A method for separating and purifying plant exosomes

Abstract

The application discloses a plant exosome separation and purification method, S1: plant juice is incubated with pectin methylesterase and cellulase, a cross-linked network wrapping exosomes is disintegrated, exosomes are released and pectin is degraded, and supernatant is obtained by centrifugation; S2: the supernatant is incubated with a photo-thermal response type affinity magnetic bead, a magnetic field is applied to obtain a complex; S3: the complex is suspended in a neutral buffer, near-infrared laser irradiation makes the photo-thermal heating of ferroferric oxide, the poly-N-isopropyl acrylamide exceeds the minimum critical solution temperature, and the poly-N-isopropyl acrylamide is forced to swell from hydrophilic to hydrophobic, so that the binding peptide is released by elastic shooting; S4: the eluate is loaded on a size exclusion chromatography column, a flow fraction signal is input into a flow fraction intelligent identification module, and a final product is automatically collected according to a classification label. The application belongs to the technical field of plant exosome separation and purification, the structure can realize high-purity and high-activity lossless separation of plant exosomes, and completely breaks the impurity wrapping and eliminates chemical elution damage.

Description

Technical Field

[0001] This invention relates to the field of plant exosome isolation and purification technology. Background Technology

[0002] Plant exosomes have great potential for cross-border communication and drug delivery, but their isolation and purification face significant technical challenges.

[0003] First, plant tissue sap contains a large amount of pectin and cellulose cross-linking networks and polysaccharides. These impurities are very easy to physically encapsulate and co-precipitate with exosomes, resulting in extremely low purity of products obtained by existing centrifugation or precipitation methods. Secondly, existing affinity purification techniques rely on chemical eluents such as low pH or high salt. The harsh chemical environment can cause perforation of the exosome lipid bilayer, denaturation of membrane proteins, and degradation of internal RNA, severely damaging its biological activity. In addition, the physical properties of different batches of plant raw materials vary greatly, and the collection of fractions during size exclusion chromatography purification is highly dependent on human experience, resulting in poor process stability and large yield fluctuations. Summary of the Invention

[0004] The purpose of this invention is to provide a method for separating and purifying plant exosomes, which can achieve high-purity and high-activity non-destructive separation of plant exosomes, completely break the impurity encapsulation and eliminate chemical elution damage, and has a simple structure and is easy to use, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for isolating and purifying plant exosomes includes the following steps: S1: Plant tissue juice is injected into a decoupling reaction vessel, and pectin methyl esterase and cellulase are added for incubation. Pectin methyl esterase removes the methoxy group on the pectin molecular backbone, and cellulase cuts the cross-linked cellulose microfilaments, causing the pectin and cellulose cross-linking network encapsulating exosomes to disintegrate, releasing exosomes and degrading large molecule pectin into small molecule oligosaccharides. After incubation, the large tissue fragments and starch granules are removed by low-speed centrifugation to obtain a supernatant containing free exosomes. S2: The supernatant is mixed and incubated with photothermal responsive magnetic beads, which include a magnetite core, a mesoporous silica shell covering the core surface, a plant phosphatidylcholine-specific binding peptide grafted onto the shell surface, and a thermosensitive polymer poly(N-isopropylacrylamide). The plant phosphatidylcholine-specific binding peptide specifically binds to phosphatidylcholine lipid molecules on the exosome membrane surface to form a magnetic bead exosome complex. An external magnetic field is applied to separate the complex from non-specifically adsorbed impurities, thereby obtaining a purified magnetic bead exosome complex. S3: The purified magnetic bead exosome complex was suspended in a neutral buffer solution. The suspension was irradiated with a near-infrared laser. The iron oxide core converted the near-infrared light energy into heat energy. The surface temperature of the magnetic beads was characterized by fluorescence nanothermometry to ensure that it was stable above the minimum critical dissolution temperature of poly(N-isopropylacrylamide). The overall system temperature was controlled to not exceed 40°C and there was no obvious heat accumulation. The exosomes remained intact. The poly(N-isopropylacrylamide) underwent a conformational change from hydrophilic swelling to hydrophobic contraction, which forced the binding peptide to separate from the exosome membrane surface and ejected the exosomes. The magnetic field was removed to separate the magnetic beads and obtain an eluent containing plant exosomes. S4: The eluent is loaded onto a Sepharose CL-2B size exclusion column for chromatography separation. The column volume is 10 mL, the flow rate is 0.5 mL / min, and the flow signal is collected in real time by an online detector and input into the flow intelligent identification module. The flow intelligent identification module outputs the classification label of the current flow. The system automatically controls the fraction collector to collect the high-purity exosome flow according to the classification label to obtain the plant exosome final product.

