Plasma metabolism composition for predicting atrial fibrillation related cerebral apoplexy and application thereof

Abstract

The invention discloses a plasma metabolism composition for predicting atrial fibrillation related cerebral apoplexy and application of the plasma metabolism composition. The plasma metabolism composition is prepared from a plasma metabolite, namely 7-hydroxy-6-methoxy-2H-chromene-2-ketone, deoxyadenosine, arachidonic acid, choline glycerophosphate, eicosapentaenoic acid, fumaric acid and thromboxane B3. The invention further discloses a preparation method of the plasma metabolism composition. The seven screened metabolites are more intuitive, so that researchers can analyze the key metabolites more intensively. The method has the advantages that the detection time is effectively saved, the detection efficiency is improved, the detection cost is reduced, the plasma metabolism composition prediction method based on the molecular level is remarkably different from a traditional clinical complication scoring strategy-based prediction mode, the prediction accuracy is higher, and the method is suitable for popularization and application. And a new technical means is provided for solving the problem of predicting the cerebral apoplexy related to the atrial fibrillation.

Description

Technical Field

[0001] The present invention relates to a plasma metabolic composition, and also relates to the application of the plasma metabolic composition in predicting atrial fibrillation-related stroke. The present invention belongs to the field of medical technology. Background Art

[0002] Atrial fibrillation (AF) is a common arrhythmia, mainly manifested as abnormal rapid and irregular contraction of atrial myocardial cells. According to the data of the Framingham Heart Study Center, due to factors such as population aging, the incidence of atrial fibrillation has tripled in the past 50 years. Atrial fibrillation-related stroke (AFS) is the most serious complication of patients with this disease, with a high fatality rate and extremely poor prognosis. This process is mainly due to the loss of atrial contraction function in patients with atrial fibrillation, resulting in a decrease in blood flow velocity, and then leading to thrombus formation. These thrombi may flow to the brain through the blood, block cerebral blood vessels, and cause stroke.

[0003] Clinically, important tools for evaluating the stroke risk of atrial fibrillation patients mainly include CHADS2 and CHA2DS2-VASc scores. These scoring systems predict the stroke risk of atrial fibrillation patients based on factors such as age, gender, and related diseases. However, these scoring systems fail to consider the underlying molecular mechanisms, resulting in relatively low prediction accuracy. Therefore, new methods based on the molecular level are needed to more accurately predict the risk of stroke in atrial fibrillation patients.

[0004] As an emerging omics technology, metabolomics has shown great potential in the early prediction of diseases. By qualitatively and quantitatively analyzing the small molecule metabolite profiles in organisms, this technology can screen out characteristic biomarkers, namely differential metabolites. These biomarkers can comprehensively reflect the physiological and pathological states of organisms, and combined with machine learning algorithms, an efficient prediction model can be constructed to significantly improve the accuracy of disease prediction. However, current research on plasma metabolites for predicting AFS is still relatively limited.

[0005] If specific metabolites related to AFS can be screened out and used as molecular markers, the accuracy of AFS risk assessment will be significantly improved. Summary of the Invention

[0006] The purpose of the present invention is to provide a plasma metabolic composition for predicting atrial fibrillation-related stroke and its application.

[0007] To achieve the above purpose, the present invention adopts the following technical means:

[0008] In this invention, blood samples of 90 AF patients and 66 AFS patients were collected, and untargeted metabolomics detection was carried out by Liquid Chromatography-Mass Spectrometry (LC-MS). A total of 671 metabolites were screened from the plasma samples of 156 patients, among which 51 metabolites were significantly correlated with AFS. These 51 metabolites were put into the Lasso model, and finally 7 metabolites were screened out as biomarkers for predicting the risk of stroke occurrence. These 7 biomarkers can accurately predict the risk of AFS (AUC = 0.94).

[0009] Based on the above research, this invention proposes a plasma metabolite composition for predicting atrial fibrillation-related stroke. The plasma metabolite composition consists of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3.

[0010] Furthermore, this invention also proposes the application of reagents or devices for detecting the contents of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3 in the preparation of reagents or devices for predicting atrial fibrillation-related stroke.

