Metabolic engineering最新文献

Prenol and isoprenol are promising advanced biofuels and serve as biosynthetic precursors for pharmaceuticals, fragrances, and other industrially relevant compounds. Despite engineering improvements that circumvent intermediate cytotoxicity and lower energy barriers, achieving high titer 'mevalonate (MVA)-derived' prenol has remained elusive. Difficulty in selective prenol production stems from the necessary isomerization of isopentenyl diphosphate (IPP) to dimethylallyl diphosphate (DMAPP) as well as the intrinsic toxicity of these diphosphate precursors. Here, the expression of specific isopentenyl monophosphate kinases with model-guided enzyme substitution of diphosphate isomerases and phosphatases enabled selective cycling of monophosphates and diphosphates, dramatically improving prenol titers and selectivity in Escherichia coli. Pairing this approach with the canonical MVA pathway resulted in 300 mg/L prenol at a 30:1 ratio with isoprenol. Further pairing with the "IPP-Bypass" pathway resulted in 526 mg/L prenol at a 72:1 ratio with isoprenol, the highest and purest MVA-derived prenol titer to date. Additionally, modifying this "IPP-Repass" for DMAPP production and coexpressing the prenyltransferase acPT1 yielded 48.3 mg/L of the potential therapeutic precursor drupanin from p-coumarate. These novel repass pathways establish a unique strategy for tuning diphosphate precursors to drive isoprenoid biosynthesis and prenylation reactions.

Acetate is a biological anion with many applications in the chemical and food industries. In addition to being a common microbial fermentative end-product, acetate can be produced by photosynthetic cyanobacteria from CO₂ using solar energy. Using wild-type cells of the unicellular model cyanobacterium Synechocystis PCC 6803 only low levels of acetate are observed outside the cells. By inserting a heterologous phosphoketolase (PKPa) in the acs locus, encoding acetyl-CoA synthetase responsible for the irreversible conversion of acetate to acetyl-CoA, an increased level of 40 times was observed. Metabolite analyses indicate an enhanced Calvin-Benson-Bassham cycle, based on increased levels of glyceraldehyde 3-phosphate and fructose-1,6-biphosphate, while the decreased levels of 3-phosphoglycerate and pyruvate suggest a quick consumption of the fixed carbon. Acetyl-P and erythrose-4-phosphate showed significantly increased levels, as products of phosphoketolase, while acetyl-CoA remained stable through the experiment. The results of intra- and extra-cellular acetate levels clearly demonstrate an efficient excretion of produced acetate from the cells in the engineered strain. Knock-out of ach and pta showed a reduction in acetate production however, it was not as low as in cells with a single knock-out of ach. Overexpressing acetyl-CoA hydrolase (Ach) and acetate kinase (AckA) did not significantly increase production. In contrast, overexpressing phosphotransacetylase (Pta) in cells containing an inserted PKPa resulted in 80 times more acetate reaching 2.3 g/L after 14 days of cultivation.

Polyhydroxyalkanoate (PHA) is an attractive bio-degradable plastic alternative to petrochemical plastics. Photosynthetic cyanobacteria accumulate biomass by fixing atmospheric CO₂, making them promising hosts for sustainable PHA production. Conventional PHA production in cyanobacteria requires prolonged cultivation under nutrient limitation to accumulate cellular PHA. In this study, we developed a system for growth-coupled production of the PHA poly-hydroxybutyrate (PHB), using the marine cyanobacterium Synechococcus sp. PCC 7002. A recombinant strain termed KB1 expressing a set of heterologous PHB biosynthesis genes (phaA/phaB from Cupriavidus necator H16 and phaE/phaC from Synechocystis sp. PCC 6803) accumulated substantial PHB during growth (11.4% of dry cell weight). To improve PHB accumulation, we introduced the Pseudomonas aeruginosa phosphoketolase gene (pk) into strain KB1, rewiring intermediates of the Calvin-Benson-Bassham (CBB) cycle (xyluose-5-phosphate, sedoheptulose 7-phosphate, and fructose-6-phosphate) to acetyl-CoA. The pk-expressing strain, KB15, accumulated 2.1-fold enhanced levels of PHB (23.8% of dried cell weight), relative to the parent strain, KB1. The highest PHB titer of KB15 strain supplemented with acetate was about 1.1 g L^-1 and the yield was further enhanced by 2.6-fold following growth at 38 °C (0.21 g L^-1 d^-1), relative to growth at 30 °C. Metabolome analysis revealed that pool sizes of CBB intermediates decreased, while levels of acetyl-CoA increased in strain KB15 compared with strain KB1, and this increase was further enhanced following growth at 38 °C. Our data demonstrate that acetyl-phosphate generated by Pk was converted into acetyl-CoA via acetate by hitherto unidentified enzymes. In conclusion, expression of heterologous PHB biosynthesis genes enabled growth-coupled PHB production in strain PCC 7002, which was increased through acetyl-CoA supplementation by bypassing acetyl-phosphate and elevating culture temperature.

