Formate-driven efficient pyruvate biosynthesis in Vibrio natriegens

Abstract

Formic acid (FA), a key one-carbon liquid compound derived directly from CO₂, can serve as a dual-purpose substrate in microbial metabolism, supplying both carbon and energy. Its potential for green biomanufacturing is immense, yet its inherent toxicity and poor metabolizability to most microbes pose a major hurdle in developing efficient microbial cell factories for value-added chemical production. Building on our prior discovery of Vibrio natriegens as a naturally proficient formic acid utilizer, we demonstrate here that formate supplementation as an auxiliary substrate can dramatically boost pyruvate production of V. natriegens from sodium gluconate, achieving a 1.9-fold increase in titer. Transcriptomic analysis revealed that formate presence induces global changes in gene expression. By subsequently downregulating the pyruvate consumption pathway, we engineered a strain that, when co-fed with formate and sodium gluconate, achieved a 49.0% improvement in pyruvate synthesis. Isotopic tracer analysis confirmed substantial formate assimilation, with approximately 9.43% incorporated into biomass. In a fed-batch fermentation, the engineered V. natriegens strain consumed 82.8 g/L sodium gluconate and 37.4 g/L formate (HCOONa·2H₂O) within 51 h, producing 56.4 g/L pyruvate at a rate of 1.1 g/L/h. This work elucidates the stimulatory role of formate in the pyruvate biosynthesis of V. natriegens and establishes a novel strategy for leveraging this feedstock in microbial production.

Graphical abstract

1. Introduction

With global carbon neutrality efforts expanding and emission reduction gaining attention, biomanufacturing is shifting from sugar-based materials to cheaper, more sustainable alternatives. Formic acid, the simplest carboxylic acid, can be produced through various pathways such as CO₂ electroreduction, biomass conversion, or industrial byproducts [1]. Given its wide availability, low cost, and suitability as a fermentation substrate for microbial utilization, formic acid has emerged as an attractive resource for green biomanufacturing in recent years [2]. Moreover, when fully oxidized, formic acid releases a large amount of energy, efficiently supporting microbial growth [2]. This makes it both a carbon source and an energy carrier during microbial utilization.

However, naturally occurring microbes that can use formic acid normally suffer from inherent limitations, including low formate tolerance and inefficient metabolism [[3], [4], [5], [6]], which restrict their practical applications. Although genetic engineering has improved formate utilization of some typical industrial microbes (e.g., Escherichia coli, Pichia pastoris, and Corynebacterium glutamicum), the achieved efficiencies remain inadequate for practical fermentation applications [[7], [8], [9], [10]]. A recent study on Vibrio natriegens has made significant progress in microbial utilization of formic acid [11]. V. natriegens, a naturally occurring marine bacterium known for its exceptionally fast growth rate, exhibits high formic acid tolerance and metabolic capacity. The wild-type V. natriegens could grow in the presence of 40 g/L of formate (HCOONa·2H₂O) and consumed 34.2 g/L within 24 h, yielding a consumption rate of 1.43 g/L/h; furthermore, the engineered V. natriegens strain exhibited higher formate consumption rate (3.29 g/L/h), surpassing all previously reported natural and engineered formate-utilizing microorganisms [11]. The discovery not only demonstrates the great potential of V. natriegens as a high-efficiency microbial chassis for formate utilization, but also suggests the need to further expand its synthetic capabilities in producing value-added products from formate.

Pyruvate is not only a key metabolic intermediate in living organisms but also an important industrial chemical widely used in chemical manufacturing, pharmaceuticals, food production, agriculture, and environmental protection [12,13]. Pyruvate can be produced through chemical synthesis and microbial fermentation. Current chemical synthesis methods like the tartaric acid process and lactic acid oxidation suffer from major limitations, including high energy consumption, substantial environmental pollution, and low yields [14]. In contrast, the pyruvate production via biological fermentation offers potential advantages in terms of environmental friendliness, sustainability, production costs, and product yield [12]. Although some engineered microbial strains can efficiently produce pyruvate (Table S1), their reliance on sugar-based feedstocks or glycerol raises production costs, reducing competitiveness with chemical synthesis. This underscores the need for exploring more cost-effective feedstocks. Recently, V. natriegens was engineered to ferment sodium gluconate for pyruvate production, achieving a titer of 56.8 g/L [15], highlighting its potential for pyruvate biosynthesis. Given V. natriegens's efficient formate utilization, a key question arises: Can it also convert formate into pyruvate? If successful, this approach could broaden feedstock options for microbial pyruvate production and further lower costs.

