Increased CO2 fixation enables high carbon-yield production of 3-hydroxypropionic acid in yeast

Abstract

CO₂ fixation plays a key role to make biobased production cost competitive. Here, we use 3-hydroxypropionic acid (3-HP) to showcase how CO₂ fixation enables approaching theoretical-yield production. Using genome-scale metabolic models to calculate the production envelope, we demonstrate that the provision of bicarbonate, formed from CO₂, restricts previous attempts for high yield production of 3-HP. We thus develop multiple strategies for bicarbonate uptake, including the identification of Sul1 as a potential bicarbonate transporter, domain swapping of malonyl-CoA reductase, identification of Esbp6 as a potential 3-HP exporter, and deletion of Uga1 to prevent 3-HP degradation. The combined rational engineering increases 3-HP production from 0.14 g/L to 11.25 g/L in shake flask using 20 g/L glucose, approaching the maximum theoretical yield with concurrent biomass formation. The engineered yeast forms the basis for commercialization of bio-acrylic acid, while our CO₂ fixation strategies pave the way for CO₂ being used as the sole carbon source.

Fig. 1. Production envelope analysis and bicarbonate metabolism rewiring for CO₂ fixation.

a The production envelope analysis for 3-HP production. Different colour represented the range of 3-HP production with its concurrent biomass under different bicarbonate flux. Reports used the malonyl-CoA pathway^,,,–, the oxaloacetate pathway, and the β-alanine pathway were labelled with dot, square, and diamond. The q_S indicated the consumed rate of glucose. The DCW was either measured directly or estimated based on OD₆₀₀ (DCW equals to 70% of OD₆₀₀). b The malonyl-CoA reductase pathway was used to produce 3-HP. Transcription factor Stb5 could realize the carbon flux rewiring from glycolysis to the oxPP pathway. c Adjusting STB5 expression by ARTp could together with the expression optimization of MCR domains, Acc1 and the PK pathway enhance 3-HP production to 0.74 g/L. d Substitution of the native promoter of STB5 with the TEF1p and an artificial promoter ARTp, respectively, and quantification of the STB5 with TEF1p and ARTp strength by qPCR. e Cellular bicarbonate derived from the potential bicarbonate transporter Sul1 (into the cell), Bor1 (outside the cell), and the endogenous synthesis from CO₂ by native carbonic anhydrase Nce103. f Providing more bicarbonate by overexpressing SUL1 enhanced 3-HP production to 0.83 g/L. g Phylogenetic analysis of SLC4 and SLC26 families identified Bor1, Sul1, and Sul2 as potential bicarbonate transporters in S. cerevisiae. (h) Rewiring the carbon flux from OAA to the high production of 3-HP. Overexpressing the phosphatase Ptc7 could significantly increase the 3-HP production based on introducing Cit1^S462A mutation, overexpressing CIT1, YHM2, and IDP2. (i) Distribution of CO₂ releasing reactions and bicarbonate utilizing reactions in S. cerevisiae. Abbreviations were defined in Supplementary Data 5 and 6. 10 mM sodium bicarbonate was added to the defined minimal medium was used for the 3-HP production. All data were presented as mean ± SD of biological triplicates. Significant comparisons of two groups were indicated in the graphs statistical analysis performed using a two-tailed unpaired Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001). Source data are provided as a Source Data file.

Fig. 2. Enzyme and pathway engineering for efficient production of 3-HP.

a Genes encoding of split MCR enzymes were either integrated in the chromosome or expressed using the plasmid system. b MCR-N domain and MCR-C*** domain co-expressed with the high copy plasmid could produce 4.12 g/L 3-HP. c Homologous modelling structure of the domains in malonate semialdehyde reductase and malonyl-CoA reductase from C. aurantiacus. d Swapping of native MCR domains with Ora1 and YdfG improved 3-HP production to 5.2 g/L and 5.3 g/L, respectively. e Carbon flux rewiring from fatty acids to 3-HP. f Downregulation of FAS1 combined with upregulation of POX2 and POX1 improved the production of 3-HP. g Nile red staining demonstrated that the size and the number of lipid droplets decreased in the fatty acid oxidation strain QLW36 compared with that of QLW26. (h) The fluorescence intensity of neutral lipids stained with Nile red in QLW26 and QLW36 was quantified in Relative Fluorescence Units (RFU). The RFU values were corrected by subtracting both the inherent autofluorescence of the samples and the fluorescence contributed by the solvent in the presence of Nile red (blank). Abbreviations were defined in Supplementary Data 5 and 6. The defined minimal medium with 10 mM NaHCO₃ was used for the 3-HP production. All data were presented as mean ± SD of biological triplicates. Significant comparisons of two groups were indicated in the graphs statistical analysis performed using a two-tailed unpaired Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001). Source data are provided as a Source Data file.

