Method and system for regulating the balance between metabolic load and product synthesis of a succinic acid fermentation

Abstract

This invention discloses a method for regulating the balance between metabolic load and product synthesis in gluconic acid fermentation, belonging to the field of biomanufacturing and process control technology. The method first integrates a heterologous enzyme gene into *E. coli* to construct a heterologous gluconic acid synthesis pathway; then, it establishes a kinetic mechanism model with key metabolites and heterologous enzyme concentrations as state variables and Miox expression parameters as regulatory inputs; based on the model, it constructs two conflicting optimization objective functions: production loss and enzyme expression burden; it designs a dynamic optimization regulation strategy to predict and optimize Miox expression regulation parameters at discrete decision moments; finally, it converts the parameters into genetic element regulation instructions, combining state monitoring to update the model to form a closed-loop control. This invention can automatically balance product synthesis and metabolic load in a dynamic fermentation environment, improving the stability, yield, and robustness to environmental disturbances in gluconic acid production, providing a solution for metabolic regulation in biomanufacturing.

Description

Technical Field

[0001] This invention belongs to the field of bio-metabolic engineering and bioprocess control technology, and in particular refers to a method and system for regulating the metabolic load and product synthesis balance of gluconic acid fermentation. Background Technology

[0002] Gluconic acid, as an important bio-based platform compound, has broad application value in industrial fields such as bio-based polymers, drug precursors, and environmental chemicals, and market demand continues to increase. Currently, the industrial production of gluconic acid mainly adopts the chemical synthesis route represented by nitric acid oxidation of glucose. This process suffers from poor atom economy and low reaction selectivity, and the production process is accompanied by the discharge of large amounts of wastewater and waste residue containing heavy metals, posing a significant pressure on the ecological environment and failing to meet the requirements of green manufacturing development. Therefore, developing environmentally friendly and sustainable biosynthetic alternative technologies has become an urgent need in this field.

[0003] Green biomanufacturing centered on microbial cell factories provides a feasible pathway for the sustainable production of gluconic acid; among these, constructing a heterologous synthesis pathway using *E. coli* as a host is considered one of the most promising industrial solutions. However, this pathway presents an inherent contradiction between product synthesis and cell growth: increasing the expression level of pathway enzymes can improve product throughput, but excessive heterologous expression will competitively occupy cellular translational and metabolic resources such as ribosomes, amino acids, and ATP, increasing the metabolic burden on the host, inhibiting cell growth, and ultimately reducing overall synthesis efficiency. Achieving a fine balance between "maximizing product synthesis" and "minimizing metabolic burden" is the core technical challenge in regulating gluconic acid biosynthesis.

[0004] Pareto multi-objective optimization is a common approach to solving the aforementioned multi-objective trade-off problems. It calculates the Pareto front and outputs a set of non-dominated compromise solutions, providing a theoretical reference for configuring control parameters. However, existing Pareto optimization methods are essentially static optimization frameworks, relying on pre-defined mathematical models with fixed parameters and assuming that the relative importance of each optimization objective remains constant throughout the fermentation cycle. In actual fermentation, key parameters such as cell physiological state, substrate concentration, culture temperature, and enzyme activity are time-varying; enzyme activity decay and shifts in metabolic network kinetics are common in the later stages of fermentation, leading to a mismatch between the optimal control parameters determined based on the initial model and fixed weights and the actual operating conditions. This results in decreased product synthesis efficiency, poor production stability, and low batch-to-batch reproducibility. The inability of static optimization to adaptively match the dynamic changes in the fermentation process is a core technical limitation restricting the large-scale application of this type of control method in industrial fermentation.

[0005] In summary, existing gluconic acid biosynthesis regulation technologies still lack intelligent regulation methods that can sense the system state online, dynamically adjust regulation parameters, and adaptively optimize the balance between product synthesis and metabolic burden, making it difficult to meet the requirements of industrial fermentation for high yield, high stability, and high reproducibility. Summary of the Invention

[0006] Therefore, this invention aims to solve the technical problem in the biosynthesis of gluconic acid Escherichia coli where traditional static optimization methods cannot adapt to the dynamic changes in cell physiology, external conditions and enzyme activity drift during fermentation, making it difficult to achieve a dynamic and precise balance between the two conflicting goals of increasing product synthesis throughput and reducing the metabolic load of host cells, resulting in low fermentation production efficiency and poor stability.

[0007] To address the aforementioned technical problems, this invention provides a method and system for regulating the metabolic load and product synthesis balance in gluconic acid fermentation. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation includes the following steps: S1: A heterologous metabolic pathway for the synthesis of gluconic acid in Escherichia coli, wherein multiple target heterologous enzyme genes are integrated into the Escherichia coli genome or expression plasmid, enabling Escherichia coli to synthesize gluconic acid stepwise from glucose-6-phosphate through this heterologous metabolic pathway; the multiple target heterologous enzyme genes include inositol-3-phosphate synthase gene ino1, inositol oxygenase gene miox, and uronic acid dehydrogenase gene udh. S2: Based on the heterologous metabolic pathway, a kinetic mechanism model for characterizing the glucose diacid fermentation process is established. The kinetic mechanism model uses the concentrations of glucose-6-phosphate, fructose-6-phosphate, inositol and the target heterologous enzyme gene as state variables, and the expression parameters of the inositol oxygenase gene miox as regulatory inputs. S3: Based on the aforementioned kinetic mechanism model, two conflicting optimization objective functions are constructed. One is used to quantify the production loss of gluconic acid production efficiency, and the other is used to quantify the enzyme expression burden of E. coli cells due to metabolic pressure caused by the expression of the target heterologous enzyme gene. S4: Based on the optimization objective function, a dynamic optimization control strategy is designed. At multiple discrete decision moments in the fermentation process, the production loss and enzyme expression burden in future periods are predicted according to the current system state and the kinetic mechanism model. Based on this, the expression control parameters of the inositol oxygenase gene miox in the next period are dynamically optimized to obtain the control strategy that makes the two optimal. S5: The expression regulation parameters obtained by optimization at each discrete decision time are transformed into actual regulation instructions for E. coli genetic elements and implemented in the fermentation process. At the same time, the kinetic mechanism model is updated by combining the actual state monitoring results during the fermentation process to form a closed-loop control. This is executed cyclically until the end of fermentation, thereby automatically maintaining the optimal balance between product synthesis and metabolic load in a dynamic fermentation environment.

[0008] In one embodiment of the present invention, the expression of the dynamic mechanism model is: , Wherein, the state vector G6P, F6P, and MI represent the concentrations of glucose-6-phosphate, fructose-6-phosphate, and inositol, respectively; Ino1 and MIOX represent the concentrations of the corresponding enzymes expressed by the heterologous enzyme genes ino1 and miox, respectively; control input Here is the regulatory function for mioxase expression, specifically expressed as follows: , It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of the heterozyme gene miox, which leads to an expression rate of 100%. Intracellular inositol mi concentration at time; It is the Hill coefficient, which characterizes the synergistic activation effect of inositol on the expression of the heterologous enzyme gene miox.

