An l-threonine metabolism disturbance optimization medium and online raman intelligent feeding control method

Abstract

The application specifically relates to a culture medium system of metabolic disturbance optimization and an intelligent feeding control method based on online Raman spectrum dynamic feedback to improve the L-threonine production efficiency of Escherichia coli Bio 59193 fermentation. The optimized feeding medium is composed of 800.00 g / L of glucose, 4.00 g / L of MgSO4.7H2O, 0.22 g / L of thiamine HCl, 0.01 g / L of nicotinamide, 1.50 g / L of betaine hydrochloride, 1.00 g / L of choline chloride and the rest is deionized water. By applying the system, the L-threonine production yield is increased from 31.57 g / L to 87.57 g / L, and the increase amplitude is 1.77 times. Through a manual pulse feeding fermentation experiment, it is confirmed that the optimal glucose concentration fluctuation range is 5-35 g / L, and based on this, the intelligent feeding control method based on online Raman spectrum dynamic feedback is developed, the stable sugar control in the fermentation process is realized, and finally the L-threonine production yield is increased to 94.44 g / L, which is about 1.99 times higher than that of the control group. The method provides an innovative solution for the industrialized fermentation of L-threonine.

Description

Technical Field

[0001] This application relates to the field of microbial fermentation, specifically to a culture medium system with optimized metabolic perturbation and an online Raman intelligent feeding control method to improve the fermentation yield of Escherichia coli Bio 59193L-threonine. Background Technology

[0002] L-Threonine (chemical name α-amino-β-hydroxybutyric acid, molecular formula C4H9NO3) is an essential amino acid with important physiological functions. Its physicochemical properties show that it is readily soluble in water but insoluble in organic solvents such as methanol, ethanol, ether, and chloroform. In terms of stereochemistry, L-threonine is the only one of the four stereoisomers that is biologically active. As one of the eight essential amino acids that the human body cannot synthesize, L-threonine is not only a key substrate for protein biosynthesis but also has wide applications in food additives, animal feed, and pharmaceutical preparations. The global market demand for L-threonine is showing a sustained and rapid growth trend. According to statistics, global production reached 907,000 tons in 2021, of which my country accounted for 93.4% (847,000 tons). From 2017 to 2023, the average annual growth rate of global demand was significant, with a cumulative increase of 76.9%, and global demand is expected to exceed 1.2 million tons in 2024. Microbial fermentation has become the main method for L-threonine production due to its readily available raw materials, relatively simple process, and environmental friendliness.

[0003] Currently, my country has reached an internationally leading level in the construction and screening of high-performance production strains. However, in the industrial application of high-performance strains, how to optimize and regulate the fermentation process efficiently, scale up the fermentation process rationally, improve process control precision and fermentation product production efficiency while reducing energy consumption has become a core technical challenge that urgently needs to be solved in the field of bio-fermentation. During fermentation, the culture medium system, as the nutrient environment for cell growth, not only directly affects productivity but also significantly impacts process economics. Synthetic culture media have unique advantages in fermentation processes due to their good reproducibility, controllability, and monitorability, as well as simpler product separation and purification processes. However, to achieve a biomass productivity comparable to that of composite culture media, it is essential to accurately determine the required nutrients and their concentrations during product synthesis, which has become a major technical challenge in the application of synthetic culture media.

[0004] In existing technologies for producing L-threonine by Escherichia coli fermentation, most methods employ composite culture media. For example, patent applications CN108048501B (Jiangsu Agricultural and Animal Husbandry Science and Technology College, 2017) by Yang Xiaozhi et al., entitled "A culture medium for L-threonine production by engineered Escherichia coli and its application," CN107760734A (Jilin University, 2017) by Wang Jian et al., entitled "A method for increasing threonine yield and its application," CN109439703B (Qiqihar Longjiang Fufeng Biotechnology Co., Ltd., 2018) by Wang Jian et al., entitled "A culture medium for threonine fermentation," and CN118853438A (Langfang Mei...) are examples of such technologies. The fermentation culture medium and method for producing L-threonine by Yuan Erying et al. (Hua Biotechnology Development Co., Ltd., 2023) and the preparation method of an optimized amino acid fermentation culture medium by Shi Shuzhi et al. (Patent CN119060912A, Hulunbuir Northeast Fufeng Biotechnology Co., Ltd., 2024) both contain corn steep liquor or corn husk hydrolysate. The culture medium system developed by Wang Zejian et al. (Patent CN116355972A, East China University of Science and Technology, 2024) for the same strain of Escherichia coli Escherichiacoli Bio 59193 as the present invention contains 16 g / L of corn steep liquor and is a composite culture medium. In contrast, the use of synthetic culture media is relatively limited. A search revealed only two inventions that designed or applied synthetic culture media systems: patent application CN107012181 (Tianjin University of Science and Technology, 2017) by Xu Qingyang et al., entitled "A Threonine Fermentation Culture Medium and a Clean Production Process for Threonine," and patent application CN115029395A (Jiangnan University, 2022) by Luo Wei et al., entitled "A Method for Increasing the Yield of L-Threonine in Escherichia coli." In contrast, the culture medium system composition in the feeding stage of this application is innovatively determined in the fermenter using a stimulus-response dynamic perturbation method, and an intelligent feeding control method is innovatively established by introducing online Raman spectroscopy dynamic feedback during the subsequent fermentation process.

[0005] Existing research indicates that during fermentation, the feeding strategy's control over the residual concentration of key nutrients, especially carbon sources, significantly impacts the production efficiency and yield of the target product. A study by Ienczak et al. from the Federal University of Santa Catarina in Brazil confirmed that feeding strategies play a decisive role in controlling the residual concentration of key nutrients. Lee et al. from the Korea Institute of Bioscience and Biotechnology further pointed out that in L-threonine fermentation, precise control of residual glucose concentration is more critical than other nutrients, directly affecting fermentation productivity. This study only improved cellular glucose utilization by optimizing culture medium components and did not address specific research on dynamic control methods and precision of glucose concentration within the fermenter. Currently, common feeding control strategies mainly include substrate concentration-based methods such as the glucose index feeding method, and physiological parameter-based methods such as the pH constant strategy used by Li Ping et al. from Zhejiang University of Commerce and Industry and Kongklom et al. from Kasetsart University in Thailand, and the DO constant strategy used by Chen Ning et al. from Tianjin University of Science and Technology. However, these strategies all require feeding to be triggered after carbon source depletion, which may lead to fluctuations in the physiological state of microorganisms, thus adversely affecting the production efficiency and final yield of the target product. Therefore, developing innovative process control strategies is of great significance in the industrial fermentation production of L-threonine.

