High-throughput screening method for anti-myocardial injury activity of catechin compounds and application of high-throughput screening method

Abstract

The invention discloses a high-throughput screening method for anti-myocardial injury activity of catechin compounds and application of the high-throughput screening method. The method is based on an AC16 human myocardial cell strain, and isoprenaline is adopted for inducing and establishing a standardized myocardial injury model; by integrating cell activity detection, apoptosis analysis, PI3K / AKT signal channel activity monitoring, myocardial injury marker quantification and cell morphology intelligent image analysis, a multi-parameter comprehensive evaluation system is established, an image analysis algorithm and a comprehensive activity scoring model based on deep learning are developed, a catechin compound structure-activity relationship database is constructed, and a catechin compound structure-activity evaluation model is established. The method has the advantages that activity prediction is realized, a liquid treatment workstation, a high-content imaging system and other automatic equipment are integrated, 480-1920 compound samples can be evaluated in a single experiment, the screening period is shortened to 72-96 hours, the screening efficiency is improved by 10-20 times, and the method is suitable for rapid screening and action mechanism research of myocardial protection drugs.

Description

Technical Field

[0001] This invention relates to the field of drug screening technology, specifically to a high-throughput screening method for the anti-myocardial injury activity of catechin compounds and its application. Background Technology

[0002] Catechins are a class of polyphenolic natural products widely found in plants such as tea, cocoa, and grapes. They mainly include various structural analogs such as catechins, epicatechins, and epigallocatechin gallate. Recent studies have shown that catechins possess significant antioxidant, anti-inflammatory, and cardioprotective activities. For example, epigallocatechin gallate can alleviate myocardial ischemia-reperfusion injury by activating the PI3K / AKT signaling pathway, reducing myocardial infarction area by more than 40%. However, the cardioprotective activities of catechins with different structures vary significantly, and their structure-activity relationship is not yet fully elucidated. This severely restricts the progress of efficiently discovering and developing novel cardioprotective drugs from natural products.

[0003] Traditional drug screening methods primarily rely on animal models and primary cells, which suffer from problems such as long cycles, high costs, low throughput, and poor reproducibility. CN108588167A discloses a high-throughput screening method for drug compounds. This method involves adding the compounds to be screened into an incubation system containing specific target receptors, liver microsomes, or liver S9. By setting up two groups—one with and one without NADPH response factor—the inhibitory activity of the compounds on the target receptors is measured. The stability and druggability of the compounds are determined based on two activity inhibition curves. This method mainly targets opioid receptors or glucagon receptors, focusing on evaluating the metabolic stability of compounds, but it cannot comprehensively evaluate the direct protective effects of compounds on cardiomyocytes, nor can it reflect the regulatory effects of compounds on myocardial injury-related signaling pathways.

[0004] Existing cardiomyocyte drug screening systems suffer from the following shortcomings: First, the choice of cell models is limited. Most studies use H9C2 rat cardiomyocytes or primary cardiomyocytes, which exhibit significant species diversity, poor passage stability, and difficulty in standardization, affecting the reliability and reproducibility of screening results. Second, evaluation indicators are limited. Traditional methods primarily assess drug activity through cell viability testing or single biochemical indicators, failing to comprehensively reflect the multidimensional protective effects of compounds against myocardial injury. Third, throughput capacity is insufficient. Traditional detection methods rely on manual operation and single readings, making rapid screening of large-scale compound libraries difficult. Fourth, there is a lack of specialized screening platforms for specific compound types. Compounds with different structural types may exert cardioprotective effects through different mechanisms, necessitating the establishment of targeted evaluation systems. Fifth, data analysis and activity prediction capabilities are weak. There is a lack of systematic methods for integrating multi-parameter data to establish structure-activity relationship models, hindering intelligent prediction of compound activity.

[0005] The PI3K / AKT signaling pathway is a key regulatory pathway for cardiomyocyte survival and function maintenance, playing a central role in diseases such as myocardial ischemia-reperfusion injury, myocardial infarction, and heart failure. PI3K activates multiple downstream effector molecules, including GSK3β, mTOR, FOXO, and eNOS, by phosphorylating AKT, thereby inhibiting apoptosis, reducing oxidative stress, promoting angiogenesis, and maintaining mitochondrial function. Studies have shown that activating the PI3K / AKT pathway can significantly reduce myocardial infarction area and improve cardiac function, with a protective effect exceeding 50%. However, high-throughput screening methods capable of simultaneously monitoring PI3K / AKT pathway activity and the degree of cardiomyocyte damage are currently lacking, limiting the efficiency of drug discovery targeting this pathway.

[0006] AC16 cell line is a human cardiomyocyte cell line derived from the fusion of adult ventricular cardiomyocytes and SV40-transformed human fibroblasts. It retains key characteristics of cardiomyocytes, including the expression of cardiomyocyte-specific transcription factors and contractile proteins, while exhibiting good proliferative capacity and passage stability. Compared to H9C2 cells, AC16 cells more closely resemble the physiological characteristics of human cardiomyocytes and offer better standardization and reproducibility compared to primary cardiomyocytes. Therefore, AC16 cell line has become an ideal model for pharmacological and toxicological studies of cardiomyocytes. However, a high-throughput screening system for the cardioprotective activity of catechin compounds based on AC16 cell line has not yet been established.

[0007] Isoproterenol is a non-selective β-adrenergic receptor agonist that can induce oxidative stress, calcium overload, mitochondrial dysfunction, and apoptosis in cardiomyocytes, making it a commonly used tool for establishing in vitro models of myocardial injury. Isoproterenol-induced myocardial injury models can simulate the pathological processes of various cardiovascular diseases such as myocardial ischemia, myocardial infarction, and heart failure, and are suitable for evaluating the efficacy of cardioprotective drugs. However, traditional isoproterenol injury models mainly rely on a single endpoint, failing to achieve simultaneous monitoring and comprehensive evaluation of multiple dimensions of myocardial injury.

[0008] In summary, there is an urgent need to develop a high-throughput screening method specifically for the cardioprotective activity of catechin compounds. This method should feature standardized cell models, multidimensional evaluation indicators, automated detection processes, and intelligent data analysis to significantly improve the efficiency of discovering and developing novel cardioprotective drugs from natural products. Summary of the Invention

[0009] The purpose of this invention is to provide a high-throughput screening method for the anti-myocardial injury activity of catechin compounds and its application, so as to solve the problems of low screening efficiency, imperfect evaluation system and weak targeting in the prior art.

[0010] This invention provides a high-throughput screening method for the anti-myocardial injury activity of catechin compounds, comprising the following steps:

[0011] Step 1: Establish a standardized AC16 cardiomyocyte injury model. AC16 human cardiomyocytes are seeded in 96-well or 384-well plates and cultured until cell confluence reaches 70% to 85%. The culture medium is then replaced with isoproterenol-containing medium to induce cardiomyocyte injury. The working concentration of isoproterenol is 50 μM to 500 μM, and the incubation time is 12 to 48 hours, establishing a stable and reproducible cardiomyocyte injury model.

[0012] Step 2: Pretreatment of the catechin compounds to be screened. One to six hours before isoproterenol-induced injury, the catechin compounds to be screened were added to the culture medium. The concentration of the compounds was set in a series of concentrations from 0.1 μM to 100 μM, with 6 to 12 parallel replicates for each concentration. A blank control group and a model control group were also set up.

[0013] Step 3: Automated detection of cell viability. After isoproterenol treatment, cell viability was detected using the CCK-8 assay or MTT assay. A multi-functional microplate reader was used to automatically read the absorbance values ​​of each well, and cell viability was calculated. Cell viability = (OD value of experimental group - OD value of blank group) / (OD value of control group - OD value of blank group) × 100%. The half-maximal effective concentration (MCP) of the compound was calculated using nonlinear fitting analysis. 50 Value, EC 50 The lower the value, the stronger the myocardial protective activity of the compound.

[0014] Step four: Multidimensional analysis of apoptosis. The apoptosis rate was detected using Annexin V-FITC / PI double staining combined with flow cytometry or a high-content imaging system. Mitochondrial function was assessed using JC-1 mitochondrial membrane potential assay; decreased mitochondrial membrane potential is an early marker of apoptosis. TUNEL staining was used to detect DNA fragmentation levels, further confirming the apoptosis level. Rapid quantitative analysis of apoptosis indicators for each well was achieved using an automated image acquisition and analysis system.