In this study, we sought to achieve the efficient synthesis of pyruvate from formate using V. natriegens. An interesting finding was that formate supplementation significantly enhanced pyruvate synthesis in V. natriegens from sodium gluconate. By optimizing downstream pyruvate metabolism, we further improved pyruvate accumulation. Using fed-batch fermentation, our engineered strain achieved efficient co-utilization of formate and sodium gluconate, reaching a pyruvate titer of 56.4 g/L at a production rate of 1.10 g/L/h.

2. Materials and methods

2.1. Bacterial strains and growth conditions

The bacterial strains employed in this study are listed in Table S2. V. natriegens ATCC 14048, preserved at −80 °C in 20% (v/v) glycerol, was used as the parental strain for formate fermentation and subsequent genetic modifications. Cultivation of V. natriegens strains was carried out in either LBv2 medium (a nutrient-rich formulation) or a modified M9 minimal medium supplemented with 2% (w/v) NaCl [16]. All the fermentations were carried out at 30 °C, and varying concentrations of formate (HCOONa·2H₂O) were added to the media as needed.

2.2. Reagents and chemicals

All the primers used in this study (Table S3) were synthesized by BioSune (Shanghai BioSune Biotechnology Co., China). HCOONa·2H₂O served as the formate source. All chemicals were purchased from Sigma-Aldrich (Sigma-Aldrich Co., USA) and Aladdin (Shanghai Aladdin Biochemical Technology Co., Ltd., China). PCR reactions were performed with KOD Plus Neo or KOD FX DNA polymerases (Toyobo Co., Japan). Multiple DNA fragments were assembled using the ClonExpress MultiS One Step Cloning Kit (Vazyme Biotech Co., Ltd., China). DNA was purified with commercial kits from Axygen Biotechnology Co., Ltd. (China).

2.3. Generation of engineered V. natriegens strains

Gene knockout in V. natriegens was performed using the multiplex genome editing by natural transformation (MuGENT) method [17]. First, the plasmid pMMB67EH-tfox-sacB was electroporated into the Vibrio natriegens strain targeted for gene knockout. Specifically, 100 ng of the plasmid was mixed with 200 μL of competent cells and incubated on ice for 10 min. The mixture was then transferred to a 2 mm gap electroporation cuvette. Electroporation was performed using the following parameters: 1600 V, 200 Ω, 25 μF, with a 2 mm gap cuvette. Following electroporation, 1 mL of LBv2 medium was added to the electroporated cell mixture. The recovery culture was incubated in a shaking incubator at 30 °C and 280 rpm for 1–2 h and then spread onto agar plates containing the antibiotic corresponding to the resistance marker on the plasmid for positive selection.

For natural transformation, V. natriegens cells harboring pMMB67EH-tfox-sacB were grown (shaking, 280 rpm) overnight at 30 °C in LBv2 medium with the addition of 100 μg/mL carbenicillin and 100 μM IPTG. After cultivating overnight, a diluted (1:100) culture was prepared with IO medium [17] containing 100 μM IPTG. Next, 50 ng of a selected product (DNA fragment with antibiotic resistance marker and dns homology arms) and 200 ng of unselected product (homology arms for the target gene without selection marker) were added. The mixture was incubated at 30 °C for 5 h. Then, 1 mL of the LBv2 medium was added, followed with incubation (shaking, 280 rpm) at 30 °C for 2 h. The culture was then plated on LBv2 agar plates containing 100 μg/mL carbenicillin and either 12.5 μg/mL chloramphenicol or 100 μg/mL kanamycin. After incubation at 30 °C for 10 h, colonies that were visible on agar plates were screened by colony PCR to identify the desired gene knockout events.

To eliminate the pMMB67EH-tfox-sacB plasmid from the mutated V. natriegens strains, colonies were harvested and resuspended in 5 mL LBv2 medium with 80 g/L sucrose, followed by shaking incubation (280 rpm, 30 °C, 10 h). The culture was then plated onto LBv2 agar plates again supplemented with 80 g/L sucrose. After 12 h at 30 °C, colonies were picked and resuspended in the LBv2 medium with and without 100 μg/mL carbenicillin to verify plasmid curing.

2.4. Analytical methods

Fermentation samples were collected at designated time points and immediately centrifuged at 10,000×g for 10 min at 4 °C. Subsequently, 200 μL of supernatants was subjected to HPLC analysis. The concentrations of formate, sodium gluconate, glucose, fructose, sucrose, pyruvate, parapyruvate, and glycerol in supernatants were quantified using an Agilent 1100 high-performance liquid chromatography (HPLC) system equipped with a Bio-Rad Aminex HPX-87H column (Bio-Rad, USA). The column was maintained at 30 °C, and the mobile phase (5 mM H₂SO₄) was delivered at a flow rate of 0.6 mL/min. A refractive index detector was employed for the detection of formate, sodium gluconate, glucose, fructose, sucrose, and glycerol, while pyruvate and parapyruvate were quantified using a UV detector at 210 nm.