[0011] Still further, this invention also proposes a reagent combination for predicting atrial fibrillation-related stroke. The reagent combination consists of reagents for detecting the contents of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3 respectively.

[0012] Still further, this invention also proposes the application of the plasma metabolite composition, the reagents or devices for detecting the contents of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3, and the reagent combination in the preparation of reagents for predicting atrial fibrillation-related stroke.

[0013] Among them, compared with AF patients, the expression levels of 7-hydroxy-6-methoxy-2H-chromen-2-one, glycerophosphocholine, and thromboxane B3 in the plasma of atrial fibrillation-related stroke patients are significantly reduced, while the expression levels of deoxyadenosine, arachidonic acid, eicosapentaenoic acid, and fumaric acid are significantly increased.

[0014] Compared with the prior art, the beneficial effects of this invention are as follows:

[0015] The 7 metabolites screened out by the present invention are more intuitive, enabling researchers to more intensively analyze these key metabolites. This advantage not only effectively saves the detection time, but also improves the detection efficiency and reduces the detection cost. Moreover, the plasma metabolite composition prediction method based on the molecular level is significantly different from the traditional prediction method based on the clinical comorbidity scoring strategy, with higher prediction accuracy, opening up new technical means for solving the AFS prediction problem. In addition, the present invention also has the following advantages:

[0016] (1) Studying the metabolic mechanism of atrial fibrillation secondary to stroke

[0017] Through metabolomics analysis, this study deeply explored the potential metabolic mechanism of atrial fibrillation patients secondary to stroke. It was found that the metabolite levels of specific lipids (such as triglycerides) and amino acids (such as tyrosine) changed significantly in patients. These changes may be closely related to pathological processes such as inflammatory response, vascular endothelial dysfunction, and oxidative stress. For example, certain metabolites may play a role in the inflammatory response, leading to thrombosis or hypoxia of brain tissue. Studying these mechanisms will help clarify the reasons for the increased risk of stroke in atrial fibrillation and provide new biomarkers for future intervention measures.

[0018] (2) Constructing a prediction model for stroke risk

[0019] By quantitatively and qualitatively analyzing high-throughput metabolomics data and combining it with machine learning, this study established a stroke risk prediction model. This model comprehensively analyzes multiple variables and identifies biomarkers closely related to the occurrence of stroke. Using this model enables doctors to more accurately assess the stroke risk of individual patients, thereby formulating corresponding management strategies.

[0020] (3) Mining intervention targets for stroke

[0021] In the present invention, by comparing the metabolomic characteristics of normal atrial fibrillation patients and those secondary to stroke, several key differential metabolites were identified. By tracking the effects of drugs at the metabolic level, new targets can be identified, improving the targeting and effectiveness of drugs. For example, the increase in certain lipid metabolites may be associated with a procoagulant state, and interventions targeting these metabolites may reduce the risk of stroke. This finding provides a theoretical basis for the future development of targeted drugs or therapies.

[0022] (4) Early detection of metabolites related to stroke

[0023] Carrying out early detection of metabolites related to stroke helps to timely detect potential stroke high-risk patients with metabolic abnormalities. By regularly monitoring the changes in these metabolites, dynamic assessment of stroke risk can be achieved, providing real-time basis for clinical decision-making and improving the quality of patient care.

[0024] (5) Early intervention for high-risk stroke patients

[0025] For stroke patients identified as high-risk after risk assessment, this study emphasizes the importance of early intervention. By implementing early lifestyle interventions, drug treatments, or regular follow-up for these patients, the incidence of stroke can be effectively reduced. This prospective intervention strategy can not only improve the prognosis of atrial fibrillation patients but also relieve the burden on the medical system.