[0006] By adopting the above technical solution, high-purity, non-destructive, and intelligent separation and purification of plant exosomes from complex plant matrices is achieved. The impurity encapsulation is broken by targeted decoupling, and chemical damage is avoided by physical ejection elution. Intelligent identification is used for precise collection, which completely solves the problems of low purity, poor activity, and unstable yield.

[0007] As a further aspect of the present invention: In step S1, the initial viscosity, initial polysaccharide concentration, and environmental pH of the plant sap are input as feature data into the enzymatic hydrolysis parameter optimization module. The enzymatic hydrolysis parameter optimization module outputs the optimal addition concentrations of pectin methyl esterase and cellulase, as well as the optimal incubation time. The enzymatic hydrolysis parameter optimization module is constructed based on a random forest regression model, using the initial viscosity, initial polysaccharide concentration, and environmental pH of the plant sap sample as feature data, and the actually determined optimal pectin methyl esterase concentration, optimal cellulase concentration, and optimal incubation time as label data. The model weights are obtained through training with historical data, and the coefficient of determination R of the trained model is... 2 >0.95.

[0008] By adopting the above technical solution, dynamic intelligent optimization and quantitative prediction of the enzymatic hydrolysis pretreatment process were realized, overcoming the fluctuation of decoupling effect caused by the difference in physical properties of different batches of plant raw materials, ensuring that the pretreatment for each extraction is in the best state, and improving the overall stability of the process and the data-driven decision-making ability.

[0009] As a further aspect of the present invention: the number of decision trees in the random forest regression model is one hundred, the tree depth is eight layers, and the mean squared error loss function is used; the incubation temperature in step S1 is 25°C to 30°C.

[0010] By adopting the above technical solutions, the architecture of the regression prediction model and the optimal enzyme activity reaction temperature range were optimized. While controlling the model complexity to prevent overfitting and ensuring high accuracy in optimization, the model was able to achieve efficient enzyme catalysis and avoid thermal damage to exosomes caused by high temperatures.

[0011] As a further aspect of the present invention: in step S2, the amino acid sequence of the plant phosphatidylcholine-specific binding peptide is serine, leucine, lysine, leucine, proline, and serine; the minimum critical dissolution temperature of the poly(N-isopropylacrylamide) is a preset critical temperature, which is hydrophilic swelling when below the temperature and hydrophobic shrinkage when above the temperature.

[0012] By adopting the above technical solution, the precise and specific affinity capture of exosomes is achieved through the combination of specific amino acid sequences and thermosensitive phase transition characteristics, providing a solid structural basis for subsequent controllable and non-destructive physical release, and ensuring the unity of capture specificity and elution controllability.

[0013] As a further aspect of the present invention: In step S2, during the incubation process of mixing the supernatant with the photothermal responsive affinity magnetic beads, a low-intensity rotating magnetic field is applied to drive the magnetic beads to tumble in the suspension to increase the probability of collision with exosomes and prevent the magnetic beads from self-aggregating.

[0014] By adopting the above technical solution and introducing a low-intensity rotating magnetic field, the diffusion limitation of static incubation is broken, the dynamic collision and binding probability of magnetic beads and exosomes is significantly improved, and the surface area loss caused by the self-aggregation of magnetic beads is effectively overcome, thus greatly improving the capture efficiency and product yield.

[0015] As a further aspect of the present invention: in step S3, a pulsed near-infrared laser is used to irradiate the suspension, causing poly(N-isopropylacrylamide) to undergo periodic hydrophilic swelling and hydrophobic contraction, generating an oscillating ejection force to release the exosomes.

[0016] By adopting the above technical solution, the periodic phase transition of the temperature-sensitive polymer is induced by pulsed laser, generating a mild and continuous mechanical oscillating ejection force. Compared with single contraction, it can more thoroughly peel off tightly bound exosomes, while effectively avoiding the damage to exosome activity caused by heat accumulation from continuous laser irradiation.

[0017] As a further aspect of the present invention: the wavelength of the pulsed near-infrared laser is 808 nanometers, the pulse frequency is 0.05 to 0.1 Hz, the duty cycle is 30% to 50%, and the photothermal conversion causes the local microenvironment temperature to fluctuate periodically between 37°C and the preset critical temperature.