[0006] In traditional industrial fermentation processes, process control relies on periodically opening and closing a closed aseptic reactor for manual sampling, followed by offline detection to obtain key parameter data. This intermittent monitoring method suffers from problems such as detection lag and untimely adjustment, failing to achieve real-time dynamic control of the fermentation process and severely restricting the improvement of production efficiency and product quality. This technological bottleneck urgently needs to be overcome through intelligent online monitoring technology. The development of modern sensing technology has provided new solutions for fermentation process monitoring, with its core advantage being the ability to perform in-situ measurements directly in the reactor environment. Modular sensor design enables the separation of the sensing probe and the measuring instrument, making simultaneous detection of multiple components possible. Among various sensing technologies, non-invasive spectroscopy and optical technologies have attracted much attention due to their non-invasive nature, specifically including ultraviolet-visible spectroscopy, infrared spectroscopy, fluorescence spectroscopy, and Raman spectroscopy. Compared with other sensing technologies, Raman spectroscopy exhibits significant technical advantages in complex fermentation systems: First, compared with the shortcomings of infrared spectroscopy, such as susceptibility to strong absorption interference from water molecules and limited ability to distinguish structural analogs; second, compared with pH and DO electrodes that require frequent calibration, and optical density sensors that are prone to signal saturation under high concentration conditions, Raman spectroscopy can not only achieve real-time online detection and dynamic feedback control, but also has outstanding characteristics such as strong anti-interference ability, non-invasiveness, high specificity, and simultaneous analysis of multiple components, making it particularly suitable for monitoring the production process of high value-added bioproducts.

[0007] The domestically produced Raman analyzer used in this application is from AsiaBio Photonics (Shenzhen) Technology Co., Ltd. Based on its independently developed optical system, it is breaking through the technological barriers of imported equipment. The widespread application of this technology will significantly improve the process monitoring level of synthetic biology and high-end biopharmaceutical production, and powerfully promote the upgrading and development of my country's biomanufacturing industry towards intelligent manufacturing. Summary of the Invention

[0008] This application provides a culture medium system optimized for metabolic perturbation and an online Raman intelligent feeding control method. The fermentation medium is a synthetic medium, and the components of the culture medium system fed during the feeding stage are glucose, MgSO4·7H2O, thiamineHCl, nicotinamide, betaine hydrochloride, and choline chloride. Based on glucose concentration control, two feeding strategies are designed: manual pulse feeding to control glucose concentration between 5-20 g / L, 0-35 g / L, and 5-35 g / L; and online Raman intelligent feedback feeding to maintain glucose concentration at 10 g / L and 20 g / L. The intelligent feedback feeding control is applied to the optimal glucose concentration control strategy, with a glucose concentration control range of 5-35 g / L, and the glucose concentration not falling below 5 g / L throughout fermentation. The online Raman intelligent feedback feeding process established in this application enables precise control and prediction of glucose concentration and the correlation between L-threonine concentration and biomass OD. 600 It has promising prospects for real-time prediction and industrial applications.

[0009] The first objective of this application is to provide a synthetic culture medium system for the fermentation production of L-threonine from Escherichia coli.

[0010] To achieve the above objectives, this application provides the following solution:

[0011] This application provides a fully synthetic fermentation medium containing the following components and their mass concentration ranges: glucose 10.00-30.00 g / L, citric acid 4.00-6.10 g / L, FeSO4·7H2O 0.05-0.10 g / L, KH2PO4 5.50-10.00 g / L, Na2HPO4·12H2O 2.00-4.00 g / L, (NH4)2SO4 0.50-1.00 g / L, (NH4)2HPO4 5.50-10.50 g / L, MgSO4·7H2O 0.60-1.00 g / L, D-calcium pantothenate 0.01-0.03 g / L, nicotinic acid 0.01-0.03 g / L, pyridoxine hydrochloride 0.01-0.03 g / L, and thiamine HCl. 0.01-0.03 g / L, Biotin 0.10-1.00 mg / L, Trace Elements 8.00-10.00 mL / L (the trace elements are denoted as: MnSO4·H2O 1.00-2.00 g / L, CoCl2·6H2O 1.00-2.00 g / L, ZnSO4·7H2O 0.20-0.50 g / L, AlCl3·6H2O 0.50-1.00 g / L, Na2MoO4·2H2O 0.40-1.00 g / L, CuSO4·5H2O 0.20-0.50 g / L, H3BO3 0.10-0.50 g / L, NH4Cl 10.00-12.50 g / L, CaCl2·2H2O) 7.00-8.00 g / L), mixed amino acids 99.00-105.00 mL / L (the amino acids are designated as: L-Ile 2.00-4.00 g / L, L-Ala 2.00-4.00 g / L, L-Cys 2.00-4.00 g / L, L-Arg 2.00-4.00 g / L, L-Leu 2.00-4.00 g / L, L-Pro 3.00-5.00 g / L, L-Asp 3.00-5.00 g / L), the remainder is deionized water, pH 6.50-7.50.

[0012] Further, the composition and optimal concentration of the culture medium are as follows: glucose 20.00 g / L, citric acid 6.00 g / L, FeSO4·7H2O 0.06 g / L, KH2PO4 6.00 g / L, Na2HPO4·12H2O 2.95 g / L, (NH4)2SO4 0.92 g / L, (NH4)2HPO4 10.00 g / L, MgSO4·7H2O 0.65 g / L, D-calcium pantothenate 0.02 g / L, nicotinic acid 0.02 g / L, pyridoxine hydrochloride 0.02 g / L, thiamine HCl 0.02 g / L, biotin 0.20 mg / L, and trace elements 8.35 mL / L (the trace elements are denoted as: MnSO4·H2O 1.92 g / L, CoCl2·6H2O). The mixture contained 1.20 g / L of ZnSO4·7H2O, 0.30 g / L of AlCl3·6H2O, 0.75 g / L of Na2MoO4·2H2O, 0.50 g / L of CuSO4·5H2O, 0.30 g / L of H3BO3, 0.16 g / L of NH4Cl, 12.00 g / L of CaCl2·2H2O, and 7.92 g / L of mixed amino acids (the amino acids were designated as: L-Ile 2.50 g / L, L-Ala 2.50 g / L, L-Cys 2.50 g / L, L-Arg 2.50 g / L, L-Leu 2.50 g / L, L-Pro 3.28 g / L, and L-Asp 3.94 g / L), with the remainder being deionized water at pH 7.00. The culture medium system in the feeding stage is determined using a stimulus-response dynamic perturbation method, which includes the following steps: real-time monitoring of the carbon dioxide release rate (CER) during the metabolic culture of the strain; adding candidate substances to the fermentation system in a single pulse when the CER reaches a stable state; and screening candidate substances that significantly improve the CER as effective components of the feeding medium. The stable state of CER refers to a CER fluctuation range of less than ±5% for at least 30 minutes, and a significant improvement in CER refers to a CER increase exceeding 15% of the stable state after the pulse.