10-hydroxy-2-decenoic acid (10-HDA), a unique unsaturated fatty acid present in royal jelly, has attracted considerable interest due to its potential medical applications. However, its low concentration in royal jelly and complex conformational structure present challenges for large-scale production. In this study, we designed and constructed a de novo biosynthetic pathway for 10-HDA in Escherichia coli. Initially, we introduced the heterologous thioesterase UaFatB1 to hydrolyze trans-2-decenoyl ACP to produce trans-2-decenoic acid, a key precursor for 10-HDA. Subsequently, we employed the bacterial cytochrome P450 enzyme CYP153AMaq to catalyze the terminal hydroxylation of trans-2-decenoic acid. Furthermore, through redox partner engineering and directed evolution, we identified the optimal combination for 10-HDA production: CYP153AMaq Q129R/V141L mutant with redox partner FdR0978/Fdx0338. Finally, we optimized the fermentation conditions and achieved a 10-HDA titer of 18.8 mg/L using glucose as primary carbon source. Our work establishes a platform for producing α,β-unsaturated fatty acids and its derivatives, facilitating more study of these compounds.

High sugar intake has become a global health concern due to its association with various diseases. Mogroside V (MG-V), a zero-calorie sweetener with multiple medical properties, is emerging as a promising sugar substitute. However, its application is hindered by low natural abundance and the inefficiency of conventional plant extraction methods. In this study, two glycosyltransferases were introduced into an engineered mogrol-producing Saccharomyces cerevisiae strain to enable the first de novo MG-V biosynthesis. Then, MG-V titer increased by 2.3 × 10⁴-fold through a series of efficient metabolic engineering strategies, including the enhancement of precursors, inhibition of the competitive pathway, and prevention of MG-V degradation. The challenges of enzyme spatial separation and high protein folding stress were addressed through lipid droplet (LD) compartmentalization and endoplasmic reticulum expansion, respectively. The ty1 transposon was employed to increase the copies of LD-targeted fusion protein AtCPR2-CYP87D18, which possessed higher CYP450 catalytic efficiency, resulting in an MG-V titer of 10.25 mg/L in shake flasks and 28.62 mg/L in a 5-L bioreactor. Overall, this study realized de novo MG-V synthesis in S. cerevisiae for the first time and provided a valuable reference for constructing microbial factories for triterpenoid saponin synthesis.