[0009] In one embodiment of the present invention, the expression of the optimization objective function is as follows: , , in, Indicates production loss, used to measure the time spent in the fermentation cycle. Within, the cumulative deviation between the actual synthesis flux of gluconic acid and the theoretical maximum synthesis flux; For a constant glucose uptake rate, This represents the rate of the mioxe enzyme-catalyzed reaction at the current time t; Indicates enzyme expression burden, used to measure fermentation cycle. The cumulative expression levels of the endogenous heterozyme gene ino1 and the heterozyme gene miox; This represents the intracellular concentration of the ino1 enzyme at the current time t. This represents the intracellular concentration of the mioxase at the current time t.

[0010] In one embodiment of the present invention, the dynamic optimization and control strategy includes a dynamic fixed-weight optimization strategy, the specific steps of which are as follows: At each discrete decision time Based on the aforementioned dynamic mechanism model, a comprehensive objective function is constructed as follows: , in, and They are in the future time period Internally, production losses and enzyme expression burden are predicted based on the current state; and These are pre-set fixed weighting coefficients that satisfy... T is a fixed time interval; Solve the optimization problem: To obtain the optimal control parameters It is applied to the fermentation control of the next stage; in, It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of the heterozyme gene miox, which leads to an expression rate of 100%. The concentration of inositol mi at that time.

[0011] In one embodiment of the present invention, the dynamic optimization and control strategy includes a dynamic adaptive weight optimization strategy, the specific steps of which are as follows: At each discrete decision time Regarding the current control parameters Apply a small relative perturbation The control parameters after adding perturbation are obtained. ; Using the current control parameters respectively and its perturbation value As a control input, the dynamic mechanism model is adjusted from the current decision moment. We begin numerical integration to predict the system dynamics over future time periods and calculate the corresponding objective function value. and the objective function value after perturbation , , This represents the objective function value corresponding to the production loss. This represents the objective function value corresponding to the enzyme expression burden; Define the i-th sub-objective at the current decision time. Influence coefficient: ,in To prevent positive numbers with a denominator of zero; Normalizing the influence yields the initial adaptive weight coefficients for each sub-objective: ; A smoothing mechanism is introduced, and its calculation formula is as follows: , ;in, It is a smoothing coefficient and This controls the response speed of weight updates; For the previous decision moment The adaptive weighting coefficients for the production loss sub-objective; Construct a comprehensive objective function with adaptive weights: Solve the optimization problem: To obtain the optimal control parameters It is applied to the fermentation control of the next stage; in, It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of the heterozyme gene miox, which leads to an expression rate of 100%. The concentration of inositol mi at that time.

[0012] In one embodiment of the present invention, the expression regulation parameters obtained by optimization at each discrete decision time are converted into actual regulatory instructions for Escherichia coli genetic elements, and the instructions are implemented in the fermentation process as follows: The expression regulation parameters include the maximum rate constant of the miox enzyme expression function. Half-maximal activation concentration of the heterozyme gene miox ;for The transcription rate of the inositol oxygenase gene miox is controlled by selecting or designing promoters with appropriate strength; for The response threshold to mi is adjusted by altering the sequence of the transcription factor IpsA or its binding site.

[0013] In one embodiment of the present invention, the inositol-3-phosphate synthase gene is derived from Saccharomyces cerevisiae, the inositol oxygenase gene is derived from mice, the uronic acid dehydrogenase gene is derived from Pseudomonas syringae, and the heterologous metabolic pathway utilizes the host's own SuhB enzyme to dephosphorylate inositol-1-phosphate to inositol.

[0014] Based on the same inventive concept, the present invention also provides a system for regulating the metabolic load and product synthesis balance of gluconic acid fermentation, comprising: a heterologous metabolic pathway construction module, a fermentation dynamic mechanism modeling module, a multi-objective optimization function construction module, a dynamic regulation parameter optimization module, and a closed-loop regulation instruction execution module. The heterologous metabolic pathway construction module is used for the heterologous metabolic pathway of gluconic acid synthesis in Escherichia coli. It integrates multiple target heterologous enzyme genes into the E. coli genome or expression plasmid, enabling E. coli to synthesize gluconic acid stepwise from glucose-6-phosphate through this heterologous metabolic pathway. The multiple target heterologous enzyme genes include inositol-3-phosphate synthase gene ino1, inositol oxygenase gene miox, and uronic acid dehydrogenase gene udh. The fermentation dynamic mechanism modeling module is used to establish a kinetic mechanism model for characterizing the glucose diacid fermentation process based on the heterologous metabolic pathway. The kinetic mechanism model uses the concentrations of glucose-6-phosphate, fructose-6-phosphate, inositol and the target heterologous enzyme gene as state variables, and the expression parameters of the inositol oxygenase gene miox as regulatory inputs. The objective function construction module is used to construct two conflicting optimization objective functions based on the kinetic mechanism model. One is used to quantify the production loss of gluconic acid production efficiency, and the other is used to quantify the enzyme expression burden of E. coli cells due to metabolic pressure caused by the expression of the target heterologous enzyme gene. The dynamic regulation parameter optimization module is used to design a dynamic optimization regulation strategy based on the optimization objective function. At multiple discrete decision moments in the fermentation process, based on the current system state and the kinetic mechanism model, it predicts the production loss and enzyme expression burden in future periods, and dynamically optimizes the expression regulation parameters of the inositol oxygenase gene miox in the next period to obtain the regulation strategy that makes both optimal. The closed-loop control instruction execution module is used to convert the expression regulation parameters obtained by optimization at each discrete decision moment into actual control instructions for the genetic elements of Escherichia coli, and implement these instructions in the fermentation process. At the same time, it updates the kinetic mechanism model by combining the actual state monitoring results during the fermentation process, forming a closed-loop control, which is executed cyclically until the end of fermentation, thereby automatically maintaining the optimal balance between product synthesis and metabolic load in a dynamic fermentation environment.

