pubmed.ncbi.nlm.nih.gov

De novo biosynthesis of bioactive isoflavonoids by engineered yeast cell factories - PubMed

  • ️Fri Jan 01 2021

De novo biosynthesis of bioactive isoflavonoids by engineered yeast cell factories

Quanli Liu et al. Nat Commun. 2021.

Abstract

Isoflavonoids comprise a class of plant natural products with great nutraceutical, pharmaceutical and agricultural significance. Their low abundance in nature and structural complexity however hampers access to these phytochemicals through traditional crop-based manufacturing or chemical synthesis. Microbial bioproduction therefore represents an attractive alternative. Here, we engineer the metabolism of Saccharomyces cerevisiae to become a platform for efficient production of daidzein, a core chemical scaffold for isoflavonoid biosynthesis, and demonstrate its application towards producing bioactive glucosides from glucose, following the screening-reconstruction-application engineering framework. First, we rebuild daidzein biosynthesis in yeast and its production is then improved by 94-fold through screening biosynthetic enzymes, identifying rate-limiting steps, implementing dynamic control, engineering substrate trafficking and fine-tuning competing metabolic processes. The optimized strain produces up to 85.4 mg L-1 of daidzein and introducing plant glycosyltransferases in this strain results in production of bioactive puerarin (72.8 mg L-1) and daidzin (73.2 mg L-1). Our work provides a promising step towards developing synthetic yeast cell factories for de novo biosynthesis of value-added isoflavonoids and the multi-phased framework may be extended to engineer pathways of complex natural products in other microbial hosts.

© 2021. The Author(s).

PubMed Disclaimer

Conflict of interest statement

J.N. is a shareholder in Biopetrolia AB and Chrysea, Inc. All other authors declare no competing interests.

Figures

Fig. 1
Fig. 1. Engineering the de novo biosynthesis of isoflavonoid chemicals in yeast.

A multi-phased metabolic engineering strategy was implemented to enable de novo production of isoflavonoids, following the pipeline of screening-reconstruction-application. In screening phase I (green box), a synthetic DEIN pathway was established and metabolic factors improving the performance of rate-limiting reactions were identified using a moderate p-HCA producing platform strain QL11. In reconstruction phase II (orange box), the DEIN production was further optimized in a wild-type strain QL179 harboring the deletion of galactose utilizing genes (GAL7/10/1), through amplifying gene expression, enhancing substrate transfer, combining effective genetic targets identified in phase I and fine-tuning the expression of a key gene involved in competing metabolic pathway (orange triangle). For application phase III (magenta box), the generated DEIN platform strain was used as the starting point for the production of bioactive glucosides PIN and DIN through introducing plant glycosyltransferases and enhancing the supply of glycosyl group donor UDP-glucose. The selected plant biosynthetic genes and overexpressed yeast native genes for isoflavonoid production were highlighted in blue boxes. Magenta arrows, designed DEIN biosynthetic pathway. Blue arrows, reactions for generating glucosides; gray arrows, byproduct pathway. At4CL1, 4-coumarate-coenzyme A ligase 1 from Arabidopsis thaliana; GmCHS8, chalcone synthase from Glycine max; GmCHR5, chalcone reductase from G. max; GmCHIB2, chalcone isomerase from G. max; Ge2-HIS, 2-hydroxyisoflavanone synthase from Glycyrrhiza echinata; GmHID, 2-hydroxyisoflavanone dehydratase from G. max; AtPAL2, phenylalanine ammonia lyase from A. thaliana; AtC4H, cinnamic acid-4-hydroxylase from A. thaliana; FAS1, beta subunit of yeast fatty acid synthetase; CrCPR2, cytochrome P450 reductase from Catharanthus roseus; STB5, yeast native transcriptional factor; EcyjfB, NAD+ kinase from Escherichia coli; GmUGT4, isoflavone 7-O-glucosyltransferase from G. max; PlUGT43, isoflavone 8-C-glucosyltransferase from Pueraria lobate; UGP1, UDP-glucose pyrophosphorylase; PGM1/2, phosphoglucomutase 1/2. In addition, yeast heme degradation was disrupted by deleting heme oxygenase-coding gene HMX1. ER endoplasmic reticulum, E4P erythrose-4-phosphate, PEP phosphoenolpyruvate, L-Phe L-phenylalanine, 5-ALA 5-aminolevulinic acid, UDP-Glc uridine diphosphate-glucose, UTP uridine triphosphate, G-6-P glucose-6-phosphate, G-1-P glucose-1-phosphate.