[0013] This application also provides a feeding control method for L-threonine production by *E. coli* based on Raman spectroscopy dynamic feedback. This includes collecting spectral data online using a Raman analyzer for each fermentation batch, processing the data, and then combining it with a PLSR model and PID control to develop an intelligent feedback feeding control process. Two feeding strategies are designed: manual pulse feeding and online Raman intelligent feedback feeding. This allows for the prediction and control of glucose concentration, L-threonine concentration, and biomass OD within the fermenter. 600 The prediction.

[0014] Furthermore, each batch of Raman signals undergoes preprocessing, using a Savitsky-Golay filter for noise reduction. A window width of 7 and a fitting order of 1 are selected to smooth the spectral data and reduce the impact of differences in excitation light intensity, detector noise, and sample fluorescence background on the acquired spectral data.

[0015] Furthermore, the culture medium, after being sterilized at 121°C for 30 minutes, is fermented in a 5L fermenter. The aeration rate is 1.5-3 VVM, the stirring speed is 400-1000 rpm, the fermentation temperature is 37°C, the pH is 7.00, and the dissolved oxygen is controlled to be no less than 30% during fermentation.

[0016] Furthermore, based on glucose concentration control, two feeding strategies were designed: manual pulse feeding to control glucose concentration fluctuations between 5-20 g / L, 0-35 g / L, and 5-35 g / L, and intelligent feedback feeding to maintain glucose concentration at 10 g / L and 20 g / L. The intelligent feedback feeding control was applied to the optimal glucose concentration control strategy, i.e., pulse feeding batch fermentation to control glucose concentration fluctuations between 5-35 g / L, with the glucose concentration not falling below 5 g / L throughout the fermentation process.

[0017] After using the fully synthetic culture medium and the culture medium system of the feeding stage described in this application to produce L-threonine from *E. coli* in a 5L fermenter, the L-threonine yield was 87.57 g / L, an increase of 1.77 times, and the OD... 600 The concentration of L-threonine was 33.13, an increase of 1.74 times, while the secretion of the byproduct acetic acid was 3.28 g / L, comparable to the pre-optimization level. Using the intelligent feeding control method based on Raman spectroscopy dynamic feedback described in this application, the optimal batch, under intelligent feedback, achieved an L-threonine concentration of 94.44 g / L, an increase of 1.99 times, and an OD... 600 The concentration was 42.13, an increase of 2.48 times, while the secretion of the byproduct acetic acid was 0.43 g / L, a decrease of 61.61%.

[0018] This application optimizes the composition of the culture medium system during the feeding phase and designs two feeding strategies: manual pulse feeding to control glucose concentration between 5-20 g / L, 0-35 g / L, and 5-35 g / L; and online Raman intelligent feedback feeding to maintain glucose concentration at 10 g / L and 20 g / L. The intelligent feedback feeding control is applied to the optimal glucose concentration control strategy, achieving a glucose concentration control range of 5-35 g / L, with the glucose concentration not falling below 5 g / L throughout fermentation. The developed online Raman intelligent feedback feeding control platform can predict and control glucose concentration during fermentation, as well as the product L-threonine and biomass OD. 600 The prediction and the provision of the optimal fed-batch fermentation strategy are of great significance for the industrial production of L-threonine by Escherichia coli. Attached Figure Description

[0019] To more clearly illustrate the embodiments and technical solutions in this application, the accompanying drawings used in the embodiments are briefly described below.

[0020] Figure 1 Cell respiration rate before and after optimization of the culture medium system during the feeding phase;

[0021] Figure 2 (A) shows the cell growth before and after optimization of the culture medium system during the feeding phase;

[0022] Figure 2 (B) shows the L-threonine production before and after optimization of the culture medium system during the feeding stage;

[0023] Figure 2 (C) shows the secretion of acetic acid, a byproduct, before and after optimization of the culture medium system during the feeding stage;

[0024] Figure 3 (A) Before Raman signal preprocessing;

[0025] Figure 3 (B) After Raman signal preprocessing;

[0026] Figure 4 (A) A batch feeding strategy based on Raman intelligent feedback to maintain glucose concentration at 20 g / L, with real-time prediction of OD. 600 The concentrations of glucose and L-threonine were compared with those obtained by spectrophotometer, biochemical analyzer and HPLC offline data;

[0027] Figure 4 (B) A batch feeding strategy based on Raman intelligent feedback to maintain glucose concentration at 10 g / L, with real-time prediction of OD. 600 The concentrations of glucose and L-threonine were compared with those obtained by spectrophotometer, biochemical analyzer and HPLC offline data;

[0028] Figure 4 (C) Optimal batch Raman feedback control after process optimization, feeding strategy with pulse control of glucose concentration fluctuations within 5-35 g / L, and real-time prediction of OD. 600 The concentrations of glucose and L-threonine were compared with those obtained by spectrophotometer, biochemical analyzer and HPLC offline data;

[0029] Figure 5 (A) Compared with the culture medium system before the feed stage optimization, the cell growth of the pulse-feed batch fermentation controlled the glucose concentration fluctuation within 5-20 g / L.

[0030] Figure 5(B) Compared with the culture medium system before the feed stage optimization, the cell growth of the batches controlled by pulse feed batch fermentation with glucose concentration fluctuations of 0-35 g / L.

[0031] Figure 5 (C) Compared with the culture medium system before the feeding stage optimization, the cell growth of the two batches of pulse feeding batch fermentation with glucose concentration fluctuations of 5-35 g / L under manual feeding and intelligent feedback feeding.

[0032] Figure 5 (D) Compared with the culture medium system before the feeding stage optimization, the cell growth of the batch with intelligent feedback feeding control maintaining glucose concentration at 10 g / L.

[0033] Figure 5 (E) Compared with the culture medium system before the feeding stage optimization, the cell growth of the batch with intelligent feedback feeding control maintaining glucose concentration at 20 g / L.