[0015] Compared with the prior art, the above-described technical solution of the present invention has the following advantages: Firstly, by introducing three heterologous enzymes to construct a specific metabolic pathway, an efficient heterologous synthesis route for the synthesis of gluconic acid in Escherichia coli was established, enabling targeted regulation of product synthesis from the source. Secondly, a kinetic mechanism model coupling key metabolites and enzyme expression levels was established, providing scientific theoretical support for the precise regulation of the fermentation process. Third, two conflicting objectives, production loss and enzyme expression burden, were defined, and two optimization strategies, dynamic fixed weight and dynamic adaptive weight, were designed to achieve dynamic online balance between product synthesis and metabolic load, breaking through the limitations of traditional static optimization. Fourth, the dynamic adaptive weight strategy achieves intelligent adjustment of weights based on sensitivity analysis. It can also be used for additional optimization and shorten decision intervals triggered by disturbances such as enzyme activity drift, which greatly improves the robustness of the fermentation process to environmental disturbances such as batch-to-batch variation and enzyme activity drift. Fifth, by transforming the optimized parameters into actual regulatory instructions for genetic elements and forming a closed-loop control, the regulatory strategy has been engineered and implemented, effectively improving the stability and yield of gluconic acid production. At the same time, the regulatory parameters of this scheme have a clear physical realization range and feasible modification methods, providing a practical and scalable solution for the metabolic regulation of complex biomanufacturing processes, and possessing important industrial application value and technical reference significance. Attached Figure Description

[0016] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0017] Figure 1 This is a schematic flowchart of a method for regulating the metabolic load and product synthesis balance of gluconic acid fermentation provided in an embodiment of the present invention. Figure 2 It is a comparison curve of adaptive weight changes under two working conditions: normal fermentation and enzyme activity drift, under a dynamic fixed weight optimization strategy. Figure 3 This is a comparison of the dynamic changes in the expression level of inositol oxygenase (miox) under two conditions: normal fermentation and enzyme activity drift, under a dynamic fixed weight optimization strategy. Figure 4 This is a dynamic evolution trend diagram of inositol (mi) concentration during fermentation under a dynamic fixed weight optimization strategy; Figure 5 This is a real-time curve showing the flux of the inositol oxygenase catalytic reaction under a dynamic fixed-weight optimization strategy. Figure 6 It is a comparison curve of adaptive weight changes under two working conditions: normal fermentation and enzyme activity drift, under the dynamic adaptive weight optimization strategy. Figure 7 This is a comparison of the dynamic changes in the expression level of inositol oxygenase (miox) under two conditions: normal fermentation and enzyme activity drift, under a dynamic adaptive weight optimization strategy. Figure 8 This is a comparison chart of the dynamic changes in the concentration of inositol (mi), an intermediate metabolite, during fermentation under a dynamic adaptive weight optimization strategy. Figure 9 This is a real-time curve showing the flux of the inositol oxygenase catalytic reaction under a dynamic adaptive weight optimization strategy. Figure 10This is a comparison chart of the performance of E. coli in synthesizing glucosidic acid under Pareto optimization, dynamic fixed weight optimization, and dynamic adaptive weight optimization under normal fermentation conditions. Figure 11 This is a comparison chart of the performance of E. coli in synthesizing gluconic acid under Pareto optimization, dynamic fixed weight optimization, and dynamic adaptive weight optimization with enzyme activity drift perturbation. Figure 12 This is a schematic diagram of a system for regulating the metabolic load and product synthesis balance of gluconic acid fermentation, provided in an embodiment of the present invention.

[0018] Explanation of the reference numerals in the accompanying drawings: 100, Heterogeneous metabolic pathway construction module; 200, Fermentation dynamic mechanism modeling module; 300, Objective function construction module; 400, Dynamic regulation parameter optimization module; 500, Closed-loop regulation command execution module. Detailed Implementation

[0019] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0020] Example 1: Reference Figure 1 and Figure 2 As shown, this invention provides a method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation, the method comprising the following steps: S1: A heterologous metabolic pathway for the synthesis of gluconic acid in Escherichia coli, in which multiple target heterologous enzyme genes are integrated into the E. coli genome or expression plasmid, enabling E. coli to synthesize gluconic acid stepwise from glucose-6-phosphate through this heterologous metabolic pathway; the multiple target heterologous enzyme genes include the inositol-3-phosphate synthase gene ino1 from Saccharomyces cerevisiae, the inositol oxygenase gene miox from mouse, and the uronic acid dehydrogenase gene udh from Pseudomonas syringae; S2: Based on the heterologous metabolic pathway, a kinetic mechanism model for characterizing the glucose diacid fermentation process is established. The kinetic mechanism model uses the concentrations of glucose-6-phosphate, fructose-6-phosphate, inositol and the target heterologous enzyme gene as state variables, and the expression parameters of the inositol oxygenase gene miox as regulatory inputs. S3: Based on the aforementioned kinetic mechanism model, two conflicting optimization objective functions are constructed. One is used to quantify the production loss of gluconic acid production efficiency, and the other is used to quantify the enzyme expression burden of E. coli cells due to metabolic pressure caused by the expression of the target heterologous enzyme gene. S4: Based on the optimization objective function, a dynamic optimization control strategy is designed. At multiple discrete decision moments in the fermentation process, the production loss and enzyme expression burden in future periods are predicted according to the current system state and the kinetic mechanism model. Based on this, the expression control parameters of the inositol oxygenase gene miox in the next period are dynamically optimized to obtain the control strategy that makes the two optimal. S5: The expression regulation parameters obtained by optimization at each discrete decision time are transformed into actual regulation instructions for E. coli genetic elements and implemented in the fermentation process. At the same time, the kinetic mechanism model is updated by combining the actual state monitoring results during the fermentation process to form a closed-loop control. This is executed cyclically until the end of fermentation, thereby automatically maintaining the optimal balance between product synthesis and metabolic load in a dynamic fermentation environment.

[0021] Furthermore, in step S1, using *E. coli* as the chassis cell, three heterologous enzymatic reaction pathways are introduced through genetic engineering to achieve the directed biosynthesis of gluconic acid. The specific process is as follows: The inositol-3-phosphate synthase gene *ino1* from *Saccharomyces cerevisiae*, the inositol oxygenase gene *miox* from mouse, and the uronic acid dehydrogenase gene *udh* from *Pseudomonas syringae* were integrated into the *Escherichia coli* genome via homologous recombination or plasmid-mediated expression, or recombinant expression plasmids were constructed to achieve intracellular heterologous expression. The gene ID of inositol-3-phosphate synthase *ino1* is YJL153C, the gene ID of inositol oxygenase *miox* is 56727, and the gene ID of uronic acid dehydrogenase *udh* is PSPTO_1053.

[0022] In the modified engineered strain, the central metabolic intermediate glucose-6-phosphate (g6p) is catalyzed by the ino1-encoded enzyme to generate inositol-1-phosphate (m1p). M1p is dephosphorylated by E. coli's own SuhB phosphatase to generate inositol (mi). mi is converted into glucuronic acid (gua) under the oxidative catalysis of the miox-encoded enzyme. Finally, gua is converted into the target product gluconic acid (ga) by the dehydrogenation catalysis of the udh-encoded enzyme. The heterologous synthesis metabolic pathway of gluconic acid with g6p as the branch node was successfully constructed.

