A dspe functionalized agarose medium for isolating evs and methods of making and using the same

Abstract

The application provides a DSPE functionalized agarose medium for separating EVs and a preparation method and application thereof, and belongs to the technical field of analytical chemistry, cell biology and biochemistry. After the NHS beads are incubated with PEG2000-NH2, the functionalized agarose medium for efficiently extracting extracellular vesicles EVs is obtained, and the functionalized agarose medium has the advantages of high stability, good specificity and multiple protein types of extracellular vesicles. The application further provides a rapid, simple and high-throughput extracellular vesicle proteomics analysis method. The method can complete sample preparation in a single tube, has good compatibility with a high-throughput automatic sample preparation workstation, and is expected to develop into an automatic sample preparation process. The application provides a convenient tool for analysis of extracellular vesicle proteomics, and has good application prospects in the fields of discovery of clinical diagnosis markers, clinical drug development and disease mechanism research.

Description

Technical Field

[0001] This invention belongs to the field of cell biology and biochemistry, specifically relating to a DSPE-functionalized agarose medium for separating EVs, its preparation method, and its application. Background Technology

[0002] Extracellular vesicles (EVs) are vesicles enclosed in a lipid bilayer secreted by various cell types. They play a crucial role in intercellular communication by "hijacking" cell membrane components and cell contents to transfer various bioactive molecules, including proteins, nucleic acids, and metabolites. 1 As our understanding of extracellular vesicles deepens, we no longer view them merely as a way for cells to excrete waste, but rather as "sentinels" in the blood, providing timely, abundant, and valuable information about abnormal states in the body. In disease research, tumor-derived extracellular vesicles are closely related to tumor progression, metastasis, and immune escape. 2 Extracellular vesicles are known to exist in almost all body fluids, and the encapsulation of their contents by lipid bilayer molecules provides a highly stable microenvironment. Studies have also shown that diagnostic accuracy, sensitivity, and specificity based on extracellular vesicle contents detection are superior to traditional liquid biopsies. Extracellular vesicles are increasingly becoming a popular research subject in the screening, diagnosis, and pathogenesis of clinical diseases, and the development of rapid and efficient extracellular vesicle extraction methods has become a hot topic in life science research.

[0003] To date, various methods have been developed to achieve highly specific separation of EVs, including ultra-high speed centrifugation (UC), size exclusion chromatography (SEC), polymer precipitation, immunoprecipitation (IP), and microfluidic chips. 3 However, these methods have drawbacks such as being time-consuming, labor-intensive, having insufficient throughput, or being expensive, limiting their application in large-scale sample studies (such as clinical samples). Tao et al. introduced amphiphilic dendritic supramolecular probes (ADSPs) onto magnetic beads or nitrocellulose membranes to capture EVs from trace amounts of clinical samples for the discovery of early diagnostic biomarkers for hepatocellular carcinoma (HCC). 4 . Wan et al. 5A biotinylate probe based on diacyl lipids (DSPE) was developed. DSPE can effectively insert into the lipid bilayer of extracellular vesicles, thus enabling specific imaging and enrichment of extracellular vesicles. These studies effectively improve the sensitivity and detection throughput of extracellular vesicle analysis.

[0004] References: 1. Kalluri, R.; LeBleu, VS, The biology, function, and biomedical applications of exosomes. Science 2020, 367 (6478) 2. Zhou, B.; Xu, K.; Zheng, X.; Chen, T.; Wang, J.; Song, Y.;Shao, Y.; Zheng, S., Application of exosomes as liquid biopsy in clinicaldiagnosis. Signal Transduction and Targeted Therapy 2020, 5 (1). 3. Cocozza, F.; Grisard, E.; Martin-Jaular, L.; Mathieu, M.;Thery, C., SnapShot: Extracellular Vesicles. Cell 2020, 182 (1), 262-262 e1. 4. Feng, X.; Iliuk, A.; Zhang, X.; Jia, S.; Shen, A.; Zhang, W.; Hu, L.; Anal Chem 2023, 95 (5), 2812-2821. 5. Wan, Y.; Cheng, G.; Liu, X.; Hao, SJ; Nisic, M.; Zhu, CD; Nat Biomed Eng 2017, 1 . Summary of the Invention

[0005] To address the problems existing in the prior art, this invention first synthesizes a DSPE-functionalized agarose medium for the effective extraction of extracellular vesicles from various body fluids. This invention also provides specific applications of the aforementioned DSPE-functionalized agarose medium, establishing a robust and high-throughput proteomics detection workflow based on rapid and efficient extracellular vesicle extraction, with promising prospects for application in areas such as the discovery of clinical diagnostic biomarkers, clinical drug development, and disease mechanism research.