2.5. RNA-seq analysis

The V. natriegens strain ATCC 14048 was cultured in the modified M9 medium using sodium gluconate (50 g/L) as the primary carbon source, with or without the supplementation of HCOONa·2H₂O (5 g/L). After 12 h of cultivation, cells were harvested by centrifugation and immediately frozen in liquid nitrogen. cDNA library construction and RNA sequencing were conducted as previously described [11]. Raw sequencing data from the Illumina platform were analyzed on the Majorbio cloud platform (Shanghai Majorbio Bio-Pharm Technology Co., Ltd., China). The abundance quantification of genes and isoforms was carried out with RSEM (RNA-Seq by Expectation-Maximization), and their expression levels were calculated using the TPM (Transcripts Per Million) method. The identification of differentially expressed genes was performed with the DESeq2 packages (http://bioconductor.org/packages/release/bioc/html/DESeq2.html).

2.6. ¹³C isotopomer analysis of amino acids and pyruvate

¹³C isotopomer analysis of intracellular amino acids was conducted on the V. natriegens P2GTG strain. Cells were cultivated in 50 mL of M9 medium supplemented with 50 g/L sodium gluconate and a 1:1 mixture of unlabeled and ¹³C-labeled HCOONa·2H₂O (2.5 g/L each). After 24 h of cultivation, cells were harvested by centrifugation (6000×g, 20 min, 4 °C). The cell pellet was subsequently hydrolyzed with 200 μL of 6 M HCl at 105 °C for 24 h. The resulting hydrolysates were dried, and the amino acids were derivatized with a mixture of 100 μL pyridine and 50 μL MSTFA at 85 °C for 1 h. Next, samples were filtered through a 0.45 μm membrane (Millipore) and analyzed using a GC-MS system equipped with a DB-5HT column (30 m × 0.25 mm, 0.1 μm). The system was operated in electron impact (EI) mode at 70 eV. Amino acids were identified by matching their mass spectra and retention times to those in an authenticated standards library.

¹³C isotopomer analysis of pyruvate was carried out as follows. Cell cultivation and harvesting were performed as same as abovementioned. Metabolites were extracted from the cell pellet using a pre-chilled acetonitrile/methanol/water mixture (2:2:1, v/v/v). The suspension was incubated at −20 °C for 20 min and then centrifuged (12,000×g, 20 min, 4 °C). The resulting supernatant was collected for analysis. Pyruvate analysis was performed using an ultra-high-performance liquid chromatography (UHPLC) system (Acquity, Waters) coupled to a Q Exactive hybrid quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific). Metabolite separation was achieved on a Luna NH₂ column (100 mm × 2 mm, 3 μm). The mass spectrometer was operated in negative electrospray ionization (ESI−) mode, with data acquired over a mass range of 70–1000 m/z at a resolution of 70,000. MS/MS spectra were collected using a normalized collision energy of 30 eV.

2.7. Analysis of ¹³C ratio in biomass

The formate assimilation ratio was determined by measuring the proportion of ¹³C in the total carbon of biomass, following a previously described method [11,18]. Briefly, the V. natriegens P2GTG cells were cultured in the M9 medium, supplemented with 50 g/L sodium gluconate, 2.5 g/L unlabeled HCOONa·2H₂O, and 2.5 g/L¹³C-labeled HCOONa·2H₂O, at 30 °C for 24 h 1 mL of the grown cells were then collected by centrifugation (6000×g, 4 °C, for 10 min). The cell pellet was washed twice and subsequently freeze-dried. The ¹³C content of the biomass was then analyzed using an isotope ratio mass spectrometer coupled with an elemental analyzer (Thermo Fisher Scientific, Inc., USA).

2.8. Fed-batch fermentation

Fed-batch fermentation in bioreactors (2-liter) was carried out with 1 L of M9 medium supplemented with 30 g/L of sodium gluconate and 5 g/L of HCOONa·2H₂O initially. In parallel, stock solutions of sodium gluconate (300 g/L) and formate (400 g/L HCOONa‧2H₂O) were prepared for subsequent feeding.

The preserved V. natriegens strain was streaked onto an agar plate, and a single colony was inoculated into fresh LBv2 medium, followed by overnight cultivation at 30 °C with shaking at 280 rpm. The resulting seed culture was inoculated into the modified M9 medium at an initial inoculation ratio of approximately 4%. Fermentation was conducted at 30 °C. The fermenter was supplied with air at a flow rate of 1.0–1.5 NL/min, while the stirring speed was controlled at 280 rpm.

When the optical density at 600 nm (OD₆₀₀) reached approximately 1.0, feeding of formate and sodium gluconate started. In addition, an automated control system was employed to regulate the pH by the precise addition of NH₃·H₂O, maintaining the pH at approximately 7.0.