Fig. 2
Fig. 2. Building and validating the biosynthetic pathway for DEIN.

a Schematic illustration of the biosynthetic pathways leading to the production of DEIN and related byproducts. To ensure efficient screening of biosynthetic enzymes for DEIN production, a p-HCA producing strain QL11, harboring overexpression of enzymes responsible for both endogenous (yellow arrows) and exogenous (blue arrows) reactions, was selected as the starting strain. ARO3, DAHP synthase; ARO4*, L-tyrosine-feedback-insensitive DAHP synthase (ARO4K229L); ARO1, pentafunctional aromatic protein; EcaroL, shikimate kinase from E. coli; ARO2, chorismate synthase; ARO7*, L-tyrosine-feedback-insensitive chorismate mutase (ARO7G141S); AtATR2, cytochrome P450 reductase from A. thaliana; CYB5, yeast native cytochrome b5. DAHP, 3-deoxy-D-arabino-2-heptulosonic acid 7-phosphate; SHIK, shikimate; S3P, shikimate-3-phosphate; EPSP, 5-enolpyruvyl-shikimate-3-phosphate; CHA, chorismic acid; PPA, prephenate. See Fig. 1 and its legend regarding abbreviations of metabolites and other gene details. b, c Production profiles of intermediates LIG and ISOLIG produced by yeast strains harboring different combinations of biosynthetic enzymes 4CL, CHS, CHR, and CHI. d DEIN production of C09 strain overexpressing different biosynthetic genes encoding 2-HIS and HID and relevant genetic characteristics of the resultant strains. For the source of selected plant genes: Mt, Medicago truncatula; Tp, Trifolium pretense. See Fig. 1 legend regarding abbreviations of other plant species. Cells were grown in a defined minimal medium with 30 g L−1 glucose as the sole carbon source, and cultures were sampled after 72 h of growth for metabolite detection. All data represent the mean of n = 3 biologically independent samples and error bars show standard deviation. The source data underlying figures (b−d) are provided in a Source Data file.

Fig. 3
Fig. 3. Tailoring the redox partner of Ge2-HIS for efficient DEIN production.

a Schematic illustration of the biosynthetic pathways leading to the production of DEIN and related byproducts. P450 enzymes are indicated in magenta. In addition, a general catalytic mechanism of the membrane-bound plant P450 is shown in the inset. See Fig. 1 and its legend regarding abbreviations of metabolites and gene details. b Different redox partners (RPs) including CPR and surrogate redox partners from self-sufficient P450s were tested to enhance the catalytic activity of P450 Ge2-HIS. GmCPR1, cytochrome P450 reductase from G. max; BM3R, the eukaryotic-like reductase domain of P450BM3 from Bacillus megaterium; RhFRED, the FMN/Fe2S2-containing reductase domain of P450RhF from Rhodococcus sp. strain NCIMB 9784; RhF-fdx, a hybrid reductase by substituting Fe2S2 domain of RhFRED with ferredoxin (Fdx) from spinach. See Fig. 1 and its legend regarding abbreviations of metabolites and other gene details. c Effect of different RPs on the production of DEIN. Cells were grown in a defined minimal medium with 30 g L−1 glucose as the sole carbon source, and cultures were sampled after 72 h of growth for metabolite detection. Statistical analysis was performed by using Student’s t test (two-tailed; two-sample unequal variance; *p < 0.05, **p < 0.01, ***p < 0.001). All data represent the mean of n = 3 biologically independent samples and error bars show standard deviation. The source data underlying panel (c) are provided in a Source Data file.