[0018] By adopting the above technical solution, the frequency and duty cycle parameters of the pulsed laser are limited, and precise periodic fluctuations of the local microenvironment temperature near the phase transition point are achieved. This ensures the effective output of the oscillating ejection force and the balance between the thermal safety of the system, further improving the thoroughness of elution and the integrity of exosomes.

[0019] As a further aspect of the present invention: in step S4, the flow signals collected in real time by the online detector include multi-wavelength ultraviolet absorbance, dynamic light scattering signals, and the first derivative features of the signals changing over time. The classification labels output by the flow intelligent identification module include high-purity exosome flow, impurity transition flow, and waste liquid flow.

[0020] By adopting the above technical solution and introducing signal change rate characteristics, a multidimensional fraction identification basis with time-series dynamic characteristics is constructed, which can keenly capture the signal inflection point in the overlapping region of chromatographic peaks, and significantly improve the accuracy and anti-interference ability of target exosome fraction segmentation in complex chromatographic patterns.

[0021] As a further aspect of the present invention: the multi-wavelength ultraviolet absorbance includes 280 nm ultraviolet absorbance and 260 nm ultraviolet absorbance, and the dynamic light scattering signal is the average hydrodynamic particle size; the intelligent flow identification module is constructed based on a gradient boosting decision tree classification model, using multi-wavelength ultraviolet absorbance, dynamic light scattering signal and their first derivative as feature data, and flow purity identification results as label data, and obtaining model weights through training with historical data; the gradient boosting decision tree classification model adopts the cross-entropy loss function, and the classification accuracy of the trained model is >99%.

[0022] By adopting the above technical solution, a classification decision model suitable for dynamic time series and multidimensional coupling features was constructed. It effectively processed protein, nucleic acid and particle size signals with derivative features, ensured the accuracy of fraction classification label output, and realized efficient and intelligent control of the purification end.

[0023] As a further aspect of the present invention: the gradient boosting decision tree classification model has 500 decision trees, a tree depth of 5 levels, and uses L2 regularization with a regularization coefficient of 0.1.

[0024] By adopting the above technical solution, the complexity suppression mechanism of the classification model is optimized. While preventing overfitting under high-dimensional feature input, it ensures the model's ability to capture subtle derivative feature changes and its generalization stability, thus ensuring the reliability of long-term automated collection.

[0025] Compared with the prior art, the beneficial effects of the present invention are: This invention achieves high-purity and high-activity non-destructive separation of plant exosomes by targeting enzymatic hydrolysis to cleave the cross-linking network of pectin and cellulose. It also utilizes near-infrared photothermal triggering of thermosensitive polymer conformational changes for physical ejection elution, completely avoiding the damage to the membrane structure caused by chemical eluents and significantly improving the purity and integrity of the product. This invention uses a machine learning model to dynamically optimize enzymatic hydrolysis parameters to adapt to different batches of raw materials, achieving intelligent and precise control and high stability of the purification process. It also accurately identifies chromatographic fraction characteristics to automatically collect target products, overcoming the fluctuations and errors caused by human experience judgment.

[0026] Other features and advantages of the present invention will be disclosed in detail in the following detailed description and accompanying drawings. Attached Figure Description

[0027] Figure 1 This is a schematic flowchart of a method for isolating and purifying plant exosomes according to an embodiment of the present invention. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] In this embodiment of the invention, a method for isolating and purifying plant exosomes is described below. Figure 1 As shown, it includes four core steps: mild enzymatic decoupling pretreatment, photothermal responsive affinity magnetic bead targeted capture, near-infrared light-triggered mild and non-destructive elution, and size exclusion chromatography combined with intelligent identification for fine purification.

[0030] Specifically, this invention introduces an enzymatic hydrolysis parameter optimization module and a flow fraction intelligent identification module to achieve intelligent and precise control of the entire process.

[0031] Step S1: Mild enzymatic decoupling of network pretreatment Plant tissue sap was obtained and injected into a decoupling reaction vessel. Pectin methyl esterase and cellulase were added for incubation. Pectin methyl esterase removed the methoxy group on the pectin molecular backbone, and cellulase cleaved the cross-linked cellulose microfilaments. The two worked together to disintegrate the pectin-cellulose cross-linking network that encapsulated the exosomes, releasing the exosomes that were physically bound by the network, and degrading the large molecule pectin into small molecule oligosaccharides.

[0032] The incubation temperature was set between 25°C and 30°C. This temperature range ensures high enzyme catalytic activity while avoiding thermal damage to the exosome membrane structure. After incubation, large tissue fragments and starch granules were removed by low-speed centrifugation to obtain a supernatant containing free exosomes.

