A method for cardiovascular disease risk assessment based on biomarkers

Abstract

The application relates to the field of biomedical detection, and particularly discloses a cardiovascular disease risk assessment method based on biomarkers, which comprises the following steps: acquiring endothelial glycocalyx integrity, mechanical sensitive ion channel activity regulation and mitochondrial energy metabolism coupling efficiency data sets; quantifying the glycocalyx-cytoskeleton mechanical transduction decoupling degree, Piezo1 phosphorylation level, TRPV4 activation threshold shift and mitochondrial respiratory chain activity under Piezo1 activation state, and constructing a mechanical perception-energy metabolism coupling function damage index; inputting the index into a decompensation early warning model, calculating a decompensation critical time, and outputting an early warning atlas containing damage subtype classification; and according to the intervention window width and damage subtype matching glycocalyx remodeling, channel activity regulation or mitochondrial energy rescue strategy. The application realizes mechanism evaluation, space-time prediction and precise intervention matching of vascular endothelial damage.

Description

Technical Field

[0001] This invention relates to the field of biomedical detection technology, and in particular to a method for cardiovascular disease risk assessment based on biomarkers. Background Technology

[0002] Cardiovascular diseases, especially atherosclerotic cardiovascular diseases, are fundamentally pathologically rooted in the instability of the vascular microenvironment caused by damage to vascular endothelial cell function. Endothelial cells, as the primary receptors of hemodynamic signals, depend on the coordinated operation of three key pathways for functional integrity: glycocalyx-mediated mechanosensing, signal transduction through mechanosensitive ion channels such as Piezo1 / TRPV4, and the response of mitochondrial energy metabolism to mechanosensing signals. These three pathways constitute a complete mechanosensing-energy metabolism coupled pathway; abnormalities in any of these pathways can lead to endothelial dysfunction.

[0003] However, existing risk assessment technologies have the following technical blind spots:

[0004] First, the detection dimensions are singular and static. Existing methods mostly rely on the detection of inflammatory cytokine concentrations such as CRP and IL-6, or lipid metabolism indicators such as LDL-C, which can only reflect systemic inflammatory burden or metabolic status, but cannot reveal the molecular mechanisms of vascular endothelial cell dysfunction. More importantly, these detections are all static indicators and cannot distinguish the functional state of proteins—for example, there is a significant difference between the expression level of Piezo1 channels and the mechanosensitive regulatory state represented by the phosphorylation modification level of serine at position 1612, but existing technologies completely ignore this dimension.

[0005] Second, there is a lack of a systemic assessment perspective. Endothelial dysfunction is a systemic imbalance in the sensing-transduction-effect cascade pathway, rather than an isolated abnormality of a single indicator. Glycocalyx degradation can lead to decreased efficiency of mechanotransmission, thereby exacerbating Piezo1 channel dysfunction; Piezo1 dysfunction can in turn directly affect mitochondrial respiratory chain activity, forming a vicious cycle. Current technologies lack assessment tools to quantify the degree of this multi-stage synergistic deviation.

[0006] Third, it is impossible to predict the critical time of decompensation. The core issue in clinical decision-making is when irreversible damage may occur in a patient, but current methods can only provide the current risk level. Recent studies have shown that the synergistic activation of TRPV4 and TGFβ1 is a key mechanism driving endothelial-mesenchymal transition by abnormal matrix stiffness, and abnormal Piezo1 phosphorylation can lead to loss of mechanosensory function. The two together determine the rate of acceleration of decompensation, but this has not yet been translated into a clinically usable predictive model.

[0007] To address the aforementioned technological gaps, this invention proposes a biomarker-based cardiovascular disease risk assessment method. This method integrates glycocalyx mechanotransduction efficiency, Piezo1 / TRPV4 channel functional status, and mitochondrial energy metabolism efficiency to construct a mechanosensory-energy metabolism coupled functional impairment index. The decompensation critical time is calculated using TRPV4-TGFβ1 co-activation intensity and the degree of Piezo1 phosphorylation abnormality as accelerating factors. Furthermore, differentiated and precise intervention strategies are matched based on the injury subtype and the intervention window width. Summary of the Invention

[0008] The purpose of this invention is to solve the technical problems in the prior art that it is impossible to assess vascular endothelial function damage from the perspective of mechanical sensing-energy metabolism coupling and that it is impossible to predict the critical time of decompensation. Therefore, this invention proposes a cardiovascular disease risk assessment method based on biomarkers.

[0009] To achieve the above objectives, the present invention adopts the following technical solution: A biomarker-based method for assessing cardiovascular disease risk includes: S1. Obtain data on endothelial glycocalyx integrity, mechanosensitive ion channel activity regulation, and mitochondrial energy metabolism coupling efficiency of the target individual, and construct the vascular endothelial mechanosensing-energy metabolism coupling function damage index through nonlinear fusion. S2. Input the damage index into the vascular microenvironment mechanical decompensation early warning model, combine local hemodynamic parameters and matrix stiffness data, identify high-risk vascular segments and calculate the decompensation critical time, and output a risk decompensation early warning map containing the location of high-risk segments and the critical time. S3. Calculate the biomechanical intervention window width based on the decompensation critical time in the early warning map, and match differentiated vascular biomechanical intervention strategies according to the window width and the damage subtype classification corresponding to the damage index, and output a precise prevention strategy plan.

[0010] As a further technical solution of the present invention, S1 specifically includes: S11. The thickness of the endothelial glycocalyx in the sublingual microcirculation was measured using lateral flow dark-field imaging technology, and the glycocalyx thickness decay rate per unit time was calculated. The concentration of heparan sulfate chain shedding in plasma was detected by immunofluorescence. The activation efficiency of endothelial nitric oxide synthase under mechanical stimulation and the colocalization coefficient of syndecan-1-actin were measured to quantify the degree of decoupling of glycocalyx-cytoskeleton mechanical transduction. S12. Calculate the shear stress response index, mechanical sensitivity adjustment index, and TRPV4 activation threshold shift index of the Piezo1 channel respectively, and calculate the mechanical sensing distortion index by combining the three indices. S13. Calculate five indices: transfer rate, mitochondrial membrane potential retention rate, metabolic rescue efficiency, complex activity change factor and ATP yield increase factor. After normalizing the five indices, perform a weighted summation to output the mitochondrial energy metabolism coupling efficiency index. S14. Input the degree of decoupling of glycocalyx-cytoskeleton mechanotransduction, mechanosensory distortion index and mitochondrial energy metabolism coupling efficiency index into a nonlinear fusion model based on random forest regression, and output the vascular endothelial mechanosensory-energy metabolism coupling function impairment index.

[0011] As a further technical solution of the present invention, S2 specifically includes: S21. The vascular endothelial mechanosensing-energy metabolism coupling function damage index is spatially mapped along the vascular tree. Computational fluid dynamics simulation is used to obtain the time average value of local wall shear stress and oscillatory shear index. Combined with the extracellular matrix stiffness distribution data measured in situ by atomic force microscopy, the gradient change rate of the damage index in different vascular segments is analyzed by gradient boosting algorithm. Segments with a gradient change rate of the damage index exceeding a preset threshold are identified as candidate high-risk vascular segments. S22. Calculate the synergistic activation intensity, the degree of loss of mechanosensitivity, and the EndMT conversion index respectively. Perform multiple linear regression analysis on the three indices to establish a quantitative scoring model for the severity of local vascular segment injury. S23. Based on time series analysis, construct a dynamic model of the evolution of the damage index over time, and calculate the critical time of decompensation from the current state to the appearance of positive expression of the mesenchymal marker α-SMA in endothelial cells; S24. Integrating the three-dimensional coordinate positioning of high-risk segments, the estimated critical time of decompensation, the change in damage index per unit time as the rate of deterioration of the mechanical microenvironment, and the damage subtype classification obtained by clustering the co-activation intensity and the degree of loss of mechanical sensitivity based on principal component analysis, a risk decompensation early warning map with spatial resolution at the micrometer level is generated.

[0012] As a further technical solution of the present invention, S3 specifically includes: S31. Calculate the difference between the estimated critical time of decompensation in the risk decompensation early warning map and the current time point to obtain the width of the biomechanical intervention window, and at the same time extract the rate of deterioration of the mechanical microenvironment in the early warning map as an indicator of intervention urgency. S32. When the width of the intervention window is greater than a preset first threshold and the rate of deterioration is less than a preset second threshold, a matching glycocalyx remodeling strategy is applied. The strategy includes intravenous infusion of the glycocalyx synthesis precursor N-acetylglucosamine to promote glycocalyx regeneration, combined application of the heparan sulfate degrading enzyme inhibitor suramin sodium to slow down glycocalyx shedding, and the Piezo1 channel mechanosensitive stabilizer glycine to maintain the channel's normal phosphorylation level. S33. When the width of the intervention window is less than the first threshold but the rate of deterioration is greater than the second threshold, differential matching is performed according to the damage subtype classification in the warning map. If it is the Piezo1 channel phosphorylation abnormality-dominant type, it is matched with protein kinase A agonist to regulate the phosphorylation state of serine at position 1612; if it is the TRPV4 channel activation threshold shift-dominant type, it is matched with the TRPV4 antagonist GSK2193874 combined with the lysyl oxidase inhibitor β-aminopropionitrile to inhibit the EndMT process. S34. When the width of the intervention window is less than the preset third threshold, a matching cell energy rescue strategy is adopted, which includes ultrasound microbubble-mediated exogenous mitochondrial targeted transplantation, tunnel nanotube formation promoter F-actin stabilizer Jasplakinolide, and Piezo1 channel agonist Yoda1 to enhance the activity of mitochondrial respiratory chain complex. S35. Integrate the above matching results and output a precise prevention strategy plan that includes the type of intervention strategy, specific drug combination, route of administration, and timing of intervention.

[0013] The beneficial effects of this invention are as follows: 1. This invention, from the perspective of the complete cascade pathway of mechanotransduction-energy metabolism response, constructs a ternary coupling assessment system by nonlinearly fusing the degree of decoupling of glycocalyx-cytoskeleton mechanotransduction, the functional status of Piezo1 / TRPV4 channels, and mitochondrial energy metabolism efficiency. By calculating Mahalanobis distance to quantify the degree of synergistic deviation of the three indicators, it can identify high-risk individuals with normal individual indicators but system coupling imbalance, filling the detection gap in the blind spot of existing technologies regarding the mechanism of vascular endothelial dysfunction.

[0014] 2. This invention introduces a survival analysis model at the cellular and molecular level, using the co-activation intensity of TRPV4-TGFβ1 and the degree of loss of mechanosensitivity due to abnormal Piezo1 phosphorylation as accelerating factors to establish a Weibull regression model to predict the critical time of decompensation. Combined with the spatial gradient distribution of the damage index and local hemodynamic parameters, it achieves millimeter-level localization of high-risk vascular segments. This technology elevates risk assessment from qualitative grading to quantitative timing, providing clinicians with two-dimensional decision-making information on where problems will occur and when they will occur.