[0006] To address the problems of the existing technology, the technical solution of the present invention is as follows: A method for preparing DSPE-functionalized agarose medium for separating EVs is as follows: 50 μL of NHS beads are washed sequentially with 1 mL of hydrochloric acid and pre-cooled PBS, then resuspended in 1 mL of PBS-Triton (pH=8) solution, and then 20 μL of 40 mM DSPE-PEG2000-NH2 reagent is added for incubation; after incubation, the supernatant is discarded by centrifugation, and excess NHS groups are blocked with ethanolamine solution; the supernatant is discarded by centrifugation, and the resulting precipitate is the DSPE-functionalized agarose medium for separating EVs.

[0007] The DSPE-functionalized agarose medium prepared by the method is shown.

[0008] A high-throughput proteomic analysis method based on DSPE-functionalized agarose medium for separating EVs includes the following steps: S1. Obtain biological samples for extracting extracellular vesicles; S2. Add DSPE-functionalized agarose medium to the biological sample used for extracting extracellular vesicles; S3. Centrifugation was used to separate and capture extracellular vesicles in DSPE-functionalized agarose medium. S4. The extracellular vesicles extracted in step S3 were lysed using EVs lysis buffer, and the extracellular vesicle proteins were purified.

[0009] S5. After pretreatment of extracellular vesicle proteins, LC-MS / MS analysis and data acquisition were performed.

[0010] In some embodiments of the present invention, the biological sample is plasma, saliva, or urine.

[0011] In some embodiments of the present invention, the EVs lysis buffer is one of RIPA, 1% SDS, 8M urea or 0.1% RapiGest SF.

[0012] In some embodiments of the present invention, the LC-MS / MS analysis and data acquisition use a data-independent acquisition mode (DIA).

[0013] Proteomic data obtained by any of the above-described high-throughput proteomic analysis methods based on DSPE-functionalized agarose media for separating EVs.

[0014] A method for discovering disease diagnostic biomarkers includes: using plasma samples from two groups, one from a healthy person and the other from a patient with a certain disease, obtaining proteomic data from the two groups using any of the methods described above, then analyzing and comparing the two groups of proteomic data using an algorithm, and using the proteins with significant differences obtained as biomarkers.

[0015] In some embodiments of the present invention, the algorithm is specifically an orthogonal partial least squares discriminant algorithm, ROC analysis, Mann-Whitney U test, random forest, and support vector machine.

[0016] In some embodiments of the present invention, the disease type is cancer, and the disease patient is a patient with precancerous lesions or a cancer patient.

[0017] A combination of biomarkers for colorectal cancer diagnosis includes proprotein convertase subtilisin 6 (PCSK6), S100-A8, anthrax toxin receptor 1 (ANTXR1), pyrophosphatase / phosphodiesterase 2 (ENPP2), GTP-binding nucleoprotein Ran (RAN), APO A4, and HAGH.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: 1) Based on the synthesis of DSPE-functionalized agarose media, this invention establishes a rapid, simple, and high-throughput method for preparing extracellular vesicle proteomics samples. This method can effectively extract and analyze extracellular vesicles from various body fluids, and has the advantages of high stability, good specificity, and the ability to obtain a wide variety of proteins from extracellular vesicles. Furthermore, this method can complete sample preparation in a single tube and has good compatibility with high-throughput automated sample preparation workstations, and is expected to develop into an automated sample preparation workflow. 2) Based on rapid and efficient extracellular vesicle extraction, this application provides a robust and high-throughput proteomics detection workflow. This workflow uses data-independent acquisition (DIA) mass spectrometry technology, which can perform full-coverage acquisition of all fragment ions, has high quantitative accuracy, and allows for retrospective proteomic quantitative analysis, making it widely applicable in large-scale clinical proteomics studies. 3) This invention provides a convenient tool for the analysis of extracellular vesicle proteome, which greatly promotes research in this field and has good application prospects in the discovery of clinical diagnostic markers, clinical drug development, and disease mechanism research. Attached Figure Description