[0026] (6) Implementing precision medicine

[0027] This invention promotes the implementation of the concept of precision medicine and emphasizes the importance of individualized medication strategies. Through metabolomics analysis, doctors can adjust treatment plans according to the specific metabolic characteristics of each patient. For example, in response to changes in metabolite concentrations, doctors can optimize drug doses, which not only improves the effectiveness of treatment but also reduces side effects. The individualized medical strategy enables patients to receive more precise treatment, ultimately achieving the goals of improving clinical outcomes and enhancing patient satisfaction. Description of the drawings

[0028] Figure 1 It is the OPLS-DA analysis diagram of metabolites (upper figure: positive ion mode; lower figure: negative ion mode);

[0029] Figure 2 It is the graph of the mean squared error of Lasso regression varying with Log(λ);

[0030] Figure 3 It is the curve graph of the regression coefficient varying with Log(λ);

[0031] Figure 4 It is the ROC curve graph of the Lasso regression model;

[0032] Figure 5 It is the expression quantity graph of 7-hydroxy-6-methoxy-2H-chromen-2-one in the AF group and the AFS group;

[0033] Figure 6 It is the expression quantity graph of deoxyadenosine in the AF group and the AFS group;

[0034] Figure 7 It is the expression quantity graph of arachidonic acid in the AF group and the AFS group;

[0035] Figure 8 It is the expression quantity graph of glycerophosphocholine in the AF group and the AFS group;

[0036] Figure 9 It is the expression quantity graph of eicosapentaenoic acid in the AF group and the AFS group;

[0037] Figure 10Expression level diagram of fumaric acid in the AF group and the AFS group;

[0038] Figure 11 Expression level diagram of thromboxane B3 in the AF group and the AFS group. Specific implementation manners

[0039] The present invention will be further described below in conjunction with specific embodiments, and the advantages and features of the present invention will become clearer as the description progresses. However, it should be understood that the described embodiments are exemplary only and do not constitute any limitation to the scope of the present invention. Those skilled in the art should understand that the details and forms of the technical solutions of the present invention can be modified or replaced without departing from the spirit and scope of the present invention, but such modifications or replacements all fall within the protection scope of the present invention.

[0040] Example 1 Screening and identification of plasma metabolites for predicting atrial fibrillation-related stroke

[0041] During the period from February 2015 to December 2017, a total of 90 patients with persistent AF without a history of stroke and 66 patients with AF with a history of stroke were collected from the First Affiliated Hospital of Harbin Medical University. Patients with malignant tumors, combined infections, or severe liver and kidney function impairments were excluded. The demographic data, clinical data, etc. of the above research subjects are shown in Table 1 for details.

[0042] Table 1 Description of the clinical characteristics of the participants

[0043]

[0044] Note: For continuous variables, normally distributed variables are expressed as mean ± standard deviation (SD), and categorical variables are expressed as frequency (percentage).

[0045] 1. Differential metabolic profile analysis of 90 AF patients and 66 AFS patients

[0046] The relevant results were obtained by metabolomics detection of the above 90 eligible AF patients and 66 AFS patients. The specific experimental methods are as follows:

[0047] 1.1 Sample extraction

[0048] After fasting for 12 hours, whole blood samples were collected from each participant using vacuum blood collection tubes. The fresh blood samples were centrifuged, and the supernatant was carefully collected and stored at -80°C for future experiments. (The samples used for the study were collected during the same period, and the sampling, aliquoting, and storage conditions were uniform)

[0049] (1) Take 100 μL of the sample, add 500 μL of the extraction solution containing an internal standard (pre-cooled 80% methanol and 0.1% formic acid), and vortex for 30 seconds;

[0050] (2) Incubate the sample on ice for 5 minutes;

[0051] (3) Centrifuge the sample at 4°C and 12,000 rpm for 20 min;

[0052] (4) Dilute some of the supernatant with LC-MS grade water to a final concentration of 53% methanol.

[0053] (5) Carefully pipette 500 μL of the supernatant into an EP tube;

[0054] (6) Centrifuge the sample at 4°C and 12,000 rpm for 20 min;

[0055] (7) Pipette 120 μL of the supernatant into a 2 mL injection vial, and take 10 μL from each sample to mix into a QC sample for on-machine detection.

[0056] 1.2 On-machine detection

[0057] The liquid chromatography-mass spectrometry system for metabolomics analysis consists of a Vanquish UHPLC system (Thermo Fisher, Germany), an ultra-high performance liquid chromatography tandem Orbitrap Q ExactiveTM HF-X mass spectrometer (Thermo Fisher, Germany), and a high-resolution mass spectrometer. The sample is injected at a flow rate of 0.2 mL / min for 17 min using a linear gradient and injected into a Hypesil Gold chromatographic column (100 * 2.1 mm, 1.9 μm).