[0015] Step 5: Quantitative monitoring of PI3K / AKT signaling pathway activity. Immunofluorescence staining combined with high-content imaging was used, employing anti-phosphorylated AKT and anti-phosphorylated PI3K antibodies to detect the activation level of the signaling pathway. Simultaneously, Western blot was used to detect the expression levels of key signaling molecules such as p-AKT, p-PI3K, and p-GSK3β. The ratio of phosphorylated protein to total protein was calculated to quantitatively evaluate the activation effect of catechin compounds on the PI3K / AKT signaling pathway. Multiplex immunofluorescence technology allows for the simultaneous detection of multiple signaling molecules in a single well, improving detection efficiency.

[0016] Step Six: Quantitative Detection of Myocardial Injury Markers. Cell culture supernatant was collected, and the release levels of myocardial injury markers such as lactate dehydrogenase, creatine kinase MB isoenzyme, and cardiac troponin I were detected using ELISA. The release levels of these markers were positively correlated with the degree of myocardial cell damage, and the degree of reduction reflected the cardioprotective effect of the compound. Simultaneously, malondialdehyde (MDA) content and superoxide dismutase (SOD) activity in the culture supernatant were measured to evaluate the antioxidant capacity of the compound. Batch testing of multiple samples was achieved using a fully automated biochemical analyzer.

[0017] Step 7: Intelligent Image Analysis of Cell Morphology. A high-content imaging system was used to automatically acquire bright-field and fluorescence images of cells from each experimental group, including multiple morphological parameters such as nuclear morphology, cell size, and cytoskeleton structure. A deep learning-based image analysis algorithm was developed to automatically identify and quantify morphological changes in cardiomyocytes, including apoptotic features such as cell shrinkage, nuclear condensation, and cell debris. The system can automatically generate a morphological score for each well, enabling an objective and quantitative evaluation of the degree of cardiomyocyte damage.

[0018] Step 8: Establishment of a comprehensive activity scoring system. A comprehensive scoring model for the cardioprotective activity of catechin compounds is established by integrating multiple parameters such as cell viability, apoptosis rate, PI3K / AKT pathway activity, damage marker levels, and morphological scores. Principal component analysis is used to reduce the dimensionality of the multidimensional data and extract key activity characteristics. The comprehensive activity score of the compounds is calculated based on the weighting coefficients of each parameter, ranging from 0 to 100 points, with higher scores indicating stronger cardioprotective activity. This scoring system enables standardized comparisons of the cardioprotective activities of catechin compounds with different structures.

[0019] Step nine: Construction of the structure-activity relationship database and activity prediction. Structural information and activity data of screened catechin compounds are collected, and molecular descriptors are calculated, including molecular weight, lipophilicity, number of hydrogen bond donors, number of hydrogen bond acceptors, molecular flexibility, number of aromatic rings, number and position of hydroxyl groups, etc. Machine learning algorithms, including support vector machines, random forests, or deep neural networks, are used to establish a quantitative structure-activity relationship model between the structure of catechin compounds and their cardioprotective activities. This model can be used to predict the cardioprotective activities of novel catechin derivatives, guiding the structural optimization and synthesis of compounds. As the database is continuously expanded, the predictive accuracy of the model will continue to improve.

[0020] Step 10: Implementation of a high-throughput automated workflow. Integrate a liquid handling workstation, automated incubator, multi-functional microplate reader, high-content imaging system, and data management system to establish a fully automated screening process. Through standardized operating procedures, a single experiment can simultaneously evaluate 480 to 1920 compound samples, each containing 6 to 12 concentration gradients, for a total of 4608 to 23040 wells. The entire screening process, from cell seeding to data analysis, can be completed within 72 to 96 hours, increasing screening efficiency by 10 to 20 times compared to traditional methods. The system automatically generates standardized data reports, including dose-response curves and EC50 data. 50 Information such as activity value, overall activity score, and compound ranking.

[0021] Furthermore, the preferred culture conditions for the AC16 cells are as follows: culture in DMEM / F12 medium containing 12.5% ​​fetal bovine serum, 2 mM L-glutamine, 100 U / mL penicillin, and 100 μg / mL streptomycin; culture temperature: 37°C; CO2 concentration: 5%; relative humidity: 95%. The cell seeding density is 5000 to 20000 cells per well, preferably 10000 cells per well for a 96-well plate and 3000 cells per well for a 384-well plate. After seeding, cells are cultured for 24 to 48 hours until they are fully adhered and reach a suitable confluence before proceeding with subsequent experiments.

[0022] Furthermore, the preferred conditions for isoproterenol-induced myocardial injury are an isoproterenol concentration of 100 μM to 300 μM and an incubation time of 24 to 36 hours. More preferably, the isoproterenol concentration is 200 μM and the incubation time is 24 hours. Under these conditions, cell viability can be reduced to 50% to 60%, establishing a moderate-severity myocardial injury model. This model has a suitable injury window, which can fully reflect the protective effect of the compound while avoiding the reduction in detection sensitivity caused by excessive damage.

[0023] Furthermore, the catechin compounds include, but are not limited to, catechins, epicatechin, catechin gallate, epicatechin gallate, gallocatechin, epigallocatechin, gallocatechin gallate, epigallocatechin gallate, proanthocyanidins, theaflavins, thearubigins, and their derivatives. The concentration gradient of the compounds is set at 0.1 μM, 0.3 μM, 1 μM, 3 μM, 10 μM, 30 μM, and 100 μM, a total of seven concentration points, covering the possible effective concentration range of the compounds. For compounds with stronger activity, the concentration gradient can be further refined to 0.03 μM, 0.1 μM, 0.3 μM, 1 μM, 3 μM, 10 μM, and 30 μM for more precise determination of EC. 50 value.

[0024] Furthermore, in the cell viability detection methods, the CCK-8 method involves adding 10 μL of CCK-8 solution to each well, incubating at 37°C in the dark for 1 to 3 hours, and measuring the absorbance at 450 nm. The MTT method involves adding 20 μL of MTT solution to each well, incubating at 37°C for 4 hours, discarding the culture medium, adding 150 μL of dimethyl sulfoxide to dissolve formazan crystals, shaking for 10 minutes, and measuring the absorbance at 490 nm. The two methods have comparable detection sensitivity; the CCK-8 method is simpler to operate as it eliminates the need for a dissolution step, while the MTT method is less expensive. The CCK-8 method is preferred for high-throughput detection.

[0025] Further, in the apoptosis detection, the Annexin V-FITC / PI double staining procedure is as follows: collect cells, wash twice with pre-cooled PBS, resuspend in 500 μL binding buffer, add 5 μL Annexin V-FITC and 5 μL PI staining solution, incubate at room temperature in the dark for 15 minutes, and immediately perform flow cytometry or high-content imaging analysis. Early apoptotic cells show Annexin V positivity and PI negativity, late apoptotic cells show Annexin V and PI double positivity, and necrotic cells show Annexin V negativity and PI positivity. The total apoptosis rate is calculated as (proportion of early apoptotic cells + proportion of late apoptotic cells) × 100%. The JC-1 staining procedure is as follows: add 10 μM JC-1 staining solution to the culture medium, incubate at 37°C for 20 minutes, wash twice with PBS, and detect the ratio of red fluorescence to green fluorescence on a fluorescence microscope or flow cytometer. A decreased ratio indicates a decrease in mitochondrial membrane potential.

[0026] Further, in the detection of the PI3K / AKT signaling pathway, the immunofluorescence staining procedure is as follows: cells are fixed with 4% paraformaldehyde for 15 minutes, permeabilized with 0.3% Triton X-100 for 10 minutes, blocked with 5% bovine serum albumin for 1 hour, primary antibody solutions including anti-p-AKT antibody diluted 1:200 and anti-p-PI3K antibody diluted 1:200 are added, incubated overnight at 4°C, washed three times with PBS, and secondary fluorescent antibody solutions including Alexa Fluor 488-labeled goat anti-rabbit IgG diluted 1:500 and Alexa Fluor 594-labeled goat anti-mouse IgG diluted 1:500 are added, incubated at room temperature for 1 hour, and the cell nuclei are counterstained with DAPI for 5 minutes. After mounting, the cells are observed under a high-content imaging system. The system automatically acquires 9 fields of view per well, analyzes 200 to 500 cells per field, and calculates the average fluorescence intensity to quantify the phosphorylation level. The Western blot procedure involves extracting total cellular protein, determining protein concentration using the BCA method, performing SDS-PAGE electrophoresis on 30 μg of protein, transferring the protein to a PVDF membrane, blocking with 5% skim milk for 1 hour, adding primary antibody solutions including anti-p-AKT antibody diluted 1:1000, anti-AKT antibody diluted 1:1000, anti-p-PI3K antibody diluted 1:1000, anti-PI3K antibody diluted 1:1000, anti-p-GSK3β antibody diluted 1:1000, and anti-GSK3β antibody diluted 1:1000, incubating overnight at 4°C, washing three times with TBST, adding HRP-labeled secondary antibody solution diluted 1:5000, incubating at room temperature for 1 hour, performing ECL chemiluminescence staining, acquiring images using a gel imaging system, quantifying band gray values ​​using ImageJ software, and calculating the ratio of phosphorylated protein to total protein.