[0033] To ensure the process stability of different batches of raw materials, the initial viscosity of the plant juice, the initial polysaccharide concentration, and the environmental pH value are used as characteristic data and input into the enzymatic hydrolysis parameter optimization module. This module outputs the optimal addition concentration of pectin methyl esterase and cellulase, as well as the optimal incubation time.

[0034] The photothermal responsive affinity magnetic beads are prepared as follows: Fe3O4@mSiO2 mesoporous silica-coated Fe3O4@mSiO2 magnetic beads are prepared by co-precipitation, followed by coating their surface with a mesoporous silica shell using the Stöber process to obtain Fe3O4@mSiO2 mesoporous silica-coated Fe3O4@mSiO2 magnetic beads with a particle size of approximately 300 nm. Fe3O4@mSiO2 is a core-shell structured nanocomposite material. 3-Aminopropyltriethoxysilane is reacted with the magnetic beads to introduce amino groups onto the surface. Subsequently, the thermosensitive polymer poly(N-isopropylacrylamide) is grafted onto the amino groups via amide bonds. Finally, a bifunctional crosslinking agent is used to condense the terminal carboxyl groups of a plant phosphatidylcholine-specific binding peptide with the remaining amino groups on the surface of the magnetic beads to obtain the photothermal responsive affinity magnetic beads. Subsequently, non-specific sites on the surface of the magnetic beads were blocked with 5% bovine serum albumin (BSA) for 30-60 minutes.

[0035] Step S2: Photothermal responsive affinity magnetic bead targeted capture Photothermal responsive affinity magnetic beads were prepared. The magnetic beads consist of an iron oxide core, a mesoporous silica shell covering the core surface, a plant phosphatidylcholine-specific binding peptide grafted onto the shell surface, and a thermosensitive polymer, poly(N-isopropylacrylamide), from the inside out. The amino acid sequence of the plant phosphatidylcholine-specific binding peptide is serine, leucine, lysine, leucine, proline, and serine. This sequence specifically recognizes and binds to phosphatidylcholine, which is abundant on the surface of plant exosomes.

[0036] The minimum critical dissolution temperature of poly(N-isopropylacrylamide) is a preset critical temperature. Below this temperature, it exhibits a hydrophilic swelling state, and above this temperature, it exhibits a hydrophobic shrinkage state.

[0037] The supernatant obtained in step S1 was mixed with photothermal responsive magnetic beads and incubated. The binding peptides specifically bound to the phosphatidylcholine lipid molecules on the surface of the exosome membrane to form a magnetic bead-exosome complex.

[0038] An external magnetic field is applied to separate the complex from non-specifically adsorbed impurities, thus obtaining a purified magnetic bead exosome complex.

[0039] Step S3: Gentle and non-destructive elution triggered by near-infrared light The purified magnetic bead exosome complex was suspended in neutral phosphate buffer and the suspension was irradiated with a near-infrared laser with a wavelength of 808 nm.

[0040] The iron oxide core converts near-infrared light energy into heat energy, and the overall system temperature is stabilized at 37°C using a fluorescent nanothermometry method, which is higher than the preset critical temperature.

[0041] Poly(N-isopropylacrylamide) undergoes a conformational change from hydrophilic swelling to hydrophobic contraction, creating steric hindrance that forces the binding peptide to separate from the exosome membrane surface and ejects the exosomes.

[0042] Remove the magnetic field to separate the magnetic beads and obtain the eluent containing plant exosomes.

[0043] Specifically, this method is suitable for small-batch, high-purity preparation in the laboratory; In one feasible embodiment, when process scale-up is required, a near-infrared LED array can be used for uniform illumination instead of a single-point laser.

[0044] Step S4: Size exclusion chromatography combined with intelligent identification for fine purification The eluent obtained in step S3 was loaded onto a size exclusion chromatography column (column volume 10 mL, flow rate 0.5 mL / min) for chromatographic separation. The flow signals were acquired in real time by an online detector. The flow signals included multi-wavelength ultraviolet absorbance and dynamic light scattering signals. The multi-wavelength ultraviolet absorbance included 280 nm and 260 nm ultraviolet absorbance, and the dynamic light scattering signal was the average hydrodynamic particle size.

[0045] The above real-time signals are input into the flow segment intelligent identification module, and the module outputs the classification label of the current flow segment. The classification label includes high-purity exosome flow segment, impurity transition flow segment, and waste liquid flow segment.