[0015] 3. This invention establishes a three-tiered progressive matching mechanism based on damage subtype classification and the width of the biomechanical intervention window: wide windows are matched with glycocalyx remodeling strategies; tight windows are matched with channel activity regulation or EndMT inhibition strategies based on subtype differences; and extremely narrow windows are matched with mitochondrial energy rescue strategies. This mechanism-guided system achieves a leap from fuzzy empirical treatment to molecularly targeted repair, providing technical support for personalized and precise prevention of cardiovascular diseases. Attached Figure Description

[0016] Figure 1 This is a flowchart of a biomarker-based cardiovascular disease risk assessment method proposed in Embodiment 1 of the present invention; Figure 2 This is a framework diagram of a cardiovascular disease risk assessment system based on biomarkers proposed in Embodiment 2 of the present invention. Detailed Implementation

[0017] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0018] Example 1 Please see the appendix Figure 1 A biomarker-based method for assessing cardiovascular disease risk includes: S1. Obtain data on endothelial glycocalyx integrity, mechanosensitive ion channel activity regulation, and mitochondrial energy metabolism coupling efficiency of the target individual, and construct a vascular endothelial mechanosensing-energy metabolism coupling functional impairment index through nonlinear fusion; specifically including: S11. Acquisition of dynamic datasets on endothelial glycocalyx integrity and quantification of the degree of decoupling between glycocalyx and cytoskeleton mechanotransduction, specifically: S111. The thickness of the glycocalyx of the sublingual microcirculation endothelium was determined using lateral flow dark field imaging technology. A sublingual microcirculation observation device equipped with a lateral flow dark field imaging light source was used. The probe was gently placed on the surface of the sublingual mucosa of the subject, and the focus was adjusted to clearly image the capillaries. According to the principle of dark field imaging, the glycocalyx layer appears as a dark edge band of blood vessel edges due to the light scattering characteristics, while the red blood cell columns appear as bright lines.

[0019] Using professional image analysis software, grayscale intensity curve analysis was performed along the vertical direction of the long axis of the blood vessel to identify the difference region between the erythrocyte column width and the total blood vessel width; assuming the erythrocyte column width is... The total vascular width is Then the instantaneous thickness of the sugar calyx The calculation is as follows: .

[0020] Microvessels were continuously observed in at least five different fields of view, with dynamic images recorded for more than ten seconds in each field of view. The decay rate of glycocalyx thickness per unit time was calculated, which reflects the degree of imbalance between dynamic shedding and regeneration of glycocalyx under physiological conditions.

[0021] S112. The concentration of heparan sulfate chain detachment in plasma was detected by immunofluorescence; 3 mL of fasting peripheral venous blood was collected from the subject and placed in a blood collection tube containing ethylenediaminetetraacetic acid (EDTA) anticoagulant, and subjected to relative centrifugation. Centrifuge for 15 minutes to separate the upper plasma layer.

[0022] Take plasma samples Add the antibody to the wells of an ELISA plate coated with heparan sulfate-specific monoclonal antibody and incubate overnight at 4°C. After washing, add horseradish peroxidase-labeled secondary antibody and incubate at room temperature for 1 hour. Add tetramethylbenzidine chromogenic solution, react in the dark for 15 minutes, then add stop solution and measure the wavelength using an ELISA reader. absorbance value at .

[0023] According to the A series of concentrations of heparan sulfate standards and their corresponding absorbance values Establishing a standard curve A four-parameter logistic regression model was used for fitting: ,in: The half-maximal effect concentration, The Hill coefficient is used. These are the minimum and maximum asymptotic absorbance values, respectively. The measured absorbance values ​​will be... Substitute into the fitted curve The plasma heparin sulfate chain shedding concentration was calculated by reverse calculation. This elevated concentration indicates disruption of the integrity of the calyx structure.

[0024] S113. Human primary umbilical vein endothelial cells were isolated and cultured, and seeded in elastic-bottom culture plates containing fibronectin at a cell density of [missing information]. cells The experimental group was subjected to laminar shear stress. Stimulation duration The control group was kept in static culture. .

[0025] Immediately after stimulation, the sample was washed with pre-cooled phosphate buffer and immunofluorescence stained with rabbit anti-human endothelial-nitric oxide synthase serine phosphorylation at position 1177 (anti-p-eNOS-S1177). Counterstaining was performed using mouse anti-human total endothelial-nitric oxide synthase antibody (anti-eNOS). Three-dimensional tomographic images were acquired using a laser scanning confocal microscope, with excitation wavelengths of [insert wavelengths here]. (Corresponding phosphorylated antibody) and (Corresponding to total protein antibody), the emission wavelengths are respectively and After background subtraction and photobleaching correction of the image, the integral value of phosphorylated endothelial nitric oxide synthase fluorescence intensity of a single cell was calculated. The fluorescence intensity integral value of total endothelial nitric oxide synthase Activation efficiency of endothelial nitric oxide synthase Defined as: ,in: To analyze the total number of cells , For the first The integral value of fluorescence intensity of phosphorylated endothelial nitric oxide synthase in individual cells. For the first The integral value of total endothelial nitric oxide synthase fluorescence intensity per cell, reference value. and The ratio of standard fluorescence intensity of healthy control cells under physiological shear stress stimulation. The diagnosis was endothelial nitric oxide synthase activation dysfunction.

[0026] S114. Simultaneously, mouse anti-human syndecan-1 extracellular domain antibody was used to label the glycocalyx core protein, and filamentous actin (F-actin) was labeled with Alexa Fluor 647-labeled phalloidin. Excitation wavelengths were as follows: (syndecan-1) and (F-actin), emission wavelengths are respectively and Spatial registration and background homogenization were performed on the dual-channel images. The Mendes co-localization coefficient algorithm was used to calculate the spatial overlap between syndecan-1 and F-actin. Let the pixel intensity distribution of the syndecan-1 channel be... The pixel intensity distribution of the F-actin channel is as follows The intensity of overlapping pixels is Then the Mendes M1 coefficient (syndecan-1 overlap ratio) is: The Mendes M2 coefficient (F-actin overlap ratio) is: Comprehensive colocation coefficient Using the geometric mean of M1 and M2: ;

[0027] The value ranges from 0 to 1. The closer the value is to 1, the greater the spatial overlap between syndecan-1 and actin, and the tighter the mechanical connection between the glycocalyx and the cytoskeleton. The diagnosis was glycocalyx-cytoskeleton decoupling.

[0028] S115. Calculate the degree of decoupling of glycosidic-cytoskeleton mechanotransduction by combining the decrease in endothelial nitric oxide synthase activation efficiency and the decrease in colocalization coefficient. ; Let the healthy control reference value be and ,but: The weighting coefficient This reflects the dominant role of endothelial nitric oxide synthase in mechanotransduction. The value ranges from 0 to 1. The closer the value is to 1, the more severe the decoupling of the glycocalyx-cytoskeleton mechanical transduction. The diagnosis indicated significant decoupling of mechanical conduction, suggesting that the loss of calyx structural integrity led to the interruption of mechanical force transmission.

[0029] S12. Acquisition of the dataset on the regulation of activity of mechanosensitive ion channels and calculation of the mechanosensing distortion index, specifically:

[0030] S121. Microfluidic chips are fabricated using polydimethylsiloxane material via soft photolithography; firstly, a master mold is fabricated on a silicon wafer using photolithography, and a parallel flat plate flow cavity structure is designed, with the cavity length... ,width ,high The polydimethylsiloxane prepolymer and curing agent were mixed in a mass ratio... After mixing and vacuum degassing, the mixture is poured onto a master mold and cured at 80 degrees Celsius for two hours. The cured polydimethylsiloxane layer is then peeled off, and inlet / outlet apertures are prepared using a punch. After being treated with oxygen plasma, it is bonded to a clean coverslip; the surface of the coverslip at the bottom of the cavity is treated with a high concentration of... Type I collagen solution was used to coat cells for two hours to promote cell adhesion. Human umbilical vein endothelial cells were then coated at a density... Inoculated at the bottom of the flow chamber, and cultured in endothelial cell culture medium (containing 5% fetal bovine serum and 1% endothelial cell growth additive) at 37°C and 5% CO2. The cells were allowed to form a fused monolayer, and the degree of fusion was confirmed to be greater than 95% by microscopic observation.

[0031] An external peristaltic pump connects the chip's inlet and outlet, precisely controlling the fluid flow to generate a specified shear stress; wall shear stress. With volumetric flow rate The relationship is: ,in: The dynamic viscosity of the culture medium. Two shear stress conditions were applied, each lasting for 600 seconds:

[0032] Physiological laminar flow conditions: steady-state laminar flow, wall shear stress ; Calculate the required volumetric flow rate: ;

[0033] Pathological turbulent flow conditions: oscillating flow, mean wall shear stress oscillation amplitude dyne oscillation frequency Volumetric flow rate changes over time: ,in: for Pathological turbulent volumetric flow rate at any given time.

[0034] Before the experiment, cells were loaded with the calcium ion fluorescent indicator Fluo-4; Fluo-4 acetoxymethyl ester working solution was prepared at the following concentration. It contains 0.02% Prönkel F-127 to increase cell permeability. The chip is placed in a 37°C incubator for loading. The extracellular fluorescent dye was then removed by washing three times with phenol red-free medium. An inverted fluorescence microscope equipped with a high-speed electron multiplication charge-coupled device camera was used, and the excitation source was an argon-ion laser. The center wavelength of the emission filter ,bandwidth Applying shear stress while maintaining frame rate Acquiring fluorescence images, exposure time Image resolution Pixels, pixel size .

[0035] Background subtraction and photobleaching correction were performed on the acquired fluorescence image sequence. The fluorescence intensity of a certain cell region at time is Baseline fluorescence intensity The normalized change in calcium ion concentration is the average fluorescence intensity in the first 10 seconds after stimulation. for: Identify the peak amplitude of calcium transients; for physiological laminar flow conditions, the peak calcium response. During the stimulation period The maximum value; for pathological turbulent conditions, due to the oscillation characteristics, the average value of the peak response within the oscillation period is calculated. : ,in The total number of oscillation periods. For the first Peak response over one cycle.

[0036] The ratio of peak calcium concentration under two stress conditions was calculated as the shear stress response index of the Piezo1 channel. : ,in To avoid small constants that are divided by zero. This indicates an imbalance in the Piezo1 channel's ability to sense physiological laminar flow relative to its sensitivity to pathological turbulence, suggesting a dysfunction in mechanosensory function.

[0037] S122. Collect the endothelial cell lysate after shear stress stimulation. Immediately after stimulation, wash the chip chamber three times with phosphate-buffered saline pre-cooled to 4°C. The volume of wash solution each time is [volume missing]. Add a mixture containing protease inhibitors (1 mM benzyl sulfonyl fluoride, aprotinin). Leucogen inhibitory peptide ) and a mixture of phosphatase inhibitors (sodium fluoride 10 mM, sodium t-fluoride 1 mM, Radioimmunoprecipitation analysis of lysis buffer volume (10 mM glycerophosphate) Ice pyrolysis time During this period, the solution was blown up every five minutes. The lysate was collected into centrifuge tubes and then... Centrifugation time Collect the supernatant and determine the protein concentration using the diquinoline carboxylic acid method. .

[0038] Immunoprecipitation was performed using a rabbit monoclonal antibody targeting the phosphorylation of serine at position 1612 of Piezo1 (anti-p-Piezo1-S1612, clone number NBP2-75563). Total protein was collected. and Protein A / G agarose beads (pre-blocked with phosphate buffer) )and Phosphorylation-specific antibodies mixed, total volume 4℃ rotational incubation time The immunoprecipitated complex was collected by centrifugation and washed three times with low-salt wash buffer (containing 150 mM sodium fluoride) and once with high-salt wash buffer (containing 500 mM sodium chloride). The volume of wash buffer for each wash was [volume missing]. .join in Sample loading buffer (containing 2%) - Mercaptoethanol, 22% sodium alkyl sulfate), boiling water bath denaturation time .

[0039] The immunoprecipitated samples and the retained 2% total lysis buffer (as a total Piezo1 internal control) were separated by 8% polyacrylamide gel electrophoresis, with a stacking gel concentration of 5% and a constant current electrophoresis current. Electrophoresis time The wet transfer method was used to transfer the film to a polyvinylidene fluoride (PVDF) membrane, with the transfer current being... Duration Sealing time with 5% skim milk powder The samples were incubated at 4°C with anti-p-Piezo1-S1612 (immunoprecipitation sample) and mouse anti-human Piezo1 total protein monoclonal antibody (total lysis buffer, clone number NBP1-78624) primary antibody. Horseradish peroxidase-labeled secondary antibody incubation time at room temperature. Chemiluminescent substrates are used for color development, and images are acquired using a gel imaging system.