3. Results and discussion

3.1. Promotion effect of formate on pyruvate synthesis in V. natriegens

Although previous studies have demonstrated the ability of V. natriegens to produce pyruvate from sugar-based substrates through fermentation [[12], [13], [14]], the potential of formate as a substrate for pyruvate production in this bacterium remains unexplored. In V. natriegens, formate is primarily metabolized via two pathways [11]: the formate-tetrahydrofolate ligase (FTL)-mediated tetrahydrofolate cycle, which facilitates formate assimilation, and formate dehydrogenase (FDH)-catalyzed oxidation, leading to formate dissimilation and NADH generation (Supplementary Fig. 1). As a result, formate can function as both a carbon and an energy source in V. natriegens. However, since V. natriegens cannot utilize formate as the sole carbon source to sustain its growth, this study investigated the use of formate as an auxiliary substrate to enhance pyruvate production in this organism.

To determine the most suitable carbon source for co-utilization with formate in V. natriegens to maximize pyruvate production, we evaluated five substrates known for their efficient utilization by this bacterium, i.e., sodium gluconate, glucose, sucrose, fructose, and glycerol. Three concentration gradients were used for each primary substrate, with the supplementation of 5 g/L of sodium formate (HCOONa·2H₂O) as the auxiliary substrate. After 24 h of cultivation, substrate consumption and pyruvate production were determined. The results showed that V. natriegens could consume formate completely under all tested conditions (Fig. 1a); in contrast, it exhibited significant differences in the co-utilization of the abovementioned five primary carbon sources, in which sucrose and sodium gluconate were consumed much better than fructose, glucose, and glycerol (Fig. 1b). In addition, the fermentation of formate plus sodium gluconate, glucose, sucrose, or fructose by V. natriegens all generated pyruvate; but co-utilization of formate and glycerol did not yield pyruvate under all the conditions (Fig. 1c). Maximal pyruvate accumulation was observed in the formate-gluconate combination, reaching 15.8 ± 0.3 g/L, which was much higher than those of the other substrate combinations (Fig. 1c). Consequently, compared to the other substrate combinations, the formate-gluconate combination attained a higher yield in pyruvate synthesis to varying degrees (Fig. 1d). Regarding the observed distinct advantage of sodium gluconate in yielding pyruvate, we propose that this advantage stems from the shorter metabolic pathway and higher energy efficiency associated with converting sodium gluconate to pyruvate. Specifically, a single sodium gluconate molecule can be transformed into one pyruvate molecule and one glyceraldehyde-3-phosphate (G3P) molecule in just three enzymatic steps. The resulting G3P can then be further converted into pyruvate via glycolysis. In general, shorter metabolic pathways minimize carbon loss, thereby enhancing the final product yield.

Fig. 1.

Comparison of the co-utilization of formate and the other five carbon sources by V. natriegens as well as the effects on pyruvate production. (a) Formate consumption by V. natriegens in the presence of different primary substrates. (b) Consumption of sodium gluconate, glucose, sucrose, fructose, and glycerol in the presence of formate. (c, d) The titer (c) and yield (d) of pyruvate. (e, f) The consumption of sodium gluconate (e) and the titer of pyruvate (f) in the modified M9 medium containing 50 g/L sodium gluconate with or without 5 g/L HCOONa·2H₂O. The concentration of 5 g/L sodium formate was chosen as the initial addition amount because V. natriegens can maintain basic growth under this concentration of sodium formate. Data are presented as mean ± standard deviation (n = 3 biologically independent replicates). Error bars show SDs. The statistical significance was assessed by a two-tailed Student's t-test. ∗∗, P < 0.01; ∗∗∗∗, P < 0.0001.

Regarding the efficient pyruvate synthesis of V. natriegens by fermenting formate and sodium gluconate, a follow-up question is whether formate acts as a booster in sodium gluconate consumption. Therefore, we further compared sodium gluconate consumption and pyruvate production of V. natriegens in the presence and absence of formate. As expected, the addition of HCOONa·2H₂O (5 g/L) did promote sodium gluconate consumption and pyruvate production (approximately 1.6 and 1.9-fold increases, respectively) compared to the levels without formate (Fig. 1e and f). Further analysis of the fermentation dynamics (Supplementary Fig. 2) demonstrates that pyruvate synthesis resulted in a gradual pH decrease. Notably, the addition of sodium formate exerted a distinct pH-stabilizing effect (Supplementary Fig. 2e), and led to an increased gluconate consumption rate (6.83 ± 0.97 g/L/h) during the exponential growth phase (8‒12 h) compared to the level without formate supplementation (Supplementary Fig. 2a). Moreover, while sodium formate supplementation exhibited slight growth inhibition during the early phase (0–12 h), cellular growth returned to normal levels in later stages as sodium formate was fully consumed (Supplementary Fig. 2d). Collectively, the significant promoting effect of formate on V. natriegens’ consumption of sodium gluconate and pyruvate synthesis has not been previously reported, providing new insights into optimizing the pyruvate production in this bacterium.

