Integrating Green Process Engineering With Whole‐Cell Catalysis to Intensify Sustainable Glucaric Acid Production

ABSTRACT

This study pioneered a whole‐cell biocatalytic process using engineered Gluconobacter oxydans to produce glucaric acid (GA), a food additive, from glucuronic acid (GlcA). From a molecular perspective, molecular dynamics simulations have confirmed that van der Waals forces and hydrogen bond interactions in acidic environments effectively stabilized the structure–activity relationship of dehydrogenase and GlcA. Leveraging this insight, multi‐strategies were implemented, including pH regulation, high‐density biocatalysis, dehydrogenase overexpression, adaptive acclimation, bioreactor scale‐up, and cell recycling technology, to significantly intensify the GA production. As a result, the recombinant G. oxydans achieved an effective production of 10.27 g/L GA with a productivity of 0.21 g/L/h, representing a 366.4% intensification over the unregulated system. In comparison with the highest level of biological methods reported to date, the process demonstrates greater economic feasibility and industrial application potential. This work elucidated the molecular mechanism of acid‐intensified biocatalysis and provided a sustainable, industrially viable GA production platform.

Cascade intensified bioproduction of gluconic acid

1. Introduction

With excessive consumption of non‐renewable fossil resources, the world faces severe energy crises and environmental degradation. (Gérardy et al. 2020). Biomass, as the only renewable carbon source, offers a critical pathway to produce bio‐based chemicals and fuels, enhancing energy security and sustainability (Queneau and Han 2022). Developing innovative biocatalytic technologies is therefore strategically vital for green development (Wohlgemuth 2021). In 2004, the Pacific Northwest National Laboratory (PNNL) and the National Renewable Energy Laboratory (NREL) jointly identified 12 biomass‐derived platform chemicals, among which glucaric acid (GA) was recognized as one of the most promising compounds from biorefining (Werpy and Petersen 2004). GA, alternatively referred to as saccharic acid, is a saturated dicarboxylic acid predominantly derived from the oxidation of glucose. Owing to the multifunctional properties of various chemical forms such as calcium salts, potassium salts, and lactones, GA can serve as a polymer precursor, building material additive, cleaning agent, food additive, and drug molecule (Kapanji et al. 2021; Zhang, Wan, et al. 2021). The global GA market is projected to increase from USD 800 million in 2020 to USD 1.47 billion by 2027, representing a compound annual growth rate (CAGR) of 7.5% during the forecast period from 2022 to 2027 (Machado et al. 2023).

The commercialization of GA has been delayed and remains costly due to the absence of economically viable technologies in industrial production. Currently, major producers continue to employ nitric acid as a homogeneous catalyst in the synthesis of GA. However, this method generates significant amounts of NO₂, leading to concerns regarding both operational efficiency and environmental impact. Additionally, other synthetic approaches, such as electrochemical (24% yield), photochemical (12% yield), and heterogeneous catalysis (82% yield), still exhibit limitations in terms of product selectivity (Deng et al. 2021; Liu et al. 2020; Moggia et al. 2021; Wu et al. 2024; Zhang, Jiang, et al. 2021). Apart from chemical methods, fermentation is currently considered by scientists as the most promising approach to overcome the bottleneck in industrial production. However, whether through genetic engineering, the synergistic action of bacteria and enzymes, or the regulation of fermentation systems, the production and the yield generally remain below 10 g/L and 20% due to competition within central metabolic pathways and the formation of by‐products (Marques et al. 2020; Petroll et al. 2020).