[0034] Figure 5 (F) Comparison of cell growth under five fed-batch fermentation glucose control strategies;

[0035] Figure 6 (A) Compared with the culture medium system before the feed stage optimization, the production of L-threonine in batches of pulse-fed batch fermentation with glucose concentration fluctuations of 5-20 g / L was controlled;

[0036] Figure 6 (B) Compared with the culture medium system before the feed stage optimization, the production of L-threonine in batches of pulse-fed batch fermentation with glucose concentration fluctuations of 0-35 g / L was controlled;

[0037] Figure 6 (C) Compared with the culture medium system before the feed stage optimization, the production of L-threonine in two batches of manual feed and intelligent feedback feed batch fermentation with glucose concentration fluctuations controlled at 5-35 g / L;

[0038] Figure 6 (D) Compared with the culture medium system before the feed stage optimization, the production of L-threonine batches with intelligent feedback feed control maintaining glucose concentration at 10 g / L.

[0039] Figure 6 (E) Compared with the culture medium system before the feed stage optimization, the production of L-threonine batches with intelligent feedback feed control maintaining glucose concentration at 20 g / L.

[0040] Figure 7 Pulsed fed batch fermentation was used to control glucose concentration fluctuations within 0-35 g / L per batch, and to study strain growth, L-threonine production, and glucose utilization.

[0041] Figure 8 (A) Compared with the culture medium system before the feed stage optimization, the secretion of acetic acid, a byproduct, was controlled in the pulse-feed batch fermentation with glucose concentration fluctuations of 5-20 g / L during the batch feeding stage.

[0042] Figure 8 (B) Compared with the culture medium system before the feed stage optimization, the secretion of acetic acid, a byproduct, was controlled in the pulse-feed batch fermentation with glucose concentration fluctuations of 0-35 g / L during the feed stage.

[0043] Figure 8 (C) Compared with the culture medium system before the feeding stage, the secretion of acetic acid, a byproduct, was controlled in the pulse feeding batch fermentation process with glucose concentration fluctuations of 5-35 g / L in both manual feeding and intelligent feedback feeding stages.

[0044] Figure 8 (D) Compared with the culture medium system before the feed stage optimization, the secretion of acetic acid by-product during the batch feed stage with intelligent feedback feed control to maintain glucose concentration at 10 g / L.

[0045] Figure 8 (E) Compared with the culture medium system before the feed stage optimization, the intelligent feedback feed control maintains the glucose concentration at 20 g / L during the batch feed stage and controls the secretion of acetic acid, a byproduct. Detailed Implementation

[0046] To provide a clearer and more complete explanation of this application, the following description is provided in conjunction with the accompanying drawings and embodiments of the technical solutions. Obviously, the described embodiments are merely some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0047] Furthermore, the numerical ranges in this application should be understood to specifically disclose each intermediate value within the upper and lower limits of each numerical range. Any stated value and the intermediate value within a stated value range, and every smaller range between any other stated value and the intermediate value within said range, also falls within the scope of this application. The upper or lower limits of these smaller ranges may be independently included or excluded from the scope.

[0048] The present application will be further illustrated below with reference to specific embodiments. It should be understood that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the scope of the application.

[0049] The strain used in this application is *Escherichia coli* Bio 59193, a strain preserved at the National Biochemical Engineering Technology Research Center (Shanghai). The compositions of the plate seed medium, liquid seed medium, initial total synthetic medium for the 5L fermenter, and fed-batch medium used before and after optimization are shown in Table 1 below.

[0050]

[0051]

[0052] Preparation of the feed culture system before and after optimization: The culture system is 500mL. First, weigh 440g of glucose monohydrate in a beaker, add an appropriate amount of deionized water, and then place it in a water bath. Stir with a glass rod to accelerate dissolution. After dissolution, pour it into a 500mL graduated cylinder and bring the volume to 420mL. Then transfer it to a 500mL feed bottle (with a rotor placed in the feed bottle). Clamp the feed tube with a water-stop clamp. Wrap the feed port with 8 layers of gauze and kraft paper and tie it tightly. Place it in an autoclave at 115℃ for 20min for sterilization. Weigh 2g of MgSO4·7H2O and 0.11g of thiamine HCl (after optimization, weigh 0.5g of choline chloride, 0.005g of nicotinamide, and 0.75g of betaine hydrochloride). Bring the volume to 80mL with deionized water, stir to dissolve, and then filter it in a laminar flow hood using a 0.22μm sterile water filter for sterilization. After the glucose has cooled to room temperature, mix the glucose with other supplementary ingredients in a clean bench. After mixing, place the mixture in a stirrable oil bath and stir until homogeneous before use.

[0053] Preparations before operating a 5L fermenter:

[0054] S1, Electrode calibration: After the dissolved oxygen electrode is placed in anhydrous sodium sulfite and the reading drops to 0, calibrate the 0 point; the pH electrode is calibrated with pH standard solution at pH 4.0 and 6.86 respectively to ensure that the parameters of the fermentation process are monitored accurately.

[0055] S2, prepare the culture medium according to Table 1 above, with a liquid volume of 40%, and sterilize at 121℃ for 30 minutes.

[0056] S3, Inoculation: After sterilization, wait for the culture medium temperature to drop to 37℃, then introduce air to maintain the pressure in the container at 0.05MPa. After igniting the flame ring, open the inoculation port and inoculate the secondary seed culture (resuspended in physiological saline) into the culture medium at a ratio of 10% (200mL). After inoculation, a sample can be taken to determine if the sugar concentration at 0 point is correct. Initial culture parameters are set as follows: aeration rate 1.5 vvm, rotation speed 400 rpm, temperature 37℃, and container pressure 0.05MPa. After stabilization, the dissolved oxygen level is calibrated to full volume.

[0057] S4, Tail Gas Spectrometry Online Monitoring: The tail gas generated during fermentation is introduced into a connected tail gas spectrometer through a pipeline, and is ionized to form charged molecular ions and fragment ions. The ionized ions are introduced into a mass analyzer for separation according to the charge-to-mass ratio (m / z). The separated ions arrive at the detector in sequence, thereby monitoring the tail gas composition and corresponding concentration data in real time. The software collects the tail gas data and calculates the real-time oxygen consumption rate (OUR) and carbon dioxide release rate (CER) of the cells according to the formula.

[0058] During fermentation, the aeration rate was set to 1.5-3 VVM, the stirring speed was 400-1000 rpm, the fermentation temperature was 37℃, the pH was 7.00, and the dissolved oxygen was controlled to be no less than 30% during fermentation.