[0019] Figure 1 A schematic diagram of an efficient EVs proteomics workflow: (A) The principle of capturing EVs using DSPE-functionalized agarose media; (B) DSPE-functionalized agarose media synthesized by one-step coupling of DSPE-PEG2000-NH2 and NHS activated ester modified agarose media; (C) A schematic diagram of an efficient EVs proteomics workflow. Figure 2 To optimize the EV proteome extraction flowchart: (A) Comparison of the number of proteins identified in EV samples extracted with different lysis buffers, (B) Comparison of the number of proteins identified in SP3 and FASP sample pretreatment methods. Figure 3 Workflow diagram for optimizing EV proteomics analysis methods: (A) Identification of protein quantity in EV samples using different liquid chromatography gradients; (B) Comparison of EV protein quantity identification and quantitative repeatability using different DIA software tools. Figure 4 Characterization of EVs separated by DSPE-functionalized agarose media: (A) Immunoblot analysis of EVs separated by DSPE-functionalized agarose media, polymer precipitation and ultracentrifugation, (B) Scanning electron microscopy of blank agarose media and the surface of agarose media after capturing EVs, (C) Silver staining of the whole proteome of plasma and EVs (separated by DSPE-functionalized agarose media). Figure 5 Performance evaluation of EVs separated by DSPE-functionalized agarose media: (A) Comparison of the number of proteins identified in samples under different EV separation methods; (B) Evaluation of the stability of protein quantification in samples under different EV separation methods; (C) Comparison of the number of EV-specific proteins identified under different EV separation methods; (D) Abundance level and identification status of high-abundance proteins in EV samples under different EV separation methods; (E) Cellular localization analysis of proteome obtained from EVs separated by DSPE-functionalized agarose media method. Figure 6 The figure shows the results of enriching EVs in different biological fluids using DSPE-functionalized agarose medium. Figure 7 Image showing the enrichment of fluorescently labeled EVs in plasma using DSPE solid-phase medium and the method disclosed in CN116990525A; Figure 8 Figure 1 shows the performance evaluation results of separating EVs using DSPE solid-phase medium and the method disclosed in CN116990525A: (A) Comparison of the number of proteins identified by the two methods; (B) Quantitative performance evaluation of the two methods; (C) Missing value evaluation of the two methods; (D) Comparison of the number of EV-specific proteins identified by the two methods; (E) Comparison of the time required for extracting extracellular vesicles by the two methods; (F) Comparison of the identification of high-abundance proteins in plasma by the two methods; (G) Comparison of the identification of EV-specific markers by the two methods. Figure 9 Figure 1 shows the results of proteomics analysis of EVs from a clinical colorectal cancer cohort: (A) Principal component analysis (PCA) of EVs proteomics in different groups during disease progression, including healthy controls (HC), adenomatous polyps (Polyp), and early colorectal cancer (CRC); (B) Supervised cluster analysis of proteins whose regulation is dysregulated during disease progression (VIP>1, p<0.05, AUC>0.8). Figure 10 Figure 1 shows the proteomics-based analysis results of EVs from a clinical colorectal cancer cohort: (A) KEGG pathway enrichment analysis of significantly altered proteins; (B) expression level distribution of seven protein markers in different groups. Statistical significance was defined using the Mann-Whitney U test (p < 0.01). ), p < 0.001 ( ), p <0.0001 ( (C) Use ROC curves and corresponding AUC values ​​to describe the diagnostic accuracy among healthy controls (HC), polyps (Polyp), and colorectal cancer (CRC) samples; (D) Validate the confusion matrix of the prediction results for the dataset, with the number of samples indicated in each box. Detailed Implementation

[0020] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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.

[0021] Example 1: Construction of a high-throughput EV proteomics analysis method based on DSPE-functionalized agarose media separation This embodiment provides a simple and efficient method for extracellular vesicle extraction and integrates it with a high-throughput proteomics sample preparation method, establishing a rapid and high-throughput extracellular vesicle proteomics analysis workflow. First, an agarose medium loaded with EVs-targeting chemical groups (DSPE) was synthesized via a one-step coupling method. Figure 1 (B) For example Figure 1 As shown in Figure A, the long-chain fatty acid structure of DSPE can insert into the lipid bilayer structure of extracellular vesicles, enabling the separation of extracellular vesicles from plasma samples within 10 minutes. Then, the DSPE solid-phase purification method was integrated with high-throughput proteomics sample preparation techniques such as single-pot, solid-phase-enhanced sample preparation (SP3) to minimize sample processing time. To establish a robust and high-throughput proteomics sample analysis workflow, data was acquired using a data-independent acquisition (DIA) mode, achieving accurate quantification of 600-800 extracellular vesicle proteins within a 40-minute analysis time. In the final simplified workflow, only 10 minutes are needed to separate extracellular vesicles, 60 minutes for protein reductive alkylation, followed by simple washing and overnight enzymatic digestion, and finally, sample detection is completed in 40 minutes. This significantly improves the efficiency of extracellular vesicle proteomics analysis. The workflow of this method is described in [link to workflow description]. Figure 1 C, the specific content is as follows: (1) Preparation of DSPE-functionalized agarose medium Prepare a 40 mM stock solution of DSPE-PEG2000-NH2 reagent (Ruixibio, R-0038) with anhydrous ethanol. Take an appropriate amount of NHS beads (Cytiva, catalog number 17037137), ensuring 50 μL of solid medium for each reaction. Wash the beads three times with 1 mM ice-cold hydrochloric acid, and wash three times with pre-chilled PBS. Resuspend the beads in 0.1% Triton PBS (pH=8) solution, add the corresponding amount of DSPE-PEG2000-NH2 reagent, and incubate at 29°C for 6 hours or at 4°C overnight. Centrifuge and discard the supernatant. Block excess NHS groups with 0.1 M ethanolamine solution and react at room temperature for 6 hours or at 4°C overnight. Centrifuge and discard the supernatant. Wash five times with PBS and resuspend the beads in 30% glycerol PBS solution. Aliquot and freeze at -20°C for later use.