[0046] The system automatically controls the fraction collector to collect high-purity exosome fractions based on classification labels, thereby obtaining plant exosome final products.

[0047] Specific explanation of the artificial intelligence model: I. Enzymatic hydrolysis parameter optimization module This module is built on a random forest regression model. It uses the initial viscosity, initial polysaccharide concentration and environmental pH of plant sap samples as feature data, and the actual determined optimal pectin methyl esterase concentration, optimal cellulase concentration and optimal incubation time as label data. The model weights are obtained by training with historical data.

[0048] The training process is as follows: Historical plant sap samples were collected to construct a dataset containing 50,000 records. This dataset was divided into a training set, a validation set, and a test set in a 7:2:1 ratio. The training set contained 35,000 records and was used for model training. The validation set contained 10,000 records and was used to adjust model parameters. The test set contained 5,000 records and was used to evaluate the final performance of the model.

[0049] The system uses 100 decision trees with a depth of eight levels to balance model accuracy and complexity. Mean squared error loss is used to measure the difference between model predictions and actual values. Early stopping is employed during training; training terminates when the mean squared error on the validation set stops decreasing after ten consecutive rounds. The coefficient of determination R0 of the trained random forest regression model on the test set is calculated. 2 >0.95, indicating high reliability.

[0050] The formula for calculating the mean squared error loss function is: Where N represents the number of samples, This represents the true label value of the i-th sample. This represents the model's prediction for the i-th sample.

[0051] The formula for the predicted output of a random forest is: Where M represents the number of decision trees, and its value is one hundred. This represents the predicted output of the m-th decision tree for the input feature vector x, where x contains three feature dimensions: initial viscosity, initial polysaccharide concentration, and environmental pH.

[0052] II. Flow Intelligent Recognition Module This module is built on a gradient boosting decision tree classification model. It uses multi-wavelength ultraviolet absorbance and dynamic light scattering signals in the size exclusion chromatography process as feature data, and fraction purity identification results as label data. The model weights are obtained by training with historical data.

[0053] The training process is as follows: Collect historical online detection signals during the tomography separation process to construct a dataset containing 100,000 records.

[0054] The dataset is divided into a training set, a validation set, and a test set in an 8:1:1 ratio. The training set contains 80,000 records and is used for model training. The validation set contains 10,000 records and is used to adjust the model's hyperparameters. The test set contains 10,000 records and is used to evaluate the model's generalization ability.

[0055] The number of decision trees is set to 500, and the tree depth is 5 layers to prevent overfitting. The cross-entropy loss function is used to measure the difference between the model's predicted classification and the true classification. An L2 regularization term is used with a regularization coefficient of 0.1 to suppress model complexity. Training is terminated when the accuracy on the validation set reaches 99% or higher and stabilizes. The classification accuracy of the gradient boosting decision tree classification model after training is >99%, and the purity of intelligent collection is ≥30% higher than that of manual collection.

[0056] The formula for calculating the cross-entropy loss function is: Where C represents the number of categories of classification labels, with a value of 3, corresponding to high-purity exosome fraction, impurity transition fraction, and waste liquid fraction, respectively; This represents the value of category c in the one-hot encoded vector of the real label; This represents the probability that the model predicts a sample belongs to class c; This represents the L2 regularization coefficient, with a value of 0.1. represents the j-th weight parameter in the model; W represents the total number of weight parameters in the model.

[0057] Model prediction probability The formula is obtained by calculating using the Softmax function: in, This represents the original score of the corresponding class c in the gradient boosting decision tree ensemble output. This represents the original score for the corresponding category j.

[0058] Example 1 This embodiment includes: isolation and purification of ginger exosomes. S1: Take fresh ginger, wash and peel it, crush it in a juicer and then filter it through a double layer of gauze.

[0059] The juice was injected into the decoupled reaction vessel, and the initial viscosity was measured to be 3.5 mPa·s, the initial polysaccharide concentration was 15 mg / mL, and the ambient pH was 6.2.

[0060] The characteristic data were input into the enzymatic hydrolysis parameter optimization module. The module output that the optimal concentration of pectin methyl esterase was 0.8 U / mL, the optimal concentration of cellulase was 1.2 U / mL, and the optimal incubation time was 45 min.

[0061] Add the enzyme according to these output parameters and incubate at 28°C.

[0062] After incubation, centrifuge at 4000 rpm for 15 min to obtain the supernatant containing free exosomes.