[0023] In step S2, based on the heterologous metabolic pathway constructed in step S1, a system dynamic state-space model coupling the expression of key metabolites and heterologous enzymes is constructed. This model uses a set of ordinary differential equations to characterize the dynamic changes of each state variable during fermentation, specifically as follows: (1) (2) (3) (4) (5) Wherein, G6P, F6P, and MI represent the intracellular concentrations of glucose-6-phosphate, fructose-6-phosphate, and inositol, respectively; Ino1 and MIOX represent the concentrations of the corresponding enzymes expressed by the heterologous enzymes ino1 and miox, respectively. This indicates the rate of glucose uptake across the membrane via the phosphotransferase system; This indicates the catalytic reaction rate of glucose-6-phosphate dehydrogenase; This indicates the catalytic reaction rate of phosphogluconose isomerase; This indicates the catalytic reaction rate of the ino1 enzyme; This indicates the rate at which inositol is effluxed into the fermentation medium; This indicates the catalytic reaction rate of the miox enzyme; Indicates the catalytic reaction rate of phosphofructokinase; It is the constitutive expression rate constant of the ino1 enzyme; It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of mioxase expression, that is, the rate at which mi induces mioxase expression to reach... Intracellular inositol mi concentration at time; It is the Hill coefficient, which characterizes the synergistic activation effect of inositol mi on miox enzyme expression; It is the cell growth rate, which characterizes the dilution effect of intracellular metabolites caused by cell proliferation during fermentation.

[0024] All the enzyme-catalyzed reaction rates and transmembrane transport rates mentioned above were quantitatively described using classical enzyme kinetic models such as the Michaelis-Menten equation and the Hill equation. All kinetic parameters in the models were obtained by combining in vitro enzyme activity experiments, in vivo fermentation experiments, and data fitting methods.

[0025] To enable engineering applications and numerical solutions for the model, the fermentation kinetics model is simplified by dimensionality reduction and uniformly represented in the standard state-space form, which is the final kinetic mechanism model: (6) Wherein, the state vector It includes the concentrations of all core metabolites and key heterologous enzymes; it controls the input. Here is the regulatory function for mioxase expression, and its specific expression is: This enables the targeted regulation of mioxase expression by intracellular MI concentration.

[0026] In the process of gluconic acid fermentation, in order to increase the synthesis yield of gluconic acid, it is necessary to increase the intracellular expression level of key enzymes in the heterologous pathway such as ino1 and miox, so as to drive more carbon flow from the central metabolic node glucose-6-phosphate (g6p) to the gluconic acid branch pathway. However, the high level of expression of heterologous proteins will excessively occupy the host cell's limited translation resources (including ribosomes, ATP and amino acids), causing a significant metabolic load, which in turn inhibits cell growth and even leads to a decrease in cell viability, ultimately restricting the overall production efficiency.

[0027] In step S3, to quantify this conflicting relationship in the system, based on the E. coli gluconic acid synthesis kinetic mechanism model constructed in step S2, the following two mutually constraining optimization objective functions are defined, with the following expressions: (7) (8) in, Indicates production loss, used to measure the time spent in the fermentation cycle. Within, the actual synthesis flux of gluconic acid (in terms of the current time t in the model). (for quantitative characterization) and theoretical maximum synthesis flux (at a constant glucose uptake rate) The cumulative deviation between (quantitative representations); Indicates enzyme expression burden, used to measure fermentation cycle. Intracellular cumulative expression levels of the endogenous heterozyme gene ino1 and the heterozyme gene miox; This represents the intracellular concentration of the ino1 enzyme at the current time t. This represents the intracellular concentration of the mioxase at the current time t.

[0028] Clearly, the two optimization objectives mentioned above are inherently conflicting: reducing production losses. To improve product synthesis efficiency, it is necessary to upregulate the expression levels of heterologous enzymes ino1 and miox in the model to increase the metabolic flux of the heterologous pathway. This process inevitably leads to an enzyme expression burden. Increase; conversely, decrease enzyme expression burden. To alleviate the host's metabolic stress, the expression level of heterologous enzymes in the model needs to be downregulated, which will directly lead to a decrease in metabolic flux through the heterologous pathway and production losses. Increase.

[0029] It should be noted that the udh enzyme involved in the kinetic mechanism model has excellent catalytic efficiency and only requires a very low intracellular expression level to meet the catalytic requirements for the conversion of glucuronic acid to gluconic acid. The metabolic load caused by its expression is negligible, therefore it is not included in the enzyme expression burden. The scope of quantification.

[0030] In step S4, to meet the dynamic balance requirement between production and metabolic load during gluconic acid fermentation, a dynamic control strategy based on rolling time-domain optimization is proposed. The core idea of ​​this strategy is to address the discrete decision-making moments of the fermentation process. (in The control parameters are re-optimized. Decision-making times can be set at fixed time intervals. Setting, i.e. The optimized regulatory parameter is the maximum expression rate constant in the inositol oxygenase (miox) expression function. and half-maximum activation concentration Specifically, it includes the following two implementation schemes: (1) Scheme 1: Dynamic fixed weight optimization strategy At each discrete decision time Construct a comprehensive objective function with fixed weights, which is used to evaluate future time periods. The overall performance of the internal system. Its mathematical expression is: (9) in, and From the current time respectively Based on the current state of the system, the calculation method for the production loss and enzyme expression burden within the predicted time interval T is similar to that of formulas (7) and (8), only the time integration interval needs to be replaced with . and These are pre-set fixed weighting coefficients that satisfy... This is used to characterize the relative importance of yield targets and metabolic load targets. Then, the following local optimization problem is solved: (10) To obtain the optimal combination of control parameters suitable for the current fermentation stage and apply it to the next control zone. The actual fermentation process is described. This strategy maintains constant weighting coefficients in a single batch of fermentation and achieves adaptability to time-varying environments through rolling optimization.

[0031] Optionally, the fixed weighting coefficient and The allocation of weighting coefficients is based on statistical analysis of historical production data or existing prior knowledge. The selection of weighting coefficients follows these principles, depending on the different control objectives of the fermentation process: When the primary goal of the process is to maximize the volumetric yield or endpoint yield of gluconic acid, the production loss function should be given a higher weight, i.e., set... , This is to ensure that the optimization algorithm prioritizes the efficient convergence of metabolic flux toward the target product, even if it means that the cell has to bear a high heterologous expression load.

[0032] If the fermentation process faces risks such as easy decline in strain viability, plasmid instability, or substrate inhibition, it is necessary to prioritize maintaining cell growth activity and metabolic network stability. In this case, the enzyme expression burden function should be assigned a higher weight, i.e., set... , This is to strictly limit the overexpression of heterologous proteins and ensure that the host cell maintains sufficient biomass accumulation capacity.

[0033] In conventional production scenarios, achieving a long-term dynamic balance between product synthesis and cell growth is a common requirement. In this case, an equal-weighted configuration can be used, i.e., setting... This ensures that the optimization process treats both production loss and metabolic load equally, thereby guaranteeing a certain product synthesis efficiency while preventing cells from experiencing physiological functional decline due to excessive burden.