[0022] (2) Extraction of EVs using DSPE-functionalized agarose medium The procedure for separating EVs using DSPE-functionalized agarose media is as follows: Taking plasma as an example, thaw 200 μL of plasma sample on ice and centrifuge at 3000 g for 10 min to remove large particles and other impurities. Add 10 μL of pre-prepared agarose media directly to the supernatant, incubate at 29 °C for 10 min, discard the supernatant and wash four times with PBS buffer to remove impurities, and store at -80 °C for further analysis.

[0023] (3) Extraction of EVs protein and preparation of LC-MS / MS analysis samples Enriched EVs samples were lysed using RIPA lysis buffer to obtain EV protein samples. The protein concentration was determined. EV protein samples containing 5 μg of protein were added to 50 μL of 8M urea / PBS and reduced with 10 mM dithiothreitol (DTT) at 37°C for 30 min, followed by alkylation with 20 mM iodoacetate (IAA) at 35°C for 30 min. Subsequently, 2 μL of pre-prepared 50 μg / μL SP3 magnetic beads and 50 μL of anhydrous ethanol were added. The sample was placed in a ThermoMixer and shaken at 1000 rpm for 5 min at 24°C to ensure sufficient binding of the protein to the SP3 magnetic beads. After binding, the EP tube was placed on a magnetic rack to effectively remove unbound impurities, and then the SP3 magnetic beads were washed three times with 80% ethanol. The SP3 magnetic beads were then resuspended in 100 μL of 30 mM NH4HCO3 and sonicated for 30 seconds to fully disperse the SP3 magnetic beads. 1 μg of trypsin (Enzyme Source Biotechnology (Beijing) Co., Ltd.) was added to the mixture, and digestion was carried out overnight at 37°C. After digestion, peptides were collected by centrifugation and magnetic separation, dried by vacuum centrifugation, and redissolved in 0.1% FA / H2O for subsequent LC-MS / MS analysis.

[0024] Following the above procedure, this experiment tested different EV lysis buffers, including RIPA, 1% SDS, 8M urea, and 0.1% RapiGest SF. Based on subsequent LC-MS / MS analysis results, RIPA performed best, identifying more than 600 proteins. Figure 2 (A)

[0025] Given the critical importance of high-throughput sample preparation workflows for clinical applications, this study compared two commonly used rapid sample preparation techniques: filter-assisted sample preparation (FASP) and single-pot solid-phase enhanced sample preparation (SP3). Subsequent LC-MS / MS analysis results indicated that SP3 was the better choice. Figure 2 (B) This method can not only identify more proteins, but also complete the entire sample preparation process in a single tube.

[0026] (4) LC-MS / MS detection and proteomic analysis of EVs protein In this experiment, a data-independent acquisition (DIA) quantitative proteomics strategy was used to characterize the EV proteome. We optimized parameters such as the liquid chromatography gradient and mass spectrometry acquisition window, and established a mass spectrometry analysis method for short gradients (effective gradient of 40 min). The number of proteins identified and the accuracy of quantification were comparable to those for long gradients. Figure 3(A). Furthermore, we compared the performance of the two DIA software programs and found that DIA-NN identified more proteins and had better reproducibility than Spectronaut. Figure 3 (B)