Advanced genome engineering enables precise and customizable modifications of bacterial species, and toolsets that exhibit broad-host compatibility are particularly valued owing to their portability. Tn5 transposon vectors have been widely used to establish random integrations of desired DNA sequences into bacterial genomes. However, the iteration of the procedure remains challenging because of the limited availability and reusability of selection markers. We addressed this challenge with CIFR, a mini-Tn5 integration system tailored for iterative genome engineering. The pCIFR vectors incorporate attP and attB sites flanking an antibiotic resistance marker used to select for the insertion. Subsequent removal of antibiotic determinants is facilitated by the Bxb1 integrase paired to a user-friendly counter-selection marker, both encoded in auxiliary plasmids. CIFR delivers engineered strains harboring stable DNA insertions and free of any antibiotic resistance cassette, allowing for the reusability of the tool. The system was validated in Pseudomonas putida, Escherichia coli, and Cupriavidus necator, underscoring its portability across diverse industrially relevant hosts. The CIFR toolbox was calibrated through combinatorial integrations of chromoprotein genes in P. putida, generating strains displaying a diverse color palette. We also introduced a carotenoid biosynthesis pathway in P. putida in a two-step engineering process, showcasing the potential of the tool for pathway balancing. The broad utility of the CIFR toolbox expands the toolkit for metabolic engineering, allowing for the construction of complex phenotypes while opening new possibilities in bacterial genetic manipulations.

Amino acid auxotrophy refers to an organism's inability to synthesize one or more amino acids that are required for cell growth. In microbiome research, co-cultures of amino acid auxotrophs are often used to investigate metabolite cross-feeding interactions and model community dynamics. Thus far, it has been implicitly assumed that amino acids are mainly cross-fed between these auxotrophs. However, this assumption has not been fully verified. For example, it could be that intermediates of amino acid biosynthesis pathways are exchanged instead, or in addition to amino acids. If true, this would significantly increase the complexity of metabolic interactions that needs to be considered. Here, we show that metabolic pathway intermediates are indeed exchanged in many co-cultures of amino acid auxotrophs. To demonstrate this, we selected 25 E. coli single gene knockouts that are auxotrophic for five different amino acids: arginine, histidine, isoleucine, proline, and tryptophan. In co-culture experiments, we paired strains that shared the same amino acid auxotrophy and monitored cell growth. We observed growth in 23 out of 55 strain pairings, indicating that pathway intermediates were exchanged between the strains. To provide further support for cross-feeding of pathway intermediates, auxotrophic E. coli strains were cultured in media supplemented with commercially available metabolic pathway intermediates at different concentrations. Supplementing media with these metabolites recovered cell growth as was predicted from the co-culture experiments. Most of these metabolites supported high growth rates, even when present at low concentrations (10 μM), suggesting the presence of high affinity transporters for these metabolites. In total, we identified eight metabolic pathway intermediates that were likely exchanged between the auxotrophic E. coli strains and verified six of these, including histidinol, N-acetyl-L-ornithine, L-ornithine, L-citrulline, keto-isoleucine and anthranilate. Taken together, this work demonstrates that exchange of metabolic pathway intermediates is more common than has been assumed so far. In future, these exchanges must be explicitly considered when constructing models of metabolite cross-feeding interactions in microbial communities and when interpreting results from microbiome studies involving auxotrophic strains.

Understanding and manipulating the cofactor preferences of NAD(P)-dependent oxidoreductases, the most widely distributed enzyme group in nature, is increasingly crucial in bioengineering. However, large-scale identification of the cofactor preferences and the design of mutants to switch cofactor specificity remain as complex tasks. Here, we introduce DISCODE (Deep learning-based Iterative pipeline to analyze Specificity of COfactors and to Design Enzyme), a novel transformer-based deep learning model to predict NAD(P) cofactor preferences. For model training, a total of 7,132 NAD(P)-dependent enzyme sequences were collected. Leveraging whole-length sequence information, DISCODE classifies the cofactor preferences of NAD(P)-dependent oxidoreductase protein sequences without structural or taxonomic limitation. The model showed 97.4% and 97.3% of accuracy and F1 score, respectively. A notable feature of DISCODE is the interpretability of its transformer layers. Analysis of attention layers in the model enables identification of several residues that showed significantly higher attention weights. They were well aligned with structurally important residues that closely interact with NAD(P), facilitating the identification of key residues for determining cofactor specificities. These key residues showed high consistency with verified cofactor switching mutants. Integrated into an enzyme design pipeline, DISCODE coupled with attention analysis, enables a fully automated approach to redesign cofactor specificity.