[0040] ImageJ software was used for band grayscale analysis. The grayscale value of the phosphorylated Piezo1 band was set as follows: The total Piezo1 band grayscale value is The background grayscale value is The corrected phosphorylation ratio for: ,in: This represents the total volume of the pyrolysis fluid. This refers to the volume of the immunoprecipitation sample.

[0041] The deviation of the phosphorylation ratio from the reference value for healthy controls was calculated as the mechanosensitivity modulation index. Let the reference value for the phosphorylation ratio in the healthy control group after being stimulated by the same physiological shear stress be . (Determined by a large sample of healthy individuals, typically) Then: Mechanosensitivity Adjustment Index ;when At this time, it indicates that the Piezo1 channel is overphosphorylated, resulting in abnormally increased mechanosensitivity; when This indicates insufficient phosphorylation of the Piezo1 channel and loss of mechanosensitivity. Both cases are considered abnormalities in mechanosensitivity regulation.

[0042] S123. Prepare mixtures of acrylamide and bisacrylamide in different proportions, and prepare a series of hardness gels by adjusting the mass ratio of the crosslinking agent bisacrylamide to the monomer acrylamide. Gel elastic modulus. The relationship with crosslinking density is as follows: ,in: For polymer density, The gas constant is Absolute temperature The average molecular weight between crosslinking points. Three hardness levels were prepared: low hardness... (Simulating the basement membrane of a healthy blood vessel), medium hardness (Simulated early hardening), high hardness (Simulating pathological fibrosis). The gel surface is treated with... Concentration of sulfosuccinimide hexahydropyridazine UV crosslinking (wavelength) ,energy ) treatment, followed by coating concentration Fibronectin.

[0043] Endothelial cells were seeded onto the three types of gels with the above-mentioned stiffness, and cultured for [time]. Continue until cells are fully spread. Prepare logarithmic gradient solutions of the TRPV4-specific agonist GSK1016790A, with concentration series as follows: Add the solutions of different concentrations to the cell culture wells sequentially, and set the treatment time for each concentration. Intermittent washing Calcium influx was measured using a fluorescence calcium imaging system. Cells were pre-loaded with Fluo-4, and the excitation wavelength was [not specified]. emission wavelength Fluorescence images were acquired during each concentration treatment period, and the peak calcium response at that concentration was calculated. (Percentage change in fluorescence intensity relative to baseline).

[0044] A four-parameter logistic model was used to fit the concentration-response data under different hardness conditions: ,in: The calcium response at concentration C, The minimum response (baseline). For maximum response, The half-maximal effect concentration, The Hill coefficient reflects the slope of the curve. Nonlinear least squares fitting is used to obtain the curve under various hardness conditions. Value: Low hardness Medium hardness High hardness .

[0045] Calculate the shift factor of the half-maximum effective concentration relative to the low-hardness healthy control matrix under medium and high hardness conditions. Define the TRPV4 activation threshold shift index. for: This index reflects the changes in the activation sensitivity of TRPV4 channels in pathological, high-hardness matrices. This indicates that the activation threshold of the TRPV4 channel is significantly reduced, meaning it can be activated at lower agonist concentrations, suggesting a structural functional abnormality in the channel and an imbalance in its sensitivity to changes in matrix stiffness.

[0046] S124. Piezo1 channel shear stress response index Mechanical sensitivity adjustment index and TRPV4 activation threshold offset index Input into the principal component analysis model.

[0047] Z-score standardization was applied to the three indices to eliminate differences in dimensions and variability. ,in and These are the group mean and standard deviation of the index (determined based on a database of healthy individuals). For the first The standardized scores of each index.

[0048] Constructing a standardized data matrix Calculate its covariance matrix Solve the eigenvalue equations: ,in For the first 1 eigenvalue, These are the corresponding eigenvectors; sort them by eigenvalue size and take the largest eigenvalue. Corresponding feature vector As the first principal component direction.

[0049] Calculate the first principal component score as the mechanical perceptual distortion index. : ,in: , , All are eigenvectors of the first principal component. This index This comprehensively reflects the overall mechanosensing function of the Piezo1 channel, reflecting its ability to sense shear stress, its phosphorylation modification status, and the TRPV4 channel's sensitivity shift to matrix hardness. A threshold is set. ,when The time interval indicates significant impairment of mechanosensory function, suggesting a comprehensive dysfunction of the endothelial cell mechanosensory system.

[0050] S13. Acquisition of the mitochondrial energy metabolism coupling efficiency dataset and construction of the mitochondrial energy metabolism coupling efficiency index are as follows:

[0051] S131. Endothelial progenitor cells labeled with DiR fluorescent dye were co-cultured with damaged endothelial cells. A confocal live-cell imaging system was used to continuously image the cells for 6 hours at 37°C and 5% CO2, with images acquired every 5 minutes. The formation of tunnel nanotubes and the process of mitochondrial transfer were tracked using ImagePro software. The percentage of endothelial progenitor cells that successfully transferred mitochondria per unit time was calculated as the transfer rate.

[0052] Collect peripheral venous blood volume of the target individual The blood was placed in a blood collection tube containing ethylenediaminetetraacetic acid (EDTA) anticoagulant. Peripheral blood mononuclear cells were separated using density gradient centrifugation with Ficoll-Paque Plus separation buffer and relative centrifugal force. Centrifugation time ,temperature Mononuclear cell layers were collected and washed twice with phosphate-buffered saline (PFS). Endothelial progenitor cells (EPCs) were isolated using immunomagnetic bead sorting. CD34 microbeads were used for positive sorting, followed by CD133 microbeads for secondary positive sorting, yielding CD34+CD133+ double-positive EPCs. The sorted EPCs were seeded in fibronectin-coated culture dishes and cultured in a dedicated EPC culture medium (containing vascular endothelial growth factor at a specific concentration). Basic fibroblast growth factor concentration Insulin-like growth factor concentration Training time The medium was changed every three days. Endothelial progenitor cell membranes were labeled with the near-infrared fluorescent dye DiR, and a DiR working solution was prepared at a concentration of [missing information]. The solvent is a mixture of dimethyl sulfoxide and culture medium (volume ratio). Endothelial progenitor cells were incubated with DiR working solution at 37°C and 5% CO2 for [duration missing]. The cells were then washed three times with pre-warmed culture medium to remove unbound dye; labeling efficiency was detected by flow cytometry, and a positive rate was required. .

[0053] Damaged endothelial cells were prepared as mitochondrial acceptor cells by seeding human umbilical vein endothelial cells into six-well plates and culturing them to confluence. Add hydrogen peroxide, an oxidative stress inducer, at a concentration of... Processing time This causes oxidative damage to cells. The degree of damage is confirmed by cell viability assay (CCK-8 assay), requiring cell viability to decrease to that of healthy controls. DiR-labeled endothelial progenitor cells and damaged endothelial cells were co-cultured in a micropatterned co-culture chip. The chip was fabricated using polydimethylsiloxane, and fibronectin stripe patterns (width) were prepared on the chip surface using microcontact printing technology. ,spacing This promotes directional contact between cells. The seeding ratio of endothelial progenitor cells to damaged endothelial cells... Total cell density cells .

[0054] Long-term live-cell imaging was performed using a rotating disk confocal microscope system equipped with a temperature-controlled stage. ) and CO2 incubator module (CO2 concentration) Maintaining cellular physiological state. A 60x oil immersion objective (numerical aperture) was used. ), excitation wavelength emission wavelength ,bandwidth Continuous imaging parameters: Total duration Time interval Single frame exposure time Image resolution Pixels, pixel size Total number of frames collected Image sequences were analyzed using ImagePro Plus software to identify tunnel nanotube structures. The criteria for identifying tunnel nanotubes were: diameter... ,length These are membranous tubular structures that connect endothelial progenitor cells to damaged endothelial cells. The movement of DiR fluorescent signals within these tunnel nanotubes is tracked to identify mitochondrial transfer events (fluorescent particles moving from donor cells to recipient cells via tunnel nanotubes).

[0055] The number of endothelial progenitor cells that successfully transferred mitochondria per unit time was counted. Let the total number of endothelial progenitor cells in the imaging field be . ,exist The number of endothelial progenitor cells that experienced at least one mitochondrial transfer event during this period was Then the transfer rate Defined as: This indicator reflects the proportion of cells with active mitochondrial translocation function in the endothelial progenitor cell population. The diagnosis was mitochondrial transfer dysfunction.

[0056] S132. Mitochondria of endothelial progenitor cells were labeled with tetramethylrhodamine methyl ester, a mitochondrial membrane potential-sensitive fluorescent dye. Tetramethylrhodamine methyl ester is a lipophilic cationic dye that specifically accumulates on the inner mitochondrial membrane, and its enrichment level is related to the membrane potential. The correlation is positive and follows the Nernst equation. Prepare a tetramethylrhodamine methyl ester working solution at the following concentration... Incubation time with endothelial progenitor cells at 37°C After washing to remove extracellular dye, the initial fluorescence intensity was detected using confocal microscopy. (excitation wavelength) emission wavelength ).

[0057] In joint training Subsequently, fluorescence quenching and recovery detection were performed on mitochondria that received intracellular transfer. High-intensity pulsed laser (wavelength...) was used. ,power Pulse duration Pulse count ) for receiving specific regions (diameter) within cells Photobleaching was performed to reduce the fluorescence intensity in the area to below 10% of its initial value. This was followed by photobleaching with a low-intensity laser (power...). Continuously monitor the fluorescence intensity recovery process in this area, with a sampling frequency of [missing information]. Monitoring duration Fluorescence recovery originates from dye diffusion from surrounding unbleached mitochondria and dye redistribution driven by membrane potential.

[0058] Single-exponential fitting was performed on the fluorescence recovery data: ,in: The fluorescence intensity at time t is The fluorescence intensity immediately after bleaching. To restore the fluorescence intensity after equilibrium, The recovery time constant is given; the fluorescence recovery half-life is calculated. : Mitochondrial membrane potential retention rate Defined as the ratio of the recovery half-life to the initial lemmatism intensity of tetramethylrhodamine methyl ester in the donor cells (normalized): ,in: and For healthy control reference values; It was determined to be a significant loss of mitochondrial membrane potential.

[0059] S133. The basal oxygen consumption, maximum respiratory volume, and ATP production of the recipient cells were measured using the XFp platform of the hippocampal energy analyzer. Cells receiving mitochondrial transfer were seeded into XFp microplates (well bottom area). In ), cell density cells well, cultivate To induce cell adhesion, add the following reagents in sequence: oligomycin (ATP synthase inhibitor), concentration... The oxygen consumption rate of ATP coupling was measured; the concentration of carbonyl cyanide m-chlorophenylhydrazone (mitochondrial uncoupling agent) was determined. Maximum respiratory volume was measured; a mixture of rotenone (a complex inhibitor) and antimycin A (a complex III inhibitor) was prepared at concentrations of [missing information]. and The oxygen consumption rate of non-mitochondria was measured.

[0060] Calculate key oxygen consumption parameters: basal oxygen consumption rate Oxygen consumption rate without drug administration; ATP production-related oxygen consumption rate. ,in: Oxygen consumption rate after oligomycin treatment; maximum respiratory capacity. ,in: The oxygen consumption rate after treatment with carbonyl cyanide m-chlorophenylhydrazone. Non-mitochondrial oxygen consumption rate after rotenone / antimycin A treatment; reserve respiratory capacity. .