[0027] Furthermore, in the detection of myocardial injury markers, cell culture supernatant was collected in 96-well or 384-well plates, and commercially available ELISA kits were used to detect lactate dehydrogenase activity, creatine kinase MB isoenzyme concentration, and cardiac troponin I concentration. The lactate dehydrogenase assay kit is based on the lactate oxidation reaction, measuring absorbance at 450 nm, and enzyme activity was calculated from a standard curve, in units of U / L. The detection range of the creatine kinase MB isoenzyme ELISA kit is 0.5 ng / mL to 50 ng / mL, and the detection range of the cardiac troponin I ELISA kit is 0.05 ng / mL to 10 ng / mL. Malondialdehyde content was determined using the thiobarbituric acid method, measuring absorbance at 532 nm, and concentration was calculated from a standard curve, in units of nmol / mg protein. Superoxide dismutase activity was determined using the WST-1 method, measuring absorbance at 450 nm, and enzyme activity was calculated from a standard curve, in units of U / mg protein. All tests were performed using a fully automated microplate reader, with a testing time of 10 to 15 minutes per 96-well plate.

[0028] Furthermore, the image analysis algorithm is based on a convolutional neural network architecture, including an input layer, multiple convolutional layers, pooling layers, fully connected layers, and an output layer. The algorithm is trained using a labeled cell image dataset, comprising four categories: normal cells, mildly damaged cells, moderately damaged cells, and severely damaged cells, with each category containing 5000 to 10000 images. The algorithm can automatically extract 89 morphological feature parameters of cells, including cell area, perimeter, roundness, aspect ratio, nucleocytoplasmic ratio, cytoskeleton integrity, cell connectivity, and number of cell processes. Through feature fusion and classifier training, the algorithm can automatically classify the degree of cell damage with an accuracy exceeding 92%. The system outputs a damage score for each well of the cell, ranging from 0 to 100 points. A higher score indicates more severe cell damage, and a better protective effect from the compound is reflected in a lower score.

[0029] Furthermore, the comprehensive activity scoring model is established by first standardizing each parameter, converting data of different dimensions into dimensionless values ​​from 0 to 100. For positive indicators such as cell viability, the standardized value = measured value / maximum value × 100. For negative indicators such as apoptosis rate and damage markers, the standardized value = (maximum value - measured value) / maximum value × 100. For positive indicators such as PI3K / AKT pathway activation, the standardized value = measured value / maximum value × 100. Then, principal component analysis is used to determine the weight coefficients of each parameter, assigning weights according to their contribution rates: cell viability weight 0.25, apoptosis rate weight 0.20, PI3K / AKT pathway activation weight 0.20, damage marker level weight 0.20, and morphological score weight 0.15. The comprehensive activity score = Σ(standardized value × weight coefficient), with a score range of 0 to 100. For EC50 Compounds with a concentration less than 1 μM and a comprehensive activity score greater than 80 are classified as highly active compounds and given priority for further research. For EC... 50 Compounds with activity values ​​between 1 μM and 10 μM and overall activity scores between 60 and 80 are classified as moderately active compounds and can be used as lead compounds for structure optimization. For EC... 50 Compounds with a value greater than 10 μM or a comprehensive activity score less than 60 are classified as low-activity compounds.

[0030] Furthermore, the structure-activity relationship model was established by using ChemDraw software to draw the molecular structures of catechin compounds and using Dragon software to calculate 3615 molecular descriptors, including topological indices, geometric indices, electronic indices, hydrophobic indices, and quantum chemical indices. Through correlation analysis and variance screening, 30 to 50 key descriptors significantly associated with cardioprotective activity were selected. The dataset was divided into training and testing sets in a 7:3 ratio, and a nonlinear regression model was established using the random forest algorithm. The model evaluation metrics included the correlation coefficient R0. 2 The root mean square error (RMSE) and mean absolute error (MAE) were measured. Model parameters, including the number of decision trees, maximum depth, and minimum number of sample splits, were optimized using five-fold cross-validation. The optimal model should have an R² value greater than 0.85 and an RMSE less than 10, indicating good fit and predictive ability. The model can output predicted activity scores and confidence intervals for compounds, providing a basis for virtual screening of new compounds. Compounds with predicted activity scores greater than 70 and confidence levels greater than 80% were prioritized for experimental validation. As experimental data accumulates, the training dataset is regularly updated and the model is retrained to continuously improve prediction accuracy.

[0031] This invention also provides the application of the high-throughput screening method for the anti-myocardial injury activity of the aforementioned catechin compounds in new drug development. This method can be used to rapidly screen catechin compounds with cardioprotective activity from natural product libraries, compound libraries, or virtual compound libraries, providing an efficient screening platform for the discovery of new cardioprotective drugs. Highly active compounds obtained through this method can be further subjected to structural optimization, pharmacodynamic evaluation, and pharmacokinetic studies, accelerating the development of innovative cardioprotective drugs.

[0032] This invention also provides the application of the screening method in the study of the mechanism of action of cardioprotective drugs. By simultaneously monitoring multiple biological indicators, especially the activation status of the PI3K / AKT signaling pathway, the molecular mechanism by which catechin compounds exert their cardioprotective effects can be elucidated in depth. This method can be used to compare the differences in the mechanisms of action of catechin compounds with different structures, reveal the relationship between structure and mechanism of action, and provide a theoretical basis for rational drug design.

[0033] This invention also provides the application of the screening method in evaluating the toxicity of drugs to cardiomyocytes. In addition to evaluating the cardioprotective activity of compounds, this method can also be used to evaluate the toxic effects of compounds on normal cardiomyocytes. By detecting the effects of compounds on cell viability, apoptosis, and signaling pathways of normal AC16 cells under isoproterenol-free conditions, the safety window of the compound can be assessed, and the therapeutic index can be calculated as the ratio of the non-toxic concentration to the effective concentration. This method can be used for early screening of cardiomyocyte toxicity, reducing the risks of drug development.

[0034] This invention also provides the application of the screening method in personalized medicine. By utilizing cardiomyocytes obtained from the differentiation of patient-derived induced pluripotent stem cells, a personalized myocardial injury model can be established. This allows for the evaluation of the protective effects of different catechin compounds on cardiomyocytes in specific patients, providing personalized guidance for clinical medication. This method is particularly suitable for special populations with genetic susceptibility or comorbidities, enabling precision medicine.

[0035] Compared with the prior art, the present invention has the following beneficial effects.

[0036] First, this invention establishes a high-throughput screening system specifically for the cardioprotective activity of catechin compounds, filling a technological gap in this field. Compared with existing general drug screening methods, this invention systematically optimizes the structural characteristics and mechanisms of action of catechin compounds, including the selection of cell models, the design of damage induction protocols, the determination of evaluation indicators, and the establishment of data analysis methods, ensuring the accuracy and reliability of the screening results. Experiments have demonstrated that the highly active compounds screened using this method also exhibit significant cardioprotective effects in animal models, validating the effectiveness of the screening system.

[0037] Secondly, this invention uses the AC16 human cardiomyocyte line as a screening model, which has significant advantages over H9C2 rat cardiomyocytes and primary cardiomyocytes. AC16 cells are human-derived, enabling them to more accurately reflect the effects of compounds in the human body and reducing false positive and false negative results caused by species differences. AC16 cells exhibit good proliferation capacity and passage stability, allowing for batch culture while maintaining consistent biological characteristics, ensuring the comparability of experimental results across different batches. AC16 cells retain key characteristics of cardiomyocytes, including the expression of cardiomyocyte-specific transcription factors GATA4 and MEF2C, and the expression of cardiac contractile proteins α-actin and myosin, thus accurately reflecting the physiological and pathological processes of cardiomyocytes.