Gluconobacter oxydans, with its unique physiological and biochemical characteristics, has emerged as a widely utilized industrial biocatalyst due to its capacity for region‐ and stereoselective catalysis mediated by a series of membrane‐bound dehydrogenases (Prust et al. 2005). Since these dehydrogenases are localized on the cell membrane, substrates can undergo reactions without entering the cytoplasm. Consequently, compared with fermentation methods, the whole‐cell catalytic technology employing G. oxydans effectively circumvents interference from metabolic pathways, thereby enhancing the yield of the target product (Da Silva et al. 2022). Moreover, G. oxydans exhibits strong tolerance to high concentrations of substrates, products, salts, and acids (Chen et al. 2025). Leveraging these advantages, G. oxydans has been successfully implemented in the industrial‐scale production of ascorbic acid, sorbose, 1,3‐dihydroxyacetone, and various sugar acids (Hua et al. 2024; Ripoll et al. 2023). Currently, the preparation of various sugar acids has been successfully achieved by G. oxydans at both industrial and laboratory scales, including gluconic acid, galactonic acid, mannonic acid, xylonic acid, and arabonic acid (Hua, Han, Liu, et al. 2023; Madadi et al. 2023; Yao et al. 2017; Zhou et al. 2018). These sugar acids are all produced through the oxidation of the aldehyde group at the first carbon position by membrane‐bound glucose dehydrogenase. However, as illustrated in Figure 1, glucuronic acid (GlcA), which also possesses an aldehyde group at the same position, has not yet been confirmed to undergo dehydrogenation to form GA. Therefore, this study aimed to overcome the production bottleneck of GA from GlcA using G. oxydans through multiple regulatory strategies, with the objective of maximizing yield and productivity, thereby providing a technically and economically viable biotechnological solution for the industrial production.

Figure 1.

The confirmed and unconfirmed bioconversion pathway of sugar to sugar acid by whole‐cell catalysis with G. oxydans.

2. Materials and Methods

2.1. Strains, Plasmids, and Media

G. oxydans NL71, derived from G. oxydans ATCC621H, was domesticated and preserved by the Institute of Biochemical Engineering at Nanjing Forestry University through long‐term adaptation in lignocellulosic hydrolysate. Escherichia coli DH5α was maintained in our laboratory and used for plasmid construction and propagation. The expression vector pBBR1MCS‐5 was obtained from Wuhan Miaoling Bioscience & Technology Co. Ltd., and all primers were synthesized by Shanghai Generay Biotech Co. Ltd.

2.2. Reagents and Chemicals

Gentamicin sulfate was purchased from Macklin Biochemical Co. Ltd. (Shanghai, China). Plasmid extraction kits, gel purification kits, high‐fidelity DNA polymerase, and the restriction enzymes BamHI and HindIII were obtained from Bioengineering (Dalian) Co. Ltd. Yeast extract was purchased from Sigma‐Aldrich (USA). Other reagents involved in biocatalysis and chromatographic elution buffers were of analytical grade and commercially available.

2.3. Plasmid Construction and Primer Design

To achieve heterologous expression of membrane‐bound glucose dehydrogenase (mGDH) in G. oxydans, the expression plasmid pBBR‐GDH was constructed based on the broad‐host‐range vector pBBR1MCS‐5 (Shen et al. 2020). The mGDH coding sequence was amplified by PCR using gene‐specific primers designed with 15–20 bp homologous overhangs for seamless cloning. The pBBR1MCS‐5 plasmid was linearized by inverse PCR using the high‐fidelity 2× Phanta Max Master Mix (Vazyme, #P515). The primer sequences are presented in Table 1.

Table 1.

The primer sequences of dehydrogenase and plasmid.

Primer	Primer sequence
GDH‐F	ATTTCACACAGGAAACAGCT
GDH‐R	TTTTAACAAAATATTAACGC
pBBR1‐F	GCGTTAATATTTTGTTAAAATTCGCGTTAAATTTTTGTTAAATCAG
pBBR1‐R	AGCTGTTTCCTGTGTGAAATTGTTATCCGC

The PCR product was digested with DpnI to remove the methylated template plasmid and purified via gel extraction. The purified mGDH fragment and linearized vector were assembled using the ClonExpress Ultra One Step Cloning Kit (Vazyme), resulting in seamless, single‐fragment homologous recombination into the pBBR1 backbone.

2.4. Transformation and Verification in E. coli

The recombinant plasmid (designated pBBR‐GDH) was introduced into chemically competent E. coli DH5α cells (100 μL) by heat‐shock transformation. Transformants were selected on LB agar plates containing 50 μg/mL gentamicin. Positive colonies were identified by colony PCR, diagnostic restriction enzyme digestion, and verified by Sanger sequencing to confirm the correct insertion and orientation of the mGDH gene.