[0027] LC-MS / MS analysis was performed using a Q Exactive™ Plus mass spectrometer (Thermo-Fisher Scientific) and an Ultimate 3000 liquid chromatograph. Peptide loading and separation were performed using reversed-phase C18 columns (trap column, 5 cm × 100 µm, particle size 3 µm; analytical column, 15 cm × 100 µm, particle size 1.9 µm). The mobile phase consisted of solvent A (containing 0.1% formic acid aqueous solution) and solvent B (80% acetonitrile-water containing 0.1% formic acid). The gradient elution procedure (effective gradient of 40 min) was as follows: 0-10 min, 4% B; 10-13 min, 4-9% B; 13-47 min, 35% B; 47-50 min, 35-45% B; 50-52 min, 45-95% B; 52-57 min, 95% B; 57-58 min, 95-5% B; 58-65 min, 5% B. The mass spectrometry scan mode was data-independent acquisition (DIA), with a full scan MS range of m / z 350-2000, a resolution of 70000, an AGC target of 3e6, and a MIT of 20 ms. We set 28 variable MS² scan windows with a resolution of 17500, and set the AGC target to 1e6 and MIT to auto MIT.

[0028] The acquired data were processed using DIA-NN 1.8.1, employing a highly advanced library-free data acquisition method. The virtual library search settings were as follows: trypsin / P, allowing a maximum of 3 missed cleavages; fixed modification as Cys reductive alkylation, variable modification as protein N-terminal acetylation, Met oxidation; peptide length range of 7-30 amino acids; primary ion charge range of 1-4; primary ion mass-to-charge ratio range of 350-2000; fragment ion mass-to-charge ratio range of 200-1800. Other search parameters were set as follows: the quantification strategy was set to "robust LC (high precision)" mode; the mass precision of MS2 and MS1 was set to 0, allowing DIA-NN to automatically determine the mass tolerance; the scan window was set to 6, corresponding to the average number of scan points per peak; and the MBR function was enabled.

[0029] Example 2: Comparison of the differences between the DSPE-functionalized agarose medium extraction method and the conventional EV extraction method The procedure for EV separation using DSPE-functionalized agarose media is described in Example 1, step "(2) Extraction of EVs using DSPE-functionalized agarose media". Two conventional EV extraction methods were selected: ultracentrifugation and polymer precipitation kit-based EV separation. Ultracentrifuged EV samples were enriched according to standard procedures. 200 μL of plasma was thawed on ice, diluted 5-fold with PBS buffer, and centrifuged at 12000 g for 30 min to remove large cell debris. The supernatant was then centrifuged at 100000 g for 70 min in a Beckman Optima™ MAX-XP ultracentrifuge. The supernatant was discarded, and the sample was resuspended in PBS and washed. It was then centrifuged again at 100000 g for 70 min to obtain the final EV sample. EV samples separated using the polymer precipitation kit were obtained using the Exosome Isolation Q3 kit (EIQ3-02001, Wayen Biotechnologies, Shanghai, China), and the procedure was performed according to the product manual.

[0030] (1) Protein immunoblotting analysis of EVs samples obtained by different separation methods EVs samples obtained using different separation methods were lysed using RIPA buffer (MACGENE, MP015), and protein concentration was measured using BCA (Pierce, Thermo Fisher Scientific). 20 μg of protein was added to loading buffer and boiled at 95°C for 5 min before immunoblotting analysis. The following antibodies were used: CD63 (Abcam, ab134045, 1:1000, rabbit antibody), HSP70 (Abcam, ab181606, 1:1000, rabbit antibody), CD9 (Abcam, ab263019, 1:1000, rabbit antibody), CD81 (Abcam, ab109201, 1:1000, rabbit antibody), TSG101 (Santa, sc-7964, 1:200, mouse antibody), GAPDH (Proteintech, HRP-60004, 1:3000), ACTB (Abcam, ab19716, 1:5000, rabbit antibody), and Syntenin (Abcam, ab133267, 1:1000, rabbit antibody).

[0031] Compare the abundance of exosome-specific biomarkers such as CD63, TSG101, HSP70 and syntenin-1 in extracellular vesicles extracted by different methods. Figure 4(A). The results of immunoblotting hybridization showed that these molecular markers were effectively enriched in samples separated by DSPE solid-phase media, with enrichment levels comparable to other separation methods such as ultracentrifugation (UC) and polymer precipitation kits.

[0032] (2) Scanning electron microscopy (SEM) observation and silver staining analysis of EVs extracted from DSPE-functionalized agarose medium The DSPE-functionalized agarose medium containing captured EVs was diluted to an appropriate concentration. The silicon wafer was then placed on the carbon conductive gel with the smooth side facing up. A small amount of sample was carefully pipetted onto the silicon wafer, and after the sample dried, EVs were characterized and imaged using scanning electron microscopy (SEM).