[0034] (2) Scheme 2: Dynamic adaptive weight optimization strategy

[0035] To further enhance the responsiveness of the control strategy to environmental fluctuations, this scheme introduces an adaptive weight adjustment mechanism based on system state. The core idea is to assess the dependence of each sub-objective on the control parameters under the current operating conditions through local sensitivity analysis at each decision-making moment, and dynamically adjust the weight allocation accordingly, enabling the optimization process to respond in real time to unexpected changes in the fermentation environment. The specific implementation steps are as follows: At each discrete decision time First, the current control parameters Apply a small relative perturbation The control parameters after adding perturbation are obtained. ; Using the current control parameters respectively and its control parameters after adding perturbations As a control input, from the current decision moment Initially, the dynamic mechanism model was used to predict the time period. Numerical integration is performed to obtain the future evolution trajectory of the system state. Based on this predicted trajectory, the objective function values ​​corresponding to the two sub-objectives (production loss and enzyme expression burden) are calculated respectively. and the objective function value after perturbation , , This represents the objective function value corresponding to the production loss. This represents the objective function value corresponding to the enzyme expression burden; Definition of the first Individual goals at the current decision-making moment Influence coefficient: ,in, To prevent positive constants with a denominator of zero. This coefficient reflects the sensitivity of the sub-objective value to parameter perturbations; the greater the influence, the more the objective is dominated by the control parameters in the current state, and it should be given a higher optimization weight.

[0036] Normalizing the influence yields the initial adaptive weight coefficients for each sub-objective: ; To avoid drastic changes in weights due to instantaneous disturbances, a first-order exponential smoothing mechanism is introduced: , ,in, It is a smoothing coefficient and This controls the response speed of weight updates; For the previous decision moment The adaptive weighting coefficients for the production loss sub-objective; Finally, the comprehensive objective function for adaptive weights is constructed as follows: (11) In the formula, and The meaning is the same as in the dynamic fixed-weight optimization strategy, which means from the current time... Based on the current state of the system, the calculation method for the production loss and enzyme expression burden within the predicted time interval T is similar to that of formulas (7) and (8), only the time integration interval needs to be replaced with .

[0037] By solving the optimization problem: To obtain the optimal control parameters for the current time period. and apply it to the next control zone. The actual fermentation process.

[0038] This scheme achieves dynamic adaptive adjustment of weight coefficients by sensing the target's sensitivity to parameter changes online. This enables the optimization strategy to more flexibly cope with uncertainties such as enzyme activity drift and substrate fluctuations during fermentation, thereby maintaining a more precise dynamic balance between product synthesis and metabolic load.

[0039] Preferably, the fixed time interval T for the decision moment ranges from 10 seconds to 3600 seconds; typically, when the fermentation process changes rapidly, a shorter interval such as T=100 seconds is selected; when the process is relatively stable, a longer interval such as T=1800 seconds is selected.

[0040] In the optimal control parameters The solution process can be approached using either gradient-based optimization algorithms or direct search algorithms, depending on the mathematical properties of the objective function and the availability of gradient information. Gradient-based optimization methods: When the gradient information of the objective function with respect to the control parameters can be obtained analytically or through numerical difference, sequential quadratic programming or quasi-Newton methods are recommended for efficient local optimization. These methods utilize gradient information to construct quadratic approximate subproblems and iteratively search along the descent direction within the feasible region of the parameters. They feature fast convergence speed and high computational accuracy, and are suitable for scenarios where the objective function is continuously differentiable and the initial point is appropriately selected.

[0041] Direct search methods: When the objective function has a complex shape (e.g., non-convex, discontinuous, or with multiple local extrema) and gradient information is difficult to obtain accurately, a direct search algorithm that does not rely on gradients can be used. Pattern search methods, through exploratory movements and pattern shifts along coordinate or pattern directions, can achieve robust local convergence within parameter constraints. Particle swarm optimization algorithms, by simulating swarm intelligence behavior, achieve a balance between global search capability and convergence speed, and are particularly suitable for complex optimization problems with moderate parameter space dimensions that need to avoid getting trapped in local optima.

[0042] Regardless of the method used, it is necessary to ensure that the optimization is carried out within the given parameter constraints and to use the satisfaction of engineering accuracy requirements (such as the relative change of the objective function being less than a preset threshold) as the convergence criterion, so as to obtain an optimal or near-optimal solution that is practically operable.

[0043] In the aforementioned dynamic optimization and control strategy, to address unexpected disturbances such as enzyme activity drift that may occur during fermentation, this invention further introduces an event-triggered decision-making mechanism to enhance the system's robust control capability. Specifically, the decision-making time... The determination of this not only relies on a preset fixed time interval T, but also incorporates online monitoring of the catalytic activity of key enzymes.

[0044] In fermentation process monitoring, the specific catalytic activity of inositol oxygenase (miox) is tracked in real time through offline sampling and measurement or online sensing technology. When the specific catalytic activity of the enzyme is detected... When the decrease in relative value (i.e., the measured value under initial or standard conditions) exceeds a preset threshold (e.g., relative residual vitality is below 80%), the system determines that a significant disturbance has occurred in the current operating condition and triggers an immediate optimization calculation. This optimization is independent of the original fixed-time scheduling, uses the measured state variables at the current moment to quickly correct the model parameters, and re-solves for the optimal control parameters in the future time domain.

[0045] Meanwhile, to address the increased uncertainty in the system's dynamic characteristics after a disturbance, the strategy will adaptively shorten subsequent fixed decision intervals (e.g., temporarily adjusting from the usual T to T / 2 or shorter), thereby increasing the control frequency during the disturbance period and ensuring that the control parameters keep pace with rapid changes in the system state. Once enzyme activity is detected to have stabilized or fermentation has entered a steady phase, the decision interval can be restored to a preset normal value.

[0046] This mechanism combines time-driven and event-driven approaches, enabling optimized regulatory strategies to sense and respond to changes in microscopic physiological states, significantly improving adaptability and robustness of control to complex conditions such as enzyme activity drift and metabolic fluctuations.

[0047] In step S5, the expression regulation parameters obtained from optimization at each discrete decision time are converted into actual regulatory instructions for the genetic elements of *E. coli*, and these instructions are implemented in the fermentation process as follows: The optimal control parameter (maximum rate constant) obtained through dynamic optimization in S4 is... and half-maximum activation concentration Interpreted as engineering instructions for specific genetic elements, and applied to the actual fermentation process, thereby realizing the physical implementation of optimization strategies.

[0048] For parameters This is physically achieved by regulating the transcription rate of the miox gene. Specifically, this is done by selecting or designing constitutive or inducible promoters with appropriate strength (such as P...). tac P BAD (and its mutant library), which is cloned upstream of the miox gene to precisely match the required transcription output level. To achieve quantitative regulation of expression intensity, the transcription initiation rate corresponding to the promoter is preferably controlled in the range of 0.01 nM / s to 10 nM / s. This range can cover the typical needs from basal expression to high expression levels and ensure compatibility with the carrying capacity of the cell translation system.