3.2. Formate globally alters gene expression in V. natriegens

To explore the mechanism by which formate enhances sodium gluconate utilization and pyruvate synthesis in V. natriegens, we analyzed the transcriptional changes of all the genes following formate induction. The results showed that a large number of genes exhibited significant changes (2-fold) in their transcriptional levels with the supplementation of formate against the control condition (no formate), in which 950 and 866 genes were up and down-regulated, respectively (Fig. 2a).

Fig. 2.

Comparative transcriptomic analysis of V. natriegens in the presence and absence of formate. (a) Volcano map of differentially expressed genes. Each dot represents one gene. The pink and blue dots indicate the genes exhibiting significantly upregulation and downregulation (FDR ≤0.05 and |log₂FC| ≥ 1), respectively, in the presence of 5 g/L of HCOONa·2H₂O. The other genes were represented by grey dots. Nosig, no significance. (b) The top 20 functional enrichment subsets generated from the data of comparative transcriptomic analysis. M, Metabolism; GIP, Genetic Information Processing; EIP, Environmental Information Processing; CP, Cellular Processes; OS, Organismal Systems. (c) Transcriptional changes in sugar assimilation and glycolysis genes in V. natriegens in response to formate. The upregulated and downregulated genes were highlighted with numbers pink and black, respectively. Data are presented as mean ± standard deviation (n = 3 biologically independent replicates). Error bars show SDs.

KEGG pathway enrichment analysis associated these genes with a spectrum of critical metabolic pathways, encompassing amino acid, carbon, and energy metabolism, as well as membrane transport and signal transduction (Fig. 2b). Given the known toxicity of formate to microorganisms, these findings suggest that V. natriegens substantially rewires its metabolic network in response to formate stress, enabling it to mitigate the stress and subsequently metabolize this organic compound. Notably, we observed that the majority of genes involved in the assimilation of various sugars (sucrose, glucose, fructose) and sodium gluconate were significantly up-regulated upon formate addition (Fig. 2c). This finding aligns with the phenotypic change (i.e., improved utilization efficiency of multiple sugars) observed in V. natriegens supplemented with formate (Fig. 1b).

In contrast, the expression of the formate dehydrogenase-encoding genes (PN96_21155, PN96_22795, PN96_05840, PN96_05845, PN96_05850, PN96_05880) responsible for formate oxidation was not up-regulated in V. natriegens by formate supplementation (Supplementary Table 4). Nevertheless, since the formate oxidation pathway in V. natriegens exhibits inherently high flux [11], it can still generate substantial amounts of NADH.

3.3. Suppressing pyruvate dehydrogenase (PDH)-encoding gene expression significantly enhances pyruvate biosynthesis in V. natriegens

To more effectively accumulate pyruvate, it is necessary to reduce its consumption as a key metabolic intermediate in cells. It is known that the PDH-mediated pyruvate-to-acetyl-CoA flux constitutes a major metabolic node. However, complete inhibition of this pyruvate-to-acetyl-CoA conversion is physiologically untenable and would severely damage cellular growth, because acetyl-CoA serves as an essential precursor for the tricarboxylic acid (TCA) cycle [15]. Therefore, we implemented a transcriptional knockdown approach targeting the PDH-encoding gene to achieve partial reduction of its enzymatic activity while maintaining basal metabolic functions (Fig. 3a).

Fig. 3.

Metabolic pathways related to pyruvate synthesis and consumption in V. natriegens and strain modification strategies. (a) Schematic diagram of the metabolic pathways and genetic engineering strategies. (b) Downregulation of the expression level of the aceE gene in V. natriegens. (c) Substrate consumption and pyruvate production by the P2GTG and P2TTG strains. (d, e) The influence of blocking the bypass pathways on the substrate consumption (d) and pyruvate production (e) of the P2GTG strain. 50 g/L of sodium gluconate 5 g/L of HCOONa·2H₂O were added into the medium when needed. The samples were harvested after 24 h of cultivation. Data are presented as mean ± standard deviation (n = 3 biologically independent replicates). Error bars show SDs. The statistical significance was assessed by a two-tailed Student's t-test. ∗, P < 0.05, ∗∗∗∗, P < 0.0001, ns, no significance.