[0033] The surface of the solid-phase medium enriched with extracellular vesicles was imaged using scanning electron microscopy (SEM). Figure 4 (B) Representative SEM images show a uniform texture on the surface of the blank medium without plasma incubation, while on the surface of the medium after EVs capture, particles with a diameter of approximately 100 nm can be clearly observed to be uniformly distributed, and these nanoparticles are consistent with the size of extracellular vesicles.

[0034] In addition, the captured EVs and plasma were analyzed by gel silver staining. After separation of the two samples by SDS-PAGE, silver staining was performed following a procedure involving fixation, washing with 30% ethanol, washing with water, sensitization, washing with water, silver staining, washing with water, development, and termination. The results showed that EVs separated by DSPE-functionalized agarose media contained richer protein band information, and high-abundance proteins such as albumin could be removed more effectively. Figure 4 (C), which will greatly promote subsequent proteomics research.

[0035] (3) Protein recovery rate and enrichment specificity of DSPE solid-phase medium and conventional EV extraction method The protein recovery and enrichment specificity of DSPE solid-phase medium were compared with those of conventional EV extraction methods. From 1 μL of plasma sample, 0.26, 0.32, and 25 μg of EV protein were extracted using DSPE solid-phase medium, ultracentrifugation, and polymer precipitation kit, respectively. The protein recovery rate was ranked as follows: polymer precipitation kit > ultracentrifugation > DSPE solid-phase medium. 1 μg of each of the three samples was taken and analyzed by mass spectrometry using the method described in Example 1. The number of proteins identified is shown in [see Example 1]. Figure 5 A( Figure 5 In the study, 50 and 43 EVs-specific proteins were identified by ultracentrifugation and agarose enrichment methods, respectively. Figure 5The concentration of high-abundance proteins in plasma (C) was significantly higher than that in the polymer precipitation kit. Meanwhile, the number and signal intensity of high-abundance proteins in the DSPE solid-phase medium group were significantly lower than in the other two groups (C). Figure 5 (D). Therefore, the specificity ranking of the three methods for enriching EVs is: DSPE solid-phase medium > ultracentrifugation > polymer precipitation kit. Cellular localization analysis of the EV proteome enriched by DSPE solid-phase medium also showed that the identified proteins were mainly located in extracellular vesicles and extracellular regions (D). Figure 5 The presence of E in the solid-phase medium further validates the specificity of EV enrichment in the DSPE solid-phase medium. Further observation revealed that 80% of the proteins in the DSPE solid-phase medium group had a coefficient of variation (CV) of less than 20% in three repeated experiments. Figure 5 The stability of the three methods (Group B) is comparable to that of the polymer precipitation kit. Therefore, the stability ranking of the three methods is: polymer precipitation kit > DSPE solid-phase medium > ultracentrifugation. In summary, the DSPE solid-phase medium method of this application is superior to the gold standard (ultracentrifugation) in terms of purity and stability for extracellular vesicle extraction. Compared with the polymer precipitation kit, it also has similar stability and higher extraction purity.

[0036] Example 3: Stability assessment of the DSPE-functionalized agarose medium extraction method for EVs In addition to plasma samples, this experiment investigated the universality of solid-phase media in other biological fluids such as urine and saliva. The results showed ( Figure 6 The DSPE-functionalized agarose medium successfully enriched extracellular vesicles labeled with Cy5 fluorescent dye, while Cy5-labeled cetuximab could not be enriched, indicating that the medium synthesized in this application is not only suitable for various types of biological fluids, but also has low non-specific adsorption.

[0037] Example 4: Comparison of the differences between DSPE-functionalized agarose media and DSPE-PEG2000-Biotin-modified streptavidin magnetic beads for extracting EVs CN116990525A discloses an integrated sensor for the isolation and detection of extracellular vesicles (EVs) and their membrane proteins. This method combines DSPE-PEG2000-Biotin and streptavidin magnetic beads to enrich EVs in biological fluids. First, we designed experiments based on fluorescently labeled antibodies and fluorescently labeled EVs to rapidly evaluate the effectiveness and specificity of the two methods for EV extraction. The results showed that, consistent with previous experimental results, the DSPE solid-phase medium could stably and specifically extract fluorescently labeled EVs, but it had no enrichment ability for fluorescently labeled cetuximab. In a parallel comparison with the DSPE-PEG2000-Biotin probe disclosed in patent CN116990525A, the medium developed in this method showed lower binding to cetuximab but higher binding to EVs. Figure 7 Therefore, it can be concluded that this method has higher enrichment efficiency and enrichment specificity for extracellular vesicles.