Fig. 4
Fig. 4. Investigation of metabolic factors affecting DEIN biosynthesis.

a Schematic illustration of the genetic modifications performed to relieve potential metabolic limitations, regarding heme metabolism (I), ER homeostasis (II), and cofactor NADPH generation (III), to improve the performance of P450 Ge2-HIS. Overexpressed genes are shown in bold and gene deletion is marked with a red cross. Red circle and arrows, transcriptional factors (TFs) activating a metabolic gene/pathway; Blue circle and flat-head arrow, TFs repressing a metabolic gene/pathway. HEM2, 5-aminolevulinic acid dehydratase; HEM3, 4-porphobilinogen deaminase; HEM13, coproporphyrinogen III oxidase; ROX1, heme-dependent repressor of hypoxic genes; PAH1, phosphatidate phosphatase; INO2/OPI1, transcription activator/repressor of phospholipid biosynthetic genes; ALD6, cytoplasmic NADP+-dependent aldehyde dehydrogenase; EcpntAB, membrane-bound transhydrogenase from E. coli; YEF1, ATP-NADH kinase. Suc-CoA, succinyl-CoA; PA, phosphatidic acid; PL, phospholipid; DAG, diacylglycerol; X-5-P, xylulose-5-phosphate; AcD, acetaldehyde; Ac, acetate; NTP, nucleoside triphosphates; NDP, nucleoside diphosphates. See Fig. 1 and its legend regarding abbreviations of metabolites and other gene details. Production of DEIN by engineered yeast strains derived from strategy groups I (b), II (c), and III (d), respectively. Cells were grown in a defined minimal medium with 30 g L−1 glucose as the sole carbon source, and cultures were sampled after 72 h of growth for metabolite analysis. Statistical analysis was performed by using Student’s t test (two-tailed; two-sample unequal variance; *p < 0.05, **p < 0.01, ***p < 0.001). All data represent the mean of n = 3 biologically independent samples and error bars show standard deviation. The source data underlying panels (b−d) are provided in a Source Data file.

Fig. 5
Fig. 5. Gene amplification and engineering of substrate trafficking improve DEIN production.

a Schematic view of the targets and strategies to improve the substrate transfer along the DEIN biosynthetic pathway. Two different oligopeptide linkers (flexible linker L1, GGGS; rigid linker L2, VDEAAAKSGR) were employed to fuse the adjacent metabolic enzymes. Strain QL179 was selected to implement GAL promoters (GALps)-mediated gene amplification. See Fig. 1 and its legend regarding abbreviations of metabolites and other gene details. b Quantification of metabolic intermediates produced by strains carrying a fused enzyme of AtC4H (E1) and At4CL1 (E2). c Comparison of the production profiles between parental strain I02 and I14 harboring additional overexpression of selected metabolic enzymes Ge2-HIS and GmHID and auxiliary CrCPR2. Cells were grown in a defined minimal medium with 30 g L−1 glucose as the sole carbon source and 10 g L−1 galactose as the inducer. Cultures were sampled after 72 h of growth for metabolite detection. Statistical analysis was performed by using Student’s t test (two-tailed; two-sample unequal variance; *p < 0.05, **p < 0.01, ***p < 0.001). All data represent the mean of n = 3 biologically independent samples and error bars show standard deviation. The source data underlying panels (b, c) are provided in a Source Data file.