[0038] Third, this invention establishes a multi-dimensional comprehensive evaluation system, overcoming the limitations of traditional single-indicator evaluation. By integrating multiple parameters such as cell viability, apoptosis, PI3K / AKT signaling pathway activity, myocardial injury markers, and cell morphology, it can comprehensively evaluate the protective effect of catechin compounds against myocardial injury. In particular, the introduction of quantitative monitoring of the PI3K / AKT signaling pathway reveals the key molecular mechanism by which the compounds exert their cardioprotective effects. Experimental data show that the comprehensive activity score is highly correlated with the cardioprotective effect of the compounds in animal models, with a correlation coefficient reaching 0.89, demonstrating the scientific validity and predictive ability of this evaluation system.

[0039] Fourth, this invention develops a deep learning-based intelligent image analysis algorithm, enabling automatic identification and quantitative analysis of morphological changes in cardiomyocytes. Traditional cell morphology evaluation relies on manual observation, which is highly subjective, inefficient, and difficult to standardize. The algorithm of this invention can automatically extract 89 morphological feature parameters to objectively and quantitatively evaluate the degree of cell damage, achieving an accuracy rate of over 92%. This algorithm significantly improves screening efficiency, reducing the image analysis time for a single 384-well plate from the traditional 4-6 hours to 15-20 minutes, an efficiency improvement of more than 15 times.

[0040] Fifth, this invention establishes a quantitative structure-activity relationship model for the cardioprotective activity of catechin compounds, enabling intelligent prediction of the activity of new compounds. Through systematic analysis of the relationship between compound structure and activity, key structural features affecting cardioprotective activity are revealed, including the number and position of hydroxyl groups, the presence of galloyl groups, and stereoconfiguration. The model's predictive accuracy (R²) reaches 0.87, making it suitable for virtual screening and rational drug design, guiding the synthesis and optimization of novel catechin derivatives, and significantly reducing R&D costs and time.

[0041] Sixth, this invention achieves high-throughput automated screening, significantly improving screening efficiency and throughput. By integrating a liquid handling workstation, automated incubator, multifunctional microplate reader, and high-content imaging system, a standardized automated workflow has been established. A single experiment can simultaneously evaluate 480 to 1920 compound samples, increasing screening throughput by 10 to 20 times compared to traditional methods. The entire screening cycle is shortened from the traditional 2 to 4 weeks to 72 to 96 hours, significantly accelerating the drug discovery process. The system also exhibits good stability and repeatability, with an intra-batch coefficient of variation of less than 10% and an inter-batch coefficient of variation of less than 15%, ensuring data quality.

[0042] Seventh, the method of this invention has broad application value. It can be used not only for screening catechin compounds but also for evaluating the cardioprotective activity of other types of natural products and synthetic compounds. By adjusting the injury induction protocol and evaluation indicators, this platform can be applied to drug screening in various cardiovascular disease models such as myocardial ischemia-reperfusion injury, myocardial infarction, myocardial hypertrophy, and myocarditis. This method can also be used for evaluating cardiomyocyte cytotoxicity, studying drug mechanisms of action, and personalized medicine, possessing significant scientific value and promising clinical application prospects.

[0043] Eighth, the method of this invention is economical, practical, and easy to promote and apply. The required cell lines, reagents, and instruments are all commercially available products. Standardized operating procedures and automated systems reduce the technical requirements for operators and minimize human error. The screening cost per sample is approximately 5 to 10 yuan, far lower than the cost of animal experiments. This method is suitable for use by pharmaceutical companies, research institutes, and third-party drug screening service organizations, helping to accelerate the development of new myocardial protective drugs and providing more treatment options for the prevention and treatment of cardiovascular diseases. Detailed Implementation

[0044] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the invention will be described in detail below with reference to specific embodiments. It should be noted that the following embodiments are for illustrative purposes only and are not intended to limit the invention. Without departing from the principles of this invention, those skilled in the art can make various improvements and modifications, and these improvements and modifications should also be considered within the scope of protection of this invention.

[0045] Example 1: Culture and passage of AC16 cardiomyocytes

[0046] The frozen AC16 human cardiomyocytes were removed from the liquid nitrogen container and immediately placed in a 37°C water bath for rapid thawing. Once the ice in the cryovials had mostly melted, they were removed and the outer walls of the cryovials were wiped with 75% ethanol for disinfection. In a clean bench, the cryovials were opened, and the cell suspension was gently aspirated with a 1mL pipette and transferred to a 15mL centrifuge tube pre-filled with 8mL of complete culture medium. The complete culture medium consisted of DMEM / F12 basal medium supplemented with 12.5% ​​fetal bovine serum, 2mM L-glutamine, 100U / mL penicillin, and 100μg / mL streptomycin. The cells were centrifuged at 300×g for 5 minutes at room temperature, the supernatant was discarded, and the cells were resuspended in 5mL of fresh complete culture medium. The cell suspension was transferred to a T-25 culture flask, and complete culture medium was added to a total volume of 10mL. The flasks were gently mixed. The culture flasks were then placed in a cell culture incubator at 37°C, 5% CO2, and 95% relative humidity.

[0047] Replace the culture medium with fresh medium 24 hours after thawing to remove non-adherent dead cells and cryopreservatives. Thereafter, change the medium every 2 to 3 days and observe cell growth. Passage is performed when cell confluence reaches 80% to 90%. The passage procedure is as follows: discard the culture medium, wash cells twice with PBS to remove serum, add 2 mL of 0.25% trypsin-EDTA digestion solution, and digest at 37°C for 3 to 5 minutes. When cells become rounded and begin to detach under a microscope, add 8 mL of complete culture medium to stop digestion. Gently pipette to completely suspend the cells and transfer them to 15 mL centrifuge tubes. Centrifuge at 300×g for 5 minutes, discard the supernatant, and resuspend the cells in 10 mL of fresh complete culture medium. Divide the cells into 3 to 5 flasks at a ratio of 1:3 to 1:5 (one flask of cells is passaged into 3 to 5 flasks), add complete culture medium to each flask to a total volume of 10 mL, and incubate in an incubator.

[0048] AC16 cells exhibit stable growth and uniform morphology from passages 5 to 20, making them suitable for experimental use. The cell cryopreservation method involves digesting and collecting cells in the logarithmic growth phase, centrifuging at 300×g for 5 minutes, discarding the supernatant, and resuspending the cells in cryopreservation solution until a cell density of 1×10⁻⁶ is achieved. 6 Cells / mL to 2×10 6 Cells / mL. The cryopreservation solution is 90% complete culture medium supplemented with 10% dimethyl sulfoxide. Aliquot 1 mL of cell suspension into cryovials, immediately place them in a programmed cooling box, and incubate overnight at -80°C. The next day, transfer them to a liquid nitrogen tank for long-term storage.

[0049] Example 2: Establishment and optimization of an isoproterenol-induced AC16 cardiomyocyte injury model

[0050] To establish a stable and reproducible model of cardiomyocyte injury, this embodiment systematically optimized the concentration and duration of action of isoproterenol.

[0051] AC16 cells in the logarithmic growth phase were digested and collected, and the cell density was adjusted to 1×10⁻⁶. 5 Cells / mL. Add 100 μL of cell suspension to each well of a 96-well cell culture plate, i.e., seed 1 × 10⁶ cells per well. 4 For each cell, sterile PBS was added to the edge wells to reduce edge effects. The culture plate was placed in an incubator and incubated for 24 hours until the cells were fully adhered.

[0052] Prepare working solutions of isoproterenol at different concentrations. Weigh isoproterenol hydrochloride powder, dissolve it in sterile PBS to prepare a 10 mM stock solution, filter it through a 0.22 μm sterile filter membrane for sterilization, aliquot it, and store it at -20°C. Before use, dilute with complete culture medium to the working concentration, including 0 μM, 50 μM, 100 μM, 200 μM, 300 μM, 400 μM, and 500 μM.

[0053] Discard the original culture medium in the 96-well plate, and add 200 μL of fresh culture medium containing different concentrations of isoproterenol to each well, with 6 parallel wells for each concentration. The blank control group is added with complete culture medium without isoproterenol. Return the culture plate to the incubator and incubate for 12 hours, 24 hours, 36 hours, and 48 hours, respectively.

[0054] At each time point, 20 μL of CCK-8 solution was added to each well, and the mixture was incubated at 37°C in the dark for 2 hours. The absorbance (OD) of each well was measured at 450 nm using a multi-mode microplate reader. 450 Cell viability is calculated using the formula: Cell viability = (Experimental group OD) / ... 450 - Blank Hole OD 450 ) / (control group OD 450 - Blank Hole OD 450 ()×100%. The relationship between isoproterenol concentration and cell viability was plotted using GraphPad Prism software.