[0059] Biomass measurement value is OD 600 Set the spectrophotometer wavelength to 600 nm, using the initial uninoculated culture medium as a blank control group and zeroing the meter. Transfer an appropriate amount of fermentation broth to a cuvette, typically filling it to 2 / 3 full. After ensuring the cuvette is free of air bubbles, place it in the spectrophotometer and record the OD. 600 Value. If the bacterial concentration in the fermentation broth is too high, the bacterial solution can be appropriately diluted before measurement, and the dilution factor should be recorded to ensure the measured OD value is accurate. 600 The biomass value is obtained by multiplying the measured value by the dilution factor when the value is in the range of 0.2-0.8.

[0060] Determination of glucose concentration: Use a syringe to draw 25 μL of a 1 g / L glucose solution to calibrate the glucose enzyme membrane of the biochemical analyzer (Shenzhen Hillman Technology Co., Ltd. S-10). After successful calibration, appropriately dilute the sample (to ensure the concentration is within the analyzer's measurement threshold of 1 g / L), record the dilution factor, invert the sample to mix thoroughly, and then inject 25 μL of the diluted sample solution into the biochemical analyzer. After waiting 20 seconds, record the reading. Multiplying this reading by the dilution factor gives the glucose concentration in the sample.

[0061] Determination of L-threonine concentration: High performance liquid chromatography (HPLC) was used. The filtered supernatant sample was serially diluted and used for OPA derivatization. The procedure was as follows:

[0062] S1, 2.47g of boric acid was weighed and dissolved in 90mL of deionized water. The pH was adjusted to 10.2 using 40% KOH solution and 10mol / L NaOH solution, respectively, and then the volume was brought to 100mL to obtain 0.4mol / L potassium borate solution and borate buffer solution.

[0063] Preparation of S2, OPA solution: Weigh 0.025g of o-phthalaldehyde powder, dissolve it in 500μL of methanol, add 50μL of 3-mercaptopropionic acid, and make up to 5mL with 0.4mol / L potassium borate solution. Filter through a 0.22μm organic filter membrane and place in a brown reagent bottle.

[0064] The derivatization reaction system for S3, OPA is: 150 μL borate buffer, 550 μL double-distilled water, 100 μL OPA solution, and 100 μL sample or standard to be tested.

[0065] S4. The mobile phase consisted of two parts, A and B. Mobile phase A consisted of 0.2 mL triethylamine, 5 g sodium acetate, 5 mL tetrahydrofuran, and 1 L ddH2O, adjusted to pH 7.2, and then filtered through a 0.22 μm aqueous membrane. Mobile phase B consisted of 5 g sodium acetate, 200 mL ddH2O, 400 mL methanol, and 400 mL acetonitrile, adjusted to pH 7.2, and then filtered through a 0.22 μm organic membrane. After filtration, the mobile phase was sonicated for 30 min to ensure the homogeneity and stability of the solution. The elution procedure for L-threonine concentration determination was as follows: Samples were analyzed using a high-performance liquid chromatography (HPLC) system (Agilent 1100) equipped with a C18 column (250 mm × 4.6 mm, 5 μm). The column temperature was 40 °C, and detection was performed at 338 nm using a UV detector. The injection volume is 20 μL, and the gradient elution program (proportion of mobile phase B) is: 0-11 min: 10%-67% (flow rate 1 mL / min). -1 ), 11-14.50 min: 67%-100% (flow rate 1-1.2 mL·min) -1 ), 14.50-15 min: 100% (flow rate 1.2 mL / min) -1 ), 15-17 min: 100%-10% (flow rate 1.2-1 mL / min) -1 ), 17-18 min: 10% (flow rate 1 mL / min) -1 ).

[0066] Determination of acetic acid concentration: Acetic acid was quantitatively analyzed using a Shimadzu LC-20A high-performance liquid chromatography system equipped with a Hi-Plex H (300 mm × 7.7 mm) and a concentration of 5 mmol / L. -1 H2SO4 was used as the mobile phase, and the flow rate was 0.4 mL·min. -1 The column temperature was 50℃, and the ultraviolet absorption was measured at 210nm.

[0067] Example 1: The effects of the fed-batch culture medium system before and after optimization on cell growth, production and by-product secretion. The specific steps are as follows: The strain was subjected to chemostat culture. When the CER stabilized (fluctuation amplitude less than ±5% for at least 30 minutes), nicotinamide, betaine hydrochloride and choline chloride were sequentially pulsed into the fermenter for discrete dynamic disturbance test. The component whose CER increase after pulse exceeded 15% of the stable state was determined to be the required component of the culture medium system.

[0068] Fed-feed batch fermentation: Before and after optimization of the fed-feed culture medium system, the feeding method implemented on the 5L fermenter was pulsed fed-feed batch fermentation, controlling the glucose concentration in the fermenter to be 5-20g / L, using a peristaltic pump of model BT100-2J and a No. 16 tubing.

[0069] The effects of optimizing the fed-batch culture medium system on cell growth, production, and byproduct secretion.

[0070] Figure 1 It can be seen that before the culture medium system was optimized during the feeding stage, the maximum CER was 60 mmol / L / h, and the CER remained constant during the subsequent feeding period. After the culture medium system was optimized during the feeding stage, the original limit of bacterial respiration intensity was broken, and the maximum CER was 276 mmol / L / h. Figure 2 (A) From the perspective of strain growth, the growth OD of the strain before optimization can be obtained. 600 The optimized OD version is 12.11. 600 The value was 33.13, which was 1.74 times higher than that of the control group. The most significant advantage of the optimized composition here is that it broke through the original growth plateau.

[0071] Figure 2 (B) Using the unoptimized batch as a control group, from the perspective of L-threonine production, the yield before optimization was 31.57 g / L, and the yield after optimization was 87.57 g / L, which is 1.77 times higher than the control group. The curve of the unoptimized batch shows that product synthesis is limited in the later stage of fed-batch synthesis, while product synthesis after optimized fed-batch composition breaks through the original limitations in the later stage of fed-batch synthesis.

[0072] Figure 2 (C) From the perspective of byproduct secretion, it can be seen that the highest acetic acid content before optimization was 2.79 g / L, and the acetic acid content at the end of fermentation after optimization was 3.28 g / L. This may be related to the shortened batch fermentation cycle and accelerated cell growth rate after optimization.

[0073] Example 2: A smart feeding control method for L-threonine production by Escherichia coli based on dynamic feedback of Raman spectroscopy.