2.5. Electroporation and Expression in G. oxydans

The verified pBBR‐GDH plasmid was introduced into G. oxydans via electroporation. Electrocompetent cells were prepared following standard protocols. After electroporation, 100 μL of transformed cells were plated on sorbitol agar supplemented with 50 μg/mL gentamicin and incubated at 30°C for 4–5 days. Colonies were further cultured in 5 mL liquid medium (30°C, 220 rpm, 12–16 h) with 50 μg/mL gentamicin. Plasmids were extracted and verified by PCR to confirm the successful generation of recombinant G. oxydans strains carrying the mGDH gene.

2.6. Adaptive Acclimation of Recombinant G. oxydans

To improve the tolerance and growth performance of recombinant G. oxydans in high‐GlcA environments, a stepwise adaptive acclimation strategy was implemented. The process involved gradually increasing GlcA concentrations in agar culture with gentamicin. The adaptive evolution proceeded through four stages, with increasing proportions of GlcA: Stage 1 (20% GlcA:80% glucose), Stage 2 (40%:60%), Stage 3 (60%:40%), and Stage 4 (80%:20%). At each stage, stable colonies exhibiting robust growth were selected and transferred to the next round of adaptation. All cultures were maintained in medium containing 50 μg/mL gentamicin to ensure plasmid stability.

2.7. Whole‐Cell Biocatalysis

The whole‐cell biocatalysis in Erlenmeyer flasks equipped with baffles was performed using either unadapted recombinant strains or recombinant strains previously adapted to various concentrations of GlcA (as described in Section 2.4). The broth consisted of 10 g/L GlcA, 5 g/L yeast extract, 0.5 g/L MgSO₄·7H₂O, 1 g/L K₂HPO₄, 2 g/L KH₂PO₄, 5 g/L (NH₄)₂SO₄. The bioconversion was conducted in a temperature‐controlled shaker at 30°C and 220 rpm for 48 h (Hua et al. 2023).

2.8. Determination of Cell Activity

The activity assays of G. oxydans under different conditions were conducted in a 96‐well microtiter plate containing a 200 μL mixture composed of 170 μL dichlorophenolindophenol (195 μM)/phenazine methosulfate (135 μM), 20 μL GlcA (10 g/L), and 10 μL broth of G. oxydans (OD₆₀₀ = 8) (Peters et al. 2013). The 96‐well plate was positioned in a Microplate Reader (Inifinite 200 PRO, Tecan Austria GmbH) configured to operate at 220 rpm and a temperature of 30°C. The absorbance at 580 nm was measured every 30 s, and the cell activity was calculated using the following formula:

$$U=\frac{V_t\frac{∆A}{∆t}}{V_{s}l\epsilon }$$ (1)

where V _t and V _s denoted the volume of mixture (200 μL) and broth of G. oxydans (10 μL); l represented the optical path length of the cuvette (0.5 cm); ε referred to the molar extinction coefficient of dichlorophenolindophenol (5800 L (mol cm)⁻¹) (Figure S1).

2.9. Molecular Dynamics Simulation

The protein structure of mGDH from G. oxydans and the molecular structures of GlcA were downloaded from UniPort and PubChem. After processing the structure of proteins and small molecules using Autodock 1.5.6, molecular docking was performed using AutoDock Vina software to explore the interactions between protein ligands. After the docking simulations were completed, select the site with the highest docking score for molecular dynamics simulations (Forli et al. 2016). Molecular dynamics simulations were performed using Amber22 software (San Francisco, CA, USA), employing the ff19SB force field to calculate system force field parameters (Tian et al. 2020). The systems with different degrees of protonation, that is, different pH environments, were constructed using Constant pH Methods of Chemistry at HARvard Macromolecular Mechanics (CHARMM) (Hwang et al. 2024). Finally, AmberTools23 was used to analyze the dynamics results, including binding free energy and hydrogen bonding (Case et al. 2023). Additionally, the detailed simulation information was provided in the Supplementary Information.

2.10. Chromatographic Analysis

High‐performance anion‐exchange chromatography (HPAEC) was used to quantify chemical compounds involved in the biotransformation process, including GA and GlcA (Hu et al. 2024). The analysis was performed using a Dionex ICS‐6000 system equipped with a PA200 analytical column, a pulsed amperometric detector, and a gold working electrode. The mobile phase consisted of sodium hydroxide (100 mmol/L) and sodium acetate (500 mmol/L), delivered at a flow rate of 0.3 mL/min. The specific elution program was shown in the corresponding Figure S2.