[0038] Next, the protein identification and quantification capabilities of the DSPE solid-phase medium and the method disclosed in CN116990525A were compared in parallel. First, extracellular vesicles were extracted from 200 μL of plasma using both methods. Based on the previously established DIA quantitative proteomics analysis method, the extracellular vesicles extracted by both methods were characterized proteomically. Mass spectrometry results showed that, under the same plasma volume conditions, the DSPE solid-phase medium identified over 700 proteins, while the patent-disclosed method identified approximately 600 proteins. Figure 8 (A) The number of EV-specific proteins identified and the intensity of quantification are also superior to the method disclosed in the patent. Figure 8 (D, G), while DSPE solid-phase media have higher quantitative repeatability ( Figure 8 (B), while having fewer missing values ​​( Figure 8 (C). Furthermore, comparing the number and signal intensity of plasma high-abundance proteins identified from extracellular vesicles extracted by the two methods, the results showed that, compared to previously reported methods, the non-specific adsorption of plasma high-abundance proteins during DSPE solid-phase extraction was lower. Figure 8 (F). A feasibility comparison of the two methods in large-scale clinical studies shows that, firstly, the DSPE solid-phase medium significantly shortens the extraction time, thereby improving sample stability during extraction and increasing the throughput of sample analysis and detection. Figure 8 (E). In summary, the DSPE solid-phase media strategy enriched extracellular vesicles and identified more proteins in a shorter time, with lower reagent costs and a more robust quantitative method.

[0039] Example 4: Application of high-throughput EV proteomics analysis based on DSPE-functionalized agarose media separation in the discovery of biomarkers for early diagnosis of CRC Plasma samples were collected from an early CRC cohort, including patients with colorectal polyps (Polyps, n=16) or early-stage colorectal cancer (Stage I or II) (CRC, n=21) and healthy volunteers (HC, n=21). EVs were then extracted from the plasma samples using DSPE-functionalized agarose media and analyzed according to the proteomics workflow optimized in Example 1. After removing missing values, a total of 633 proteins were stably quantified. Unsupervised principal component analysis (PCA) was first used to investigate the overall differences among the three groups, revealing a significant clustering separation between early CRC patients and healthy volunteers. However, polyp patients also showed a clear clustering trend, falling between healthy and cancerous states, suggesting that polyps may be in an inflammatory state and have the potential to become cancerous over time. Figure 9 In order to further find potential biomarkers related to the diagnosis of colorectal cancer, differentially expressed proteins were screened using a combination of multivariate and univariate statistical analysis: (1) variables with a contribution greater than 1 to the orthogonal partial least squares discriminant analysis model (OPLS-DA) grouping (VIP (Variable importance inprojection) > 1) were extracted; (2) Mann-Whitney U test analysis was performed on the differentially expressed variables obtained from the initial screening to find differentially expressed proteins with significant changes between the two groups (P < 0.05); (3) In order to further evaluate the diagnostic ability of the above differentially expressed proteins, their clinical diagnostic value was ranked using receiver operating characteristic (ROC) curves, and protein features with an area under the receiver operating characteristic curve (AUC) greater than 0.85 were selected as potential biomarkers. After the above statistical analysis, 184 proteins with significant dysregulation were found, and their expression levels showed significant changes at various stages of disease progression ( Figure 9Of the proteins identified, 42 showed significant changes from healthy individuals to polyps, 145 showed significant changes during the transition from healthy individuals to colorectal cancer, and 70 showed significant changes during the transition from polyps to colorectal cancer. Among the identified proteins, some previously reported CRC-related biomarkers also showed statistical differences in our dataset, such as inflammation-related C-reactive protein (CRP), potential transforming growth factor β-binding protein 1 (LTBP1), and matrix metalloproteinase-9 (MMP9). These proteins exhibited aberrant changes during disease progression, demonstrating the reliability of the EV proteomics analysis method established in this invention for biomarker discovery. Pathway enrichment analysis showed that significantly dysregulated proteins were closely related to functions such as cell matrix adhesion, toxin transport, complement activation, and protein or RNA localization. Figure 10 (A). In summary, the above analysis reveals changes in the EV proteome profile during early malignant transformation in the gut, and these changes in protein signatures have the potential to serve as effective biomarkers for predicting early disease progression.