[0055] The experimental results showed that after 12 hours of incubation, the cell survival rate was 72.3% when the isoproterenol concentration reached 500 μM, indicating relatively mild damage. After 24 hours of incubation, 100 μM isoproterenol reduced cell survival to 78.5%, 200 μM to 56.8%, 300 μM to 42.1%, 400 μM to 28.7%, and 500 μM to 18.2%. After 36 hours of incubation, 100 μM isoproterenol reduced cell survival to 68.3%, 200 μM to 48.2%, and 300 μM to 32.5%. After 48 hours of incubation, even 50 μM isoproterenol reduced cell survival to below 60%, but the damage was too severe, leading to the death of most cells.

[0056] Considering both the damage window and detection sensitivity, the optimal damage model conditions were determined to be an isoproterenol concentration of 200 μM and an incubation time of 24 hours. Under these conditions, the cell survival rate was 56.8%, establishing a moderate-severity myocardial injury model. This model effectively reflects the protective effect of the compound while avoiding the reduction in detection sensitivity caused by excessive damage, making it suitable for high-throughput drug screening.

[0057] Further validation of the biological characteristics of this injury model was conducted. Annexin V-FITC / PI double staining flow cytometry was used to detect cell apoptosis. The results showed that the apoptosis rate in the control group was 4.2%, while after 24 hours of treatment with 200 μM isoproterenol, the apoptosis rate increased to 38.6%, with early apoptotic cells accounting for 23.4% and late apoptotic cells accounting for 15.2%. JC-1 staining was used to detect mitochondrial membrane potential. The results showed that the red-green fluorescence ratio in the isoproterenol-treated group was 62.1% lower than that in the control group, indicating a significant decrease in mitochondrial membrane potential. ELISA was used to detect the release of lactate dehydrogenase in the culture supernatant. The control group was 86.3 U / L, while the isoproterenol-treated group increased to 478.5 U / L, a 4.5-fold increase. These results confirm that the model successfully induced cardiomyocyte injury, including pathological features such as apoptosis, mitochondrial dysfunction, and disruption of cell membrane integrity.

[0058] Example 3: Systematic evaluation of the cardioprotective activity of epigallocatechin gallate.

[0059] Epigallocatechin gallate is the most abundant catechin compound in tea. This embodiment systematically evaluates its cardioprotective activity and verifies the effectiveness of the screening method.

[0060] AC16 cells were seeded into 96-well plates at a density of 1 × 10⁶ cells per well. 4 Cells were cultured for 24 hours. A series of epigallocatechin gallate solutions were prepared, including 0 μM, 0.1 μM, 0.3 μM, 1 μM, 3 μM, 10 μM, 30 μM, and 100 μM, with six parallel wells for each concentration. Culture medium containing different concentrations of epigallocatechin gallate was added to the cells for 2 hours of pretreatment. Isoproterenol was then added to the medium to a final concentration of 200 μM, and incubation continued for another 24 hours. A normal control group was established without isoproterenol and epigallocatechin gallate, while a model control group was established with only isoproterenol and no epigallocatechin gallate.

[0061] After incubation, cell viability was assessed. 20 μL of CCK-8 solution was added to each well, and the cells were incubated at 37°C in the dark for 2 hours. The absorbance was measured at 450 nm to calculate cell viability. Results showed that the cell viability in the model control group was 55.7%, while the viability in the 0.1 μM epigallocatechin gallate group was 59.2%, 64.8% in the 0.3 μM group, 72.5% in the 1 μM group, 81.3% in the 3 μM group, 88.6% in the 10 μM group, 90.2% in the 30 μM group, and 89.7% in the 100 μM group. Nonlinear fitting analysis was performed using GraphPad Prism software to calculate the EC50 of epigallocatechin gallate. 50The value was 1.25 μM, indicating that it has strong cardioprotective activity.

[0062] Apoptosis was detected. Cells from each experimental group were collected and analyzed by flow cytometry after double staining with Annexin V-FITC and PI. The results showed that the apoptosis rate in the normal control group was 3.8%, the apoptosis rate in the model control group was 39.2%, and the apoptosis rate in the 10 μM epigallocatechin gallate treatment group decreased to 16.5%, with an apoptosis inhibition rate of 58.0%. Mitochondrial membrane potential was detected by JC-1 staining. The red-green fluorescence ratio in the normal control group was 2.35, which decreased to 0.87 in the model control group, and increased to 1.92 in the 10 μM epigallocatechin gallate treatment group, indicating that it could effectively maintain mitochondrial membrane potential.

[0063] The PI3K / AKT signaling pathway was detected. Total protein was extracted from cells in each experimental group, and the expression levels of p-AKT, AKT, p-PI3K, PI3K, p-GSK3β, and GSK3β were detected by Western blot. The results showed that the p-AKT / AKT ratio was 1.00 in the normal control group, decreased to 0.42 in the model control group, and increased to 0.88 in the 10 μM epigallocatechin gallate treatment group, indicating that it could activate AKT phosphorylation. The p-PI3K / PI3K ratio was 1.00 in the normal control group, decreased to 0.38 in the model control group, and increased to 0.82 in the 10 μM epigallocatechin gallate treatment group. The p-GSK3β / GSK3β ratio was 1.00 in the normal control group, decreased to 0.45 in the model control group, and increased to 0.91 in the 10 μM epigallocatechin gallate treatment group. These results confirm that epigallocatechin gallate exerts its cardioprotective effect by activating the PI3K / AKT signaling pathway.

[0064] Myocardial injury markers were detected. Culture supernatants from each experimental group were collected, and the release levels of lactate dehydrogenase, creatine kinase MB isoenzyme, and cardiac troponin I were detected by ELISA. Results showed that lactate dehydrogenase activity was 82.5 U / L in the normal control group, increased to 485.3 U / L in the model control group, and decreased to 178.6 U / L in the 10 μM epigallocatechin gallate treatment group. Creatine kinase MB isoenzyme concentration was 1.2 ng / mL in the normal control group, increased to 18.7 ng / mL in the model control group, and decreased to 5.8 ng / mL in the 10 μM epigallocatechin gallate treatment group. Cardiac troponin I concentration was 0.08 ng / mL in the normal control group, increased to 2.35 ng / mL in the model control group, and decreased to 0.62 ng / mL in the 10 μM epigallocatechin gallate treatment group. These results indicate that epigallocatechin gallate can significantly reduce cardiomyocyte injury.

[0065] Antioxidant capacity was evaluated. Malondialdehyde (MDA) content and superoxide dismutase (SOD) activity in the culture supernatant were measured. Results showed that MDA content in the normal control group was 2.8 nmol / mg protein, increasing to 15.6 nmol / mg protein in the model control group, and decreasing to 6.2 nmol / mg protein in the 10 μM epigallocatechin gallate treatment group. SOD activity was 86.3 U / mg protein in the normal control group, decreasing to 38.5 U / mg protein in the model control group, and recovering to 72.8 U / mg protein in the 10 μM epigallocatechin gallate treatment group. These results indicate that epigallocatechin gallate possesses significant antioxidant capacity.

[0066] Cell morphology analysis was performed using a high-content imaging system. Cell nuclei were stained with DAPI, and the cytoskeleton was stained with Phalloidin-TRITC. Cell images of each experimental group were automatically acquired, with nine fields of view per group, analyzing 300 to 500 cells per field. Image analysis algorithms automatically extracted morphological parameters such as cell area, perimeter, roundness, and nucleocytoplasmic ratio. Results showed that cells in the normal control group had regular morphology, with a cell area of ​​1285 μm², a cell perimeter of 148 μm, a roundness of 0.73, and a nucleocytoplasmic ratio of 0.18. Cells in the model control group showed significant shrinkage, with a cell area reduced to 687 μm², a cell perimeter reduced to 98 μm, a roundness decreased to 0.52, and a nucleocytoplasmic ratio increased to 0.35, exhibiting disordered cytoskeleton structure. The 10 μM epigallocatechin gallate treatment group showed significant improvement in cell morphology, with a cell area of ​​1078 μm². 2 The cells had a perimeter of 132 μm, a roundness of 0.68, a nucleocytoplasmic ratio of 0.22, and a basically normal cytoskeleton structure. The system automatically calculated the cell damage score: 5.2 points for the normal control group, 78.5 points for the model control group, and 28.3 points for the 10 μM epigallocatechin gallate treatment group, with a protection rate of 64.0%.