All experiments were performed in at least three independent biological replicates (n = 3). Data were presented as mean ± standard deviation (SD) at each time point. For kinetic time course data, error bars represented SD at each time point. For multiple group comparisons, one‐way analysis of variance (ANOVA) was performed using IBM SPSS Statistics 26. followed by Tukey's post hoc test for pairwise comparisons. Statistically significant differences (p < 0.05) were indicated by different superscript letters (a, b, c, …) in the figures; groups sharing the same letter are not significantly different, whereas groups with different letters differ significantly (p < 0.05). For two group comparisons, a two tailed Student's t test was performed, and significance was indicated by an asterisk (*p < 0.05). All statistical tests were performed with a significance threshold of α = 0.05.

3. Results and Discussion

3.1. The Whole‐Cell Catalysis of GA With Different pH Conditions

As one of the most valuable biorefining products, GA necessitates the development of green and efficient preparation technologies to replace traditional, outdated methods, which is an urgent priority. Previous studies have demonstrated that G. oxydans is capable of oxidizing the aldehyde group at the primary position of aldose to a carboxylic acid group (Yao et al. 2017). However, the bioconversion pathway involved in the transformation of GlcA into GA through whole‐cell catalysis has not been fully elucidated. To evaluate the dehydrogenation capability of G. oxydans toward GlcA, the whole‐cell catalysis were initially performed under conventional conditions, specifically at neutral pH (Figure 2a,b). The results indicated that under pH conditions of 6 and 7, only 0.43 and 0.22 g/L of GA were produced, respectively, rendering the process economically unviable compared to the chemical method. Nevertheless, despite the unsatisfactory production, the experiment successfully demonstrated the feasibility of the conversion pathway.

Figure 2.

The bio‐performance of whole‐cell catalysis of glucuronic acid by G. oxydans in different pH environments. (a–f) The kinetic process of whole‐cell catalysis. (g) The growth of G. oxydans in different pH environments. (h) The comparison of cell activity using glucuronic acid as a substrate in different pH environments.

According to previous studies by Pyo et al. and Hua et al. under acidic environmental conditions, the membrane‐bound alcohol dehydrogenase of G. oxydans exhibited notable catalytic activity toward substrates that were not typically dehydrogenated (Hua et al. 2025; Pyo et al. 2020). Research indicated that although acidic conditions inhibited the catalytic activity of dehydrogenase, the protonated state of enzyme proteins could significantly enhance their binding ability to substrates, thereby improving catalytic efficiency. Therefore, whole‐cell catalysis in acidic environments was further conducted, and the results are shown in Figure 2c–f. It was found that as the acidity of the broth increased, the accumulation of GA significantly rose. When the pH was 4 (Figure 2d), the production of GA was the highest, reaching 2.26 g/L, which was a 425.6% increase compared to the unregulated experimental group (Figure 2b). When the broth was in a strongly acidic condition, the production of GA showed a downward trend. This was because the excessively low pH caused instability in the enzyme structure, leading to partial inactivation, and also damaged the integrity of the cell membrane, affecting the regeneration ability of coenzymes (such as PQQ). The growth of G. oxydans in Figure 2g confirmed the reason for the decrease in GA production in acidic environments. The results showed that in an extremely acidic environment (pH = 2), the cell structure would be impaired, and some bacteria would initiate autolysis programs, resulting a reduction in bacterial concentration from the initial OD₆₀₀ = 8 to below 5. As the pH environment neared neutrality, even if G. oxydans did not metabolize GlcA for growth, the maintenance of bacterial concentration also improved, which contradicted the highest GA production performance at pH 4. Moreover, Figure 2h elucidated this perspective by comparing cell activity (serving as an indicator of overall dehydrogenase activity) in different pH environments. The results demonstrated that when using GlcA as a substrate, cell activity initially increased with the rise in pH, reaching the maximum viability value of 0.34 U at pH 4, and then decreased as the pH increased. Therefore, even though the maintenance of bacterial cells was not as favorable as in a neutral environment, the cell activity exhibited a higher level at pH 4, and this pH‐dependent difference accounted for the phenomenon of the highest yield at pH 4. Overall, the pH optimization experiment not only confirmed the feasibility of GA generation but also offered technical support for enhancing its production.