[0063] S2: Add BSA-blocked photothermal-responsive magnetic beads to the supernatant, with a final concentration of 10 mg / mL, and incubate at 4°C for 2 h. Plant phosphatidylcholine-specific binding peptides specifically bind to phosphatidylcholine lipid molecules on the surface of ginger exosomes to form a complex.

[0064] An external magnetic field was applied for 2 minutes, the supernatant was discarded, and the complex was washed three times with phosphate buffer at pH 7.4 to obtain the purified magnetic bead exosome complex.

[0065] In one feasible embodiment, the magnetic bead particle size is 200-400 nm; wherein, the Fe3O4@mSiO2 core-shell structure particle size is about 300 nm.

[0066] S3: The purified magnetic bead exosome complex was suspended in 5 mL of phosphate buffer at pH 7.4 and subjected to an 808 nm near-infrared laser at 2 W / cm². 2 Irradiate the suspension with the power for 4 minutes.

[0067] Photothermal conversion of the iron oxide core was achieved, and the overall system temperature was stabilized at 37°C using fluorescence nanothermometry, exceeding the minimum critical dissolution temperature of poly(N-isopropylacrylamide) at 35°C. The polymer then contracted and ejected exosomes. The magnetic field was removed to separate the magnetic beads, and the eluent was obtained.

[0068] S4: Load the eluent onto a Sepharose CL-2B size exclusion column (column volume 10 mL, flow rate 0.5 mL / min).

[0069] The online detector collects flow signals in real time. When the ultraviolet absorbance of a certain flow at a wavelength of 280nm is 0.65, the ultraviolet absorbance at a wavelength of 260nm is 0.38, and the average hydrodynamic particle size is 85nm, the set of signals is input into the flow intelligent identification module.

[0070] The module outputs a high-purity exosome fraction, which is then automatically collected by a fraction collector to obtain ginger exosomes as the final product.

[0071] According to NTA analysis, the purity of the final product was 3.5 times higher than that of the traditional ultracentrifugation method, and the particle size distribution was between 30-150 nm, which met the exosome standard. The exosome morphology integrity rate was 97%, and transmission electron microscopy (TEM) showed that the exosomes had a typical cup-shaped morphology and intact structure.

[0072] Example 2 The technical feature that distinguishes this embodiment from Embodiment 1 is that it includes: the isolation and purification of grapefruit exosomes. S1: Extract juice from fresh grapefruit pulp, and filter through double-layered gauze. The juice was injected into the decoupled reaction vessel, and the initial viscosity was measured to be 2.8 mPa·s, the initial polysaccharide concentration was 22 mg / mL, and the ambient pH was 4.5.

[0073] The feature data was input into the enzymatic hydrolysis parameter optimization module, which output that the optimal concentration of pectin methyl esterase was 1.5 U / mL, the optimal concentration of cellulase was 0.6 U / mL, and the optimal incubation time was 60 min.

[0074] Add the enzyme according to these parameters and incubate at 25°C. After incubation, centrifuge at 4000 rpm for 15 minutes and obtain the supernatant.

[0075] S2: Add BSA-blocked photothermal-responsive magnetic beads to the supernatant, with a final concentration of 12 mg / mL, and incubate at 4°C for 2.5 h. An external magnetic field is applied for separation, and the purified magnetic bead exosome complex is obtained by washing.

[0076] S3: The complex was suspended in phosphate buffer at pH 7.4 and subjected to an 808 nm near-infrared laser at 2 W / cm². 2 Irradiate with high power for 5 minutes, then locally heat to 37°C to trigger ejection release and obtain the eluent. S4: Load the eluent onto a size exclusion column (column volume 10 mL, flow rate 0.5 mL / min). When the online detector detects that the absorbance of a certain fraction at 280nm wavelength is 0.82, the absorbance at 260nm wavelength is 0.51, and the average hydrodynamic particle size is 105nm, the fraction is input into the fraction intelligent identification module. The module outputs a classification label as high-purity exosome fraction, and the system automatically collects it to obtain grapefruit exosome final products.

[0077] Testing revealed that the product integrity and purity were significantly better than those of traditional processes. Transmission electron microscopy (TEM) showed that the exosomes exhibited a typical cup-shaped morphology and intact structure.

[0078] Example 3 The difference between this embodiment and Embodiment 1 is as follows: In step S2, BSA-blocked photothermal affinity magnetic beads are added to the supernatant, with a final concentration of 10 mg / mL. Then, the incubation system is placed in a low-intensity rotating magnetic field generator, with the rotating magnetic field intensity set to 15 mT and the rotation speed set to 30 rpm. The mixture is then incubated at 4°C for 2 h.