It has been known that the aceE gene encodes PDH in V. natriegens [15]. To down-regulate the expressional level of aceE, we replaced its native promoter with a previously reported weaker P2 promoter [19], and moreover, changed the start codon of aceE from ATG to GTG/TTG (Fig. 3b). The yielding engineered strains, P2GTG and P2TTG, were then compared with the wild-type strain to evaluate their pyruvate biosynthesis capabilities using sodium gluconate (50 g/L) and formate (5 g/L HCOONa·2H₂O) as dual substrates. The results demonstrated that P2GTG, P2TTG, and the wild-type strain consumed formate completely while also utilizing comparable amounts of sodium gluconate (27.0 ± 2.0, 29.0 ± 2.0 and 28.8 ± 0.9 g/L for the wild-type, P2GTG, and P2TTG, respectively) (Fig. 3c). Interestingly, the P2GTG and P2TTG strains exhibited significantly enhanced pyruvate production (21.9 ± 0.3 g/L and 22.1 ± 0.1 g/L), a 49.0% increase over the wild-type strain (14.7 ± 0.1 g/L) (Fig. 3c). Consequently, the carbon yields of P2GTG and P2TTG in pyruvate synthesis were also markedly improved (reaching 88.5% and 89.2%, respectively), while that of the wild-type strain was 63.2%. These findings indicate the efficacy of the metabolic strategy in suppressing pyruvate-to-acetyl-CoA flux, which improved pyruvate production in V. natriegens during co-fermentation of formate and sodium gluconate. One noteworthy point is that previous studies found that acetate supplementation could also significantly promote glucose utilization and pyruvate synthesis in V. natriegens [20]. Given that acetate metabolism can feed the TCA cycle with acetyl-CoA, supplementing with a certain amount of acetate during the fermentation by the engineered strain P2GTG in this study may compensate for its insufficient TCA cycle flux caused by the intentional suppression of the aceE gene, thereby enhancing the energy supply of the strain. This strategy may further contribute to improving pyruvate synthesis and warrants further investigation.

Next, we selected the engineered strain P2GTG to block the bypass pathways that may affect pyruvate production, aiming to further enhance its pyruvate synthesis capacity (Fig. 3a). These modifications included: (1) deletion of the lactate dehydrogenase genes (ldh1-ldh2-ldh) to prevent the conversion of pyruvate into lactate; (2) deletion of the ppsA (ppsA1-ppsA2) genes to prevent the conversion of pyruvate to phosphoenolpyruvate (PEP); (3) deletion of the atpFH (atpF1-atpF2-atpH) genes to reduce oxidative phosphorylation of pyruvate, a strategy which has been proven effective in enhancing pyruvate synthesis in some microorganisms. Three derivative strains, i.e., P2GTGΔldh, P2GTGΔldhΔppsA, and P2GTGΔldhΔppsAΔatpFH were then compared to the parental strain P2GTG to evaluate their pyruvate production and substrate utilization. However, none of these new strains demonstrated further improvement compared to P2GTG (Fig. 3d and e).

Next, we employed ¹³C-labeled formate to analyze its metabolic flux in the V. natriegens P2GTG strain. Analysis of the ¹³C isotopic enrichment in pyruvate and eleven associated amino acids revealed no detectable labeling in pyruvate; in contrast, substantial ¹³C enrichment was observed in aspartate, lysine, threonine, glutamate, and methionine, while the remaining amino acids showed minimal or no incorporation (Fig. 4a and b). This finding indicates that formate is not directly converted to pyruvate; rather, its enhancement of pyruvate synthesis likely involves indirect regulatory mechanisms. We propose two explanations. First, the protons (H⁺) generated from formate oxidation may augment the proton motive force, thereby driving ATP synthesis to energize the uptake and catabolism of sugars and sodium gluconate, a premise supported by our transcriptomic data (Fig. 2c). Second, since a high NADH level is known to inhibit the pyruvate dehydrogenase complex (PDHC) [21,22], the excess NADH derived from formate oxidation would thus impede the conversion of pyruvate to acetyl-CoA, resulting in pyruvate accumulation in the P2GTG strain.

Fig. 4.

Analysis of formate metabolism in the V. natriegens P2GTG strain using ¹³C-formate. (a and b) ¹³C enrichment in pyruvate and in amino acids associated with the direct (a) and indirect (b) formate assimilation pathways. Cells were cultured in the modified M9 minimal medium with 50 g/L sodium gluconate and 5 g/L HCOONa·2H₂O (HCOONa·2H₂O). The formate was a 1:1 M mixture of ¹³C-labeled and unlabeled formate. (c) Total ¹³C incorporation into the cellular biomass of the P2GTG strain. The ¹³C atomic fraction in biomass was determined by elemental analysis coupled with isotope ratio mass spectrometry according to the protocol in Materials and Methods section. The assimilated formate is not a directly measured value, but rather calculated using the following formula: (¹³C ratio in biomass) × 2 = (assimilated carbon moles from formate in biomass)/(total carbon moles in biomass). Using Sample No.1 as an example: assimilated carbon moles from formate in biomass = 0.0164 × 2 = 0.0328 mol/mol_{total carbon}. Biomass typically consists of 40-50% carbon (w/w). Therefore, 1 g DCW contains approximately 0.45 g carbon, equivalent to 0.0375 mol carbon. Thus, the amount of carbon derived from formate in the biomass is: 0.0375 × 0.0328 = 0.00123 mol/g DCW. Given that the molecular weight of formate is 45.02 g/mol, the corresponding assimilated formate is: 0.00123 × 45.02 = 0.056 g/g DCW, which corresponds to approximately 9.08% (assimilated formate/consumed formate). Data are presented as the mean ± SD (n = 3 biologically independent samples).