3.2. Molecular Dynamics Simulation in Different Protonation Degree Systems

The notable variations in the activity of G. oxydans across different pH environments could be attributed to alterations in the structure–activity relationship between mGDH and GlcA. To gain insights into the molecular‐level interactions between the protein and the ligand, molecular dynamics simulations were performed, with the results presented in Figure 3. Figure 3a,b illustrates the 2D and 3D interaction profiles between the compound and key amino acid residues. The results indicated that the ARG265 and LYS498 residues on the protein receptor engage in hydrogen bonding interactions with GlcA, whereas the other 14 residues participated in van der Waals interactions with the GlcA. It was evident that hydrogen bonds and van der Waals forces serve as the primary interaction forces responsible for stabilizing the conformation of the complex.

Figure 3.

The structure–activity relationship between membrane‐bound glucose dehydrogenase and glucuronic acid at the molecular level. (a, b) The 2D and 3D schematic diagrams of the interaction between key amino acid residues of membrane‐bound glucose dehydrogenase and glucuronic acid. (c) The trend of changes in the total binding free energy and van der Waals forces in different pH environments. (d) The trend of changes in the number of hydrogen bonds in different pH environments.

Therefore, molecular dynamics simulations were employed to further analyze the interaction between mGAD and GlcA using the MM/PBSA method, with a focus on the contributions of van der Waals forces and hydrogen bonds. Figure 3c illustrates the variations in total binding free energy and van der Waals interactions between the receptor and ligand under different protonation states. The total binding free energy increased progressively with rising pH levels, because in acidic environments, the amino acid residues at the receptor active site are protonated and positively charged, which can then form stable hydrogen bonds and electrostatic interactions with negatively charged GlcA (Hua et al. 2025). It was worth noting that as the pH increased from 5 to 6, the binding free energy of synthesis jumped from −5.13 to 0.56 kcal/mol. Generally speaking, when the binding energy is less than −5.0 kcal/mol, it indicates good binding activity between the two, and less than −7.0 kcal/mol indicates strong binding activity (Fu et al. 2022). Therefore, only in acidic environments was the binding between mGDH and GlcA more stable, which confirmed the macroscopic biocatalytic results. Moreover, the trend of changes between van der Waals force and pH was consistent with the total free energy, because the increase in pH led to deprotonation of the polar region of the substrate, which in turn increased the contact area between the non‐polar part and the non‐polar pocket of the enzyme, thereby reducing the binding ability of ligand and protein under neutral conditions. It should be noted that this sharp pH‐dependent binding transition was observed only for the specific mGDH‐GlcA pair investigated in this study. Whether similar behavior existed for other G. oxydans dehydrogenases remained unknown and required further investigation. The number of hydrogen bonds, another key factor contributing to the specificity and stability of the interaction between mGDH and GlcA, varied with pH, as illustrated in Figure 3d. Due to the challenges in achieving a stable binding state between the ligand and the protein under strongly acidic conditions, only the hydrogen bond data for the pH range of 4 to 7 are presented. The results indicated a general decrease in the number of hydrogen bonds as pH increased, which could be attributed to changes in the protonation states of amino acid residues. Overall, the variation patterns of the two primary interaction forces that facilitated the formation of a stable mGDH‐ GlcA complex with respect to pH suggested that the binding affinity and complex stability were greater under acidic conditions, thereby favoring the biotransformation of GlcA.