[0074] Calibration and sterilization of the Raman electrode: The Raman electrode was calibrated with anhydrous ethanol and cyclohexane under the following conditions: 300 ms integration time, 20 scans, and 65% laser intensity. After successful calibration, the Raman electrode was inserted into the fermenter along with the fermenter and culture medium and sterilized at 121°C for 30 min. After cooling to room temperature, the Raman spectrometer was connected, and spectral data were acquired. During acquisition, the Raman laser power was 0.45 W, the integration time was manually adjusted to 2000 ms, the scans were averaged to 10, and the laser intensity was 70%. The acquired Raman spectral data were processed using RAMAN software.

[0075] Raman spectroscopy preprocessing: Acquired Raman spectra are often affected by various factors, including differences in excitation light intensity, detector noise, and sample fluorescence background. Therefore, preprocessing of the acquired spectra is necessary. During preprocessing, a Savitsky-Golay filter is used for noise reduction, with a window width of 7 and a fitting order of 1, to smooth the spectral data and reduce the loss of important features.

[0076] Establishment of Raman intelligent feedback feeding platform: Raman spectral data collected during multiple fermentation processes were processed and constructed using partial least squares regression (PLSR) model and PID control.

[0077] The five glucose concentration control strategies designed in this application are as follows:

[0078] (1) First method: When the initial glucose concentration in the fermenter drops from 20g / L to 5g / L, manually pulse feed is added from the feed bottle to 20g / L by the pump and then the feeding is stopped. When the glucose concentration in the fermenter drops to 5g / L again, this operation is repeated. After all the liquid in the feed bottle is replenished and the glucose concentration in the fermenter is depleted, the fermentation ends.

[0079] (2) The second method: When the initial glucose concentration in the fermenter drops from 20 g / L to 0 g / L (the pH value in the fermenter rises and the dissolved oxygen rises, at which point feeding is required), the feed is manually pulsed from the feed bottle to 35 g / L and then feeding is stopped. When the glucose concentration in the fermenter drops to 0 g / L again, this operation is repeated. After all the feed liquid in the feed bottle is replenished and the glucose concentration in the fermenter is depleted, the fermentation ends.

[0080] (3) The third method: When the initial glucose concentration in the fermenter drops from 20 g / L to 5 g / L, manually pulse feed is added from the feed bottle to 35 g / L and then feeding is stopped. When the glucose concentration in the fermenter drops to 0 g / L again, this operation is repeated. After all the liquid in the feed bottle is replenished and the glucose concentration in the fermenter is depleted, the fermentation ends.

[0081] (4) Fourth method: When the initial glucose concentration in the fermenter drops from 20 g / L to 10 g / L, the glucose concentration in the fermenter is set to a constant value of 10 g / L through the Raman intelligent feedback feeding platform. Fermentation ends when all the liquid in the feeding bottle is replenished and the glucose concentration in the fermenter is depleted.

[0082] (5) Fifth method: When the initial glucose concentration in the fermenter drops from 20 g / L to 10 g / L, the glucose concentration in the fermenter is set to a constant value of 20 g / L through the Raman intelligent feedback feeding platform. Fermentation ends when all the liquid in the feeding bottle is replenished and the glucose concentration in the fermenter is depleted.

[0083] 1. Raman spectroscopy preprocessing.

[0084] Figure 3 (A) Raman signal before pretreatment. As can be seen, the Raman signal is affected by interference from detector differences or impurities in the culture medium.

[0085] Figure 3 (B) Preprocessed Raman signal. Denoising was performed using a Savitsky-Golay filter with a window width of 7 and a fitting order of 1, and the spectral data was smoothed.

[0086] 2. L-Threonine Fermentation Process Based on Raman Intelligent Feedback Control

[0087] Figure 4 (A) A batch feeding strategy based on Raman feedback to maintain glucose concentration at 20 g / L, with real-time prediction of OD. 600 The concentrations of glucose and L-threonine were compared with those obtained from spectrophotometer, biochemical analyzer, and HPLC offline data. The predicted error for glucose concentration was 2.80 g / L; the controlled error for glucose concentration was 2.92 g / L; and the L-threonine concentration and OD... 600 The prediction errors were 2.39 g / L and 4.70 g / L, respectively.

[0088] Figure 4 (B) A batch feeding strategy based on Raman feedback to maintain glucose concentration at 10 g / L, with real-time prediction of OD. 600 The concentrations of glucose and L-threonine were compared with those obtained from spectrophotometer, biochemical analyzer, and HPLC offline data. The predicted error for glucose concentration was 2.06 g / L; the controlled error for glucose concentration was 2.32 g / L; and the L-threonine concentration and OD... 600 The prediction errors were 1.7 g / L and 2.73 g / L, respectively.

[0089] Figure 4(C) Optimal batch Raman feedback control after process optimization, feeding strategy to control glucose concentration fluctuations within 5-35 g / L, and real-time prediction of OD. 600 The concentrations of glucose and L-threonine were compared with those obtained from spectrophotometer, biochemical analyzer, and HPLC offline data. The predicted error for glucose concentration was 2.87 g / L; the control error for glucose concentration was 3.5 g / L; and the L-threonine concentration and OD... 600 The prediction errors were 4.02 g / L and 2.92 g / L, respectively.

[0090] 3. Cell growth under five fed-batch fermentation strategies

[0091] Figure 5 (A) Compared to the culture medium system before optimization during the fed-batch fermentation phase, the cell growth of cells in batches with glucose concentration fluctuations of 5-20 g / L under pulsed fed-batch fermentation was compared to the highest OD of cells in the control group before optimization. 600 The value is 12.11. Under this feeding strategy, the OD... 600 It is 33.13, an increase of 1.74 times.

[0092] Figure 5 (B) Compared to the culture medium system before optimization during the fed-batch fermentation phase, the cell growth of cells in batches with glucose concentration fluctuations ranging from 0 to 35 g / L under pulsed fed-batch fermentation was compared. The highest cell biomass OD was observed in the control group under fed-batch fermentation before optimization. 600 The value is 12.11. Under this feeding strategy, the OD... 600 It reached 33.78, an increase of 1.79 times.

[0093] Figure 5 (C) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the cell growth of two batches under manual and intelligent feedback fed-batch fermentation with glucose concentration fluctuations controlled at 5-35 g / L was compared. The highest biomass OD of the cells under the fed-batch fermentation control group before optimization was also compared. 600 The value is 12.11. Under this replenishment strategy, the OD (Original Displacement) of manual replenishment and intelligent feedback replenishment is... 600 The values ​​were 34.5 and 42.13 respectively, representing increases of 1.85 and 2.48 times respectively.

[0094] Figure 5 (D) Compared to the culture medium system before optimization during the fed-batch fermentation phase, the cell growth under intelligent feedback fed-batch control with glucose concentration maintained at 10 g / L was compared to the control group before optimization, showing the highest cell biomass OD. 600 The value is 12.11. Under this feeding strategy, the OD... 600 It is 33.85, an increase of 1.80 times.

[0095] Figure 5(E) Compared to the culture medium system before optimization during the fed-batch fermentation phase, the intelligent feedback fed-batch control maintained the glucose concentration at 20 g / L for cell growth. The highest biomass OD of the cells under the fed-batch fermentation control group before optimization. 600 The value is 12.11. Under this feeding strategy, the OD... 600 It reached 39.58, an increase of 2.27 times.

[0096] Figure 5 (F) Comparison of cell growth under five fed-batch fermentation strategies. In terms of growth, the optimal feeding strategy is pulsed fed-batch fermentation, controlling the sugar concentration fluctuation in the fermenter between 5-35 g / L. Intelligent feedback feeding shows even greater advantages compared to manual feeding, with lower OD... 600 It increased from 34.50 to 42.13, a rise of 22.12%.

[0097] L-Threonine Production under 4-5 Fed-batch Fermentation Strategies

[0098] Figure 6 (A) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the production of L-threonine was controlled with glucose concentration fluctuations of 5-20 g / L in batches. Before optimization, the product yield under the fed-batch fermentation control group was 31.57 g / L, while under this feeding strategy, the L-threonine yield was 87.57 g / L, an increase of 1.77 times.

[0099] Figure 6 (B) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the production of L-threonine was controlled by controlling the glucose concentration fluctuation within 0-35 g / L in batches. Before optimization, the product yield under the fed-batch fermentation control group was 31.57 g / L, while under this fed-batch fermentation strategy, the L-threonine yield was 87.73 g / L, an increase of 1.78 times.

[0100] Figure 6 (C) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the production of L-threonine was controlled in two batches under manual and intelligent feedback feeding, with glucose concentration fluctuations ranging from 5 to 35 g / L. Before optimization, the product yield under the control group fed-batch fermentation was 31.57 g / L. Under manual feeding and Raman intelligent feedback feeding, the L-threonine yields were 91.3 g / L and 94.44 g / L, respectively, representing increases of 1.89 and 1.99 times.

[0101] Figure 6(D) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the production of L-threonine under intelligent feedback fed-batch control with glucose concentration maintained at 10 g / L was as follows: Before optimization, the product yield under the fed-batch fermentation control group was 31.57 g / L, while under this fed-batch strategy, the L-threonine yield was 88.49 g / L, an increase of 1.80 times.

[0102] Figure 6 (E) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the production of L-threonine was controlled by intelligent feedback feeding to maintain the glucose concentration at 20 g / L. Before optimization, the product yield under the fed-batch fermentation control group was 31.57 g / L, while under this feeding strategy, the L-threonine yield was 86.46 g / L, an increase of 1.74 times.

[0103] Figure 7 It can be seen that when the sugar concentration is controlled to fluctuate between 0-35 g / L in pulsed fed-batch fermentation, the product will be consumed by the cells when the glucose concentration is below about 5 g / L. Moreover, as the feeding continues and the cell mass increases, the product will be consumed by the cells more and more. This can be seen at 22.91 h, 30.73 h, 34.48 h, 37.7 h and 40.70 h of the fermentation cycle curve. At these five time points, the product was consumed at 0.7 g / L, 2.3 g / L, 4.3 g / L, 11.6 g / L and 34 g / L respectively.

[0104] 5. Acetic acid secretion as a byproduct during the feeding stage under five fed-batch fermentation strategies.

[0105] Figure 8 (A) Compared with the culture medium system before optimization during the fed-batch fermentation stage, the secretion of acetic acid, a byproduct, was controlled by the pulsed fed-batch fermentation, with glucose concentration fluctuations ranging from 5 to 20 g / L. At the fermentation endpoint, the acetic acid concentration in the control group was 1.12 g / L, while after optimization of the fed-batch culture medium, the acetic acid concentration at the fermentation endpoint was 3.09 g / L.

[0106] Figure 8 (B) Compared with the culture medium system before the feed-feeding phase optimization, the secretion of acetic acid, a byproduct, was controlled during the feed-feeding phase of the pulsed feed-feeding batch fermentation, with glucose concentration fluctuations ranging from 0 to 35 g / L. At the fermentation endpoint, the acetic acid concentration in the control group was 1.12 g / L, while under this feed-feeding strategy, the acetic acid concentration at the fermentation endpoint was 1.32 g / L.

[0107] Figure 8(C) Compared to the culture medium system before optimization during the fed-batch fermentation stage, the secretion of acetic acid, a byproduct, was controlled in two batches during both manual and intelligent feedback feeding stages, with glucose concentration fluctuations ranging from 5 to 35 g / L. At the fermentation endpoint, the acetic acid concentration in the control group was 1.12 g / L. Under this feeding strategy, the acetic acid concentrations at the fermentation endpoint were 0.16 g / L and 0.43 g / L under manual and Raman intelligent feedback feeding, respectively, representing decreases of 85.71% and 61.61%.

[0108] Figure 8 (D) Compared to the culture medium system before optimization during the fed-batch feeding phase, the intelligent feedback feeding control maintained the glucose concentration at 10 g / L, controlling the secretion of the byproduct acetic acid during the batch feeding phase. At the fermentation endpoint, the acetic acid concentration in the control group was 1.12 g / L, while under this feeding strategy, the acetic acid concentration at the fermentation endpoint was 2 g / L.

[0109] Figure 8 (E) Compared to the culture medium system before optimization during the fed-batch feeding phase, the intelligent feedback feeding control maintained the glucose concentration at 20 g / L and controlled the secretion of the byproduct acetic acid during the batch feeding phase. At the fermentation endpoint, the acetic acid concentration in the control group was 1.12 g / L, while under this feeding strategy, the acetic acid concentration at the fermentation endpoint was 0.44 g / L.

[0110] The present application has been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present application. Therefore, the content of this specification should not be construed as a limitation of the present application.

Claims

1. A culture medium system for the fermentation production of L-threonine by Escherichia coli, characterized in that, The culture medium system is a fully synthetic culture medium, with the following composition: glucose 10.00-30.00 g / L, citric acid 4.00-6.10 g / L, FeSO4·7H2O 0.05-0.10 g / L, KH2PO4 5.50-10.00 g / L, Na2HPO4·12H2O 2.00-4.00 g / L, (NH4)2SO4 0.50-1.00 g / L, (NH4)2HPO4 5.50-10.50 g / L, MgSO4·7H2O 0.60-1.00 g / L, D-calcium pantothenate 0.01-0.03 g / L, nicotinic acid 0.01-0.03 g / L, pyridoxine hydrochloride 0.01-0.03 g / L, and thiamine HCl. 0.01-0.03 g / L, Biotin 0.10-1.00 mg / L, Trace Elements 8.00-10.00 mL / L (the trace elements are denoted as: MnSO4·H2O 1.00-2.00 g / L, CoCl2·6H2O 1.00-2.00 g / L, ZnSO4·7H2O 0.20-0.50 g / L, AlCl3·6H2O 0.50-1.00 g / L, Na2MoO4·2H2O 0.40-1.00 g / L, CuSO4·5H2O 0.20-0.50 g / L, H3BO3 0.10-0.50 g / L, NH4Cl 10.00-12.50 g / L, CaCl2·2H2O) 7.00-8.00 g / L), mixed amino acids 99.00-105.00 mL / L (the amino acids are designated as: L-Ile 2.00-4.00 g / L, L-Ala 2.00-4.00 g / L, L-Cys 2.00-4.00 g / L, L-Arg 2.00-4.00 g / L, L-Leu 2.00-4.00 g / L, L-Pro 3.00-5.00 g / L, L-Asp 3.00-5.00 g / L), the remainder is deionized water, pH 6.50-7.

50.

2. The method as described in claim 1, characterized in that, The culture medium used consisted of: glucose 20.00 g / L, citric acid 6.00 g / L, FeSO4·7H2O 0.06 g / L, KH2PO4 6.00 g / L, Na2HPO4·12H2O 2.95 g / L, (NH4)2SO4 0.92 g / L, (NH4)2HPO4 10.00 g / L, MgSO4·7H2O 0.65 g / L, D-calcium pantothenate 0.02 g / L, nicotinic acid 0.02 g / L, pyridoxine hydrochloride 0.02 g / L, thiamine HCl 0.02 g / L, biotin 0.20 mg / L, and trace elements 8.35 mL / L (the trace elements were designated as: MnSO4·H2O 1.92 g / L, CoCl2·6H2O). The mixture contained 1.20 g / L of ZnSO4·7H2O, 0.30 g / L of AlCl3·6H2O, 0.75 g / L of Na2MoO4·2H2O, 0.50 g / L of CuSO4·5H2O, 0.30 g / L of H3BO3, 12.00 g / L of NH4Cl, and 7.92 g / L of CaCl2·2H2O, 100.00 mL / L of mixed amino acids (the amino acids were designated as: L-Ile 2.50 g / L, L-Ala 2.50 g / L, L-Cys 2.50 g / L, L-Arg 2.50 g / L, L-Leu 2.50 g / L, L-Pro 3.28 g / L, and L-Asp 3.94 g / L), with the remainder being deionized water at pH 7.

00.

3. A fed-batch culture medium system for L-threonine fermentation production, characterized in that it includes inoculating activated Escherichia coli culture at an inoculation rate of 10% into the culture medium system according to any one of claims 1-2 for batch fermentation, including the following components after entering the feeding stage: glucose 800.00 g / L, MgSO4·7H2O 4.00 g / L, thiamine HCl 0.22 g / L, nicotinamide 0.01 g / L, betaine hydrochloride 1.50 g / L, choline chloride 1.00 g / L, with the remainder being deionized water.

4. The culture medium system for the feeding stage as described in claim 3, characterized in that the composition of the culture medium system is optimized by a metabolic perturbation method, including the following steps: adding candidate substances to the fermentation system when the culture process reaches metabolic homeostasis (CER homeostasis), and screening candidate substances that significantly improve CER as effective components of the feeding medium.

5. The method for screening the components of the culture medium system during the feeding stage as described in claim 4, characterized in that the CER stable state refers to a CER fluctuation amplitude of less than ±5% lasting for at least 30 minutes.

6. The method for screening the components of the culture medium system during the feeding stage as described in claim 4, characterized in that a significant increase in CER refers to an increase in CER exceeding 15% of the steady state after the pulse.

7. The fed-batch fermentation method as described in claim 3, characterized in that the culture medium described in any one of claims 1-2 is fermented in a 5L fermenter. The aeration rate is 1.5-3 VVM, the stirring speed is 400-1000 rpm, the fermentation temperature is 37℃, the pH is 7.00, and the dissolved oxygen is controlled to be not lower than 30% during fermentation.

8. A smart fed-batch control method for L-threonine fermentation, comprising starting batch fermentation using any one of the culture medium systems described in claims 1-2, and feeding using the fed-batch culture medium system described in claim 3, wherein the entire fermentation process follows claim 7, characterized in that, For each fermentation process, spectral data was collected using a Raman analyzer. After processing, a smart feedback feeding control method was developed by combining a partial least squares regression (PLSR) model with PID control. This method was used to predict and control glucose concentration, L-threonine concentration, and biomass OD within the fermenter. 600 The prediction.

9. A method for controlling the feeding of L-threonine fermentation as described in any one of claims 3, 7, and 8, based on the control of glucose concentration, characterized in that... The following two strategies are used: manual pulse feeding to control glucose concentration between 5-20 g / L, 0-35 g / L, and 5-35 g / L, and online Raman intelligent feedback feeding to control glucose concentration at 10 g / L and 20 g / L.

10. A smart feeding control method for L-threonine fermentation, utilizing the optimal glucose concentration control strategy of claim 9, and feeding the culture medium system of claim 3, wherein the entire fermentation process follows the procedures of claims 7-8, characterized in that... The glucose concentration is controlled within the range of 5-35 g / L, and the glucose concentration should not be lower than 5 g / L throughout the fermentation process.

11. A culture medium system as described in any one of claims 1-3, characterized in that... A strain of Escherichia coli (Bio 59193) that produces L-threonine was cultured.

12. A culture medium system as described in any one of claims 1-3, characterized in that... Application in L-threonine fermentation production.

13. A method for screening the components of a culture medium system in the feeding stage as described in claims 4-6, characterized in that... Application in L-threonine fermentation production.

14. A method according to any one of claims 7-10, characterized in that... Application in L-threonine fermentation production.

Images