[0061] Let the basal oxygen consumption of damaged endothelial cells that did not receive mitochondrial transfer (control group) be 0.5%. The basal oxygen consumption of cells receiving mitochondrial transfer (experimental group) was The basal oxygen consumption of healthy control cells was The oxygen consumption rate will increase by a factor of [number]. for: ATP production recovery percentage for: ,in: To determine the oxygen consumption rate associated with ATP production in cells receiving mitochondrial transfer, Oxygen consumption rate related to ATP production in healthy control cells; metabolic rescue efficiency. Overall increase in oxygen consumption and recovery of ATP production: ,in: To increase the oxygen consumption rate by a factor of 10, This represents the percentage of ATP production recovery.

[0062] S134. Endothelial cells were treated with the Piezo1-specific agonist Yoda1, Yoda concentration... Processing time This activates Piezo1 channel-mediated calcium signaling. The activities of mitochondrial respiratory chain complexes I-IV before and after treatment were measured using the high-resolution Oxygraph-2k respiration assay system. Mitochondria were isolated, and mitochondrial protein concentrations were determined. reaction temperature reaction volume Specific substrates were added sequentially to activate each complex. Complex I: glutamate concentration. malic acid concentration Determine the NADH-driven oxygen consumption rate Complex II: Succinic acid concentration Rotenone concentration (Inhibit complex I), determine FADH2-driven oxygen consumption rate. Complex III: Ascorbic acid concentration - Tetramethyl-p-phenylenediamine concentration Determine the oxygen consumption rate driven by cytochrome c reduction. Complex IV: Reduced cytochrome c concentration Determine the oxygen consumption rate driven by cytochrome c oxidation. .

[0063] Calculate the fold change in the activity of each complex after Yoda1 treatment relative to the untreated state. Let the activity of complex I before treatment be... After processing, it becomes The change factor of the complex activity for: Similarly, calculate the fold change in activity of complexes II, III, and IV. ; Change in the overall activity of the complex Geometric mean .

[0064] ATP production was determined using a bioluminescence assay kit at different time points after the addition of Yoda1. Cells were lysed, and ATP content was determined using a luciferin-luciferase reaction. Luminescence intensity. With ATP concentration The relationship is: ,in Here is the luciferase reaction rate constant; calculate the ATP yield (ATP production per unit time): ,in The percentage increase in ATP production per unit time; the factorial increase in ATP production after Yoda1 treatment. for: ,in and The figures represent ATP yields before and after treatment, respectively.

[0065] S135. Transfer rate Mitochondrial membrane potential retention rate Metabolic rescue efficiency , fold change in complex activity and ATP production rate increase Perform Min-Max normalization to unify its value range. Interval. For each index Its normalized value for: ,in and These are the 5th and 95th percentiles of the index in the healthy population, respectively (determined based on a large sample database).

[0066] The weights of each index are determined using the analytic hierarchy process (AHP), a judgment matrix is ​​constructed, and the weight vector is calculated using the eigenvalue method to obtain the transition rate weights. Mitochondrial membrane potential retention rate weighting Metabolic rescue efficiency weight Weighted fold change in complex activity ATP yield increase weighting .

[0067] The five normalized indices are summed according to their weights to output the mitochondrial energy metabolism coupling efficiency index. : The index ranges from 100 to 100. The closer the value is to 1, the more robust the mitochondrial energy metabolism coupling function; a threshold is set. ,when It was determined to be a mitochondrial energy metabolism coupling dysfunction.

[0068] S14. The fusion calculation of the functional impairment index of vascular endothelial mechanosensing-energy metabolism coupling is as follows:

[0069] S141. Perform Z-score standardization on the degree of decoupling of glycocalyx-cytoskeleton mechanotransduction output in step S11, the mechanosensory distortion index output in step S12, and the mitochondrial energy metabolism coupling efficiency index output in step S13. Calculate the mean and standard deviation of each index, convert the original values ​​into standard scores, eliminate differences in dimensions and degree of variation, and obtain the standardized feature vector X.

[0070] S142. Establish a benchmark covariance matrix based on a database of healthy individuals. Calculate a standardized data matrix of three characteristics of healthy individuals. Its covariance matrix is: Covariance matrix elements Reflecting characteristics With features The degree of linear correlation: ,in and This is the standardized sample mean (theoretically 0). This is the standardized value of the i-th feature of the k-th healthy individual.

[0071] Calculate the standardized eigenvector Vector of mean from healthy control group (After standardization) Mahalanobis distance : The calculation is as follows: ,in The inverse matrix (precision matrix) of the covariance matrix has the following elements: .

[0072] Mahalanobis distance The method quantifies the joint deviation of three indicators from the health benchmark, taking into account the correlation between features, and is superior to simple Euclidean distance. The larger the value, the more severe the imbalance in the three indicators, namely, the higher the degree of coupling disorder between glycocalyx mechanotransduction, ion channel sensing, and mitochondrial energy metabolism.

[0073] S143. A random forest regression model was used to analyze the contribution weights of the three indicators to vascular endothelial function impairment, using the training dataset. ,in For the first The three feature vectors of each sample The corresponding label for the degree of cardiovascular endothelial function impairment (determined by clinical outcome events or gold standard testing). This represents the total number of training samples.

[0074] Build includes A random forest of decision trees, where each tree is sampled from the training set via bootstrap sampling. One sample is used, and the remaining samples are treated as out-of-bag (OOB) data; at the split node of each tree, a random selection is made. The optimal split point is searched using each feature.

[0075] The reduction in Gini impurity for each feature at the decision tree node split is calculated. To assess the importance of features, let the features be... In the Nodes of a tree The Gini impurity before splitting is [value missing]. The weighted average of the Gini impurities of the left and right child nodes after splitting is: and The reduction in Gini impurity during the split is: ,in These represent the number of samples for the split node and its left and right child nodes, respectively.

[0076] feature Importance rating The sum of the Gini impurity reductions for this feature split across all trees: ,in For the first Using features in a tree The set of nodes to be split.

[0077] The weights of each feature are calculated based on the feature importance score. : ,satisfy Calculate the weighted collaborative deviation index. Weighted deviation and Mahalanobis distance correction based on integrated standardized features: ,in The degree of absolute deviation of the standardized features. This is a logarithmic correction term for the Mahalanobis distance, enhancing sensitivity to extreme deviations.

[0078] S144. Mahalanobis distance Deviation from weighted co-occurrence index Input a nonlinear activation function, which is mapped to the endothelial mechanosensory-energy metabolism coupling functional impairment index with values ​​in the range [0, 100]. Using a composite Sigmoid activation function, the intermediate variables are first calculated. : ,in The coefficients optimized using training data, the cross term Capture the synergistic effect of the two exponents. Then map it to the double sigmoid function. : ,in: The center position parameter of the Sigmoid function. The sigmoid slope parameter. This is the correction amplitude for the hyperbolic tangent. This represents the threshold position before decompensation. This is the hyperbolic tangent temperature parameter.

[0079] Risk classification is performed by setting damage index thresholds:

[0080] Mild damage; endothelial mechanosensory-energy metabolism coupling function is basically normal.

[0081] Moderate damage, with compensable dysfunction in coupling function;

[0082] Severe injury, the injury has reached a pre-decompensation state, and the cardiovascular risk is significantly increased;

[0083] When the index is greater than 60, it indicates that the synergistic coupling function of glycocalyx mechanotransduction, Piezo1 / TRPV4 ion channel sensing and mitochondrial energy metabolism is severely impaired, and the endothelial cells are in a critical state of transition from physiological mechanical adaptation to pathological mechanical stress, requiring immediate targeted intervention.

[0084] S2. Input the damage index into the vascular microenvironment mechanical decompensation early warning model, combine it with local hemodynamic parameters and matrix stiffness data, identify high-risk vascular segments and calculate the decompensation critical time, and output a risk decompensation early warning map containing the location of high-risk segments and the critical time; specifically including:

[0085] S21. Spatial mapping of the endothelial mechanosensing-energy metabolism coupling functional impairment index and identification of high-risk vascular segments:

[0086] S211. The endothelial mechanosensory-energy metabolism coupling functional impairment index output in step S14 is spatially mapped along the vascular tree. First, based on the patient's coronary artery computed tomography angiography image data, three-dimensional reconstruction of the blood vessels is performed using medical image processing software (such as Mimics or 3D Slicer). A semi-automatic segmentation algorithm is used to extract the vascular lumen contour, and combined with a centerline extraction algorithm to generate the vascular tree topology. The vascular tree is divided into several discrete segments, each with a length of L. segment = 5 mm, assigned a unique spatial coordinate identifier (x, y, z).

[0087] The damage index is assigned to the corresponding vascular segment. For segments without direct measurement data, Kriging space interpolation based on anatomical location is used for estimation, taking into account blood flow direction, vascular branch angle and inter-segment distance weights to generate a continuous spatial distribution field of the damage index.

[0088] S212. Computational fluid dynamics (CFD) simulation was used to obtain the time-averaged local wall shear stress and the oscillatory shear index. The 3D model of the blood vessel was imported into CFD software (such as ANSYS Fluent or OpenFOAM), meshed, and the boundary layer mesh height was set to y. + ≤1, to ensure the accuracy of near-wall flow analysis.

[0089] Assume blood is an incompressible Newtonian fluid with density ρ. blood = 1060 kg / m³, dynamic viscosity =0.0035 Pa·s. The inlet boundary condition is set as a pulsatile velocity inlet, based on the blood flow velocity waveform measured by individualized Doppler ultrasound of the patient, and the Womersley number is used to characterize the pulsation characteristics. The outlet boundary condition is set as a pressure outlet, and the outlet pressure is distributed according to Murray's law.

[0090] Solve the Reynolds-averaged Navier-Stokes equations using the SST k-ω turbulence model. Time step

[0091] Δt CFD = 0.001s, the total simulation duration covers 3 cardiac cycles. The convergence criterion is set at a residual of less than 10. －5 .

[0092] Extracting the average time value of wall shear stress ,in: This refers to the instantaneous wall shear stress. The duration of the cardiac cycle, Heart rate.

[0093] Calculate the oscillatory shear index : .

[0094] S213. Combine extracellular matrix stiffness distribution data obtained from in-situ measurements using atomic force microscopy. Perform nanoindentation testing in ex vivo vascular samples or biopsy tissue using atomic force microscopy (such as the Bruker BioScope Resolve). Employ silicon nitride probes with a tip curvature radius R. tip =20 nm, spring constant k cantilever = 0.1 N / m.

[0095] Test points were selected at the corresponding segments of the vascular tree, with a point spacing of d. AFM = 50μm, forming a gridded test array. Force curves were acquired at an approximation velocity of 1μm / s, with the maximum indentation depth h. max =500nm, ensuring it remains within the elastic deformation range. The Hertz model is used to fit the force-displacement curve to calculate the local elastic modulus.

[0096] ,in Let F be the Poisson's ratio (taken as 0.5), F be the applied force, and δ be the indentation depth.

[0097] S214. Analyze the gradient change rate of the damage index in different vascular segments using the gradient boosting algorithm. Construct a feature matrix, where each row represents a vascular segment, and the features include: damage index, ... , , Blood vessel diameter D vessel curvature κ vessel bifurcation angle θ bifurcation .

[0098] The XGBoost gradient boosting regression algorithm was employed, with the target variable being the spatial rate of change of the damage index, Δdamage / Δs, where Δdamage is the change in the damage index and Δs is the change in arc length along the vessel centerline. The learning rate η was set. XGB =0.05, tree depth max depth =6, number of iterations n round =200, L2 regularization coefficient λ XGB =1.0.

[0099] Calculate feature importance and identify the feature combinations that contribute most to the gradient change of the damage index. Calculate the predicted gradient change rate for each segment and set a threshold θ. gradient =15, identifying segments whose damage index gradient change rate exceeds a preset threshold as candidate regions for high-risk vascular segments.