[0058] Positive ion mode (POS): Mobile phase A: 0.1% formic acid aqueous solution; Mobile phase B: methanol

[0059] Negative ion mode (NEG): Mobile phase A: 5 mM ammonium acetate (pH 9.0) solution; Mobile phase B: methanol.

[0060] Create the gradient of the solvent in the following manner: 2% B for 1.5 minutes: At the beginning, use 2% of solvent B and hold for 1.5 minutes. 2% B - 100% B for 12.0 minutes: Over the next 12 minutes, gradually increase the concentration of solvent B from 2% to 100%. 100% B for 14.0 minutes: Keep the concentration of solvent B at 100% for 14 minutes. 100% B - 2% B for 14.1 minutes: Then, within 14.1 minutes, gradually decrease the concentration of solvent B from 100% back to 2%. 2% B until 17 minutes: Finally, keep the concentration of solvent B at 2% until the total experimental duration reaches 17 minutes.

[0061] The Q ExactiveTM HF-X mass spectrometer was used in positive / negative ion mode with a spray voltage of 3.2 kV, a capillary temperature of 320 °C, a sheath gas flow rate of 40 Arb, and an auxiliary gas flow rate of 10 Arb.

[0062] 1.3 Qualitative and quantitative analysis of metabolites

[0063] The raw data files generated by UHPLC-MS / MS were processed using Compound Discoverer 3.1 (CD3.1, Thermo Fisher), and peak alignment, peak picking, and quantification were performed for each metabolite. The set parameters included a retention time tolerance of 0.2 minutes, an actual mass tolerance of 5 ppm, a signal intensity tolerance of 30%, a signal-to-noise ratio of 3, and a minimum intensity, etc. Subsequently, the peak intensity was normalized according to the overall spectral intensity. The normalized data was used to predict the molecular formula based on the adduct ions, molecular ion peaks, and fragment ions. After that, the peaks were matched with the mzCloud (https: / / www.mzcloud.org / ), mzVault, and MassList databases to obtain accurate qualitative and relative quantitative results.

[0064] 1.4 Results of metabolomics analysis

[0065] The relative standard deviation (RSD) of metabolites was calculated according to QC, and all metabolites met the requirement of RSD < 30%, and the distribution is shown in Table 2. Subsequently, a t-test was performed on the plasma metabolic characteristics of 90 AF patients and 66 AFS patients to calculate their Fold Change (FC) values, and orthogonal partial least squares-discriminant analysis (OPLS-DA) was carried out. The established OPLS-DA model ( Figure 1 ) was used to determine the importance of each metabolite in the projection (Variable Importance in Projection, VIP). Metabolites meeting the following criteria were regarded as differentially expressed metabolites (DEMs): p < 0.05, FC > 1.2 or < 0.83, and VIP > 1. We screened out 51 DEMs in 90 AF patients and 66 AFS (Table 3).

[0066] Table 2 RSD distribution of 671 metabolites

[0067] RSD(%) Positive ion mode Positive and negative ion modes RSD < 5 127 139 5 ≤ RSD < 10 87 70 10 ≤ RSD < 15 69 40 15 ≤ RSD < 20 43 15 20 ≤ RSD < 25 39 14 25 ≤ RSD < 30 23 5 Total 388 283

[0068] Table 3 51 plasma differential metabolites between the two groups of AF and AFS

[0069]

[0070]

[0071]

[0072]

[0073]

[0074]

[0075] 2. Construction and validation of the AFS prediction model

[0076] 2.1 Construction of the AFS prediction model

[0077] Construct a Lasso regression model to determine the AFS predictors. In order to estimate the individual-specific occurrence risk of AFS. Plot the receiver operating characteristic (ROC) curve to evaluate the prediction accuracy of the model.

[0078] In this example, the samples were randomly divided into a training set and a validation set at a ratio of 1:1 (that is, each of the training set and the test set contains 45 AF patients and 33 AFS patients), without repetition. Set the random seed number, and use the glmnet package in the R software package for Lasso regression analysis. Adopt 5-fold cross-validation to select the predictor variables from 51 plasma differential metabolites with significant differences( Figure 2 ), Lambda.1se gives a model with good performance but the least number of independent variables( Figure 3 ). 7 out of the initial 51 variables were included in the prediction model, namely 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, thromboxane B3.