3.3. The Intensification of Biocatalysis by Optimizing Inoculation and Overexpressing Dehydrogenase

Through molecular‐level analysis, it has been confirmed that an acidic environment facilitated the conversion of GlcA to GA by G. oxydans. Therefore, a variety of strategies should be implemented to intensify bioprocess, thereby enhancing production efficiency and establishing a more reliable biotechnological pathway for industrial applications. Optimization of inoculation amount is the core link in the development of biocatalytic processes, which requires comprehensive consideration of biological characteristics, kinetic parameters, and economic factors to achieve efficient, stable, and low‐cost industrial production (Zhang, Jiang, et al. 2021). Figure 4 illustrates the kinetic profile of GlcA being catalyzed to GA by whole‐cell catalysis of G. oxydans under a constant pH of 4.0 and varying inoculation levels (OD₆₀₀ 2–8). The optimization results showed that the inoculation amount had a significant effect on yield improvement. The GA production under the condition of OD₆₀₀ of 10 was 4.01 g/L, which was 83.9% higher than the production of 2.18 g/L under the condition of OD₆₀₀ of 2. However, increasing the initial inoculation amount from an OD₆₀₀ of 8 to 10 resulted in only a marginal increase in production, from 3.96 to 4.01 g/L (an increase of 1.3%). This phenomenon might be attributed to the negative feedback inhibition of catalytic activity associated with high cell density, such as substrate competition and oxygen limitation. Therefore, taking into account both the cost and time required for bacterial culture, an OD₆₀₀ value of 8 represented the optimal inoculation condition. Moreover, to fully understand the relationship between inoculum amount and process performance, cell growth and activity were measured under the conditions of Figure S3. Over the entire reaction course, no significant increase in OD₆₀₀ was observed for any inoculum size (Figure S3a), confirming that GlcA was not utilized for cell growth but was specifically converted to GA by mGDH. Additionally, when the specific mGDH activity was measured after normalizing all samples to the same cell density, no notable difference was found among the different inoculum amount groups (Figure S3b), indicating that the per‐cell catalytic capacity was independent of the initial cell density. Therefore, the positive correlation between inoculum amount and GA titer (Figure 4) arose solely from the increased total biocatalyst mass, which accelerated substrate consumption and product formation without altering growth or specific activity.

Figure 4.

The kinetic process of whole‐cell catalysis of glucuronic acid by G. oxydans with different inoculation amounts. (a–e) OD₆₀₀ of 2–10. (f) The comparison of GA production with different inoculation amounts.

The current production capacity still fell short of meeting industrial requirements. This limitation might primarily stem from the fact that the expression of dehydrogenase in the wild‐type G. oxydans is naturally regulated at the genomic level, which constitutes a key rate‐limiting step in the biotransformation process (Prust et al. 2005). Therefore, considering that mGDH is a key enzyme in GA production, this study aimed to increase its expression through genetic engineering techniques to intensify the catalytic efficiency. The growth performance of the recombinant strain was firstly evaluated, and the results are presented in Figure 5a. From the growth curve of the recombinant strain, it could be seen that 0–6 h was the lag period, 6–18 h was the logarithmic growth period, and 18–24 h was the stable period. Following 24 h of cultivation under nutrient‐rich conditions, the final biomass reached an OD₆₀₀ of 3.6, comparable to the growth level observed in the wild‐type strain. This suggested that the metabolic burden and physiological stress associated with dehydrogenase overexpression did not adversely affect the growth performance of the recombinant strain. Figure 5b illustrates the dehydrogenation activity of the recombinant strain toward GlcA under various pH conditions, aiming to rapidly identify the optimal pH condition. The results demonstrated that the recombinant strain exhibited high activity within a pH range of 3.5 to 5, which aligned closely with the catalytic kinetic behavior observed in the wild‐type strain. This indicated that the overexpression of dehydrogenase in the recombinant strain did not affect the pH sensitivity of the enzyme's active site, thereby reflecting, to some extent, the structural and functional stability of the recombinant system. At pH 4, the recombinant strain demonstrated the highest activity, achieving a value of 0.57 U. Consequently, under identical experimental conditions, whole‐cell catalysis was conducted, and the corresponding results are presented in Figure 5c. After 48 h of biocatalysis, 5.36 g/L of GA was simultaneously produced, and the efficiency of recombinant strain was further improved by 35.4% compared to the original strain, thus demonstrating the effectiveness of the gene editing‐based reinforcement strategy. For direct comparison, quantitative summary of the differences in the performance of the wild‐type strain and recombinant strain is provided in Supplementary Table S1, which confirmed that the enhancement in GA production was specifically attributed to mGDH overexpression.

Figure 5.

Physiological characteristics and biocatalytic performance of recombinant G. oxydans by overexpressing membrane‐bound glucose dehydrogenase. (a) Growth performance of recombinant G. oxydans. (b) The catalytic activity of recombinant G. oxydans towards glucuronic acid in different pH environments. (c) The kinetic process of whole‐cell catalysis of glucuronic acid by recombinant G. oxydans under pH 4. (d) The comparison of catalytic performance between recombinant and original G. oxydans.