[0100] S22. Construction of a quantitative scoring model for the severity of local vascular segment injury, specifically as follows:

[0101] S221. Endothelial cell samples were collected from candidate areas of high-risk vascular segments. Endothelial cells were isolated by enzymatic digestion using collagenase type I at a concentration of 1 mg / ml at 37°C for 15 minutes. The isolated cells were seeded in 35 mm confocal culture dishes and cultured in complete endothelial cell culture medium until the cells adhered.

[0102] A fluorescence resonance energy transfer (FRET) detection system was constructed to detect the binding of the TRPV4 channel C-terminus to calmodulin. The TRPV4-cyan fluorescent protein fusion expression plasmid and the calmodulin-yellow fluorescent protein fusion expression plasmid were co-transfected into endothelial cells using liposome transfection reagent. After transfection, the cells were cultured for 48 h to allow for full expression of the fluorescent protein. The transfection efficiency was detected by flow cytometry, with a positive rate of greater than 80%.

[0103] The fluorescence lifetime imaging microscopy was used for detection, with a picosecond pulsed laser at a wavelength of 440 nm and a pulse frequency of 40 MHz as the excitation source. Fluorescence decay curves were acquired using time-correlated single-photon counting mode, with an acquisition time of 60 s and 4096 time channels. The intrinsic fluorescence lifetime of the donor cyan fluorescent protein in the absence of energy transfer and the apparent fluorescence lifetime in the presence of energy transfer were determined.

[0104] The intensity of calcium influx signal was quantified by the decay of donor fluorescence lifetime. The fluorescence resonance energy transfer efficiency was calculated by the fluorescence lifetime method. The formula is: energy transfer efficiency equals donor intrinsic lifetime minus donor apparent lifetime divided by donor intrinsic lifetime. The higher the efficiency value, the tighter the binding between the C-terminus of TRPV4 channel and calmodulin, and the stronger the calcium influx signal.

[0105] Immunofluorescence staining was used to detect the nuclear translocation level of phosphorylated Smad2 / 3 proteins to quantify the activity of the TGFβ1 signaling pathway. A rabbit monoclonal antibody specifically recognizing the serine phosphorylation sites at positions 465 / 467 of the Smad2 / 3 protein was used as the primary antibody, and Alexa Fluor 568-labeled donkey anti-rabbit immunoglobulin G was used as the secondary antibody. Z-axis tomographic images were acquired using a laser scanning confocal microscope with a layer thickness of 0.5 μm, for a total of 20 layers. Image analysis software was used to automatically segment the cell nucleus and cytoplasm, and the ratio of the fluorescence intensity of phosphorylated Smad2 / 3 in the cell nucleus to the total fluorescence intensity of the entire cell was calculated. This ratio is the indicator of TGFβ1 signaling pathway activity.

[0106] The Pearson correlation coefficient between the two is calculated as the coactivation intensity. Let the fluorescence resonance energy transfer efficiency be E, the phosphorylation level of Smad2 / 3 nuclear translocation be S, and the sample size be N. The coactivation intensity ρ is then calculated using the following formula: ; in: and These represent the luminescence resonance energy transfer efficiency and phosphorylation value of the i-th cell, respectively. nuclear translocation level, and These are the respective sample means; a correlation coefficient greater than 0.7 indicates a significant synergistic amplification effect between the TRPV4-mediated calcium influx signal and the TGFβ1 signaling pathway.

[0107] S222. The phosphorylation level of serine at position 1612 of the Piezo1 channel in high-risk segment endothelial cells was determined using a phosphorylation-specific antibody combined with immunofluorescence. A rabbit monoclonal antibody specifically recognizing the phosphorylation site of serine at position 1612 of the Piezo1 protein was used as the primary antibody, and donkey anti-rabbit immunoglobulin G labeled with Alexa Fluor 488 was used as the secondary antibody.

[0108] Imaging parameters were standardized for healthy control and high-risk areas using a laser scanning confocal microscope. The excitation wavelength was 488 nm, and the emission wavelength was 519 nm. Excitation power, photomultiplier tube gain, and pinhole diameter were kept consistent between the two groups. Ten fields of view were randomly selected from each region, and 20 cells were analyzed in each field, for a total of 200 cells, for quantitative fluorescence intensity analysis.

[0109] Using the average fluorescence intensity of the healthy control area as a baseline, the relative fluorescence intensity ratio of the high-risk area was calculated as the degree of loss of mechanosensitivity. Let the average fluorescence intensity of the healthy control area be... The average fluorescence intensity in high-risk areas is The degree of loss of mechanical sensitivity Calculated using the following formula: A ratio less than 0.6 or greater than 1.4 indicates abnormal phosphorylation regulation of the Piezo1 channel. A ratio less than 0.6 suggests insufficient phosphorylation leading to loss of mechanosensitivity, while a ratio greater than 1.4 suggests excessive phosphorylation leading to abnormally increased mechanosensitivity.

[0110] S223. The expression levels of α-smooth muscle actin, a mesenchymal marker of endothelial cells, and vascular endothelial cadherin, a marker of endothelial cells, were detected by Western blotting. Cell lysates were collected, and protein concentrations were determined using the diquinoline carboxylic acid method. Equal volumes of protein samples were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis. The separating gel concentration was 10%, and the stacking gel concentration was 5%. The electrophoresis conditions were a constant current of 25 mA and an electrophoresis time of 90 min.

[0111] Proteins were transferred to polyvinylidene fluoride (PVDF) membranes via wet transfer under constant current of 300 mA for 90 min. Blocking was performed with 5% skim milk powder at room temperature for 1 h. Primary antibodies were incubated overnight at 4°C using mouse anti-human α-smooth muscle actin monoclonal antibody and rabbit anti-human vascular endothelial cadherin polyclonal antibody, respectively. Secondary antibodies labeled with horseradish peroxidase were incubated at room temperature for 1 h. Enhanced chemiluminescence substrate was used for color development, and images were acquired using a gel imaging system.

[0112] The grayscale values ​​of the bands were quantified using image analysis software, with the grayscale value of the α-smooth muscle actin band set as G. α The gray value of the endothelial cadherin band is G. VE The gray value of the internal reference protein glyceraldehyde-3-phosphate dehydrogenase band was G. GAPDH The relative expression level of α-smooth muscle actin after internal reference correction is G. α Divide by G GAPDH The relative expression level of vascular endothelial cadherin was G. VE Divide by G GAPDH .

[0113] The ratio of α-smooth muscle actin to vascular endothelial cadherin expression was calculated as the endothelial-mesenchymal transition index. The index is calculated using the following formula: This index assesses the degree of endothelial cell phenotypic transformation in high-risk segments, with an index greater than 0.5 indicating a significant endothelial-mesenchymal transition process.

[0114] S224. Increase the co-activation strength Degree of loss of mechanical sensitivity With endothelial-mesenchymal transition index Multiple linear regression analysis was performed to establish a quantitative scoring model for the severity of local vascular segment injury.

[0115] Let the severity of local vascular segment injury be scored as follows: Establish the following multiple linear regression equation: ; in: For the intercept term, The regression coefficient is taken as the absolute value of the loss of mechanical sensitivity, deviating from 1. The absolute value minus 1 is used to uniformly quantify the damage contribution of both underphosphorylation and hyperphosphorylation abnormalities.

[0116] The regression coefficients are estimated using the least squares method. Let the training sample size be M, and the true severity of the injury in the j-th sample be... The predicted value is The objective function is to minimize the sum of squared residuals: ; The regression coefficients are estimated by solving the normal equations, and the model outputs... The value ranges from 0 to 100, with higher scores indicating more severe damage to local vascular segments.

[0117] S23. Calculation of the critical time for decompensation, specifically as follows:

[0118] S231. A kinetic model of the injury index of endothelial cells in high-risk vascular segments is constructed based on time series analysis, using the monitored values ​​of the injury index output in step S21 at three consecutive time points, with a time interval of 3 months (90 days). Let the injury indices at the three time points be... , , The corresponding time point is , , ,in - = - =90 days.

[0119] A first-order differential equation is used to fit the trajectory of the damage index over time, and a Logistic growth model is established to describe the growth of the damage index within a finite range: ,in: The damage index, For time, The damage growth rate constant is The damage repair rate constant is The theoretical maximum damage index is set to 100.

[0120] The differential equation is discretized using the finite difference method, for each time point to The forward difference approximation is used: ; For time points to The backward difference approximation is adopted: ; Solve the above two equations simultaneously and substitute the known value. and =90 days, forming about and The system of two linear equations in two variables can be solved by matrix inversion or least squares method to obtain estimates of the damage growth rate constant and the damage repair rate constant.

[0121] S232. Substitute the co-activation intensity of TRPV4 and TGFβ1 calculated in step S22 as the damage acceleration factor into the modified kinetic equation. Define the damage acceleration factor. It is a linear function of the cooperative activation intensity ρ: ,in: The acceleration factor is calibrated to 0.5 based on clinical data.

[0122] The corrected kinetic equations are as follows: Meanwhile, the degree of loss of mechanosensitivity caused by abnormal phosphorylation of Piezo1 channels was used as a threshold adjustment factor and substituted into the damage threshold function.

[0123] Define the effective damage index The value of the actual damage index after correction for the degree of loss of mechanosensitivity: ,in: The threshold adjustment coefficient is set to 0.3, and R represents the degree of loss of mechanosensitivity. This correction makes the effective damage more severe in individuals with abnormal mechanosensitivity regulation under the same actual damage index.

[0124] S233. The decompensation critical time was calculated using the Weibull regression model in survival analysis. The endpoint event was defined as the appearance of positive α-smooth muscle actin expression in endothelial cells, i.e., an endothelial-mesenchymal transition index greater than 0.5, denoted as event = 1. Samples that have not yet experienced this event were censored, with event = 0.

[0125] Establish risk function , indicating at a point in time The probability that an individual who has not yet experienced the endpoint event will experience the endpoint event in the immediate following instant: , , These are the regression coefficients of co-activation intensity and the degree of loss of mechanosensitivity on risk, respectively.

[0126] Where: baseline risk function Follows a Weibull distribution: , The shape parameter determines the trend of risk over time. A value greater than 1 indicates increasing risk. An equal value of 1 indicates constant risk. A value less than 1 indicates decreasing risk; The scale parameter determines the time scale of the risk.

[0127] Cumulative risk function The integral of the risk function from 0 to t: ; Solve when the accumulated risk reaches The corresponding time point, i.e., the median time of decompensation. : ;

[0128] Substitute into the cumulative risk function expression: ;

[0129] Solving for tcritical using algebraic transformations: Among them, the regression coefficient and The maximum likelihood estimation is obtained by constructing and maximizing the log-likelihood function based on the observation time and event state of the training samples.

[0130] S234. Based on the damage index kinetic equation and Weibull regression results, output a comprehensive estimate of the critical time for decompensation, taking the time predicted by the kinetic model to reach an effective damage index of 60. Weibull model predicts time Weighted average: The weight The value of 0.6 reflects the dominant role of the dynamic model in short-term forecasting and the supplementary role of the Weibull model in long-term risk assessment.

[0131] The kinetic prediction time is obtained by solving the following equations: ; in Given the current damage index, the integral equation is obtained by solving the equation using numerical integration (such as the Runge-Kutta method). The final output is the comprehensive decompensation critical time. The timeframe is in days, representing the period from the current moment until the appearance of endothelial cells. Positive expression of smooth muscle actin indicates the expected time for significant endothelial-mesenchymal transition.