[0079] During this process, the magnetic beads continuously tumble in the suspension under the drive of the rotating magnetic field, which greatly increases the probability of collision with ginger exosomes. At the same time, the tumbling of the magnetic beads overcomes the static electricity and magnetic self-aggregation between magnetic particles, thereby increasing the binding capacity.

[0080] After incubation, the magnetic bead exosome complex was adsorbed by switching to a static external magnetic field for 2 minutes. The complex was then washed three times with phosphate buffer at pH 7.4 to obtain purified magnetic bead exosome complex.

[0081] In step S3, the purified magnetic bead exosome complex was suspended in 5 mL of phosphate buffer at pH 7.4. The suspension was then irradiated with an 808 nm pulsed near-infrared laser at a pulse frequency of 1 Hz and a duty cycle of 40%, i.e., irradiation for 5-10 seconds followed by a 5-10 second off-light cycle, with an average power density of 2 W / cm². 2 The total irradiation time is 5 minutes.

[0082] During each laser irradiation period, the photothermal conversion of the iron oxide core causes the local temperature to rise rapidly to 37°C, which is higher than the minimum critical dissolution temperature of 35°C, causing the poly(N-isopropylacrylamide) to undergo hydrophobic shrinkage. During the laser shutdown period, heat dissipation causes the local temperature to drop back below 35°C, and the polymer re-swells in a hydrophilic manner.

[0083] This periodic contraction and relaxation generates a mechanical oscillating ejection force, which completely separates the bound peptide from the exosome membrane surface and ejects it.

[0084] The magnetic field was removed to separate the magnetic beads, and an eluent containing plant exosomes was obtained. This pulsed oscillating ejection method improved the exosome elution rate by 25% compared to continuous laser, and the suspension system showed no significant heat accumulation.

[0085] In step S4, the eluent is loaded onto a Sepharose CL-2B size exclusion column for chromatographic separation.

[0086] The online detector acquires flow signals in real time, including not only the 280nm UV absorbance, 260nm UV absorbance, and average hydrodynamic particle size at the current moment, but also calculates the first derivative characteristics of the above three parameters over time in real time through the signal processing module.

[0087] When a fraction has an absorbance of 0.65 at 280 nm, an absorbance of 0.38 at 260 nm, an average hydrodynamic particle size of 85 nm, and the first derivative of the absorbance at 280 nm changes from positive to negative (indicating the transition region at the chromatographic peak), the multidimensional feature vector is input into the fraction intelligent identification module.

[0088] The module uses a gradient boosting decision tree to determine that it is at the boundary between high-purity exosome fractions and impurity transition fractions, accurately outputting the classification label of the high-purity exosome fractions. The system automatically controls the fraction collector to accurately cut and collect the fractions, effectively avoiding the mixing of tailing impurities, and obtaining ginger exosome final products.

[0089] According to NTA and electron microscopy, the purity of the final product was further improved by 1.2 times compared with Example 1, and the morphological integrity rate of exosomes reached 99%. Transmission electron microscopy (TEM) showed that the exosomes had a typical cup-shaped morphology and intact structure.

[0090] This invention provides a method for separating and purifying plant exosomes, which can achieve high purity and high activity of plant exosomes without damage, completely breaking the impurity encapsulation and eliminating chemical elution damage.

[0091] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0092] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0093] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A method for isolating and purifying plant exosomes, characterized in that, Includes the following steps: S1: Plant tissue juice is injected into a reaction vessel, pectin methyl esterase and cellulase are added for incubation to degrade macromolecular pectin into small molecule oligosaccharides. After incubation, the large tissue fragments and starch granules are removed by low-speed centrifugation to obtain a supernatant containing free exosomes. S2: The supernatant is mixed and incubated with photothermal responsive magnetic beads, which include a magnetite core, a mesoporous silica shell covering the core surface, a plant phosphatidylcholine-specific binding peptide grafted onto the shell surface, and a thermosensitive polymer poly(N-isopropylacrylamide). The plant phosphatidylcholine-specific binding peptide specifically binds to phosphatidylcholine lipid molecules on the exosome membrane surface to form a magnetic bead exosome complex. An external magnetic field is applied to separate the complex from non-specifically adsorbed impurities, thereby obtaining a purified magnetic bead exosome complex. S3: The purified magnetic bead exosome complex is suspended in a neutral buffer solution. The suspension is irradiated with a near-infrared laser. The iron oxide core converts the near-infrared light energy into heat energy. The surface temperature of the magnetic beads is characterized by fluorescence nanothermometry to ensure that it is stable above the minimum critical dissolution temperature of poly(N-isopropylacrylamide) and the overall system temperature is controlled not to exceed 40°C. Poly(N-isopropylacrylamide) undergoes a conformational change from hydrophilic swelling to hydrophobic contraction, which forces the binding peptide to separate from the exosome membrane surface and ejects the exosomes. The magnetic field is removed to separate the magnetic beads and obtain an eluent containing plant exosomes. S4: The eluent is loaded onto a chromatographic column for chromatographic separation. The sample is collected in real time by a detector and input into the identification module. The identification module outputs the classification label of the current fraction. The system automatically controls the collector to collect the high-purity exosome fraction according to the classification label to obtain the final plant exosome product.