[0067] Based on the comprehensive activity scoring model, the comprehensive activity score of table gallic acid catechin gallate was calculated by integrating the above parameters. The standardized value for cell viability was 94.3 points, the standardized value for apoptosis rate was 86.5 points, the standardized value for PI3K / AKT pathway activation was 88.0 points, the standardized value for damage marker level was 82.7 points, and the standardized value for morphological score was 78.6 points. Based on the weighted coefficients, the comprehensive activity score = 94.3 × 0.25 + 86.5 × 0.20 + 88.0 × 0.20 + 82.7 × 0.20 + 78.6 × 0.15 = 86.4 points, classifying it as a highly active compound.

[0068] Example 4: Comparative Study of the Cardioprotective Activities of Catechins with Different Structures

[0069] To systematically study the structure-activity relationship of catechin compounds, this example compares the cardioprotective activities of seven catechin compounds with different structures, including catechin, epicatechin, catechin gallate, epicatechin gallate, gallocatechin, epigallocatechin, and epigallocatechin gallate.

[0070] Using the same experimental method as in Example 3, AC16 cells were seeded in 96-well plates and pretreated for 2 hours with different compounds ranging from 0.1 μM to 100 μM. Then, they were treated with 200 μM isoproterenol for 24 hours. Cell viability, apoptosis rate, PI3K / AKT pathway activity, damage markers, and changes in cell morphology were measured. EC50 was calculated. 50 Value and overall activity score.

[0071] Cell viability assay results showed that the EC50 of the seven compounds... 50 The concentrations of the compounds were 18.6 μM for catechin, 12.4 μM for epicatechin, 6.8 μM for catechin gallate, 4.2 μM for epicatechin gallate, 5.1 μM for gallocatechin, 2.9 μM for epigallocatechin, and 1.25 μM for epigallocatechin gallate. At a concentration of 10 μM, each compound increased cell viability to 65.8%, 70.2%, 77.5%, 82.3%, 80.1%, 85.7%, and 88.6%, respectively, all of which were significantly higher than the 55.7% in the model control group.

[0072] Cell apoptosis assay results showed that at a concentration of 10 μM, the seven compounds reduced the apoptosis rate to 31.2%, 28.5%, 24.3%, 20.8%, 22.1%, 18.9%, and 16.5%, respectively, all significantly lower than the 39.2% in the model control group. The apoptosis inhibition rates were 20.4%, 27.3%, 38.1%, 47.0%, 43.6%, 51.8%, and 58.0%, respectively.

[0073] The results of PI3K / AKT signaling pathway detection showed that at a concentration of 10 μM, the p-AKT / AKT ratios of the seven compounds were 0.58, 0.63, 0.71, 0.78, 0.74, 0.83, and 0.88, respectively; the p-PI3K / PI3K ratios were 0.54, 0.59, 0.68, 0.75, 0.71, 0.80, and 0.82, respectively; and the p-GSK3β / GSK3β ratios were 0.61, 0.66, 0.73, 0.81, 0.78, 0.86, and 0.91, respectively. All these values ​​were significantly higher than those of the model control group, indicating that these compounds can activate the PI3K / AKT signaling pathway, but the degree of activation varies.

[0074] The results of myocardial injury marker detection showed that, at a concentration of 10 μM, the seven compounds reduced the release of lactate dehydrogenase to 356.2 U / L, 318.5 U / L, 276.3 U / L, 232.8 U / L, 248.5 U / L, 205.3 U / L, and 178.6 U / L, respectively; reduced the concentration of creatine kinase MB isoenzyme to 12.3 ng / mL, 10.8 ng / mL, 9.2 ng / mL, 7.5 ng / mL, 8.1 ng / mL, 6.4 ng / mL, and 5.8 ng / mL, respectively; and reduced the concentration of cardiac troponin I to 1.48 ng / mL, 1.26 ng / mL, 1.05 ng / mL, 0.83 ng / mL, 0.92 ng / mL, 0.71 ng / mL, and 0.62 ng / mL, respectively. All of these levels were significantly lower than those in the model control group.

[0075] The comprehensive activity score calculation results showed that the scores of the seven compounds were 62.3, 68.5, 74.2, 80.8, 78.5, 84.2, and 86.4, respectively, with the ranking order being epigallocatechin gallate > epigallocatechin > epigallocatechin gallate > gallatechin > catechin gallate > epigallocatechin > catechin.

[0076] Structure-activity relationship analysis showed that the introduction of the galloyl group significantly improved the cardioprotective activity of the compounds, with compounds containing the galloyl group exhibiting significantly higher activity than those without. The presence of the 3-hydroxyl group also contributed to the enhanced activity, with phenotypic compounds showing higher activity than their corresponding non-phenotypic counterparts. The 3′,4′,5′-trihydroxy structure further enhanced activity, with gallocatechin compounds showing higher activity than their corresponding catechin compounds. Epigallocatechin gallate esters, possessing three favorable structures—3-hydroxyl, 3′,4′,5′-trihydroxyl, and galloyl—exhibited the highest cardioprotective activity.

[0077] Example 5: Establishment and Validation of a High-Throughput Automated Screening Process

[0078] To achieve high-throughput automated screening, this embodiment establishes a standardized workflow that integrates multiple automated devices.

[0079] The equipment configuration includes a Hamilton Microlab STARlet eight-channel liquid handling workstation, a CellCelector automated cell culture system, a BioTek Synergy H1 multi-functional microplate reader, a MolecularDevices ImageXpress Micro Confocal high-content imaging system, a BD FACSCanto II flow cytometer, a ProteinSimple Jess fully automated Western blot system, and a LIMS data management system.

[0080] The standardized operating procedure is as follows. On the first day, the liquid handling workstation automatically seeded AC16 cells, adding 100 μL of cell suspension to each well of 10 96-well plates, at a cell density of 1 × 10⁶ cells / well. 5 1 × 10⁶ cells / mL, inoculated per well 4 960 wells containing 100 cells were cultured. The culture plates were automatically transferred to an automated incubator and incubated for 24 hours. On the second day, the liquid handling workstation automatically added compounds, adding 480 compounds to be screened according to the set concentration gradient to the corresponding wells. Seven concentrations were tested for each compound, with two parallel wells for each concentration, occupying a total of 672 wells. The remaining 288 wells were used for the control group and standards. After 2 hours of compound pretreatment, isoproterenol was automatically added to bring the final concentration to 200 μM. The culture plates were automatically transferred to the incubator for another 24 hours of incubation. On the third day, the liquid handling workstation automatically added cell viability assay reagents, adding 20 μL of CCK-8 solution to each well. After 2 hours of incubation, a multi-mode microplate reader automatically read the absorbance values, and the data was automatically imported into the analysis software to calculate cell viability and EC50. 50 Value. On the fourth day, a selective second test was conducted on EC. 50 Compounds with concentrations less than 10 μM were used to detect apoptosis, PI3K / AKT pathway activity, and damage markers using a high-content imaging system and ELISA. The entire screening process, from cell seeding to data analysis, was completed within 96 hours.

[0081] To verify the stability and reproducibility of the system, ten catechin compounds with known activities were selected as standards and repeatedly tested in different batches, with each compound tested 10 times. Statistical analysis showed that the intra-batch coefficient of variation for cell viability ranged from 6.8% to 9.2%, with an average of 8.1%, and the inter-batch coefficient of variation ranged from 11.3% to 14.7%, with an average of 12.8%, both meeting the quality standards for high-throughput screening. EC 50The intra-batch coefficient of variation for the values ​​ranged from 8.5% to 11.2%, with an average of 9.7%, while the inter-batch coefficient of variation ranged from 12.8% to 15.6%, with an average of 14.1%. The intra-batch coefficient of variation for the overall activity score ranged from 7.2% to 10.5%, with an average of 8.9%, while the inter-batch coefficient of variation ranged from 10.6% to 14.2%, with an average of 12.5%.

[0082] To verify the accuracy of the system, 20 catechin compounds were selected and tested in parallel using both the screening system and traditional manual methods, and the results of the two methods were compared. Correlation analysis showed that the EC values ​​determined by the two methods were similar. 50 The correlation coefficient for the analytical parameters was 0.94, and the correlation coefficient for the overall activity score was 0.91, indicating good consistency between the automated system and the traditional method. The automated system's operation time was 96 hours, while the traditional method required 240 to 336 hours, representing a 2.5 to 3.5-fold increase in efficiency. The detection cost per sample for the automated system was 8.5 yuan, compared to 35 to 50 yuan for the traditional method, a cost reduction of 75% to 80%.