[0040] To enable early diagnosis of polyps or CRC, the classification performance of each protein was evaluated using a random forest algorithm and ROC analysis. Proteins with different expression levels were ranked in importance, and the top seven proteins were selected. These seven proteins were used to form a biomarker combination for disease prediction. Figure 10(B) Apolipoprotein A-IV (APOA4) has previously been identified as a candidate biomarker for the diagnosis of colorectal cancer. Proprotein convertase subtilisin 6 (PCSK6) is a secreted serine protease that cleaves various growth factors, signaling molecules, hormones, and adhesion proteins, playing a crucial role in various physiological and pathological processes. Current research primarily reveals a close association between high PCSK6 expression and atherosclerotic plaque instability, hypertension, and other cardiovascular diseases. No direct link between PCSK6 and CRC has yet been reported; therefore, PCSK6 may serve as a novel biomarker for early CRC screening. Hydroxyacylglutathione hydrolase (HAGH) is a key enzyme involved in the glutathione metabolic cycle. It hydrolyzes glutathione containing hydroxy acid residues into free glutathione, allowing it to continue participating in antioxidant and detoxification processes. Therefore, HAGH plays an important role in maintaining glutathione homeostasis, the antioxidant defense system, and normal cellular physiological functions. Current research has identified HAGHL as a novel metabolic oncogene that regulates colorectal cancer development and could serve as a diagnostic and prognostic factor for colorectal cancer. However, the role of HAGH in colorectal cancer has been rarely reported. Therefore, the HAGH identified in the above analysis may also serve as a novel biomarker for early CRC screening. Unlike the results of this invention, GTP-binding nucleoprotein Ran (RAN), pyrophosphatase / phosphodiesterase 2 (ENPP2), and anthrax toxin receptor 1 (ANTXR1) have previously been proposed as potential biomarkers for CRC prognosis and liver metastasis, but their functional roles and specific molecular mechanisms in CRC pathogenesis and metastasis remain undetermined. S100A8 protein (S100A8) is considered a universal biomarker associated with inflammation; however, our data show that its level was significantly downregulated in the polyp and CRC groups compared to healthy controls. However, we believe these differences may be due to many complex factors, such as increased tumor cell uptake in tumor tissue leading to downregulation of plasma EVs levels, or inconsistencies in biomarker alterations at different disease stages. To evaluate the diagnostic ability of the aforementioned differentially expressed proteins, receiver operating characteristic (ROC) curves were used to assess their clinical diagnostic value. We constructed a potential biomarker group from the seven differentially expressed proteins using binary logistic regression and performed ROC analysis. The results showed that the AUC value of the biomarker group was 1, and both the diagnostic sensitivity and specificity were 100%. Figure 10Next, based on the seven biomarkers discovered, a machine learning (ML) model was constructed using a support vector machine (SVM) to predict non-malignant polyps and early-stage CRC. The entire dataset was divided into a training set (50%) and a validation set (50%), used to train the SVM classifier and evaluate its classification performance, respectively. The results showed that when this classifier was applied to the validation set to evaluate the diagnostic potential of the predictive model, it achieved disease classification with 100% accuracy. Figure 10 (D).

Claims

1. A high-throughput proteomic analysis method for separating EVs based on DSPE-functionalized agarose media, characterized in that, Includes the following steps: S1. Obtain biological samples for extracting extracellular vesicles; S2. Add the DSPE-functionalized agarose medium to the biological sample used for extracting extracellular vesicles; S3. Centrifugation was used to separate and capture extracellular vesicles in DSPE-functionalized agarose medium. S4. Use EVs lysis buffer to lyse the extracellular vesicles extracted in step S3 to obtain the extracellular vesicle proteins. S5. After processing the proteins of extracellular vesicles, perform LC-MS / MS analysis and data acquisition; The biological sample is plasma, saliva, or urine; the EVs lysis buffer is one of RIPA, 1% SDS, 8M urea, or 0.1% RapiGest SF. The LC-MS / MS analysis and data acquisition use a data-independent acquisition mode; The preparation method of DSPE-functionalized agarose medium for separating extracellular vesicles (EVs) includes: washing a certain amount of NHS beads sequentially with hydrochloric acid and pre-cooled PBS, then resuspending them in PBS-Triton solution, adding the corresponding amount of DSPE-PEG2000-NH2 reagent for incubation; after incubation, centrifuging and discarding the supernatant, blocking excess NHS groups with ethanolamine solution; centrifuging and discarding the supernatant again, the resulting precipitate is the bisacyl lipid DSPE-functionalized agarose medium for separating extracellular vesicles (EVs); The structural formula of the bisacyl lipid DSPE-functionalized agarose medium includes: 。 2. Proteomic data obtained by the high-throughput proteomic analysis method for separating EVs based on DSPE-functionalized agarose media as described in claim 1.

3. A combination of biomarkers for colorectal cancer diagnosis, characterized in that, It includes the proprotein convertase subtilisin 6, S100A8 protein, anthrax toxin receptor 1, ENPP2, GTP-binding nucleoprotein Ran, apolipoprotein A-IV, and hydroxyacylglutathione hydrolase.

Images