2. The method for isolating and purifying plant exosomes according to claim 1, characterized in that, In step S1, the initial viscosity, initial polysaccharide concentration, and environmental pH of the plant sap are input as feature data into the enzymatic hydrolysis parameter optimization module. The enzymatic hydrolysis parameter optimization module outputs the optimal addition concentrations of pectin methyl esterase and cellulase, as well as the optimal incubation time. The enzymatic hydrolysis parameter optimization module is constructed based on a random forest regression model. It uses the initial viscosity, initial polysaccharide concentration, and environmental pH of the plant sap sample as feature data, and the actually determined optimal pectin methyl esterase concentration, optimal cellulase concentration, and optimal incubation time as label data. The model weights are obtained through training with historical data.

3. The method for isolating and purifying plant exosomes according to claim 2, characterized in that, The random forest regression model has one hundred decision trees with a tree depth of eight layers, and uses the mean squared error loss function; the incubation temperature in step S1 is 25℃ to 30℃.

4. The method for isolating and purifying plant exosomes according to claim 1, characterized in that, In step S2, the amino acid sequence of the plant phosphatidylcholine-specific binding peptide is serine, leucine, lysine, leucine, proline, and serine; the minimum critical dissolution temperature of the poly(N-isopropylacrylamide) is a preset critical temperature, which is hydrophilic swelling below the temperature and hydrophobic shrinkage above the temperature.

5. The method for isolating and purifying plant exosomes according to claim 1, characterized in that, In step S2, during the incubation process of mixing the supernatant with the photothermal responsive affinity magnetic beads, a low-intensity rotating magnetic field is applied to drive the magnetic beads to tumble in the suspension to increase the probability of collision with exosomes and prevent the magnetic beads from self-aggregating.

6. The method for isolating and purifying plant exosomes according to claim 1, characterized in that, In step S3, the suspension is irradiated with a pulsed near-infrared laser to cause poly(N-isopropylacrylamide) to undergo periodic hydrophilic swelling and hydrophobic contraction, generating an oscillating ejection force to release the exosomes.

7. The method for isolating and purifying plant exosomes according to claim 6, characterized in that, The pulsed near-infrared laser has a wavelength of 808 nanometers, a pulse frequency of 0.05 to 0.1 Hz, and a duty cycle of 30% to 50%. The photothermal conversion causes the local microenvironment temperature to fluctuate periodically between 37°C and the preset critical temperature.

8. The method for isolating and purifying plant exosomes according to claim 1, characterized in that, In step S4, the flow signals collected in real time by the detector include multi-wavelength ultraviolet absorbance, dynamic light scattering signals, and the first derivative features of the signals changing over time. The classification labels output by the identification module include high-purity exosome flow, impurity transition flow, and waste liquid flow.

9. The method for isolating and purifying plant exosomes according to claim 8, characterized in that, The multi-wavelength ultraviolet absorbance includes 280 nm ultraviolet absorbance and 260 nm ultraviolet absorbance, and the dynamic light scattering signal is the average hydrodynamic particle size; The identification module is built based on a gradient boosting decision tree classification model. It uses multi-wavelength ultraviolet absorbance, dynamic light scattering signal and its first derivative as feature data, and the purity identification result of the fraction as label data. The model weights are obtained by training with historical data. The gradient boosting decision tree classification model adopts the cross-entropy loss function.

10. The method for isolating and purifying plant exosomes according to claim 9, characterized in that, The gradient boosting decision tree classification model has 500 decision trees with a depth of five levels, and uses L2 regularization with a regularization coefficient of 0.1.

Images