[0049] For parameters The physical implementation of this is achieved through the modification of inositol-responsive transcriptional regulatory elements. Specifically, amino acid residues or base substitutions are made to the ligand-binding domain of the transcription factor IpsA or its recognized DNA operon sequence to adjust its response threshold to the effector molecule inositol (mi). After modification, transcriptional activation of the miox gene should occur within a preset inositol concentration window, corresponding to a half-maximum activation concentration. The concentration is preferably controlled within the range of 0.1 mM to 50 mM, which matches the typical concentration profile of inositol accumulation during fermentation, ensuring the sensing sensitivity and dynamic response range of the control system.

[0050] After completing the above genetic modification, the system enters the next decision-making cycle: at the new decision moment The fermentation state variables are re-collected, and dynamic optimization is performed again based on the updated system state to obtain the optimal control parameters for the next time period, which are then implemented again. This process is repeated to form a rolling time-domain closed-loop control architecture of "online optimization-implementation-re-optimization".

[0051] Through the aforementioned closed-loop optimized control strategy, the fermentation system can sense changes in internal state (such as enzyme activity drift and metabolite accumulation) and fluctuations in the external environment (such as changes in substrate concentration) in real time, and automatically adjust the expression kinetic parameters of the Miox enzyme, thereby maintaining the optimal balance between product synthesis efficiency and cellular metabolic load during the dynamically changing fermentation process. This method can significantly improve the robustness, endpoint yield, and batch-to-batch reproducibility of the gluconic acid fermentation process, laying a technical foundation for achieving high-level industrial production.

[0052] To verify the effectiveness and robustness of the proposed regulation method, the following simulation verification system was constructed. The simulation experiment was based on the kinetic mechanism model of the Escherichia coli gluconic acid synthesis pathway described in step S2, and the model parameters were assigned values ​​based on experimental data.

[0053] 1. Model parameter configuration

[0054] Cell specific growth rate set to glucose uptake rate Constitutive expression rate constant of ino1 enzyme Hill coefficient Rates of each enzyme-catalyzed reaction , , , , and All reactions were described using the Michaelis-Menten or Hill kinetic equations, and the relevant kinetic parameters are as follows: The maximum reaction rate of glucose-6-phosphate dehydrogenase is... The maximum reaction rate of phosphoglucose isomerase is The maximum reaction rate of ino1 enzyme The maximum rate of inositol efflux is The maximum reaction rate of mioxase is The maximum reaction rate of phosphofructokinase is Hill coefficient of phosphofructokinase Set the total fermentation time. At hour, the initial state vector is .

[0055] 2. Dynamically optimize framework configuration

[0056] Set decision interval The entire fermentation process involves 360 decision points, measured in seconds. The optimization solution utilizes the `fmincon` function in MATLAB as the numerical solver, with a relative tolerance set to [value missing]. The absolute tolerance is set to To ensure calculation accuracy, adjust the parameters. and The optimized search range is set to its corresponding initial reference value. The interval provides a sufficient feasible region for parameter optimization.

[0057] 3. Execution of Dynamic Fixed Weight Optimization (Scheme 1)

[0058] In Scheme 1, fixed weighting coefficients need to be preset. When maximizing gluconic acid production is the primary objective, a fixed weighting coefficient can be set. , If a balance is desired between production loss and metabolic burden, then we can assume... At every decision-making moment Based on the current system state As initial conditions for numerical integration, within the parameter constraints, the comprehensive objective function shown in equation (9) is solved. The goal is to minimize this problem and obtain a set of optimal control parameters. And apply this set of parameters to the next time period. Fermentation process control.

[0059] 4. Execution of Dynamic Adaptive Weight Optimization (Scheme 2)

[0060] Option 2 requires setting parameters related to sensitivity calculation and weight update: relative perturbation ratio. Numerical stability constant used to prevent the denominator from being zero when calculating sensitivity. And the smoothing coefficient that controls the smoothness of weight changes. At the first decision moment initialize the weights to At each subsequent decision-making moment. Based on the current control parameters and the value after applying a small relative perturbation. As input, numerical integration is performed on the mechanistic model to predict the system's performance in the next time period. The dynamic behavior within the target area is analyzed, and the corresponding sub-target prediction values ​​are calculated accordingly. , Its corresponding perturbation value , .

[0061] Next, according to the formula Calculate the sensitivity coefficient of each sub-objective to parameter changes. Then, based on... and The rules are used to update the adaptive weights. Finally, the integrated objective function shown in formula (11) is constructed using the updated weights. And solve its minimization problem to obtain and apply the new optimal control parameters. And apply this set of parameters to the next time period. Fermentation process control.

[0062] To verify the effectiveness of the method of the present invention in dealing with common disturbances in actual fermentation processes, a typical disturbance scenario of enzyme activity drift was introduced into the simulation: at the 5th hour of fermentation (i.e., the mid-stage), the maximum reaction rate of inositol oxygenase (miox) was measured. The value is instantly reduced to 50% of its original value to simulate the condition of enzyme protein inactivation or catalytic efficiency decline.

[0063] Dynamic fixed-weight optimization exhibits good performance in dealing with environmental disturbances, as shown in the simulation results. Figures 2-5 As shown. Figure 2 It shows that the total objective function varies with fixed weights. The curves show that under normal conditions and after the enzyme activity shifts, the two curves essentially overlap, with minimal performance difference, indicating that this scheme has strong anti-interference capabilities. Meanwhile, Figures 4 to 5 The dynamic trajectories of inositol (mi) concentration, miox enzyme expression level, and miox catalytic flux were presented, remaining stable under both conditions, further validating the effectiveness of the proposed scheme in maintaining system homeostasis. It should be noted that... Figure 2 This also reflects the key characteristics of the scheme: system performance and weighting coefficients. The choice is closely related, when Larger values ​​result in better performance, highlighting the importance of pre-setting the weights for Scheme 1.

[0064] The dynamic adaptive weight optimization method exhibits more intelligent adjustment characteristics, as shown in the simulation results. Figures 6-9 As shown. Figure 6 The dynamic change curve of the weighting shows that when enzyme activity drift occurs at the 5th hour of fermentation, the weighting... A slight increase was observed. This change demonstrates the effectiveness of the sensitivity-driven weight update mechanism—the system can automatically sense disturbances and dynamically adjust target priorities. Similar to Scheme 1, Figures 7 to 9 The results show that key indicators such as mi concentration, miox enzyme expression level, and reaction flux under scheme 2 remain stable under normal fermentation and disturbance conditions, indicating that the method has stronger robustness.

[0065] Under normal fermentation conditions, the performance comparison of the three optimization methods is as follows: Figure 10 As shown in the figure. Comprehensive evaluation indicates that the dynamic adaptive weight optimization method performs best, with its total objective function value reduced by 30.55% compared to static Pareto optimization, demonstrating a significant performance advantage. For the dynamic fixed weight optimization method, considering the possibility of randomly set weights in practical applications, its average performance under all possible weights is evaluated. The results show that the average performance is 7.63% higher than static Pareto optimization, indicating that the dynamic re-optimization mechanism itself has significant value.