Fig. 6
Fig. 6. Combinatorial optimization to increase the production of DEIN.

a Effect of deleting genes involved in the regulation of heme metabolism on DEIN biosynthesis. Production of DEIN by strains fed with the heme biosynthetic precursor 5-ALA (b) or expressing different copies of Ge2-HIS and GmHID genes (c). d Process optimization for DEIN production. Cells were grown in a defined minimal medium with 30 g L−1 glucose (batch) or with six tablets of FeedBeads (FB) as the sole carbon source and 10 g L−1 galactose as the inducer. Cultures were sampled after 72 h (batch) or 90 h (FB) of growth for metabolite analysis. e Schematic view of the interplay between isoflavonoid biosynthesis and yeast cellular metabolism connected by the branchpoint malonyl-CoA. See Fig. 1 and its legend regarding abbreviations of metabolites and gene details. f Fine-tuning the expression of gene FAS1 via promoter engineering improves DEIN formation under optimized cultivation conditions. g Effect of genetic modifications altering the regulation of GAL induction on DEIN production under optimized cultivation conditions. The constitutive mutant of galactose sensor Gal3 (GAL3S509P) was overexpressed from a multi-copy plasmid (2 µm) under the control of GAL10p and gene ELP3, encoding a histone acetyltransferase, was deleted. Cells were grown in a defined minimal medium with six tablets of FB as the sole carbon source and 10 g L−1 galactose as the inducer. Cultures were sampled after 90 h of growth for metabolite detection. Statistical analysis was performed by using Student’s t test (two-tailed; two-sample unequal variance; *p < 0.05, **p < 0.01, ***p < 0.001). All data represent the mean of n = 3 biologically independent samples and error bars show standard deviation. The source data underlying panels (a−d) and (f, g) are provided in a Source Data file.

Fig. 7
Fig. 7. Production of DEIN-derived glucosides.

a Schematic view of the engineered metabolic pathways for the biosynthesis of glucosides PIN and DIN (pink box), and relevant byproducts (gray box). See Fig. 1 legend for gene details. b Characterization of metabolic enzymes responsible for glucoside biosynthesis. Three copies of PlUGT43 and GmUGT4 under the control of constitutive promoters were integrated into the DEIN producer C28, resulting in strains E03 and E06, respectively. Cells were grown in a defined minimal medium with 30 g L−1 glucose as the sole carbon source, and cultures were sampled after 72 h of growth for LC-MS analysis. c Production profiles of PIN and DIN in DEIN hyper-producing strain I34 background with or without increased UDP-glucose supply. Combined overexpression of genes PGM1/2 with UPG1 was implemented to enhance the generation of glycosyl group donor UDP-glucose. See Fig. 1 legend for gene details. Cells were grown in a defined minimal medium with six tablets of FB as the sole carbon source and 10 g L−1 galactose as the inducer. Cultures were sampled after 90 h of growth for metabolite detection. Statistical analysis was performed by using Student’s t test (two-tailed; two-sample unequal variance; *p < 0.05, **p < 0.01, ***p < 0.001). All data represent the mean of n = 3 biologically independent samples and error bars show standard deviation. The source data underlying figure c are provided in a Source Data file.

Similar articles

Cited by

References

    1. Veitch NC. Isoflavonoids of the leguminosae. Nat. Prod. Rep. 2013;30:988–1027. doi: 10.1039/c3np70024k. - DOI - PubMed
    1. Dixon RA, Steele CL. Flavonoids and isoflavonoids—a gold mine for metabolic engineering. Trends Plant Sci. 1999;4:394–400. doi: 10.1016/S1360-1385(99)01471-5. - DOI - PubMed
    1. Sharma, V. & Ramawat, K. G. In Natural Products (eds Ramawat, K. G. & Mérillon, J.-M.) 1849−1865 (Springer-Verlag, Berlin Heidelberg, 2013).
    1. Dixon RA, Pasinetti GM. Flavonoids and isoflavonoids: from plant biology to agriculture and neuroscience. Plant Physiol. 2010;154:453–457. doi: 10.1104/pp.110.161430. - DOI - PMC - PubMed
    1. Blicharski T, Oniszczuk A. Extraction methods for the isolation of isoflavonoids from plant material. Open Chem. 2017;15:34–45. doi: 10.1515/chem-2017-0005. - DOI

Publication types

MeSH terms

Substances

LinkOut - more resources