3.4. The Intensification of Biocatalysis by Adaptive Acclimating

To facilitate the gradual adaptation of the recombinant G. oxydans to elevated enzyme expression pressure, thus enhancing the system's resilience against functional disruptions and improving its industrial robustness, adaptive acclimation experiments were carried out. The growth of cells with an increase in GlcA concentration from 0% to 80% could be found in Figure S4. Furthermore, colonies that exhibited stability and robust growth under each concentration ratio condition were selected for amplification, followed by the implementation of whole‐cell catalysis, and the results are presented in Figure 6. The kinetic results indicated that during the early adaptive acclimation phase, the biocatalytic performance of G. oxydans progressively improved. The strain isolated from the solid medium containing 60% GlcA (30 g/L) demonstrated the highest efficiency, achieving a GA production of 6.50 g/L within 48 h. However, when the GlcA concentration was further increased, the catalytic activity of the selected strain markedly declined, resulting in only 2.53 g/L of GA. Potential explanations for this decline included the possibility that a GlcA concentration of 30 g/L might have reached the substrate inhibition threshold for dehydrogenase activity, or that the elevated salt concentration might have induced osmotic stress, adversely affecting cell function (Chen et al. 2025). Additionally, a non‐linear curve fitting was performed between the GA accumulated by whole cell catalysis of the strains selected in each domestication round and the acclimation concentration, as shown in Figure 6f. The results indicated that the relationship between the two followed a polynomial curve, with the equation y = −5.98 ×10⁻⁷ x ⁴ + 4.33 × 10⁻⁵ x ³ + 4.13 × 10⁻⁵ x ²−0.006x + 5.04, and the correlation coefficient R ² was 0.998. Based on the equation, it was further calculated that when the proportion of GlcA in the solid culture medium was 54.08% (27 g/L), the acclimated strain had the best catalytic effect. Therefore, 10 generations of acclimation were conducted on the recombinant G. oxydans at this concentration, and finally the acclimated strains were amplified and preserved for future use. The stability of the adapted phenotype was then evaluated by serial subculture for 10 generations in the absence of selection pressure. As shown in Figure S5, the relative GA conversion ability gradually declined from generation 1 (100%, reference) to generation 10 (46%). The values for generations 2–9 were 94%, 87%, 81%, 74%, 68%, 62%, 56%, and 51%, respectively. These results indicated that the improved GA production capacity acquired through adaptive acclimation is only partially stable and tends to revert under non‐selective conditions. Therefore, to maintain high GA productivity in long‐term or repeated‐batch processes, periodic application of selection pressure (e.g., maintaining a minimum GlcA proportion in the culture medium) was recommended.

Figure 6.

The biocatalytic performance of G. oxydans with different degrees of acclimation. (a–e) The kinetic process of whole‐cell catalysis by G. oxydans screened on culture media with different ratios of glucuronic acid concentration (0%, 20%, 40%, 60%, 80%). (f) Nonlinear curve fitting of the catalytic performance of G. oxydans with different degrees of acclimation.

3.5. The Intensification of Biocatalysis by Improving Oxygen Transfer in Bioreactor

To ensure process robustness and economic viability, it was essential to carry out scale‐up validation experiments in bioreactors. Notably, G. oxydans is a strictly aerobic microorganism that utilizes oxygen as the terminal electron acceptor during the whole‐cell catalysis (Hua and Xu 2025). Consequently, the efficiency of oxygen transfer significantly influences the catalytic performance, and the shake flask system, which relies solely on surface aeration, is evidently inadequate to meet the oxygen demand. Building upon the precise regulation of oxygen partial pressure with respect to cell respiratory intensity and the oxygen transfer coefficient, our team previously developed the SOS‐BR system (Figure S6), which enabled automated and efficient oxygen supply in real time. Using SOS‐BR, a scale‐up experiment was conducted for the whole‐cell catalysis of GlcA by G. oxydans, and the results are presented in Figure 7. After 48 h of biocatalytic reaction, the production of GA reached 10.27 g/L, with an average productivity of 0.21 g/L/h. However, real‐time productivity declined significantly from an initial rate of 0.89–0.03 g/L/h, suggesting that prolonging the reaction time negatively affected the economic viability of the process. Therefore, cell recycling experiments were carried out in the SOS‐BR system, and the results are presented in Figure 7b. The results indicated that the catalytic performance of G. oxydans decreased as the number of recycling cycles increased. Notably, in the fifth cycle, GA production reached only 2.98 g/L, representing a significant decline of 58.8% compared to the previous cycle. To investigate the cause of the productivity drop after the fifth recycling cycle, cell viability was assessed by flow cytometry at the end of each cycle. As shown in Figure S7, viability remained relatively high (≥ 85%) for the first four cycles but declined precipitously to 46% after cycle 5. This strong correlation between viability loss and reduced GA production suggested that membrane integrity damage and cell death were primarily responsible for the diminished performance in later cycles. From an engineering economics perspective, it was therefore advisable to limit cell recycling to four cycles. As a result, in each of these cycles, GA production of 10.27, 9.66, 8.89, and 7.23 g/L were obtained, respectively, resulting in a total mass of 108.2 g (3 L broth volume).