Fig. 3. Genotype identification of 3-HP tolerance and increase in the production of 3-HP.

a Schematic representation of transporters and permeases in the cell membrane. b Overexpression of ESBP6, deletion of SAM2, GSF2, or ERF2 could significantly improve the production of 3-HP. c Growth profiling showed that overexpression of ESBP6 could remarkably increase cell tolerance to high concentrations of 3-HP, compared with the WT strain. Abbreviations were defined in Supplementary Data 5 and 6. The defined minimal medium with 10 mM NaHCO₃ was used for the 3-HP production. All data were presented as mean ± SD of biological triplicates. Significant comparisons of two groups were indicated in the graphs statistical analysis performed using a two-tailed unpaired Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001). Source data are provided as a Source Data file.

Fig. 4. Alleviating bicarbonate competitions from the de novo pyridine ring synthesis enhanced 3-HP production.

a Carbamoyl-P was synthesized by Cpa1/2 from ATP, bicarbonate, and glutamate. Atom sources for purine, pyrimidine, arginine, and histidine were marked. b Substitution of URA3 marker by HIS3 marker in the same plasmid background increased the 3-HP titre to 8.22 g/L in the QLW53 strain. c ¹³C isotope labelling indicated that less M + 1 uracil was produced in QLW53 strain corresponding to the higher 3-HP titre. d CaCO₃ was used to provide more bicarbonate recycling CO₂, 0, 10, 50, 75, 100, and 150 mM represent 0 g/flask, 0.02 g/flask, 0.10 g/flask, 0.15 g/flask, 0.20 g/flask, and 0.30 g/flask. Abbreviations were defined in Supplementary Data 5 and 6. The defined minimal medium with 30 mM ¹³C isotope labelling NaHCO₃ or the corresponding concentrations of CaCO₃ as shown in the figures was used for the 3-HP production. All data were presented as mean ± SD of biological triplicates. Significant comparisons of two groups were indicated in the graphs statistical analysis performed using a two-tailed unpaired Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001). Source data are provided as a Source Data file.

Fig. 5. Utilization of inositol pyrophosphate signalling to provide more ATP for 3-HP production.

a Schematic representation of regulation mechanisms involved in glycolysis and respiration in yeast. b OCA5 deletion improved the production of 3-HP to 10.15 g/L. The defined minimal medium with 0.15 g CaCO₃ for each flask was used for the 3-HP production. Abbreviations were defined in Supplementary Data 5 and 6. All data were presented as mean ± SD of biological triplicates. Significant comparisons of two groups were indicated in the graphs statistical analysis performed using a two-tailed unpaired Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001). Source data are provided as a Source Data file.

Fig. 6. Abolishing potential degradations of MSA improved 3-HP production.

a Candidate enzymes that could potentially degrade 3-HP. b Deletion of UGA1 combined with overexpression of the MCR-C domain using CCW12 promoter could improve the 3-HP production to 11.25 g/L. Abbreviations were defined in Supplementary Data 5 and 6. The defined minimal medium with 0.15 g CaCO₃ for each flask was used for the 3-HP production. All data were presented as mean ± SD of biological triplicates. Significant comparisons of two groups were indicated in the graphs statistical analysis performed using a two-tailed unpaired Student’s t-test (*p < 0.05, **p < 0.01, ***p < 0.001). Source data are provided as a Source Data file.

Fig. 7. The final 3-HP production and its conversion to acrylic acid.

a The production envelope incorporated with key 3-HP production results reported in this study. Collectively, four of our constructed strains exceeded the 3-HP production boundary with native metabolisms, with QLW71 approaching the inaccessible one to one ratio of bicarbonate influx versus glucose uptake. Qs was substrate (glucose) consumption rate. b After a simple centrifugation step to remove cell pellets, the medium could be used directly for dehydration to generated acrylic acid. Source data are provided as a Source Data file.