chemicals.

Figure 1. Proof-of-concept: Successful metabolic engineering strategy guided by genome-scale metabolic modeling.

Panel a shows the central carbon metabolism of S. cerevisiae, and the model-guided metabolic engineering strategy for succinate over-production. Succinate production is directly coupled to biomass formation based on three gene deletions: sdh3 (cytochrombe b subunit of succinate dehydrogenase complex), and ser3/ser33 (3-phosphoglycerate dehydrogenase isoenzymes). The remodeling of central carbon flux towards succinate requires minimizing the conversion of succinate to fumarate, and forcing the biomass-required amino acids L-glycine and L-serine to be produced from glyoxylate pools. Production of glyoxylate results from isocitrate conversion by Icl1p, producing equimolar succinate. As the biomass yield increases, the demand for L-glycine and L-serine increase proportionally, driving biomass-coupled succinate production. Legend: native reactions (blue solid line), lumped native reactions (blue dashed line), interrupted reactions (red solid line), up-regulated reactions (green solid line). Panel b demonstrates the proof of concept. The reference strain and genetically engineered mutant strain, 8D, supplemented with 500 mg L^{− 1} glycine were physiologically characterized in 2L well-controlled stirred-tank fermentations. There was a 13.3X improvement in succinate titer.

Figure 2. Metabolic engineering enhanced by directed evolutions.

Cell populations were transferred across six shake flask cultures until a glycine prototroph was isolated. Subsequently, successive cultures were used to select for faster growth. From the final shake flask (SF3) the strain Evolved 8D was isolated. The succinate yield on biomass is plotted for each shake flask culture, demonstrating a 7.8X increase. The right plot shows the profile of specific growth rate and succinate yield on biomass for the final selection of faster growing cells.

Figure 3. Transcriptome guided metabolic engineering – Analysis.

Affymetrix Yeast 2.0 DNA microarrays were used for transcriptome analysis of each strain cultured in well-controlled glucose batch fermentations. The top 2000 differentially expressed genes had an adjusted p.value<0.01 and log-fold change (lfc)>0.5. A carbon-limited and nitrogen-limited chemostat transcriptome data set using the reference strain, surveyed at dilution rates (D) of 0.03, 0.1, and 0.2 h^{− 1} was used to determine which genes are growth-related under each condition. A total of 36 unique growth-related genes were identified from statistical analysis of each data set and with a total of 8 growth-related genes being among the top 20 differentially expressed genes between the reference and evolved 8D strain. After removal of the 36 genes, a total of 1964 genes were carried further for pathway analysis.

Figure 4. Summary of succinate microbial cell factory construction.

The specific growth rate (1/h), maximum succinate titer (g/L), maximum succinate yield on biomass (g/g-biomass), and maximum yield on glucose (g/g-glucose) are reported for the reference strain, 8D, 8D evolved, and 8D evolved with pICL1. A 43-fold improvement in succinate yield on biomass was observed across the full cycle of metabolic engienering that included in silico guided approaches, directed evolution, and transcriptome based identification of a 2^nd round of metabolic engineering targets.