To roughly quantify the assimilated formate, we further measured the total ¹³C content in the cellular biomass. The results demonstrated that ¹³C accounted for approximately 1.69% ± 0.07% of the total cellular carbon (mol), implying that about 9.43% ± 0.32% of formate was assimilated into biomass via anabolism (Fig. 4c).

Collectively, these findings indicate that a portion of formate is assimilated and metabolized by the cells, albeit not through direct utilization for pyruvate synthesis.

3.4. High pyruvate production of the P2GTG strain with continuous supplementation of formate

As shown in Fig. 3c, the P2GTG strain, when grown with formate and sodium gluconate as dual substrates, could completely consume formate but left residual sodium gluconate. This suggests that the capacity of P2GTG for formate utilization remains underexploited.

Therefore, fed-batch fermentation with formate supplementation in shake flasks was used to assess P2GTG's potential in formate utilization. The fermentation was initiated with the supplementation of 5 g/L of formate (HCOONa·2H₂O) and 50 g/L of sodium gluconate. Additional 3 g/L of formate was supplemented at either 18 h only or at both 18 h and 24 h. As shown in Table 1, after 33 h of fermentation, the P2GTG strain consumed 34.86 ± 0.01 g/L of sodium gluconate and all the initially supplemented formate (FA + condition), producing 24.15 ± 0.10 g/L of pyruvate. When an additional 3 g/L of HCOONa·2H₂O was supplemented at 18 h (FA++ condition), P2GTG utilized 38.08 ± 0.05 g/L of sodium gluconate along with all the formate, resulting in a higher pyruvate yield (29.36 ± 0.09 g/L). Further formate supplementation at both 18 and 24 h (FA+++ condition) led to complete consumption of all the sodium gluconate and formate, with pyruvate production reaching 34.85 ± 0.18 g/L, approximately 44.3% higher than the control (FA+). These results further confirm that formate as an auxiliary substrate can significantly promote sodium gluconate metabolism in V. natriegens. Fed-batch strategies could unlock P2GTG's formate assimilation capacity, leading to enhanced pyruvate biosynthesis.

Table 1.

The effect of formate supplementation on pyruvate production by the engineered V. natriegens strain P2GTG.

Samplesa	Total formate added (g/L)	Total sodium gluconate added (g/L)	Residual formate (g/L)	Residual sodium gluconate (g/L)	Pyruvate production (g/L)
FA+	5.0	50	0	15.14 ± 0.01	24.15 ± 0.10
FA++	8.0	50	0	11.92 ± 0.05	29.36 ± 0.09
FA+++	11.0	50	0	0	34.85 ± 0.18

^aSymbols “+”, “++”, and “+++” denote different sodium formate supplementation strategies: “+”, addition of 5 g/L at 0 h; “++”, addition of 5 g/L at 0 h and 3 g/L at 18 h; “+++”, addition of 5 g/L at 0 h, and 3 g/L at both 18 h and 24 h. Data are presented as mean ± standard deviation (n = two independent samples). Formate used here was HCOONa·2H₂O.