[0132] S24. Generation of Risk Decompensation Early Warning Map:

[0133] Four key information categories were extracted to construct the foundation of the atlas data. The first category was the three-dimensional coordinates of high-risk segments. Based on the discretization results of the vascular tree in step S21, spatial identifiers of candidate segments identified by the gradient boosting algorithm were extracted, and their absolute positions were recorded using the medical imaging coordinate system as a reference. The second category was the decompensation critical time. The comprehensive estimate output in step S23 was extracted and color-coded for risk grading: green for greater than 365 days, yellow for 180 to 365 days, orange for 90 to 180 days, and red for less than 90 days. The third category was the rate of deterioration of the mechanical microenvironment. Based on the kinetic model in step S231, the change in the damage index per unit time was calculated, and the increment over the next 90 days was predicted using the current damage index as a benchmark, and the daily average rate was calculated. The fourth category was damage subtype classification. Based on the co-activation intensity in step S221 and the degree of loss of mechanosensitivity in step S222, a two-dimensional feature space was constructed, and the K-means algorithm was used to cluster the subtypes into three categories: co-activation dominant, mechanosensitivity loss dominant, and dual damage type.

[0134] The original vascular tree model was refined into a mesh to achieve micrometer-level spatial accuracy. Each segment was subdivided into 20 to 50 sub-segments along its long axis, with sub-segment lengths controlled between 100 and 250 micrometers. The cross-section was layered using concentric circles, and the intima layer thickness was approximately 200 micrometers. Kriging interpolation was used to map macroscopic parameters onto the refined mesh. The damage index was weighted by blood flow direction, the critical time was calculated using nearest neighbor interpolation, the deterioration rate was calculated using linear interpolation, and subtype classification was directly assigned. Key microscopic feature locations were marked using co-localization analysis, achieving a spatial accuracy of 10 to 20 micrometers at the single-cell scale.

[0135] Interactive 3D maps are generated using scientific visualization software. The decompensation critical time is color-coded to represent the vascular surface texture, forming a risk heatmap. The rate of deterioration of the mechanical microenvironment is visualized using surface roughness; the higher the rate, the more convex the vascular surface. Subtype classification is represented by overlays of translucent blue, purple, and magenta blocks with adjustable transparency. A time-based animation module is constructed, extrapolating the spatial evolution of the damage index over 365 days based on modified kinetic equations, with a frame rate of 12 frames per 30 days. Interactions such as multi-angle rotation, arbitrary planar slicing, and local zooming are supported. Mouse hover displays precise coordinates, damage index, critical time, deterioration rate, subtype classification, and recommended strategies.

[0136] The generated standardized report includes a panoramic 3D view, close-ups of key segments, time-evolution animation, a summary table of quantitative parameters, and clinical decision support recommendations. The summary table lists the segment number, coordinates, damage index, critical time, rate of deterioration, and subtype classification for each high-risk segment. Decision recommendations are automatically generated based on subtype: for synergistic activation-dominant subtypes, a TRPV4 antagonist combined with a TGFβ1 inhibitor is recommended; for mechanosensitive loss-dominant subtypes, a Piezo1 phosphorylation modulator is recommended; and for dual-damage subtypes, a combination regimen is recommended. The atlas is stored in DICOM format, supporting import into hospital image archiving and communication systems, and interface with electronic medical record systems to achieve clinical workflow integration, providing spatial navigation support for precise intervention timing and individualized treatment strategy development.

[0137] S3. Calculate the biomechanical intervention window width based on the decompensation critical time in the early warning map, and match differentiated vascular biomechanical intervention strategies according to the window width and the damage subtype classification corresponding to the damage index, outputting a precise prevention strategy plan; specifically including:

[0138] S31. Calculation of the biomechanical intervention window width and assessment of intervention urgency, specifically:

[0139] S311. Extract the estimated decompensation critical time value for high-risk vascular segments from the risk decompensation early warning map. This value is stored in the form of a timestamp, representing the predicted time point of irreversible endothelial injury, in the format of year, month, day, hour, and minute. Simultaneously, obtain the current detection time point, and calculate the time difference between the two using the time difference calculation module as the initial biomechanical intervention window width. Let the decompensation critical timestamp be... The current timestamp is The initial intervention window width The calculation is as follows: The result is converted to days and rounded to one decimal place; this width represents the remaining interventional time before endothelial decompensation.

[0140] S312. Extract the mechanical microenvironment deterioration rate parameter from the risk decompensation early warning map. This parameter is the change in the damage index per unit time, calculated by the kinetic model in step S231, and the unit is damage index points per day, representing the rate of vascular function deterioration. Simultaneously, extract damage subtype classification labels for high-risk segments. These labels are generated by cluster analysis in step S244, including three types: Piezo1 phosphorylation abnormality-dominated, TRPV4 activation threshold shift-dominated, and dual-damage type.

[0141] S313. The initial intervention window width is adjusted by incorporating patient-specific factors and establishing a correction coefficient matrix, including age correction coefficient, complication correction coefficient, and medication history correction coefficient. Age correction coefficient. The coefficients were set according to patient age groups: 1.0 for under 45 years old, 0.9 for 45-60 years old, 0.8 for 60-75 years old, and 0.7 for over 75 years old. Complication correction factor. Based on the Charlson Complication Index, the coefficient decreases by 0.05 for each additional complication, with a minimum of 0.6. Medication history correction coefficient. Based on current cardiovascular medication use, the following values ​​were set: 0.95 for statins, 0.9 for angiotensin-converting enzyme inhibitors, 0.85 for combined use, and 1.0 for no related medications.

[0142] Calculate the overall correction factor : Corrected effective intervention window width for: ,in This is the uncorrected width of the intervention window.

[0143] S314. Package the effective intervention window width, the rate of deterioration of the mechanical microenvironment, and the damage subtype classification label into a decision matrix, which will serve as input parameters for intervention strategy matching in subsequent steps S32 to S34. Simultaneously, set intervention window grading thresholds: first threshold θ1 = 90 days, second threshold θ2 = 30 days. Compare the effective intervention window width with the thresholds to generate intervention urgency level indicators: greater than 90 days indicates low urgency, 30 to 90 days indicates medium urgency, and less than 30 days indicates high urgency.

[0144] S32. When the effective intervention window width is greater than the first threshold of 90 days and the rate of deterioration of the mechanical microenvironment is less than the second threshold of 0.5 damage index points per day, it is determined to be a low-urgency state, and a glycocalyx remodeling strategy with the core of restoring glycocalyx integrity and mechanosensory function is matched.

[0145] The specific implementation of the strategy includes: intravenous infusion of N-acetylglucosamine, a precursor to glycocalyx synthesis, at a dose of 10 mg per kilogram of body weight daily, dissolved in 250 ml of normal saline, continuously infused over 2 hours for 14 days, to promote the regeneration of heparan sulfate chains, the main component of glycocalyx; combined use of sulamine sodium, an inhibitor of heparan sulfate degrading enzymes, at a dose of 20 mg daily, administered intravenously in two divided doses, to inhibit the degradation of glycocalyx by matrix metalloproteinases and heparinase; and glycine, a Piezo1 channel mechanosensitive stabilizer, at a dose of 3 g daily, administered orally in three divided doses, to maintain the normal phosphorylation level of serine at position 1612 of the Piezo1 channel and stabilize the channel's ability to sense physiological laminar shear stress.

[0146] S33. When the effective intervention window width is less than the first threshold of 90 days but greater than the third threshold of 30 days, and the rate of deterioration of the mechanical microenvironment is greater than 0.5 damage index points per day, it is judged as a medium urgency state, and differential matching is performed according to the damage subtype classification in the early warning map.

[0147] If the dominant phosphorylation pattern is abnormal Piezo1 channel phosphorylation, a channel reactivation strategy targeting the phosphorylation state of serine at position 1612 of Piezo1 is employed. Specifically, the protein kinase A agonist foscorine is administered orally at a dose of 10 mg twice daily to promote Piezo1 channel phosphorylation by activating the protein kinase A signaling pathway; this is combined with the protein phosphatase 2A inhibitor okadaic acid, administered subcutaneously at a dose of 50 μg daily, to reduce phosphorylation removal. The two pathways work synergistically to restore normal mechanosensitivity of the Piezo1 channel.

[0148] If the TRPV4 channel activation threshold shift is dominant, a matrix stiffness softening strategy targeting the inhibition of endothelial-mesenchymal transition is employed. Specifically, the TRPV4 channel-specific antagonist GSK2193874 is used at a dose of 10 mg orally daily to block calcium influx induced by pathologically high matrix stiffness; combined with the lysyl oxidase inhibitor β-aminopropionitrile at a dose of 50 mg / kg body weight daily via intraperitoneal injection, to inhibit collagen cross-linking and reduce matrix stiffness, thus disrupting the positive feedback loop between increased stiffness and endothelial-mesenchymal transition.

[0149] For cases of dual-type injury, the two strategies mentioned above should be used in combination, and the dosage should be adjusted according to the patient's tolerance.

[0150] S34. When the effective intervention window width is less than the third threshold of 30 days, it is determined to be a high-urgency state, and a cellular energy rescue strategy that uses the reconstruction of the mitochondrial energy metabolism network is matched.

[0151] The specific implementation of the strategy includes: ultrasound microbubble-mediated exogenous mitochondrial targeted transplantation, in which mitochondria are extracted from healthy donor cells, mixed with lipid microbubbles, and infused via peripheral veins to achieve targeted delivery of mitochondria to high-risk vascular segments under ultrasound-targeted irradiation, once every two weeks for a total of three times; Jasplakinolide, a tunnel nanotube formation promoter and F-actin stabilizer, is administered at a dose of 200 nanomoles daily via local intravascular perfusion to promote tunnel nanotube formation by stabilizing the actin backbone; and Yoda1, a Piezo1 channel-specific agonist, is administered orally at a dose of 1 mg daily to enhance mitochondrial respiratory chain complex activity and increase adenosine triphosphate (ATP) yield by moderately activating Piezo1 channel-mediated calcium signaling.

[0152] S35. Integrate the above matching results and output a precise prevention strategy plan that includes the intervention strategy type, specific drug combination, route of administration, and intervention sequence. The plan document includes: strategy type identifier (glycocalyx remodeling, channel activity remodeling, matrix stiffness softening, or cellular energy rescue); specific drug names, dosages, routes of administration, frequencies, and treatment durations; intervention sequence arrangement, showing the start and end times of each drug in Gantt chart format; monitoring indicators and follow-up plans, including injury index re-examination time points, imaging assessment nodes, and endpoint event monitoring; and adverse reaction warnings and dosage adjustment plans. The plan is output in structured electronic document format, supporting import into clinical decision support systems and association with patients' electronic medical records for full-cycle management.

[0153] Example 2

[0154] Please see the appendix Figure 2 A biomarker-based cardiovascular disease risk assessment system, used to implement a biomarker-based cardiovascular disease risk assessment method, comprising:

[0155] The data acquisition module is used to acquire dynamic datasets of endothelial glycocalyx integrity, mechanosensitive ion channel activity regulation, and mitochondrial energy metabolism coupling efficiency of the target individual.

[0156] A damage index construction module, connected to the data acquisition module, is used for multi-dimensional analysis of the dynamic dataset of endothelial glycocalyx integrity to quantify the glycocalyx thickness decay rate, heparan sulfate chain shedding concentration, and the degree of decoupling of glycocalyx-cytoskeleton mechanotransduction. Based on a mechanosensitive ion channel activity regulation dataset, the difference in the dynamic response of Piezo1 channel to calcium ion influx under physiological laminar shear stress and pathological turbulent shear stress was determined. The phosphorylation modification level of serine residue 1612 in Piezo1 channel was analyzed to quantify its mechanosensitive regulatory state. The activation threshold shift of TRPV4 channel under abnormally increased matrix stiffness was measured to comprehensively calculate... The mechanosensory distortion index was calculated. Based on the mitochondrial energy metabolism coupling efficiency dataset, the transfer rate of mitochondria from endothelial progenitor cells to endothelial cells via tunnel nanotubes, the retention rate of mitochondrial membrane potential, the metabolic rescue efficiency of receiving cells, and the activity of mitochondrial respiratory chain complexes and ATP production under Piezo1 channel activation were quantified to construct the mitochondrial energy metabolism coupling efficiency index. The degree of decoupling of glycocalyx-cytoskeleton mechanotransduction, the mechanosensory distortion index, and the mitochondrial energy metabolism coupling efficiency index were input into a nonlinear fusion model. By analyzing the degree of synergistic deviation among the three, the vascular endothelial mechanosensory-energy metabolism coupling function impairment index was output.