[0083] Example 6: Construction of Structure-Activity Relationship Database and Establishment of Activity Prediction Model

[0084] To achieve intelligent prediction of the cardioprotective activity of catechin compounds, this embodiment constructs a structure-activity relationship database and establishes a machine learning prediction model.

[0085] We collected structural and activity data from 126 catechin compounds reported in the literature and screened in this study, including EC. 50 Values ​​and overall activity scores were calculated. Compound structures were drawn using ChemDraw software and saved in MOL format. Dragon software was used to calculate 3615 molecular descriptors, including 0-dimensional descriptors (molecular weight, number of atoms, number of bonds), 1-dimensional descriptors (connectivity index, charge index), 2-dimensional descriptors (topological index, hydrophobicity index), 3-dimensional descriptors (geometric index, surface area, volume), and quantum chemical descriptors (HOMO energy, LUMO energy, dipole moment, etc.).

[0086] Correlation analysis was used to screen descriptors significantly associated with cardioprotective activity, with Pearson correlation coefficients greater than 0.6 and p-values ​​less than 0.01. Analysis of variance was performed to remove descriptors with variances below a threshold to avoid overfitting. Collinearity analysis was used to remove highly correlated descriptors with correlation coefficients greater than 0.9, retaining independent characteristic variables. Finally, 38 key descriptors were selected, including molecular weight, number of hydrogen bond donors, number of hydrogen bond acceptors, number of rotatable bonds, topological polar surface area, octanol-water partition coefficient logP, total number of hydroxyl groups, number of aromatic hydroxyl groups, number of aliphatic hydroxyl groups, number of galloyl groups, number of stereochemical centers, number of aromatic rings, maximum interatomic distance, molecular dipole moment, HOMO energy, and LUMO energy.

[0087] 126 compounds were randomly divided into a training set of 88 compounds and a test set of 38 compounds in a 7:3 ratio. A nonlinear regression model was built using the random forest algorithm to predict the comprehensive activity score of the compounds. Random forest is an ensemble learning method that constructs multiple decision trees and votes or averages their predictions, exhibiting strong generalization ability and resistance to overfitting.

[0088] Model parameter optimization employed a grid search combined with five-fold cross-validation. The search space included a decision tree count of 100 to 500 with a step size of 50, a maximum depth of 5 to 20 with a step size of 5, a minimum number of splits of 2 to 10 with a step size of 2, and a minimum number of leaf nodes of 1 to 5 with a step size of 1. Cross-validation results showed that the optimal parameter combination was a decision tree count of 300, a maximum depth of 15, a minimum number of splits of 4, and a minimum number of leaf nodes of 2.

[0089] The model's fit on the training set was as follows: correlation coefficient R² = 0.92, root mean square error (RMSE) = 5.8, and mean absolute error (MAE) = 4.3. The model's prediction results on the test set were: correlation coefficient R² = 0.87, root mean square error (RMSE) = 7.2, and mean absolute error (MAE) = 5.6. These results indicate that the model has good fitting and predictive capabilities.

[0090] Feature importance analysis revealed that the 10 descriptors contributing most to the model were, in descending order: total hydroxyl number (11.8%), galloyl group number (9.6%), aromatic hydroxyl number (8.2%), hydrogen bond donor number (7.5%), topological polar surface area (6.9%), molecular weight (5.8%), logP (5.3%), aromatic ring number (4.7%), LUMO energy (4.2%), and rotatable bond number (3.9%). These results reveal key structural features influencing the cardioprotective activity of catechin compounds.

[0091] The established model was used to predict the activity of 20 unverified catechin derivatives. The model predicted that 5 compounds would have a comprehensive activity score greater than 80, 10 compounds would have a score between 60 and 80, and 5 compounds would have a score less than 60. Experimental verification was performed on the 5 compounds predicted to have high activity. The measured comprehensive activity scores were 83.2, 81.5, 78.6, 76.3, and 74.8, respectively, with errors of 2.1%, 3.5%, 6.8%, 8.2%, and 10.5% compared to the predictions. The average error was 6.2%, demonstrating that the model has good predictive accuracy.

[0092] The established structure-activity relationship database includes compound names, chemical structures, molecular formulas, molecular weights, CAS numbers, 38 molecular descriptors, and experimental activity data (EC). 50The database provides information such as activity scores, predicted activity data, 95% confidence intervals, and literature sources. It utilizes a MySQL relational database management system and offers a web-based query and retrieval interface, supporting structure search, substructure search, similarity search, and multi-condition combined search. The database is open to authorized users and provides data support for the research and development of catechin compounds.

[0093] Example 7 Confirmatory Study: Correlation between Screening Methods and Animal Models

[0094] To verify the effectiveness of the screening method of the present invention, this embodiment evaluates the efficacy of the highly active compounds obtained by screening in an animal model and compares the correlation between the in vitro screening results and the in vivo efficacy.

[0095] Seven catechin compounds with different activities from Example 4 were selected to evaluate their cardioprotective effects in an isoproterenol-induced mouse myocardial injury model. Eight-week-old male ICR mice, weighing 30-35g, were randomly divided into nine groups of ten mice each, including a normal control group, a model control group, and seven compound-administered groups. Mice in the compound-administered groups were administered the corresponding compound by gavage daily at a dose of 50 mg / kg body weight for seven consecutive days. On days 6 and 7, mice in the model control group and compound-administered groups were subcutaneously injected with isoproterenol at a dose of 85 mg / kg body weight, 24 hours apart, to establish the myocardial injury model. The normal control group was injected with an equal volume of physiological saline.

[0096] On day 8, serum was collected from mice, and the levels of lactate dehydrogenase, creatine kinase MB isoenzyme, and cardiac troponin I were detected by ELISA. The results showed that the lactate dehydrogenase activity in the model control group was 2856 U / L, while in the seven compound groups it decreased to 2385 U / L, 2156 U / L, 1923 U / L, 1658 U / L, 1782 U / L, 1521 U / L, and 1389 U / L, respectively. The concentration of creatine kinase MB isoenzyme in the model control group was 158.6 ng / mL, while in the seven compound groups it decreased to 135.2 ng / mL, 121.8 ng / mL, 106.5 ng / mL, 91.2 ng / mL, 98.6 ng / mL, 83.5 ng / mL, and 76.8 ng / mL, respectively. The concentration of cardiac troponin I was 12.35 ng / mL in the model control group, and decreased to 10.58 ng / mL, 9.42 ng / mL, 8.15 ng / mL, 6.93 ng / mL, 7.48 ng / mL, 6.21 ng / mL, and 5.67 ng / mL in the seven compound groups, respectively.

[0097] Mice were euthanized, and their hearts were removed for histological examination. Heart tissue was fixed in 10% neutral formaldehyde, embedded in paraffin, sectioned, stained with hematoxylin and eosin (HE), and observed under a light microscope. The myocardial tissue in the model control group showed significant pathological changes, including disordered myocardial fiber arrangement, myocardial fiber breakage, inflammatory cell infiltration, and hemorrhage. The degree of myocardial pathological damage was reduced to varying degrees in the seven compound-treated groups. The epigallocatechin gallate group showed essentially normal myocardial tissue structure with only mild inflammatory cell infiltration. Quantitative evaluation was performed using a pathological scoring system. The model control group scored 8.2 points, while the scores for the seven compound groups were 6.8, 6.1, 5.3, 4.5, 4.9, 3.8, and 3.2 points, respectively.

[0098] Myocardial infarction area was determined using the TTC staining method. The heart was transversely sliced ​​into five pieces, each approximately 2 mm thick, and placed in a 1% TTC solution for staining at 37°C for 20 minutes. Normal myocardial tissue stained red, while the infarcted area appeared pale white. After photographing, the percentage of infarct area to total ventricular area was measured using ImageJ software. The myocardial infarction area in the model control group was 41.8%, which decreased to 36.2%, 32.5%, 28.7%, 24.3%, 26.1%, 21.5%, and 18.9% in the seven compound groups, respectively.