[0066] For the more challenging enzyme activity drift scenario, the performance comparison results of the three methods are shown in [reference needed]. Figure 11 Static Pareto optimization suffers significant performance degradation due to the inability to adjust parameters online, with the overall objective function value decreasing by 40.57%. In contrast, the two dynamic optimization methods proposed in this invention demonstrate robustness: the dynamic fixed-weight optimization method experiences only a 2.02% performance decrease, while the dynamic adaptive-weight optimization method shows an even smaller decrease of only 0.79%. Under this perturbation condition, the two dynamic optimization methods achieve performance improvements of 32.96% and 50.2% respectively compared to static Pareto optimization, fully demonstrating the effectiveness and superiority of the dynamic optimization framework proposed in this invention in addressing the uncertainties of actual fermentation processes.

[0067] In summary, the gluconic acid biosynthesis regulation method based on dynamic multi-objective optimization provided by this invention effectively overcomes the inherent limitations of traditional static Pareto optimization in dealing with dynamic perturbations during fermentation by constructing a control framework that integrates dynamic re-optimization and adaptive weight adjustment. Implementation effect analysis shows that both the proposed dynamic fixed-weight optimization and dynamic adaptive weight optimization can significantly improve the robustness and overall performance of the system under normal and perturbed conditions, with the adaptive weight strategy exhibiting superior online regulation capability and stability. This invention provides a practical solution for the precise and robust regulation of metabolic pathways in complex biomanufacturing processes and has significant industrial application value.

[0068] Example 2: Based on the same inventive concept as Embodiment 1, such as Figure 12 As shown, the present invention also provides a system for regulating the metabolic load and product synthesis balance of gluconic acid fermentation, comprising: a heterologous metabolic pathway construction module 100, a fermentation dynamic mechanism modeling module 200, an objective function construction module 300, a dynamic regulation parameter optimization module 400, and a closed-loop regulation instruction execution module 500. The heterologous metabolic pathway construction module 100 is used for the heterologous metabolic pathway of gluconic acid synthesis in Escherichia coli. It integrates multiple target heterologous enzyme genes into the E. coli genome or expression plasmid, enabling E. coli to synthesize gluconic acid stepwise from glucose-6-phosphate through this heterologous metabolic pathway. The multiple target heterologous enzyme genes include the inositol-3-phosphate synthase gene ino1 from Saccharomyces cerevisiae, the inositol oxygenase gene miox from mouse, and the uronic acid dehydrogenase gene udh from Pseudomonas syringae. The fermentation dynamic mechanism modeling module 200 is used to establish a kinetic mechanism model for characterizing the glucose diacid fermentation process based on the heterologous metabolic pathway. The kinetic mechanism model uses the concentrations of glucose-6-phosphate, fructose-6-phosphate, inositol and the target heterologous enzyme gene as state variables, and the expression parameters of the inositol oxygenase gene miox as regulatory inputs. The objective function construction module 300 is used to construct two conflicting optimization objective functions based on the kinetic mechanism model. One is used to quantify the production loss of gluconic acid production efficiency, and the other is used to quantify the enzyme expression burden of E. coli cells due to metabolic pressure caused by the expression of the target heterologous enzyme gene. The dynamic regulation parameter optimization module 400 is used to design a dynamic optimization regulation strategy based on the optimization objective function. At multiple discrete decision moments in the fermentation process, based on the current system state and the kinetic mechanism model, it predicts the production loss and enzyme expression burden in future periods, and dynamically optimizes the expression regulation parameters of the inositol oxygenase gene miox in the next period to obtain a regulation strategy that makes the two optimal. The closed-loop control instruction execution module 500 is used to convert the expression regulation parameters obtained by optimization at each discrete decision moment into actual control instructions for the genetic elements of Escherichia coli, and implement the instructions in the fermentation process. At the same time, it updates the kinetic mechanism model by combining the actual state monitoring results during the fermentation process, forming a closed-loop control, which is executed cyclically until the end of fermentation, thereby automatically maintaining the optimal balance between product synthesis and metabolic load in a dynamic fermentation environment.

[0069] This embodiment proposes a system for balancing the metabolic load and product synthesis in gluconic acid fermentation, which is used to implement the aforementioned method for balancing the metabolic load and product synthesis in gluconic acid fermentation. Therefore, the specific implementation of the system for balancing the metabolic load and product synthesis in gluconic acid fermentation can be found in the embodiment section of the aforementioned method for balancing the metabolic load and product synthesis in gluconic acid fermentation. For example, the heterologous metabolic pathway construction module 100, the fermentation dynamic mechanism modeling module 200, the objective function construction module 300, the dynamic regulation parameter optimization module 400, and the closed-loop regulation instruction execution module 500 are respectively used to implement steps S1 to S5 in the method described in Embodiment 1. Therefore, the specific implementation can be referred to the description of the corresponding embodiments. To avoid redundancy, it will not be repeated here.

[0070] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0071] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0072] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0073] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0074] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation, characterized in that, include: S1: A heterologous metabolic pathway for the synthesis of gluconic acid in Escherichia coli, wherein multiple target heterologous enzyme genes are integrated into the Escherichia coli genome or expression plasmid, enabling Escherichia coli to synthesize gluconic acid stepwise from glucose-6-phosphate through this heterologous metabolic pathway; the multiple target heterologous enzyme genes include inositol-3-phosphate synthase gene ino1, inositol oxygenase gene miox, and uronic acid dehydrogenase gene udh. S2: Based on the heterologous metabolic pathway, a kinetic mechanism model for characterizing the glucose diacid fermentation process is established. The kinetic mechanism model uses the concentrations of glucose-6-phosphate, fructose-6-phosphate, inositol and the target heterologous enzyme gene as state variables, and the expression parameters of the inositol oxygenase gene miox as regulatory inputs. S3: Based on the aforementioned kinetic mechanism model, two conflicting optimization objective functions are constructed. One is used to quantify the production loss of gluconic acid production efficiency, and the other is used to quantify the enzyme expression burden of E. coli cells due to metabolic pressure caused by the expression of the target heterologous enzyme gene. S4: Based on the optimization objective function, a dynamic optimization control strategy is designed. At multiple discrete decision moments in the fermentation process, the production loss and enzyme expression burden in future periods are predicted according to the current system state and the kinetic mechanism model. Based on this, the expression control parameters of the inositol oxygenase gene miox in the next period are dynamically optimized to obtain the control strategy that makes the two optimal. S5: The expression regulation parameters obtained by optimization at each discrete decision time are transformed into actual regulation instructions for E. coli genetic elements and implemented in the fermentation process. At the same time, the kinetic mechanism model is updated by combining the actual state monitoring results during the fermentation process to form a closed-loop control. This is executed cyclically until the end of fermentation, thereby automatically maintaining the optimal balance between product synthesis and metabolic load in a dynamic fermentation environment.

2. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation according to claim 1, characterized in that: The expression for the dynamic mechanism model is: ， Wherein, the state vector G6P, F6P, and MI represent the concentrations of glucose-6-phosphate, fructose-6-phosphate, and inositol, respectively; Ino1 and MIOX represent the concentrations of the corresponding enzymes expressed by the heterologous enzyme genes ino1 and miox, respectively; control input Here is the regulatory function for mioxase expression, specifically expressed as follows: , It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of the heterozyme gene miox, which leads to an expression rate of 100%. Intracellular inositol mi concentration at time; It is the Hill coefficient, which characterizes the synergistic activation effect of inositol on the expression of the heterologous enzyme gene miox.

3. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation according to claim 1, characterized in that: The expression for the optimization objective function is as follows: ， ， in, Indicates production loss, used to measure the time spent in the fermentation cycle. Within, the cumulative deviation between the actual synthesis flux of gluconic acid and the theoretical maximum synthesis flux; For a constant glucose uptake rate, This represents the rate of the mioxe enzyme-catalyzed reaction at the current time t; Indicates enzyme expression burden, used to measure fermentation cycle. The cumulative expression levels of the endogenous heterozyme gene ino1 and the heterozyme gene miox; This represents the intracellular concentration of the ino1 enzyme at time t. This represents the intracellular concentration of the mioxase at the current time t.

4. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation according to claim 3, characterized in that: The dynamic optimization and control strategy includes a dynamic fixed-weight optimization strategy, the specific steps of which are as follows: At each discrete decision time Based on the aforementioned dynamic mechanism model, a comprehensive objective function is constructed as follows: ， in, and They are in the future time period Internally, production losses and enzyme expression burden are predicted based on the current state; and These are pre-set fixed weighting coefficients that satisfy... T is a fixed time interval; Solve the optimization problem: To obtain the optimal control parameters It is applied to the fermentation control of the next stage; in, It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of the heterozyme gene miox, which leads to an expression rate of 100%. The concentration of inositol mi at that time.

5. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation according to claim 3, characterized in that: The dynamic optimization and control strategy includes a dynamic adaptive weight optimization strategy, the specific steps of which are as follows: At each discrete decision time Regarding the current control parameters Apply a small relative perturbation The control parameters after adding perturbation are obtained. ; Using the current control parameters respectively and its perturbation value As a control input, the dynamic mechanism model is adjusted from the current decision moment. We begin numerical integration to predict the system dynamics over future time periods and calculate the corresponding objective function value. and the objective function value after perturbation , , This represents the objective function value corresponding to the production loss. This represents the objective function value corresponding to the enzyme expression burden; Define the i-th sub-objective at the current decision time. Influence coefficient: ,in To prevent positive numbers with a denominator of zero; Normalizing the influence yields the initial adaptive weight coefficients for each sub-objective: ; A smoothing mechanism is introduced, and its calculation formula is as follows: , ;in, It is a smoothing coefficient and This controls the response speed of weight updates; For the previous decision moment The adaptive weighting coefficients for the production loss sub-objective; Construct a comprehensive objective function with adaptive weights: Solve the optimization problem: To obtain the optimal control parameters It is applied to the fermentation control of the next stage; in, It is the maximum rate constant of the miox enzyme expression function; It is the half-maximal activation concentration of the heterozyme gene miox, which leads to an expression rate of 100%. The concentration of inositol mi at that time.

6. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation according to claim 1, characterized in that: The method for converting the expression regulation parameters obtained from optimization at each discrete decision time point into actual regulatory instructions for E. coli genetic elements and implementing these instructions in the fermentation process is as follows: The expression regulation parameters include the maximum rate constant of the miox enzyme expression function. Half-maximal activation concentration of the heterozyme gene miox ;for The transcription rate of the inositol oxygenase gene miox is controlled by selecting or designing promoters with appropriate strength; for The response threshold to mi is adjusted by altering the sequence of the transcription factor IpsA or its binding site.

7. The method for regulating the metabolic load and product synthesis balance in gluconic acid fermentation according to claim 1, characterized in that: The inositol-3-phosphate synthase gene is derived from Saccharomyces cerevisiae, the inositol oxygenase gene is derived from mice, the uronic acid dehydrogenase gene is derived from Pseudomonas syringae, and the heterologous metabolic pathway utilizes the host's own SuhB enzyme to dephosphorylate inositol-1-phosphate to inositol.

8. A system for regulating the metabolic load and product synthesis balance in gluconic acid fermentation, characterized in that, Includes the following modules: A heterologous metabolic pathway construction module is used for the heterologous metabolic pathway of gluconic acid synthesis in Escherichia coli. Multiple target heterologous enzyme genes are integrated into the E. coli genome or expression plasmid, enabling E. coli to synthesize gluconic acid stepwise from glucose-6-phosphate through this heterologous metabolic pathway. The multiple target heterologous enzyme genes include inositol-3-phosphate synthase gene ino1, inositol oxygenase gene miox, and uronic acid dehydrogenase gene udh. The fermentation dynamic mechanism modeling module is used to establish a kinetic mechanism model for characterizing the glucose diacid fermentation process based on the heterologous metabolic pathway. The kinetic mechanism model uses the concentrations of glucose-6-phosphate, fructose-6-phosphate, inositol and the target heterologous enzyme gene as state variables, and the expression parameters of the inositol oxygenase gene miox as regulatory inputs. The objective function construction module is used to construct two conflicting optimization objective functions based on the aforementioned kinetic mechanism model. One is used to quantify the production loss of gluconic acid production efficiency, and the other is used to quantify the enzyme expression burden of E. coli cells due to metabolic pressure caused by the expression of the target heterologous enzyme gene. The dynamic regulation parameter optimization module is used to design a dynamic optimization regulation strategy based on the optimization objective function. At multiple discrete decision moments in the fermentation process, based on the current system state and the kinetic mechanism model, it predicts the production loss and enzyme expression burden in future periods, and dynamically optimizes the expression regulation parameters of the inositol oxygenase gene miox in the next period to obtain the regulation strategy that makes the two optimal. The module also includes a closed-loop control instruction execution module, which converts the expression regulation parameters obtained from optimization at each discrete decision moment into actual control instructions for the genetic elements of E. coli and implements these instructions in the fermentation process. At the same time, it updates the kinetic mechanism model by combining the actual state monitoring results during the fermentation process, forming a closed-loop control that is executed cyclically until the end of fermentation, thereby automatically maintaining the optimal balance between product synthesis and metabolic load in a dynamic fermentation environment.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for regulating the metabolic load and product synthesis balance of gluconic acid fermentation as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for regulating the metabolic load and product synthesis balance of gluconic acid fermentation as described in any one of claims 1 to 7.

Images