[0079] 2.2 Validation of the AFS prediction model

[0080] To further verify the stability of the model, the relative content values of these 7 metabolites (7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, thromboxane B3) were re-entered into the training set of the Lasso model for training to obtain the final model. Among them, AUC = 0.942 (88.5%-99.8%)( Figure 4 ), indicating that these 7 biomarkers can accurately predict the risk of AFS. The expression levels of the 7 metabolites in the two groups of AF patients and AFS patients are as Figures 5 to 11As shown, it can be seen from the results that compared with patients with atrial fibrillation, the expression levels of 7-hydroxy-6-methoxy-2H-chromen-2-one, glycerophosphocholine, and thromboxane B3 in the plasma of patients with atrial fibrillation-related stroke were significantly decreased, while the expression levels of deoxyadenosine, arachidonic acid, eicosapentaenoic acid, and fumaric acid were significantly increased.

[0081] 3. Conclusion

[0082] In summary, the present invention quantitatively and qualitatively analyzed the plasma metabolite profiles of 90 AF patients and 66 AFS patients by LC-MS method, and a total of 671 metabolites were identified. By calculating the p-value through the T-test, calculating the FC value, and calculating the VIP value through OPLS-DA analysis, 51 DEMs (FC > 1.2 or < 0.83, p < 0.05, VIP > 1) were finally screened out, and these 51 DEMs were incorporated into the Lasso regression model to identify 7 biomarkers - 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3. These 7 biomarkers can accurately predict the risk of AFS (AUC = 0.94), thus helping clinicians to take targeted preventive measures in a timely manner. This will help improve the prognosis of patients with atrial fibrillation and enhance their quality of life. In addition, the present invention can also be applied to the study of the metabolic mechanism of secondary stroke in atrial fibrillation, the exploration of intervention targets for stroke, the implementation of precision medicine, and other aspects. Therefore, the present invention has important clinical application value and practical significance.

Claims

1. A plasma metabolic composition for predicting atrial fibrillation-related stroke, characterized in that, The plasma metabolic composition described above is composed of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3.

2. Use of the plasma metabolic composition according to claim 1 in the preparation of a reagent for predicting atrial fibrillation-related stroke.

3. The application according to claim 2, wherein Compared with patients with atrial fibrillation, the expression levels of 7-hydroxy-6-methoxy-2H-chromen-2-one, glycerophosphocholine, and thromboxane B3 in the plasma of patients with atrial fibrillation-related stroke are significantly decreased, while the expression levels of deoxyadenosine, arachidonic acid, eicosapentaenoic acid, and fumaric acid are significantly increased.

4. Use of a reagent or device for detecting the contents of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3 in the preparation of a reagent or device for predicting atrial fibrillation-related stroke.

5. The application according to claim 4, characterized in that Compared with patients with atrial fibrillation, the expression levels of 7-hydroxy-6-methoxy-2H-chromen-2-one, glycerophosphocholine, and thromboxane B3 in the plasma of patients with atrial fibrillation-related stroke are significantly decreased, while the expression levels of deoxyadenosine, arachidonic acid, eicosapentaenoic acid, and fumaric acid are significantly increased.

6. A reagent combination for predicting atrial fibrillation-related stroke, characterized in that, The reagent combination described above is composed of reagents for respectively detecting the contents of plasma metabolites 7-hydroxy-6-methoxy-2H-chromen-2-one, deoxyadenosine, arachidonic acid, glycerophosphocholine, eicosapentaenoic acid, fumaric acid, and thromboxane B3.

7. Use of the reagent combination according to claim 6 in the preparation of a reagent for predicting atrial fibrillation-related stroke.

8. The application according to claim 7, wherein Compared with patients with atrial fibrillation, the expression levels of 7-hydroxy-6-methoxy-2H-chromen-2-one, glycerophosphocholine, and thromboxane B3 in the plasma of patients with atrial fibrillation-related stroke are significantly decreased, while the expression levels of deoxyadenosine, arachidonic acid, eicosapentaenoic acid, and fumaric acid are significantly increased.

Images