[0099] Total protein was extracted from cardiac tissue, and the expression of proteins related to the PI3K / AKT signaling pathway was detected by Western blot. The results showed that the expression levels of p-AKT, p-PI3K, and p-GSK3β were significantly reduced in the model control group. The expression of these phosphorylated proteins was upregulated to varying degrees in the seven compound administration groups, with the epigallocatechin gallate group showing the most significant upregulation, consistent with the results of in vitro experiments.

[0100] Correlation analysis showed that the Pearson correlation coefficient between the comprehensive activity score obtained from in vitro screening and the reduction rate of myocardial infarction area in vivo was 0.89 (P < 0.001); the correlation coefficient with the reduction rate of serum lactate dehydrogenase was 0.86; the correlation coefficient with the reduction rate of creatine kinase MB isoenzyme was 0.88; the correlation coefficient with the reduction rate of cardiac troponin I was 0.87; and the correlation coefficient with the reduction rate of pathological score was 0.85. These results indicate that the in vitro high-throughput screening method established in this invention has a good correlation with in vivo efficacy and can effectively predict the cardioprotective effect of compounds in animal models, verifying the reliability and practical value of the screening method.

[0101] The above embodiments demonstrate that the high-throughput screening method for the anti-myocardial injury activity of catechin compounds established in this invention has good sensitivity, specificity, repeatability, and predictability, and can effectively screen out compounds with cardioprotective activity, providing an efficient technical platform for the development of new cardioprotective drugs.

Claims

1. A high-throughput screening method for the anti-myocardial injury activity of catechin compounds, characterized in that... Includes the following steps: AC16 human cardiomyocytes were used at 5 × 10 3 Up to 1×10 4 Cells were seeded at a density of 10 cells / well in 96-well cell culture plates and cultured in DMEM / F12 medium containing 12.5% ​​fetal bovine serum, 2 mM glutamine and 1% penicillin-streptomycin at 37°C in a 5% CO2 incubator for 24 hours to allow cell adhesion. After culturing for another 24 hours to reach the stable growth phase, the cells were subjected to serum starvation for 2 hours. The catechin compounds to be screened were diluted to working concentrations with serum-free DMEM / F12 medium and added to cell wells for pretreatment for 2 hours. The final concentrations of the compounds were one or more of the following: 0.1 μM, 0.3 μM, 1 μM, 3 μM, 10 μM, 30 μM, and 100 μM. After pretreatment, serum-free culture medium containing 100 μM isoproterenol and corresponding concentrations of compounds was added, and the treatment lasted for 24 hours to establish a myocardial injury model. A normal control group and an injury model control group were also set up. Cell viability was detected using the CCK-8 assay. Cell viability was calculated by measuring absorbance at 450 nm. Nonlinear regression analysis was used to establish dose-response curves and calculate the half-maximal effective concentration (MCP). 50 value; Apoptosis was detected using Annexin V-FITC / PI double staining. Fluorescence images were acquired using a fluorescence microscope or high-content imaging system. Annexin V-positive and PI-positive cells were automatically identified and counted using image analysis software, and the apoptosis rate and apoptosis inhibition rate were calculated. The mRNA expression levels of atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and β-myosin heavy chain β-MHC were detected using real-time quantitative PCR. - The relative expression level was calculated using the ΔΔCt method, and the inhibition rate of the compound on the expression of myocardial injury markers was calculated. Western blot was used to detect the phosphorylation levels of PI3K and AKT proteins, and the p-PI3K / PI3K and p-AKT / AKT ratios were calculated to evaluate the activation effect of the compounds on the PI3K / AKT signaling pathway. Immunofluorescence staining was performed using wheat germ lectin (WGA) to label cell membranes and DAPI to label cell nuclei. Fluorescence images were acquired using a high-content imaging system. A deep learning-based intelligent image analysis algorithm was used to automatically identify cell outlines and calculate cell surface area to evaluate the inhibitory effect of the compound on cell hypertrophy. The JC-1 staining method was used to detect mitochondrial membrane potential, and the ratio of red fluorescence intensity to green fluorescence intensity was calculated to evaluate the protective effect of the compound on mitochondrial membrane potential. Intracellular reactive oxygen species (ROS) levels were detected using the DCFH-DA fluorescent probe method, and the inhibitory effect of compounds on ROS generation was evaluated by measuring fluorescence intensity. Based on the above multi-parameter detection results, a comprehensive activity scoring system was established to calculate the comprehensive protection score of the compound. The comprehensive protection score is equal to 0.25 multiplied by the normalized value of cell survival rate plus 0.20 multiplied by the normalized value of apoptosis inhibition rate plus 0.15 multiplied by the normalized value of myocardial injury marker inhibition rate plus 0.20 multiplied by the normalized value of PI3K / AKT pathway activation plus 0.10 multiplied by the normalized value of cell morphology protection plus 0.05 multiplied by the normalized value of mitochondrial membrane potential protection rate plus 0.05 multiplied by the normalized value of ROS inhibition rate. The activity level of the compound is determined according to the comprehensive protection score. The comprehensive protection score is greater than 70 points and is determined to be a high-activity compound, a score of 40 to 70 points and is determined to be a medium-activity compound, and a score less than 40 points and is determined to be a low-activity compound.

2. The method according to claim 1, characterized in that, The catechin compounds are selected from one or more of catechin C, epicatechin EC, gallocatechin GC, epigallocatechin EGC, catechin gallate CG, epicatechin gallate ECG, gallocatechin gallate GCG, and epigallocatechin gallate EGCG.

3. The method according to claim 1, characterized in that, Further, structural analysis is performed on the screened active compounds, and molecular descriptors are calculated, including one or more of the following: molecular weight, lipid-water partition coefficient logP, number of hydrogen bond donors, number of hydrogen bond acceptors, number of rotatable bonds, polar surface area, number of aromatic rings, number of hydroxyl groups, number of gallic ester groups, and stereoisomer types. A quantitative structure-activity relationship (QSAR) model is established between the compound structural parameters and the activity score. The QSAR model is established using one of the following methods: multiple linear regression, partial least squares regression, or support vector machine regression.

4. The method according to claim 1, characterized in that, The AC16 cells were passaged from the 5th to the 10th generation, and passaged when the cell confluence reached 80% to 90%. The cells were digested with 0.25% trypsin-EDTA solution for 3 to 5 minutes during passage.

5. The method according to claim 1, characterized in that, The conditions for establishing the isoproterenol-induced myocardial injury model were as follows: the final concentration of isoproterenol was 100 μM, the treatment time was 24 hours, and under these conditions, cell viability decreased to 40% to 50% of the normal control group, the apoptosis rate increased to 20% to 30%, the mRNA expression of ANP, BNP, and β-MHC was upregulated by 3 to 5 times, and the phosphorylation levels of PI3K and AKT decreased by 50% to 60%.

6. The method according to claim 1, characterized in that, The formula for calculating cell viability is: Cell viability equals experimental group OD 450 Subtract blank hole OD 450 The difference, divided by the OD of the normal control group 450 Subtract blank hole OD 450 The difference is multiplied by 100%; the apoptosis rate is calculated as follows: the apoptosis rate equals the sum of the number of Annexin V positive cells and the number of Annexin V / PI double positive cells, divided by the total number of cells, and then multiplied by 100%.

7. The method according to claim 1, characterized in that, The intelligent image analysis algorithm is based on a deep learning convolutional neural network architecture and can automatically identify cell nuclei, delineate cell membrane outlines, measure cell surface area, and identify abnormally shaped cells, with an accuracy rate of over 95%.

8. The method according to claim 1, characterized in that, The method employs a liquid workstation for automated sample addition, liquid exchange, and washing; a multi-functional microplate reader for automated absorbance and fluorescence intensity detection; and a high-content imaging system for automated image acquisition. The screening process for a single 96-well plate, from cell seeding to data analysis, takes 5 days, and a single experiment can simultaneously evaluate 480 compound samples.

9. The method according to claim 1, characterized in that, The screening system of the method has a quality evaluation index Z' factor greater than 0.5, an intra-batch coefficient of variation (CV) less than 10%, and an inter-batch coefficient of variation (CV) less than 15%.

10. The application of the high-throughput screening method for the anti-myocardial injury activity of catechin compounds according to any one of claims 1 to 9 in the development of cardiovascular protective drugs, wherein the application includes rapidly screening lead compounds with cardioprotective activity from natural product extracts, synthetic compound libraries or virtual screening hit compounds, evaluating the mechanism of action of the compounds, optimizing the compound structure to improve activity and selectivity, and providing candidate compounds and development strategies for the development of therapeutic drugs targeting myocardial injury related to β-adrenergic overactivation.