Based on the above findings, we further investigated the potential of the P2GTG strain in formate utilization and pyruvate production with continuous supplementation of substrates in bioreactors. The fermentation was carried out in triplicate (Fig. 5), with initial supplementation of 30 g/L of sodium gluconate and 5 g/L of formate (HCOONa·2H₂O) in the medium. When the cell density reached OD₆₀₀ ∼1.0, additional formate and sodium gluconate was added with a rate of 2 g/L/h and 6 g/L/h, respectively. In subsequent fermentation stage, the feeding rates of formate and sodium gluconate were controlled to maintain their concentrations below 10 g/L and 30 g/L, respectively (Fig. 5). In the fermentation experiments, the P2GTG strain was able to consume all of the formate (approximately 37.4 g/L) and more than 80% (approximately 82.8 g/L) of the sodium gluconate in each round (Fig. 5 and Table S5). Furthermore, across the three experiments, the P2GTG strain produced 61.1 g/L, 54.5 g/L and 53.5 g/L (average is 56.4 g/L) of pyruvate, with corresponding pyruvate productivity rates of 1.2 g/L/h, 1.1 g/L/h and 1.0 g/L/h (average is 1.1 g/L/h), and yields of 0.7 mol/mol, 0.7 mol/mol and 0.5 mol/mol, respectively (Table S5), demonstrating a stable pyruvate production capacity. Additionally, the by-products lactate and acetate were detected in the fermentation broth in all three rounds (Fig. 5), among which the lactate concentration was relatively high, reaching 1.2, 5.5 and 17.1 g/L (Table S5). Considerable variability in the maximum biomass (Maximum OD₆₀₀) and lactate production was observed across the three rounds, which may be attributed to unstable dissolved oxygen control during different fermentation rounds. This provides valuable insights for the subsequent optimization of pyruvate production by the P2GTG strain.

Fig. 5.

Fermentation profiles of the P2GTG strain with continuous supplementation of formate and sodium gluconate. (a‒c) First (a), second (b), and third (c) rounds of fermentation. Left panels showed the consumption of formate and sodium gluconate. Right panels showed cell growth and the production of pyruvate, acetate, and lactate.

Moreover, previous studies have revealed that V. natriegens produced a polymeric byproduct, parapyruvate, during its pyruvate synthesis [20]. Accordingly, we analyzed the parapyruvate content in the fermentation broth of the P2GTG strain. The results revealed the presence of approximately 10.6 g/L of parapyruvate (Supplementary Fig. 3). Evidently, the formation of this byproduct reduces pyruvate yield and leads to carbon loss, representing an issue that should be addressed in future strain engineering efforts.

Formate represents an important one-carbon organic resource. While V. natriegens has been reported to exhibit relatively efficient formate metabolism [11], its capability to effectively convert formate for targeted product biosynthesis remained unclear. A key finding of this study reveals that V. natriegens not only efficiently utilizes formate as a substrate but also employs it as a potentiator to enhance the consumption of the primary carbon source, sodium gluconate, thereby enabling highly efficient dual-substrate fermentation for pyruvate production. This intriguing phenomenon could be attributed to V. natriegens’ high energy requirements during formate assimilation, which potentially upregulates glycolytic metabolism of sodium gluconate to meet ATP demands.

Although this study demonstrates efficient pyruvate production in V. natriegens using formate and sodium gluconate as dual substrates, considerable potential remains for further optimization. Firstly, the formate utilization efficiency in the dual-substrate fermentation remains suboptimal. Formate consumption was markedly lower than sodium gluconate consumption (Table S5). To further increase the proportion of formate in dual-substrate fermentation, laboratory adaptive evolution of the aforementioned V. natriegens P2GTG strain under formate stress could be a highly effective strategy. In addition, a more challenging goal is to enable the V. natriegens P2GTG strain to efficiently grow and produce pyruvate using formate as the sole carbon source (rather than as a co-substrate), which hinges on fundamentally modulating its formate dissimilation and assimilation. To this end, a feasible strategy involves attenuating formate dissimilation to CO₂ while enhancing formate assimilation pathways to establish optimal metabolic equilibrium in V. natriegens. Achieving this goal will require both metabolic engineering of formate utilization pathways and laboratory adaptive evolution under formate-selective pressure.

4. Conclusions

This study demonstrates the potential of V. natriegens to utilize formate as an auxiliary substrate for efficient pyruvate production. Formate metabolism not only supplied an additional carbon source but also accelerated sodium gluconate (the primary substrate) consumption and pyruvate formation. The pathways associated with pyruvate metabolism in V. natriegens were analyzed and repressing the conversion of pyruvate to acetyl-CoA significantly enhanced pyruvate accumulation. Isotopic tracing and RNA-seq analyses indicate that formate enhances pyruvate synthesis in V. natriegens indirectly, rather than serving as a direct precursor. Specifically, the additional energy derived from formate oxidation appears to facilitate sugar uptake and glycolytic flux, thereby increasing pyruvate synthesis efficiency. The findings and strain modification strategies in this study will be useful for formate bioutilization. Future work should focus on increasing the substitution ratio of formate for sodium gluconate and ultimately using formate as the sole substrate for V. natriegens in pyruvate production.

CRediT authorship contribution statement

Wangshuying Deng: Writing – original draft, Validation, Methodology, Funding acquisition, Conceptualization. Xinyu Zhang: Writing – original draft, Validation, Data curation. Jiayin Liu: Validation. Wei Ye: Validation. Xianfeng Zhu: Supervision, Resources. Weihong Jiang: Supervision, Project administration. Yang Gu: Writing – review & editing, Project administration, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no conflicts of interest or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: