A method for preparing oyster fresh-taste peptide

Abstract

The present application relates to the technical field of bioactive peptides, and discloses a directional preparation method of oyster umami peptides, which comprises the following steps: (1) microfluidic high-pressure shearing; (2) superhigh-pressure assisted glycogen decomposition; (3) synergistic enzymolysis; (4) peptidomics analysis; (5) batch molecular docking of oyster peptides and T1R1 / T1R3 taste receptors; (6) synergistic ultrafiltration separation through sonochemistry; and (7) microwave-assisted freeze-drying to obtain oyster umami peptide powder. The microfluidic high-pressure shearing technology combined with the superhigh-pressure glycogen pre-decomposition technology significantly improves the utilization efficiency of raw materials: the fluid mechanical force generated by the microfluidic high-pressure shearing can gently regulate the conformation of oyster protein molecules, improve the solubility and enzymolysis accessibility thereof, and the superhigh-pressure treatment destroys the dense structure of glycogen particles, and in combination with special enzymes, the directional removal of glycogen is realized, thus effectively solving the problem of high glycogen interference specific to oysters and laying a foundation for subsequent efficient enzymolysis.

Description

Technical Field

[0001] This invention relates to the field of bioactive peptide technology, specifically to a method for the targeted preparation of oyster umami peptides. Background Technology

[0002] Oysters (Crassostrea gigas), an important marine economic shellfish worldwide, are not only rich in high-quality protein, various amino acids, and trace elements, but also highly favored by consumers for their unique sweet and savory taste. Studies have shown that oysters contain 10-15% protein, with a highly distinctive amino acid composition. Glutamic acid, a core amino acid for umami, accounts for 12-18% of the total amino acids, far exceeding the 8-12% found in most marine organisms; glycine accounts for 15-22%, and alanine for 8-12%. This high content of umami and sweet amino acids lays the material foundation for the umami characteristics of subsequent peptides. In addition to abundant umami and sweet amino acids, oysters also contain flavor nucleotides such as 5'-inosinic acid and 5'-guanylic acid, as well as organic acids such as taurine and succinic acid, which can form a synergistic umami network with amino acids, making them an ideal raw material for developing high-quality marine-derived umami peptides. With the increasing demand from consumers for natural and healthy food additives, oyster umami peptides, as a novel natural seasoning, demonstrate enormous market potential and application prospects.

[0003] However, the existing oyster umami peptide preparation technology has many bottlenecks, which seriously restrict the product quality and its industrial development. (1) Glycogen particles (up to 5-8%) in oysters aggregate to form gel-like substances during enzymatic hydrolysis, which not only reduces the contact efficiency between the enzyme and the oyster substrate, but also encapsulates the target peptides, resulting in a significant decrease in yield and seriously interfering with enzymatic hydrolysis and separation and purification. However, the current traditional glycogen removal methods such as acid treatment or high temperature treatment are prone to denaturation of oyster proteins and loss of nutrients, affecting the final functional characteristics of the product. (2) Oyster proteins have a dense structure, rich in hydrophobic amino acid residues and complex disulfide bond networks, while traditional enzymatic hydrolysis mostly uses empirical process parameters, which makes it difficult to achieve efficient and precise enzymatic hydrolysis, resulting in low enzymatic hydrolysis efficiency. (3) Existing processes mostly use blind molecular weight classification or simple sensory evaluation, which cannot accurately identify and directionally enrich peptides with strong umami activity. This extensive separation strategy results in low umami peptide content in the product, unstable umami intensity, and significant batch-to-batch differences, making it difficult to meet the strict requirements of the food industry for standardized raw materials.

[0004] In recent years, microfluidic high-pressure shearing technology has shown promising prospects in food processing fields such as protein modification, sterilization, and quality improvement. Microfluidic high-pressure shearing technology generates high-intensity shear forces through precise fluid control, inducing directional changes in protein structure and selectively modifying specific active groups, thereby effectively improving the functional properties of proteins. However, research on the application of microfluidic high-pressure shearing technology in oyster protein pretreatment is still lacking. Meanwhile, molecular docking, as an important branch of computational biology, has been widely used in drug design and bioactive substance screening. Currently, studies have applied molecular docking technology to the interaction between single umami peptides and T1R1 / T1R3 receptors. However, these studies are limited to docking analysis of a few compounds. Oysters, after enzymatic hydrolysis, typically produce 2,000-5,000 peptides with different sequences. Traditional single-molecule docking methods are not only extremely time-consuming (docking analysis of a single peptide takes 2-4 hours), but also difficult to ensure the consistency of docking parameters, resulting in low reliability of screening results. This cannot meet the high-throughput screening requirements of thousands of different flavor peptides in oyster enzymatic hydrolysis products, and has become a key bottleneck restricting the accurate screening of marine-derived umami peptides.

[0005] Therefore, there is an urgent need to develop a new process for the targeted preparation of oyster umami peptides that integrates efficient extraction and enzymatic hydrolysis technologies. This process can effectively solve the problems of raw material pretreatment and enzymatic hydrolysis efficiency, and can also achieve precise screening and efficient separation of umami peptides based on the scientific guidance of batch molecular docking, thereby obtaining high-quality oyster umami peptide products with stable quality and strong functionality. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology and provide a method for the directional preparation of oyster umami peptides based on microfluidic high-pressure shearing and computer-aided screening.

[0007] To achieve the above objectives, the technical solution of the present invention is: a method for the targeted preparation of oyster umami peptides, comprising the following steps:

[0008] (1) Microfluidic high-pressure shearing: Oyster meat is added to water and homogenized. The oyster homogenate is injected into the inlet of the microfluidic chip and high-pressure shearing is performed through the micron-contraction channel. The micron-contraction channel has a contraction size of 28 µm × 25 µm, and the inlet flow rate is 1~3 mL / min and the inlet pressure is 4~6 MPa.

[0009] (2) Ultra-high pressure assisted glycogenolysis: The pH of the oyster paste after high-pressure shearing was adjusted to 6.2~6.8, and then ultra-high pressure treatment was carried out. After treatment, pressure-resistant α-amylase and glucosylamylase were added to the oyster paste to carry out glycogenolysis.

[0010] (3) Synergistic enzymatic hydrolysis: Based on AlphaFold 2 protein structure prediction technology, a three-dimensional structural model of the main proteins of oyster was constructed. The optimal enzyme cleavage sites were analyzed using machine learning algorithms. The enzyme system ratio and process parameters were determined in advance. A three-stage synergistic enzymatic hydrolysis strategy was adopted: First, neutral protease was added to the oyster slurry after glycogen hydrolysis for crude enzymatic hydrolysis. Then, pepsin was added for enzymatic hydrolysis. Finally, aminopeptidase was added. The mass ratio of neutral protease, pepsin and aminopeptidase was 1.6:0.8:0.5. After the enzymatic hydrolysis was completed, the mixture was heated to inactivate the enzymes, cooled and centrifuged, and the supernatant was collected.

[0011] (4) Peptidomics analysis: Take an appropriate amount of enzymatic hydrolysis supernatant, desalt and concentrate it using ultrafiltration centrifuge tubes, separate and purify the concentrate using reversed-phase high performance liquid chromatography, and then perform mass spectrometry detection using a high-resolution quadrupole-orbit trap mass spectrometer with nanospray ionization source and data-dependent acquisition mode. The obtained mass spectrometry data are processed using MaxQuant software, and peptide identification is completed by searching the UniProt protein database.

[0012] (5) Batch molecular docking of oyster peptides with T1R1 / T1R3 taste receptors: A three-dimensional structural model of human T1R1 / T1R3 heterodimeric umami receptors was constructed. The conformation of peptides identified by peptidomics analysis was predicted. After prediction, batch molecular docking simulation was performed using AutoDockVina 1.2.0. Peptides with molecular weight of 450-750 Da were screened out to have the strongest average binding affinity.

[0013] (6) Sonochemical-assisted ultrafiltration separation: The enzymatic hydrolysis supernatant obtained in step (3) was subjected to ultrasonic-assisted membrane ultrafiltration separation. The ultrasonic frequency was 20-40 kHz and the acoustic power density was 0.8-1.5 W / cm³. 2 The membrane ultrafiltration separation process involves sequentially passing the membrane through a 10 kDa molecular weight cutoff ultrafiltration membrane, a 3 kDa molecular weight cutoff ultrafiltration membrane, and a 1 kDa molecular weight cutoff composite nanofiltration membrane to enrich peptides in the range of 450-750 Da.

[0014] (7) Microwave-assisted freeze drying: Add a protective agent to the enriched solution obtained in step (6), mix evenly, and then perform microwave pretreatment. The microwave power density is 1.8-2.2 W / g, the treatment time is 30-45 s, and then freeze-dry to obtain oyster umami peptide powder.

[0015] Furthermore, the temperature of the microfluidic chip in step (1) is 25±1℃.

[0016] Further; in step (2), the ultra-high pressure treatment is performed at a pressure of 500~750 MPa, a time of 2~4 min, a temperature of 20~25℃, the amount of pressure-resistant α-amylase added is 0.5~2% of the oyster meat, the amount of glucose amylase added is 0.3~1.5% of the oyster meat, the enzymatic hydrolysis temperature is 45~60℃, and the enzymatic hydrolysis time is 60 min.

[0017] Further; in step (3), add 0.6-1.0% (w / w) neutral protease to the oyster slurry, hydrolyze at 45-48℃, pH 6.8-7.2, for 1.5-2.0 h; then add 0.3-0.6% (w / w) pepsin, hydrolyze at 50-55℃, pH 5.5-6.0, for 1.0-1.5 h; finally add 0.2-0.4% (w / w) aminopeptidase, hydrolyze at 48-52℃, pH 6.5-7.0, for 0.8-1.2 h. After hydrolysis, rapidly heat to 85-90℃ and hold for 10 min to completely inactivate all enzyme activity, then cool to room temperature, centrifuge at 8000-10000 g for 20-30 min, and collect the supernatant.

[0018] Further; the ultrafiltration centrifuge tube in step (4) has a molecular weight cutoff of 3 kDa, and the chromatographic conditions of high performance liquid chromatography are: a C18 reversed-phase column with an inner diameter of 75 μm, a length of 25 cm, and a packing particle size of 1.9 μm; mobile phase A is ultrapure water containing 0.1% formic acid; mobile phase B is acetonitrile solution containing 0.1% formic acid; and the gradient elution program is set to gradient elution with a total time of 120 min. The mass spectrometry detection conditions are: MS1 scan range is set to 350-1800 m / z, resolution is set to 70000, and MS2 fragmentation adopts high-energy collision dissociation mode.

[0019] Further; in step (5), the mGluR1 glutamate receptor crystal structure PDB ID: 5X2M and the T1R2 / T1R3 sweet taste receptor ligand binding domain PDB ID: 6N51 were used as homology modeling templates. A three-dimensional structural model of the human T1R1 / T1R3 heterodimeric umami receptor was constructed using the SWISS-MODEL online platform. The initial model was optimized using the AMBER force field. The three-dimensional conformation of each peptide was predicted using the PEP-FOLD3 protein structure prediction software. Twenty candidate conformations were generated for each peptide. The most representative conformation was selected for subsequent molecular docking analysis through cluster analysis.

[0020] Furthermore; the 10 kDa molecular weight cutoff ultrafiltration membrane material mentioned in step (6) is polyethersulfone, and the operating parameters are: operating pressure 0.25 MPa, feed flow rate 120 L / h;

[0021] The 3 kDa molecular weight cutoff ultrafiltration membrane is made of polyvinylidene fluoride, and the operating parameters are: operating pressure 0.35 MPa, feed flow rate 80 L / h;

[0022] The 1 kDa molecular weight cutoff composite nanofiltration membrane has the following operating parameters: operating pressure 0.6 MPa, feed flow rate 50 L / h.

[0023] Further; the protective agent in step (7) includes trehalose and ascorbic acid, and the mass-volume ratio of trehalose, ascorbic acid and enrichment solution is (2-4) g : (0.1-0.3) g : 100 mL.

[0024] The beneficial effects of this invention are:

[0025] (1) Microfluidic high-pressure shearing technology combined with ultra-high pressure glycogen pre-decomposition technology significantly improves the utilization efficiency of raw materials: The fluid mechanical force generated by microfluidic high-pressure shearing can gently regulate the conformation of oyster protein molecules, improve their solubility and enzymatic accessibility; ultra-high pressure treatment destroys the dense structure of glycogen particles, and combined with a special enzyme system, it realizes the targeted removal of glycogen, effectively solving the problem of high glycogen interference unique to oysters, and laying the foundation for subsequent efficient enzymatic hydrolysis.

[0026] (2) Machine learning-guided enzymatic hydrolysis optimization and batch molecular docking screening achieve precise targeted preparation: Enzyme system design based on protein structure prediction improves the targeting and efficiency of enzymatic hydrolysis; a peptide database is established through comprehensive peptidomics analysis, and high-throughput virtual screening of thousands of peptides is achieved by combining batch molecular docking technology, breaking through the processing capacity limitation of traditional single molecular docking, and realizing the first precise screening of umami peptides based on receptor-ligand interaction mechanism.

[0027] (3) Acoustochemical synergistic membrane separation technology improves separation efficiency and product purity: ultrasonic cavitation effect effectively prevents membrane fouling and extends membrane lifespan, while acoustic flow effect enhances mass transfer process; a three-stage progressive separation strategy designed based on the target molecular weight range determined by molecular docking results achieves targeted enrichment of specific umami peptides, significantly improving the content and purity of target peptides in the product. Attached Figure Description

[0028] Figure 1 This is a graph showing the effect of microfluidic high-pressure shearing treatment on the solubility of oyster proteins.

[0029] Figure 2 It is the kinetic curve of ultra-high pressure assisted glycogenolysis;

[0030] Figure 3 This is a graph showing the changes in free amino nitrogen content during the three-stage synergistic enzymatic hydrolysis process;

[0031] Figure 4 This is a graph showing the changes in the molecular weight distribution of peptides before and after the three-stage synergistic enzymatic hydrolysis.

[0032] Figure 5 (A) is a molecular weight distribution diagram of peptides identified by peptidomics analysis; (B) is an amino acid composition diagram of peptides identified by peptidomics analysis.

[0033] Figure 6 (A) shows the results of the bulk molecular docking of oyster peptides with T1R1 / T1R3 taste receptors; (B) shows the results of the molecular dynamics simulation verification of the first 50 peptides.

[0034] Figure 7 This is a diagram of the product's storage stability test. Detailed Implementation

[0035] The present invention will be described below through specific embodiments. Unless otherwise specified, the methods used in the embodiments are conventional methods in the art.

[0036] Example:

[0037] A method for the targeted preparation of oyster umami peptides, comprising the following steps:

[0038] (1) Microfluidic high-pressure shearing pretreatment: 2 kg of fresh Pacific oysters were selected, and about 1.6 kg of oyster meat was obtained after treatment. Deionized water was added to adjust the solid-liquid ratio to 1:5 (w / v), and shell fragments were removed through a 200 µm filter. The solid content was adjusted to 10% (w / w), resulting in 12.5 L of homogenate. The homogenate was degassed under vacuum for 5 min to remove air bubbles. The microfluidic chip (purchased from Suzhou Wenhao Microfluidic Technology Co., Ltd.) tubing was rinsed sequentially with 70% ethanol, deionized water, and nitrogen to remove particles and air bubbles. Then, the homogenate was injected into the inlet of the microfluidic chip at a rate of 2 mL / min using a high-pressure injection pump. Microfluidic chip parameter settings: microchannel contraction zone size 28 µm × 25 µm; inlet pressure 5 MPa; temperature 25℃. A high-speed Phantom V641 microscope camera with a 20× objective lens, a frame rate of 20000 fps, and a pixel size of 0.67 µm was used to observe cell rupture and cavitation clouds in real time. The outlet temperature after sampling is below 5℃ to prevent thermal denaturation. Protein solubility assays showed that, compared to the traditional enzymatic pretreatment method (control group), the solubility of oyster protein increased from 62.3% in the control group to 85.7%, an increase of 37.5% (see...). Figure 1Traditional enzymatic hydrolysis relies on the site-specific cleavage of peptide bonds by proteases. Its efficiency is limited by the contact between the enzyme and the substrate, the reaction time, and the final inactivation process. Furthermore, the heat inactivation step itself carries the risk of denaturation of heat-sensitive proteins, which is why its protein solubility stops at 62.3%. In contrast, microfluidic high-pressure shearing delivers oyster cells into a micrometer-sized slit. Under the combined action of megapascal pressure and mega-shear rate, intracellular protein particles are sheared into submicrometer-sized fragments. The accompanying cavitation impact and stretching flow field within the microfluidic channel disintegrate protein aggregates that would otherwise form due to metal ion or hydrophobic interactions, preventing re-aggregation at low temperatures. At the same time, the entire process is completed at a constant temperature of 25°C, eliminating the risk of thermal denaturation and precipitation.

[0039] The traditional enzymatic pretreatment method is as follows: 12.5 L of oyster homogenate was prepared according to the above method and placed in a constant temperature water bath at 50°C and pH 7.0 to simulate the optimal environment for neutral protease. Then, 1.5% (by weight) of neutral protease was added to the homogenate, and the enzymatic hydrolysis reaction was carried out for 3 hours with stirring. After the reaction, the system was rapidly heated to 90°C and maintained for 15 minutes to ensure complete inactivation of the protease, thus terminating the reaction. Finally, the sample was cooled to room temperature and centrifuged; the supernatant was the enzymatically hydrolyzed protein solution, and its protein solubility was determined to be 62.3% (based on standard).

[0040] (2) Ultra-high pressure assisted glycogenolysis: The pH of the oyster slurry after high-pressure shearing was adjusted to 6.5 using a 0.5M sodium hydroxide solution, and then transferred in batches to an ultra-high pressure treatment device (Avure HPP600, Avure Corporation, USA) for ultra-high pressure treatment. Each batch had a processing capacity of 12.5 L, and a total of 5 batches were processed. Ultra-high pressure treatment parameters: pressure set at 500 MPa, pressurization rate at 300 MPa / min, holding time at 3 min, depressurization rate at 200 MPa / min, and treatment temperature controlled at 22±1℃ by a circulating cooling system.

[0041] A total of 62.5 L of oyster slurry was collected from five batches after high-pressure treatment. 20 g of pressure-resistant α-amylase (Novozymes Termalyl SC, 120 KNU / g) and 12 g of glucoamylase (Novozymes AMG 300L, 300 AGU / g) were added to the mixture. The enzyme-containing mixture was transferred to a 50℃ constant-temperature water bath shaker for glycogen hydrolysis at 120 rpm for 60 min. Every 15 min during hydrolysis, 5 mL samples were accurately taken using a pipette, and the change in glycogen content was detected using the iodine-potassium iodide colorimetric method. The kinetic analysis of glycogen degradation showed that: in the initial stage of the reaction (0 min), the glycogen content was 6.8±0.2%; during the rapid degradation phase (0-30 min), the glycogen content decreased sharply, dropping to 4.2±0.1% at 15 min and 2.1±0.1% at 30 min; during the slow degradation phase (30-60 min), the degradation rate gradually slowed down, with the glycogen content at 0.9±0.1% at 45 min and stabilizing at 0.7±0.1% at 60 min (see...). Figure 2 The glycogen removal rate reached 89.7%, far exceeding the 52% of traditional glycogen decomposition methods, meeting the requirements of subsequent processes;

[0042] The traditional glycogenolysis method differs from the ultra-high pressure assisted glycogenolysis method mentioned above in that it does not include ultra-high pressure treatment. After the pH of the oyster slurry is adjusted to 6.5, pressure-resistant α-amylase and glucosylamylase are directly added for enzymatic hydrolysis, and the hydrolysis time is 2 hours.

[0043] (3) Synergistic enzymatic hydrolysis: Based on AlphaFold 2 protein structure prediction technology, a three-dimensional structural model of the main proteins of oysters was constructed. The optimal enzyme cleavage sites were analyzed using machine learning algorithms, and the enzyme system ratio and process parameters were determined in advance. A three-stage synergistic enzymatic hydrolysis strategy was adopted: First, neutral protease was added to the oyster slurry after glycogen hydrolysis for crude enzymatic hydrolysis, then pepsin was added for enzymatic hydrolysis, and finally aminopeptidase was added. After the enzymatic hydrolysis was completed, the oysters were heated to inactivate the enzymes, cooled and centrifuged, and the supernatant was collected.

[0044] High-confidence three-dimensional structural information of major oyster proteins was obtained from the AlphaFold 2 protein structure database, including myosin heavy chain (UniProt ID: A0A219G5J8), actin (UniProt ID: A0A219FXK3), and collagen α chain (UniProt ID: A0A219G2E8).

[0045] Machine learning algorithm construction and training: A restriction enzyme cleavage site prediction model was built using Python 3.8, primarily relying on libraries including scikit-learn 1.0.2, XGBoost 1.5.1, pandas 1.4.2, and NumPy 1.21.5. The algorithm architecture employed an ensemble learning strategy, combining a random forest algorithm (n_estimators=500, max_depth=15) and an XGBoost gradient boosting algorithm (learning_rate=0.1, max_depth=8) for model fusion. The training dataset was constructed by extracting 15,847 validated protease-substrate interaction records from the MEROPS protease database (Release 12.4), systematically dividing them into training, validation, and test sets in an 8:1:1 ratio, covering the cleavage specificity of neutral proteases, pepsin, and aminopeptidase. Each sample included its local sequence fragment and three-dimensional structural environment information, and a positive sample label was assigned to the peptide bond position where cleavage occurred.

[0046] During feature engineering design, multi-dimensional molecular descriptors were extracted: (1) Physicochemical properties of amino acid residues, including hydrophobicity index (Kyte-Doolittle scale), isoelectric point, molecular volume, polarity index, hydrophilicity, side chain pKa, charge, accessible surface area, reversed-phase chromatography retention time, refractive index, optical rotation, and 12 basic features of chemical descriptors; (2) Protein secondary structure information, calculated by the DSSP algorithm to determine the proportions of α-helices, β-sheets, β-turns, and random coils; (3) Local microenvironment features, calculated by the CASTp algorithm to determine spatial features such as binding pocket depth, solvent-accessible surface area, and local charge distribution; (4) Sequence conservation features, analyzed by ClustalW multiple sequence alignment to determine the evolutionary conservation of enzyme cleavage sites. To optimize feature quality, we performed Z-score normalization on 46 continuous features, and then used principal component analysis for dimensionality reduction, retaining 32 principal components with a cumulative variance contribution rate of over 95% as the final input features for model training. This series of preprocessing operations effectively improves the expressive power of features and the training efficiency of the model.

[0047] The model architecture employs an ensemble learning strategy, simultaneously constructing two base models: Random Forest and XGBoost. Hyperparameters were optimized using a grid search method, determining the key parameters for the Random Forest model as n_estimators=500 and max_depth=15, while setting the learning rate for the XGBoost model to 0.1 and the maximum tree depth to 8. During training, hierarchical ten-fold cross-validation was used to ensure the model could fully learn data features and maintain stability. Finally, a weighted average method was used to fuse the two base models into an ensemble model, with the superior-performing XGBoost model assigned a higher weight of 0.6.

[0048] During the model performance evaluation phase, the ensemble model achieved satisfactory performance metrics on the independent test set. Its prediction accuracy reached 88.9%, precision and recall reached 90.8% and 89.5% respectively, the F1 score (reflecting overall model performance) was 90.1%, and the AUC (assessing classification ability) reached 0.951. These evaluation results fully demonstrate that the constructed model not only possesses excellent prediction accuracy and generalization ability but also fully meets the practical requirements for subsequent targeted enzymatic digestion process design.

[0049] Protein structure analysis and enzyme cleavage site prediction: The three-dimensional structure of oyster proteins was analyzed in detail using PyMOL 2.5.2 software and the BioPython 1.79 library. The myosin heavy chain (molecular weight 223 kDa) showed a typical biheaded structure, with a globular head (S1 region) rich in ATP-binding sites and actin-binding domains, and a rod-shaped tail (S2 region) mainly consisting of α-helical structures. Structural analysis revealed a large number of exposed hydrophobic amino acid clusters in the head region. The machine learning model identified 68 potential neutral protease cleavage sites, mainly concentrated in dipeptide bonds such as Leu-Ala (16 sites), Phe-Gly (12 sites), Val-Ser (14 sites), and Ile-Thr (11 sites). These sites are mostly located in the loop region on the protein surface, with good enzyme accessibility.

[0050] Actin (molecular weight 42 kDa) exhibits a compact globular structure containing four main domains. Its molecular surface is relatively smooth, resulting in a relatively limited number of exposed cleavage sites. The prediction model identified 34 viable cleavage sites, primarily located in the flexible junctions between domains, including Ala-Gly (8 sites), Leu-Val (7 sites), and Phe-Ala (6 sites). These cleavage sites have relatively low accessibility and require appropriate denaturation conditions for effective cleavage.

[0051] Collagen α-chains (molecular weight 138 kDa) have a typical triple helix structure, rich in glycine (33%) and proline (12%), exhibiting a Gly-XY repetitive sequence pattern, where X and Y positions are often proline and hydroxyproline, respectively. This structural characteristic determines that collagen is highly resistant to common proteases, but exhibits specific sensitivity to pepsin. The predictive model identified 52 pepsin-specific cleavage sites, mainly located at Phe-Pro (15), Leu-Tyr (12), Ala-Phe (10), and Val-Leu (8) positions. These sites are mostly distributed in the telopeptide and cross-linked regions of the collagen triple helix structure.

[0052] Based on oyster protein composition analysis (myosin 45.2%, actin 31.8%, collagen 22.3%, other proteins 0.7%) and enzyme cleavage site density predicted by machine learning, a multi-objective optimization algorithm (NSGA-II) was used to design the enzyme system ratio. The objective function was constructed by comprehensively considering three aspects: maximizing peptide release rate, minimizing enzyme cost, and optimizing the proportion of target molecular weight peptides. Constraints were set such as the total enzyme dosage not exceeding 2% of the protein weight, and the pH and temperature adaptability range of each enzyme. After 1000 generations of evolutionary calculations, with a population size of 100, a crossover probability of 0.9, and a mutation probability of 0.1, the optimal enzyme system ratio was finally determined to be neutral protease: pepsin: aminopeptidase = 1.6:0.8:0.5. Compared with the traditional empirical ratio (1:1:1), this optimized ratio increases the theoretical peptide yield by 32.6%, increases the proportion of peptides in the target molecular weight range (450-750 Da) from 23.4% to 31.8%, and reduces enzyme cost by 18.7%.

[0053] First stage crude enzymatic hydrolysis: Based on the optimization results of the machine learning algorithm, 16g of neutral protease (Novozymes Alcalase 2.4L, enzyme activity 2.4AU / g, where 1AU is defined as the release of 1μmol of tyrosine equivalent per min at 50℃ and pH 8.0) was added to 12.5L of glycogen-depleted oyster slurry. The temperature was adjusted to 45℃, the pH was controlled at 7.0, and the enzymatic hydrolysis time was 2.0 h.

[0054] The second stage of refined enzymatic hydrolysis: Based on the optimal action conditions and substrate specificity of pepsin predicted by machine learning algorithms, 8 g of pepsin (Sigma P7000, derived from porcine gastric mucosa, with an enzyme activity ≥2500 units / mg protein, where 1 unit is defined as producing 1 μg of tyrosine equivalent per minute at 37℃ and pH 1.5) was added to the first-stage hydrolysate, resulting in an actual enzyme activity of 25,000 units. The temperature was raised to 52℃, the pH was adjusted to 5.7, and enzymatic hydrolysis continued for 1.5 h.

[0055] The third stage of modified enzymatic hydrolysis: Based on the unique enzymatic properties of aminopeptidase and optimization suggestions from machine learning algorithms, 5 g of aminopeptidase (Sigma L2632, enzyme activity ≥30 units / mg protein, where 1 unit is defined as the release of 1 μmol of leucine β-naphthylamine per min at 25℃ and pH 8.0) was added to the reaction system, resulting in an actual enzyme activity of 180 units. The temperature was adjusted to 48℃, the pH was adjusted back to 7.0, and the hydrolysis time was 1 h. During this period, changes in peptide composition were monitored in real time using high-performance liquid chromatography-mass spectrometry. The results showed that the free amino nitrogen content steadily increased from 0.16 g / L to 1.56 g / L, a 9.4-fold increase compared to the initial value, indicating that the three-stage synergistic hydrolysis strategy effectively achieved deep hydrolysis of oyster protein. The changes in free amino nitrogen content during the three-stage synergistic hydrolysis process are shown in [reference needed]. Figure 3 .

[0056] After enzymatic hydrolysis, the temperature was rapidly raised to 90 °C and held for 10 min to completely inactivate all enzyme activity. The mixture was then cooled to room temperature. Product separation was performed using high-speed centrifugation in a Beckman Avanti J-26S XP centrifuge at 9000 g and 4 °C for 25 min. Centrifugation parameters were optimized to ensure effective separation of insoluble residues while avoiding damage to peptide molecules from excessive centrifugal force. 11.2 L of clear supernatant was obtained after centrifugation. The supernatant was pale yellow, with a transparency >95%, and no visible suspended matter. The molecular weight distribution of peptides in the supernatant was obtained by gel chromatography (see figure). Figure 4 The solids recovery rate reached 89.6% by dry weight method, and the protein recovery rate reached 91.3%, providing a high-quality, high-purity sample basis for subsequent peptidomics analysis and molecular docking screening. Molecular weight distribution of peptides before enzymatic hydrolysis: The glycogen-dehydrated oyster slurry was centrifuged at 9000g and 4℃ for 25 min before enzymatic hydrolysis. The supernatant was then subjected to gel chromatography to obtain the molecular weight distribution of peptides. (See figure). Figure 4 .

[0057] (4) Peptidomics analysis:

[0058] Comprehensive peptidomics analysis of the enzymatic hydrolysis supernatant was performed using high-resolution liquid chromatography-tandem mass spectrometry. First, the total peptide concentration in the enzymatic hydrolysis supernatant was determined. Then, a sample solution containing 10 μg of total peptides was precisely transferred for desalting and concentration.

[0059] Desalting and concentration were performed using ultrafiltration centrifuge tubes with a molecular weight cutoff of 3 kDa: the ultrafiltration tubes were first pretreated by centrifugation with pure water, then the sample solution was added, and the tubes were concentrated by centrifugation at 4°C and 14,000 times the acceleration due to gravity; then, ultrapure water containing 0.1% formic acid (v / v) was added for washing and desalting, and this step was repeated twice to fully remove salt; finally, the ultrafiltration tubes were reverse-centrifuged to recover 18 μL of concentrated peptide solution.

[0060] The concentrated peptide solution was separated by chromatography using a C18 reversed-phase column with an inner diameter of 75 μm, a length of 25 cm, and a packing particle size of 1.9 μm. Mobile phase A was ultrapure water containing 0.1% formic acid, and mobile phase B was acetonitrile solution containing 0.1% formic acid. The peptides were effectively separated by a gradient elution program with a total duration of 120 min.

[0061] Mass spectrometry was performed on a high-resolution quadrupole-orbit trap mass spectrometer equipped with a nanospray ionization source, using a data-dependent acquisition mode. The first-stage mass spectrometry (MS1) scan range was set to 350–1800 m / z with a resolution of 70,000 m / z; the second-stage fragmentation (MS2) selected the most intense precursor ions and generated fragment ions using a high-energy collisional dissociation mode.

[0062] Finally, the obtained raw mass spectrometry data were processed using MaxQuant software, and peptide identification was performed by searching the UniProt protein database, with the false discovery rate strictly controlled below 1%. Comprehensive identification of peptide sequences, molecular weight distribution, and relative abundance information in the enzymatic digests was conducted.

[0063] Peptidomics analysis revealed 3,247 different peptide sequences, with molecular weights ranging from 286 to 2,845 Da. Peptides in the 450-750 Da range accounted for 34.6% of the total, those in the 750-1500 Da range accounted for 42.8%, and those above 1500 Da accounted for 22.6%. (See attached image) Figure 5 (A) Of the identified peptides, 25.0% contained glutamic acid, 28.0% contained glycine, and 22.0% contained alanine. (See [reference needed]). Figure 5 (B)

[0064] (5) Batch molecular docking of oyster peptides with T1R1 / T1R3 taste receptors: Using the SWISS-MODEL online platform, a human T1R1 / T1R3 heterodimeric receptor model was constructed using the mGluR1 receptor (PDB ID: 5X2M) as the primary template and the T1R2 / T1R3 receptor (PDB ID: 6N51) as the secondary template. Model quality assessment: QMEAN score -2.15, Ramachandran plot showed that 92.3% of the residues were located in the allowed region, indicating high model reliability. The AMBER 20 software package was used to optimize the receptor model using the ff14SB force field. Simulation conditions: TIP3P water model, 150mM NaCl, NPT ensemble, temperature 300K, pressure 1atm, simulation time 10ns. The results showed that the receptor structure reached equilibrium after 5ns, and the RMSD value stabilized at 2.1±0.3Å.

[0065] Based on 3,247 peptide sequences identified by peptidomics, three-dimensional conformation prediction was performed using the PEP-FOLD3 server. Twenty candidate conformations were generated for each peptide, and the most representative conformations were selected through RMSD clustering analysis (cluster radius 2.0 Å). A total of 3,247 high-quality peptide three-dimensional structures were obtained. A high-throughput batch molecular docking workflow based on AutoDock Vina 1.2.0 was established, and automated processing was achieved using Python scripts. Docking parameters: the search space center coordinates were determined based on the known glutamate binding sites of the T1R1 receptor (x=45.2, y=12.8, z=38.7), the search range was 22 Å × 22 Å × 22 Å, exhaustiveness=16, and the number of output patterns was 9.

[0066] Batch docking analysis took 72 hours (using a 24-core CPU for parallel computation), and the results are statistically presented (see [link]). Figure 6 (A)

[0067] Peptides with binding energy <-6.5 kcal / mol: 847 types (accounting for 26.1%).

[0068] Peptides with binding energy <-7.0 kcal / mol: 312 types (9.6%).

[0069] Peptides with binding energy <-7.5 kcal / mol: 89 types (2.7%).

[0070] Statistical analysis of high-throughput docking results and molecular mechanism analysis of stable complex interactions revealed that the molecular weights of highly active peptides (binding energy <-7.0 kcal / mol) were mainly concentrated in the 450-750 Da range, accounting for 76.3% of this group. These peptides mostly contain acidic amino acids such as glutamic acid and aspartic acid, as well as hydrophobic amino acids such as phenylalanine and leucine. 20 ns molecular dynamics (MD) validation was performed on the top 50 highly active peptides, and 42 peptides (84%) maintained stable binding at the receptor binding site. Figure 6 In the middle (B), the binding free energy is between -25 and -45 kJ / mol.

[0071] (6) Acoustochemical synergistic ultrafiltration separation: The enzymatic hydrolysis supernatant obtained in step (3) is subjected to ultrasonic synergistic membrane ultrafiltration separation through an ultrasonic membrane separation device (CLMB-6-1, Shaoxing Haina Membrane Technology Co., Ltd., with built-in three-stage membrane separation). The ultrasonic frequency is 30kHz and the acoustic power density is 1.2W / cm².

[0072] First-stage membrane separation: A polyethersulfone ultrafiltration membrane (Millipore Pellicon XL) with a molecular weight cutoff of 10 kDa was used to process 11.2 L of feed. The membrane area was 0.5 m², the operating pressure was 0.25 MPa, and the feed flow rate was 120 L / h. After continuous operation for 4 h with ultrasonic assistance, the permeate volume was 9.8 L, the concentrate volume was 1.4 L, and the retention rate reached 87.1%.

[0073] Second-stage membrane separation: A 3 kDa polyvinylidene fluoride ultrafiltration membrane (Sartorius Hydrosart) was used, with an operating pressure of 0.35 MPa and a feed flow rate of 80 L / h. After 5 hours of operation, 8.2 L of permeate and 1.6 L of concentrate were obtained. The content of the 3-10 kDa component in the permeate decreased from 26.4% to 4.1%.

[0074] Third-stage membrane separation: A 1 kDa composite nanofiltration membrane (DOW Filmtec NF270) was used at an operating pressure of 0.6 MPa and a feed flow rate of 50 L / h. After 8 hours of separation, 5.8 L of permeate and 2.4 L of concentrate were obtained. The purity of peptides in the target molecular weight range (450-750 Da) increased from 34.6% before membrane separation to 78.9%, with an enrichment factor of 2.28 times.

[0075] Monitoring of indicators during membrane separation showed that: the flux of the first-stage membrane stabilized at 38 L / (m²·h) from the initial 45 L / (m²·h), with a decay rate of only 15.6%; the flux of the second-stage membrane stabilized at 34 L / (m²·h) from 42 L / (m²·h), with a decay rate of 19.0%; and the flux of the third-stage membrane stabilized at 21 L / (m²·h) from 28 L / (m²·h), with a decay rate of 25.0%.

[0076] (7) Microwave-assisted freeze drying: 72 g of trehalose and 4.8 g of ascorbic acid were added to the 2.4 L enrichment solution obtained in step (6). After thorough mixing, microwave pretreatment was performed using a microwave device (Galanz G80F23CN3XV-Q6). The treatment was carried out in 5 batches with a microwave power density of 2.0 W / g and a treatment time of 35 s. No obvious peptide degradation was observed after treatment. Then, freeze-drying was performed using a Christ Alpha 2-4 LSCplus apparatus with the following parameters: Pre-freezing stage: the enriched sample was pre-frozen at -45℃ for 7 h to ensure complete freezing; Main drying stage: vacuum degree 10 Pa, cold trap temperature -55℃, sublimation temperature program: -40℃ (6 h) → -25℃ (4 h) → -10℃ (4 h) → 20℃ (8 h); Desorption drying stage: maintained at 20℃ for 2 h to ensure complete removal of residual moisture; Monitoring data during drying showed that the sample temperature reached -40℃ at 8 h, -25℃ at 14 h, -10℃ at 18 h, and 20℃ at 22 h. The total moisture collected in the cold trap was 2.28 L, close to the theoretical value, and 218 g of oyster umami peptide powder was finally obtained.

[0077] The oyster umami peptide powder is a pale yellow powder with a rich seafood flavor. Product quality indicators show the following: moisture content 3.8%, water activity 0.21, 450-750 Da target peptide content 78.9%, and total amino acid content 68.2%, including glutamic acid 15.3%, glycine 12.7%, and alanine 8.9%. The umami intensity was assessed using an electronic tongue test (Alpha MOSASTREE). Using monosodium glutamate (MSG) as a control, the umami intensity of the product was equivalent to 12.3 mg / mL MSG solution, significantly higher than that of oyster peptide products prepared by traditional processes (equivalent to 7.8 mg / mL MSG). [Qin, Y., Li, R., Liao, Q., Shi, G., Zhou, Y., Wan, W., Li, J., Ma, H., Zhang, Y., & Yu, Z. (2023). Comparison of biochemical composition, nutritional quality, and metals concentrations between males and females of three different Crassostrea sp.] Food Chemistry , 398 , 133868.https: / / doi.org / 10.1016 / j.foodchem.2022.133868). Accelerated stability testing (40℃, 75%RH, 6 months) showed that the product retained 94.6% of its umami intensity and maintained 76.1% of the 450-750 Da target peptide content, indicating good storage stability (see 133868.https: / / doi.org / 10.1016 / j.foodchem.2022.133868). Figure 7 ).

[0078] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Other variations and modifications are possible without departing from the technical solutions described in the claims.

Claims

1. A method for the targeted preparation of oyster umami peptides, characterized in that, The steps are as follows: (1) Microfluidic high-pressure shearing: Oyster meat is added to water and homogenized. The oyster homogenate is injected into the inlet of the microfluidic chip and high-pressure shearing is performed through the micron-contraction channel. The micron-contraction channel has a contraction size of 28 µm × 25 µm. The inlet flow rate is 1~3 mL / min and the inlet pressure is 4~6 MPa. The temperature of the microfluidic chip is 25±1℃. (2) Ultra-high pressure assisted glycogenolysis: The pH of the oyster paste after high-pressure shearing was adjusted to 6.2~6.8, and then ultra-high pressure treatment was carried out. After treatment, pressure-resistant α-amylase and glucosylamylase were added to the oyster paste to carry out glycogenolysis. (3) Synergistic enzymatic hydrolysis: Based on AlphaFold 2 protein structure prediction technology, a three-dimensional structural model of the main oyster proteins was constructed. Machine learning algorithms were used to analyze the optimal restriction enzyme sites, and the enzyme system ratio and process parameters were pre-determined. A three-stage synergistic enzymatic hydrolysis strategy was adopted: First, 0.6-1.0% (w / w) neutral protease was added to the oyster slurry after glycogen hydrolysis for crude enzymatic hydrolysis. The hydrolysis temperature was 45-48℃, the pH was 6.8-7.2, and the hydrolysis time was 1.5-2.0 h. Then, 0.3-0.6% (w / w) pepsin was added for enzymatic hydrolysis. The hydrolysis temperature was raised to 50-55℃, the pH was adjusted to 5.5-6.0, and the hydrolysis time was 1.0-1.5 h. Finally, 0.2-0.4% (w / w) aminopeptidase was added. The hydrolysis temperature was adjusted to 48-52℃, the pH was adjusted back to 6.5-7.0, and the hydrolysis time was 0.8-1.2 h. h, the mass ratio of neutral protease, pepsin and aminopeptidase is 1.6:0.8:0.

5. After enzymatic hydrolysis, the temperature is rapidly raised to 85-90℃ and held for 10 min to completely inactivate all enzyme activities. Then, it is cooled to room temperature and centrifuged at 8000-10000g for 20-30 min. The supernatant is collected. (4) Peptidomics analysis: Take an appropriate amount of enzymatic hydrolysis supernatant, desalt and concentrate it using ultrafiltration centrifuge tubes, separate and purify the concentrate using reversed-phase high performance liquid chromatography, and then perform mass spectrometry detection using a high-resolution quadrupole-orbit trap mass spectrometer with nanospray ionization source and data-dependent acquisition mode. The obtained mass spectrometry data are processed using MaxQuant software, and peptide identification is completed by searching the UniProt protein database. (5) Batch molecular docking of oyster peptides with T1R1 / T1R3 taste receptors: Using the crystal structure PDBID of mGluR1 glutamate receptor: 5X2M and the ligand binding domain PDBID of T1R2 / T1R3 sweet taste receptor: 6N51 as homology modeling templates, a three-dimensional structural model of human T1R1 / T1R3 heterodimeric umami receptors was constructed using the SWISS-MODEL online platform. The initial model was optimized using the AMBER force field. The three-dimensional conformation prediction of peptides identified by peptidomics analysis was performed using the PEP-FOLD3 protein structure prediction software. Twenty candidate conformations were generated for each peptide. The most representative conformation was selected by cluster analysis for subsequent molecular docking analysis. After prediction, batch molecular docking simulation was performed using AutoDock Vina 1.2.

0. Peptides with molecular weight of 450-750 Da were screened to have the strongest average binding affinity. (6) Sonochemical-assisted ultrafiltration separation: The enzymatic hydrolysis supernatant obtained in step (3) was subjected to ultrasonic-assisted membrane ultrafiltration separation. The ultrasonic frequency was 20-40 kHz and the acoustic power density was 0.8-1.5 W / cm³. 2 The membrane ultrafiltration separation process involves sequentially passing the membrane through a 10 kDa molecular weight cutoff ultrafiltration membrane, a 3 kDa molecular weight cutoff ultrafiltration membrane, and a 1 kDa molecular weight cutoff composite nanofiltration membrane to enrich peptides in the range of 450-750 Da. (7) Microwave-assisted freeze drying: Add a protective agent to the enriched solution obtained in step (6), mix evenly, and then perform microwave pretreatment. The microwave power density is 1.8-2.2 W / g, the treatment time is 30-45 s, and then freeze-dry to obtain oyster umami peptide powder.

2. The method for targeted preparation of oyster umami peptides according to claim 1, characterized in that: In step (2), the ultra-high pressure treatment is performed at a pressure of 500-750 MPa, a time of 2-4 min, and a temperature of 20-25℃. The amount of pressure-resistant α-amylase added is 0.5-2% of the oyster meat, the amount of glucose amylase added is 0.3-1.5% of the oyster meat, the enzymatic hydrolysis temperature is 45-60℃, and the enzymatic hydrolysis time is 60 min.

3. The method for targeted preparation of oyster umami peptides according to claim 1, characterized in that: The ultrafiltration centrifuge tubes mentioned in step (4) have a molecular weight cutoff of 3 kDa. The chromatographic conditions for high performance liquid chromatography are as follows: a C18 reversed-phase column with an inner diameter of 75 μm, a length of 25 cm, and a packing particle size of 1.9 μm; mobile phase A is ultrapure water containing 0.1% formic acid; mobile phase B is acetonitrile solution containing 0.1% formic acid; and the gradient elution program is set to a total time of 120 min. The mass spectrometry detection conditions are as follows: the MS1 scan range is set to 350-1800 m / z, the resolution is set to 70000, and MS2 fragmentation adopts high-energy collision dissociation mode.

4. The method for targeted preparation of oyster umami peptides according to claim 1, characterized in that: The 10 kDa molecular weight cutoff ultrafiltration membrane material mentioned in step (6) is polyethersulfone, and the operating parameters are: operating pressure 0.25 MPa, feed flow rate 120 L / h; The 3 kDa molecular weight cutoff ultrafiltration membrane is made of polyvinylidene fluoride, and the operating parameters are: operating pressure 0.35 MPa, feed flow rate 80 L / h; The 1 kDa molecular weight cutoff composite nanofiltration membrane has the following operating parameters: operating pressure 0.6 MPa, feed flow rate 50 L / h.

5. The method for targeted preparation of oyster umami peptides according to claim 1, characterized in that: The protective agent in step (7) includes trehalose and ascorbic acid, and the mass-volume ratio of trehalose, ascorbic acid and enrichment solution is (2-4) g: (0.1-0.3) g: 100 mL.

Images