[0157] The risk decompensation early warning module, connected to the damage index construction module, is used to input the endothelial mechanosensing-energy metabolism coupling function damage index into a pre-trained vascular microenvironment mechanical decompensation early warning model. By analyzing the gradient distribution characteristics of the damage index in different segments of the vascular tree, combined with the spatiotemporal fluctuation pattern of local hemodynamic wall shear stress and in-situ detection data of extracellular matrix stiffness, it identifies high-risk vascular segments in which endothelial cells transform from physiological mechanical adaptation to pathological mechanical stress. Based on the extracellular matrix stiffness detection data of high-risk vascular segments and the expression level of EndMT phenotypic transformation markers in endothelial cells, it analyzes the synergistic activation intensity of calcium influx signal mediated by TRPV4 channel and TGFβ1 signaling pathway through abnormal increase in matrix stiffness, combined with the degree of loss of mechanosensitivity caused by abnormal phosphorylation of Piezo1 channel, and calculates the decompensation critical time for local vascular segments to evolve from mechanosensing-energy metabolism coupling imbalance to irreversible endothelial damage. The output includes a risk decompensation early warning map containing spatial location tags of high-risk segments, estimated decompensation critical time, mechanical microenvironment deterioration rate, and mechanosensing-energy metabolism coupling damage subtype classification.

[0158] The intervention strategy matching module, connected to the risk decompensation early warning module, calculates the biomechanical intervention window width from the current moment to the occurrence of irreversible endothelial injury based on the estimated decompensation critical time and the rate of deterioration of the mechanical microenvironment in the risk decompensation early warning map. When the biomechanical intervention window width is greater than a first threshold and the rate of deterioration of the mechanical microenvironment is less than a second threshold, a glycocalyx remodeling strategy centered on restoring glycocalyx integrity and mechanosensory function is matched. This strategy includes combined intervention of glycocalyx synthesis precursor supplementation, heparan sulfate degradation enzyme inhibitors, and Piezo1 channel mechanosensitivity stabilizers. When the biomechanical intervention window width is less than the first threshold but the rate of deterioration of the mechanical microenvironment is greater than the second threshold, further differentiated matching is performed based on the mechanosensory-energy metabolism coupled injury subtype classification. For cases dominated by abnormal Piezo1 channel phosphorylation, a channel activity remodeling strategy targeting the phosphorylation state of serine at position 1612 of Piezo1 is employed. For cases dominated by TRPV4 channel activation threshold shift, a combined intervention of TRPV4 channel antagonists and matrix crosslinking inhibitors targeting the EndMT process is employed. When the biomechanical intervention window is less than the third threshold, a cellular energy rescue strategy targeting the reconstruction of the mitochondrial energy metabolism network is employed, including a combined intervention of exogenous mitochondrial transplantation, tunnel nanotube formation promoters, and Piezo1 channel-mediated mitochondrial respiration enhancers. The output is a precise prevention strategy that matches the current vascular microenvironment mechanical characteristics, mechanosensory-energy metabolism coupled injury subtype, and biomechanical intervention window.

[0159] The central control module is connected to the data acquisition module, the damage index construction module, the risk decompensation early warning module, and the intervention strategy matching module, respectively, and is used to coordinate the working sequence and data transmission of each module.

[0160] The data storage module, connected to the central control module, is used to store dynamic datasets of endothelial glycocalyx integrity, datasets of mechanosensitive ion channel activity regulation, datasets of mitochondrial energy metabolism coupling efficiency, vascular endothelial mechanosensing-energy metabolism coupling function damage index, risk decompensation early warning maps, and precise prevention strategies.

[0161] As can be seen from the above description, the embodiments of the present invention achieve the following technical effects:

[0162] This invention constructs a ternary coupling assessment system from the perspective of the complete cascade pathway of mechanosensing-transduction-energy metabolism response: By detecting the glycocalyx thickness decay rate, heparan sulfate chain shedding concentration, and the degree of decoupling between the glycocalyx and cytoskeleton mechanotransduction, the functional state of the first threshold for mechanosensing signal transmission into the cell is quantified. By measuring the difference in calcium response of the Piezo1 channel under physiological and pathological shear stress, the phosphorylation modification level of serine at position 1612 of Piezo1, and the degree of activation threshold shift of the TRPV4 channel, the functional state of ion channels is incorporated into the assessment system, revealing the molecular nature of mechanosensing distortion. By tracking mitochondrial transfer rate, membrane potential retention rate, metabolic rescue efficiency, and mitochondrial respiratory chain activity under Piezo1 activation, a coupling efficiency index between mechanosensing and energy metabolism is established. This invention quantifies the degree of synergistic imbalance among the three indicators by calculating the Mahalanobis distance and weighted synergistic deviation index using a nonlinear fusion model. This system-coupled assessment paradigm can identify high-risk individuals whose individual indicators may be within the critical range, but whose overall performance deviates significantly from the norm. This significantly improves the sensitivity and specificity of risk assessment and fills the gap in the detection of vascular endothelial function damage mechanisms in existing technologies.

[0163] This invention introduces survival analysis models at the cellular and molecular level, constructing a dynamic early warning system with spatiotemporal dimensions. By analyzing the gradient distribution characteristics of the damage index in different segments of the vascular tree, combined with the spatiotemporal fluctuation patterns of local hemodynamic wall shear stress and in-situ detection data of extracellular matrix stiffness, high-risk vascular segments transitioning from physiological mechanical adaptation to pathological mechanical stress are accurately identified, achieving millimeter-level spatial localization. The synergistic activation intensity of the TRPV4 and TGFβ1 signaling pathways and the degree of mechanosensitivity loss due to abnormal Piezo1 channel phosphorylation are used as accelerating factors, introduced into a Weibull regression model to quantitatively predict the decompensation threshold time for the evolution of local vascular segments from mechanosensory-energy metabolism coupling imbalance to irreversible endothelial damage. This elevates cardiovascular risk assessment from qualitative grading to quantitative timing, providing clinicians with two key decision-making dimensions: where the problem will occur and when it will occur. Simultaneously, based on principal component clustering analysis of synergistic activation intensity and the degree of mechanosensitivity loss, damage subtype classification is achieved, providing a mechanistic basis for subsequent precise intervention.

[0164] This invention establishes a three-tiered, progressive precision intervention matching mechanism: The first tier is based on intervention urgency stratification: by calculating the biomechanical intervention window width, it distinguishes three groups of people—those eligible for elective intervention, those requiring active intervention, and those requiring urgent intervention—to achieve precise timing of intervention; The second tier is based on damage subtype stratification: for those dominated by Piezo1 phosphorylation abnormalities, it matches channel activity remodeling strategies; for those dominated by TRPV4 activation threshold shifts, it matches combined intervention with TRPV4 antagonists and matrix crosslinking inhibitors; and for those dominated by mitochondrial metabolic disorders, it matches combined intervention with exogenous mitochondrial transplantation and tunnel nanotube promoters; The third tier is based on joint analysis of intervention window and damage subtype, outputting a complete protocol including intervention strategy type, specific drug combination, route of administration, and intervention sequence. This approach represents a leap from vague treatment to targeted repair: for patients with glycocalyx damage, N-acetylglucosamine supplementation can promote glycocalyx regeneration; for patients with Piezo1 phosphorylation abnormalities, regulating protein kinase A activity can restore mechanosensitivity; for patients with TRPV4 activation threshold deviations, GSK2193874 can be used to block the EndMT process; and for patients with mitochondrial metabolic disorders, Yoda1 can enhance mitochondrial respiratory chain activity. This precise matching strategy based on molecular mechanisms holds promise for significantly improving intervention efficacy, avoiding unnecessary drug exposure, and truly achieving personalized and precise prevention of cardiovascular diseases.

[0165] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of the invention is limited to these examples; within the framework of the invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of the different aspects of the invention as described above, which are not provided in detail for the sake of brevity.

[0166] This invention is intended to cover all such substitutions, modifications, and variations that fall within the broad scope of this specification. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for assessing cardiovascular disease risk based on biomarkers, characterized in that, include: We acquired data on the integrity of the endothelial glycocalyx, the activity regulation data of mechanosensitive ion channels, and the coupling efficiency data of mitochondrial energy metabolism in the target individuals, and constructed an index of vascular endothelial mechanosensing-energy metabolism coupling function damage through nonlinear fusion. The damage index is input into the vascular microenvironment mechanical decompensation early warning model. Combined with local hemodynamic parameters and matrix stiffness data, high-risk vascular segments are identified and the decompensation critical time is calculated. A risk decompensation early warning map containing the location of high-risk segments and the critical time is output. The biomechanical intervention window width is calculated based on the decompensation critical time in the early warning map. Based on the window width and the damage subtype classification corresponding to the damage index, differentiated vascular biomechanical intervention strategies are matched, and precise prevention strategy plans are output.

2. The method for assessing cardiovascular disease risk based on biomarkers according to claim 1, characterized in that, The acquisition of endothelial glycocalyx integrity data, mechanosensitive ion channel activity regulation data, and mitochondrial energy metabolism coupling efficiency data of the target individual, followed by nonlinear fusion to construct a vascular endothelial mechanosensing-energy metabolism coupling functional impairment index, specifically includes: The thickness of the endothelial glycocalyx in the sublingual microcirculation was measured using lateral flow dark-field imaging, and the glycocalyx thickness decay rate per unit time was calculated. The concentration of heparan sulfate chain shedding in plasma was detected by immunofluorescence. The activation efficiency of endothelial nitric oxide synthase under mechanical stimulation and the colocalization coefficient of syndecan-1-actin were measured to quantify the degree of decoupling of glycocalyx-cytoskeleton mechanotransduction. The shear stress response index, mechanical sensitivity modulation index, and TRPV4 activation threshold shift index of the Piezo1 channel were calculated separately, and the mechanical sensing distortion index was calculated by combining the three indices. Five indices were calculated: transfer rate, mitochondrial membrane potential retention rate, metabolic rescue efficiency, complex activity change factor, and ATP yield increase factor. After normalization of the five indices, they were weighted and summed to output the mitochondrial energy metabolism coupling efficiency index. The degree of decoupling of glycocalyx-cytoskeleton mechanotransduction, mechanosensory distortion index, and mitochondrial energy metabolism coupling efficiency index are input into a nonlinear fusion model based on random forest regression to output the vascular endothelial mechanosensory-energy metabolism coupling function impairment index.