Figure 7.

The whole‐cell catalysis of glucuronic acid by G. oxydans in a sealed oxygen‐supplied bioreactor. (a) Changes in the concentrations of various substances during the 48‐h batch whole‐cell catalysis. (b) The trend of changes in the biocatalytic performance of G. oxydans during cell recycling process.

In this study, GA production was significantly intensified through multiple improvement strategies, and the individual intensification effects of these methods were 75.2%, 137.2%, 187.6%, and 366.4%, respectively (Figure 8). Currently, there are two notable cases reporting the highest levels of GA production via biological methods. Zhao et al. employed Saccharomyces cerevisiae as the fermentation strain and achieved a GA production of 15.6 g/L over 192 h through the fused expression of myo‐inositol oxygenase and uronate dehydrogenase, resulting in an average productivity of 0.08 g/L/h (Zhao et al. 2023). Krishna et al. utilized Komagataella phaffii as the fermentation strain, overexpressed uronate dehydrogenase, and applied a mixed substrate (myo‐inositol + monosodium glutamate) to obtain a GA production of 17.6 g/L within 120 h, with an average productivity of 0.15 g/L/h (Krishna et al. 2025). Although the production achieved in this study was slightly lower than the highest levels currently reported, this difference could be attributed to the deliberate limitation of reaction time. Notably, the average productivity of the process developed in this study reached 0.21 g/L/h, representing an increase of 162.5% and 40% compared to the aforementioned studies, respectively. However, all existing reports have focused on the fermentation or biocatalytic production of GA using glucose as the initial substrate, in which the conversion of GlcA to GA constitutes only one step within a multi‐step pathway. As such, a direct comparison fails to adequately demonstrate the economic advantages of the process technology developed in this study. Overall, this process demonstrates considerable economic potential and offers a promising industrial approach for the biological green synthesis of GA.

Figure 8.

Visual diagram of the effect of multiple intensification strategies on GA production improvement.

4. Conclusions

This study demonstrated a sustainable and efficient process to produce industrially important biomass‐derived chemical GA by combining mechanistic investigations with multiple intensification strategies. At the molecular level, the key factors contributing to the stable production of GA were identified by visualizing the structure–activity relationship between dehydrogenase and GlcA, as well as by numerically solving the binding energy. At the macro level, a significant increase in both the production and productivity of GA was achieved through the synergistic implementation of multiple intensification strategies. The integrated approach not only overcame key limitations of current biological and chemical methods but also established a sustainable platform for industrial‐scale GA synthesis. These advances highlighted the potential of combining fundamental research with applied biotechnology to address challenges in biomass valorization, thereby paving the way for broader applications of GA.

Author Contributions

X.H. and Y.X. designed the project, and X.H. performed experiments, data analysis, and manuscript writing. Y.H. and J.G. conducted experiments and data analysis. W.H. and J.G. helped to revise the manuscript. Y.X. supervised the project. All authors read and approve the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File

Data Availability Statement

The data supporting the findings of this study are available within the Article and its Supplementary Information or from the corresponding authors upon reasonable request. The protein structure can be found in the UniProt database (https://www.uniprot.org/uniprotkb/P27175/entry). The ligands structure can be found in PubChem database (https://pubchem.ncbi.nlm.nih.gov/).