3. The method for assessing cardiovascular disease risk based on biomarkers according to claim 2, characterized in that, The calculation of the Piezo1 channel shear stress response index, mechanosensitivity modulation index, and TRPV4 activation threshold shift index, respectively, and the calculation of the mechanosensitivity distortion index by combining the three indices, specifically includes: An endothelial cell microfluidic chip model was constructed, and physiological laminar shear stress and pathological turbulent shear stress were applied respectively. The dynamic changes of calcium influx mediated by Piezo1 channel were monitored in real time using the calcium ion fluorescent indicator Fluo-4. The ratio of peak calcium concentration under the two stress conditions was calculated as the Piezo1 channel shear stress response index. Endothelial cell lysates were collected after shear stress stimulation and immunoprecipitated with a rabbit monoclonal antibody targeting phosphorylated serine at position 1612 of Piezo1. The proportion of phosphorylated Piezo1 in total Piezo1 protein was determined by Western blotting, and the deviation of the proportion from the healthy control value was calculated as the mechanosensitivity regulation index. Endothelial cells were seeded on polyacrylamide hydrogel matrices of different hardness, and gradient concentration solutions of TRPV4 specific agonist GSK1016790A were added. The dose-response curves of calcium influx under different hardness conditions were measured using a fluorescence calcium imaging system. The deviation factor of the half maximum effect concentration value relative to the 2 kPa healthy control matrix was calculated as the TRPV4 activation threshold deviation index. The Piezo1 channel shear stress response index, mechanical sensitivity adjustment index, and TRPV4 activation threshold shift index are input into the principal component analysis model, and the first principal component score is extracted as the mechanical perception distortion index.

4. The method for assessing cardiovascular disease risk based on biomarkers according to claim 3, characterized in that, The process involves calculating five indices: transfer rate, mitochondrial membrane potential retention rate, metabolic rescue efficiency, fold change in complex activity, and fold increase in ATP production. These indices are then normalized and weighted summed to output the mitochondrial energy metabolism coupling efficiency index, which specifically includes: Endothelial progenitor cells labeled with DiR fluorescent dye were co-cultured with damaged endothelial cells and continuously imaged for 6 hours using a confocal live-cell imaging system, acquiring images every 5 minutes. The formation of tunnel nanotubes and the process of mitochondrial transfer were tracked using ImagePro software, and the percentage of endothelial progenitor cells that successfully transferred mitochondria per unit time was calculated as the transfer rate. Mitochondria of endothelial progenitor cells were labeled with the mitochondrial membrane potential-sensitive fluorescent dye TMRM. After transfer, the fluorescence recovery half-life of labeled mitochondria in the recipient cells was measured using fluorescence quenching recovery technique. The ratio of the fluorescence intensity of TMRM to that of the donor cells was calculated as the mitochondrial membrane potential retention rate. The basal oxygen consumption, maximum respiratory volume, and ATP production of the receiving cells were measured using the hippocampal energy analyzer XFp platform. These were compared with damaged endothelial cells that did not receive mitochondrial transfer. The increase in oxygen consumption and the percentage recovery of ATP production were calculated as metabolic rescue efficiency. Endothelial cells were treated with the Piezo1-specific agonist Yoda1. The activity of mitochondrial respiratory chain complexes I-IV before and after treatment was measured using the high-resolution Oxygraph-2k respiratory assay system. ATP production was measured using bioluminescence assay, and the fold change in complex activity and the fold increase in ATP yield were calculated. The mitochondrial energy metabolism coupling efficiency index is output by normalizing the transfer rate, mitochondrial membrane potential retention rate, metabolic rescue efficiency, complex activity change factor, and ATP yield increase factor, and then weighting and summing them.

5. The method for assessing cardiovascular disease risk based on biomarkers according to claim 4, characterized in that, The degree of decoupling of the glycocalyx-cytoskeleton mechanotransduction, the mechanosensory distortion index, and the mitochondrial energy metabolism coupling efficiency index are input into a nonlinear fusion model based on random forest regression to output a vascular endothelial mechanosensory-energy metabolism coupling function impairment index, specifically including: The degree of decoupling of the glycocalyx-cytoskeleton mechanotransduction, the mechanosensory distortion index, and the mitochondrial energy metabolism coupling efficiency index were Z-score normalized to obtain the normalized feature vector X. A benchmark covariance matrix is ​​established based on a database of healthy individuals. The Mahalanobis distance D between the standardized eigenvector X and the mean vector μ of the healthy control group is calculated to quantify the joint deviation of the three indicators from the healthy benchmark. The contribution weights of three indicators to vascular endothelial function injury were analyzed using a random forest regression model, and the weighted co-deviation index W was calculated. The Mahalanobis distance D and the weighted co-displacement index W are input into a nonlinear activation function to output the vascular endothelial mechanosensory-energy metabolism coupling function impairment index.

6. The method for assessing cardiovascular disease risk based on biomarkers according to claim 5, characterized in that, The damage index is input into the vascular microenvironment mechanical decompensation early warning model. Combined with local hemodynamic parameters and matrix stiffness data, high-risk vascular segments are identified and the decompensation critical time is calculated. The model outputs a risk decompensation early warning map that includes the location of high-risk segments and the critical time. Specifically, it includes: The endothelial mechanosensing-energy metabolism coupling function damage index is spatially mapped along the vascular tree. Computational fluid dynamics simulation is used to obtain the time average value of local wall shear stress and oscillatory shear index. Combined with extracellular matrix stiffness distribution data measured in situ by atomic force microscopy, the gradient change rate of the damage index in different vascular segments is analyzed by gradient boosting algorithm. Segments with a gradient change rate of the damage index exceeding a preset threshold are identified as candidate high-risk vascular segments. The synergistic activation intensity, the degree of loss of mechanosensitivity, and the EndMT conversion index were calculated separately. The three indices were then subjected to multiple linear regression analysis to establish a quantitative scoring model for the severity of local vascular segment injury. A dynamic model of the evolution of the damage index over time was constructed based on time series analysis, and the critical time of decompensation from the current state to the appearance of positive expression of the mesenchymal marker α-SMA in endothelial cells was calculated. By integrating the three-dimensional coordinate positioning of high-risk segments, the estimated critical time of decompensation, the change in damage index per unit time as the rate of deterioration of the mechanical microenvironment, and the damage subtype classification obtained by clustering the co-activation intensity and the degree of loss of mechanical sensitivity based on principal component analysis, a risk decompensation early warning map with spatial resolution at the micrometer level is generated.

7. The method for assessing cardiovascular disease risk based on biomarkers according to claim 6, characterized in that, The calculation of synergistic activation intensity, loss of mechanosensitivity, and EndMT conversion index, respectively, followed by multiple linear regression analysis of these three indices, establishes a quantitative scoring model for the severity of local vascular segment injury. Specifically, this includes: Endothelial cell samples were collected from candidate regions of high-risk vascular segments. The binding efficiency of TRPV4 channel C-terminus to calmodulin was determined by fluorescence resonance energy transfer technology. The intensity of calcium influx signal was quantified by the decay of donor fluorescence lifetime. At the same time, the nuclear translocation level of phosphorylated Smad2 / 3 was detected by immunofluorescence staining to quantify the activity of TGFβ1 signaling pathway. The Pearson correlation coefficient between the two was calculated as the synergistic activation intensity. The phosphorylation level of serine at position 1612 of the Piezo1 channel in endothelial cells of high-risk segments was determined by phosphorylation-specific antibody combined with immunofluorescence. The relative fluorescence intensity ratio of the high-risk area was calculated as the degree of loss of mechanosensitivity, with the fluorescence intensity of the healthy control area as the benchmark. The expression levels of α-SMA, a marker of endothelial cell mesenchymal matrix, and VE-cadherin, a marker of endothelial cell, were detected by Western blot. The α-SMA / VE-cadherin expression ratio was calculated as the EndMT conversion index to assess the degree of phenotypic transformation of endothelial cells in high-risk segments. The synergistic activation intensity, the degree of loss of mechanosensitivity, and the EndMT conversion index were subjected to multiple linear regression analysis to establish a quantitative scoring model for the severity of local vascular segment injury.

8. The method for assessing cardiovascular disease risk based on biomarkers according to claim 7, characterized in that, The dynamic model of the damage index evolution over time, constructed based on time series analysis, calculates the critical time for decompensation from the current state to the appearance of positive expression of the mesenchymal marker α-SMA in endothelial cells, specifically including: A dynamic model of the damage index of endothelial cells in high-risk vascular segments was constructed based on time series analysis. The trajectory of the damage index over time was fitted by a first-order differential equation based on the monitoring values ​​of the damage index at three consecutive time points. The co-activation intensity of TRPV4 and TGFβ1 was used as a damage acceleration factor and substituted into the modified kinetic equation. At the same time, the degree of loss of mechanosensitivity caused by abnormal phosphorylation of Piezo1 channel was used as a threshold regulation factor and substituted into the damage threshold function. The Weibull regression model in survival analysis was used to calculate the critical time of decompensation. The positive expression of α-SMA in endothelial cells was defined as the endpoint event. Co-activation intensity and the degree of loss of mechanosensitivity were used as covariates to establish a risk function and solve for the time point when the cumulative risk reaches 50%. Based on the damage index dynamic equation and Weibull regression results, a comprehensive estimate of the critical time for decompensation is output.

9. The method for assessing cardiovascular disease risk based on biomarkers according to claim 8, characterized in that, The biomechanical intervention window width is calculated based on the decompensation critical time in the early warning map. Based on the window width and the damage subtype classification corresponding to the damage index, differentiated vascular biomechanical intervention strategies are matched, and a precise prevention strategy is output, specifically including: The difference between the estimated critical time of decompensation in the risk decompensation early warning map and the current time point is calculated to obtain the width of the biomechanical intervention window. At the same time, the rate of deterioration of the mechanical microenvironment in the early warning map is extracted as an indicator of the urgency of intervention. When the width of the intervention window is greater than a preset first threshold and the rate of deterioration is less than a preset second threshold, a matching glycocalyx remodeling strategy is employed. The strategy includes intravenous infusion of the glycocalyx synthesis precursor N-acetylglucosamine, combined with the application of the heparan sulfate degrading enzyme inhibitor suramin sodium, and the Piezo1 channel mechanosensitive stabilizer glycine. When the intervention window width is less than the first threshold but the deterioration rate is greater than the second threshold, differential matching is performed according to the damage subtype classification in the early warning map. If it is the Piezo1 channel phosphorylation abnormality-dominant type, it is matched with protein kinase A agonist to regulate the phosphorylation state of serine at position 1612; if it is the TRPV4 channel activation threshold shift-dominant type, it is matched with the TRPV4 antagonist GSK2193874 combined with the lysyl oxidase inhibitor β-aminopropionitrile to inhibit the EndMT process. When the width of the intervention window is less than the preset third threshold, a matching cell energy rescue strategy is used. The strategy includes ultrasound microbubble-mediated exogenous mitochondrial targeted transplantation, tunnel nanotube formation promoter F-actin stabilizer Jasplakinolide, and Piezo1 channel agonist Yoda1 to enhance the activity of mitochondrial respiratory chain complex. By integrating the above matching results, a precise prevention strategy plan is output, which includes the type of intervention strategy, specific drug combination, route of administration, and timing of intervention.

10. A method for assessing cardiovascular disease risk based on biomarkers according to claim 9, characterized in that, The difference between the estimated critical time for decompensation in the aforementioned risk decompensation early warning map and the current time point is calculated to obtain the width of the biomechanical intervention window. Simultaneously, the rate of deterioration of the mechanical microenvironment in the early warning map is extracted as an indicator of intervention urgency, specifically including: The estimated critical time of decompensation for high-risk vascular segments is extracted from the risk decompensation early warning map, and the current detection time point is obtained. The time difference between the two is calculated as the initial biomechanical intervention window width. The mechanical microenvironment deterioration rate parameter was extracted from the risk decompensation early warning map, and damage subtype classification labels of high-risk segments were extracted, including Piezo1 phosphorylation abnormality type, TRPV4 activation threshold shift type and mitochondrial metabolic disorder type. The initial intervention window width is corrected by introducing weighting coefficients for patient age, complications, and medication history, and the corrected effective intervention window width is calculated. The effective intervention window width, mechanical microenvironment deterioration rate, and damage subtype classification label are packaged and output to the decision matrix as input parameters for subsequent intervention strategy matching. At the same time, the effective intervention window width is compared with the preset intervention window grading threshold to generate an intervention urgency level identifier.

Images