Application of 24S-hydroxycholesterol as cerebral arterial thrombosis biomarker

Abstract

The invention belongs to the technical field of biomedicine, and particularly discloses application of 24S-hydroxycholesterol as a biomarker in preparation of an in-vitro diagnosis preparation for auxiliary diagnosis of cerebral arterial thrombosis. Through targeted cholesterol lipidomics analysis, it is found that the expression level of 24S-hydroxycholesterol in peripheral blood of a patient suffering from acute ischemic stroke is remarkably increased. On the basis, the invention claims application of 24S-hydroxycholesterol in preparation of in-vitro diagnostic reagents or monitoring products for cerebral arterial thrombosis diagnosis, risk assessment, severity grading, curative effect evaluation and prognosis judgment. The marker provides a brand new quantification means for early diagnosis and disease assessment of cerebral apoplexy, can also be used as a potential action target for drug screening and development, and has an important clinical transformation value.

Description

Technical Field

[0001] This invention belongs to the field of biomedical technology, specifically relating to a novel biomarker and its application, particularly the application of 24S-hydroxycholesterol (24S-HC) in the preparation of in vitro diagnostic agents for the auxiliary diagnosis of ischemic stroke (especially acute ischemic stroke [AIS]). Background Technology

[0002] Ischemic stroke is one of the leading causes of death and disability worldwide. Its treatment is highly dependent on a specific "time window," thus requiring extremely high accuracy and timeliness in early identification. Currently, clinical diagnosis primarily relies on neuroimaging examinations such as computed tomography (CT) and magnetic resonance imaging (MRI). However, CT has limited sensitivity for early ischemic changes (approximately 16%), while MRI, although more sensitive, still has a false negative rate of about 20%, and imaging equipment is not readily available in all medical institutions (especially in pre-hospital emergency settings). This makes it difficult for some patients with atypical symptoms or those undergoing transport to receive a rapid and reliable diagnosis, delaying optimal treatment.

[0003] Therefore, there is an urgent clinical need to develop peripheral blood biomarkers that are deeply linked to disease mechanisms and suitable for rapid detection, serving as a powerful complement to existing imaging tools. In recent years, the development of lipidomics technology has provided a powerful tool for understanding disease mechanisms and discovering novel biomarkers at the metabolic level. In the pathological process of ischemic stroke, lipid metabolism in the central nervous system undergoes significant remodeling, with cholesterol oxidative metabolites receiving increasing attention as key signaling molecules and metabolic indicators.

[0004] 24S-HC is a major cholesterol metabolite produced in the brain by the neuron-specific enzyme CYP46A1, and is crucial for maintaining cholesterol homeostasis. Previous studies have suggested changes in 24S-HC in neurological diseases such as Alzheimer's disease and multiple sclerosis, indicating its potential as a neurological-specific biomarker. However, in the field of acute ischemic stroke, systematic research and definitive conclusions regarding the dynamic changes of 24S-HC in the peripheral blood of patients, the strength of its clinical association with stroke risk, and its potential causal mechanisms are lacking. Current technologies have not yet revealed the specific elevation of 24S-HC in the peripheral circulation of AIS patients and its excellent early diagnostic value, nor have they combined it with machine learning models to construct high-precision auxiliary diagnostic tools, nor have they explored its potential causal relationship with stroke risk from a genetic perspective (such as Mendelian randomization analysis). Therefore, exploring the systematic application of 24S-HC in ischemic stroke fills a gap in this field. Summary of the Invention

[0005] (a) Purpose of the invention The primary objective of this invention is to provide 24S-HC as a novel, highly specific and sensitive biomarker for multi-stage management of ischemic stroke, in order to address the problems of insufficient timeliness and limited accessibility of existing diagnostic technologies in certain scenarios.

[0006] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: Firstly, this invention provides the application of 24S-HC as a biomarker for ischemic stroke. Specifically, it relates to the application of 24S-HC in the preparation of in vitro diagnostic agents for the auxiliary diagnosis of ischemic stroke. Based on targeted lipidomics analysis of a clinical cohort of AIS patients, this invention is the first to discover and demonstrate that the level of 24S-HC in the peripheral blood of AIS patients is statistically significantly upregulated compared to healthy controls (HC). This change is closely related to the occurrence and development of the disease. Based on this finding, 24S-HC can be used to prepare in vitro diagnostic reagents, kits, or monitoring products for the following purposes: early auxiliary diagnosis of ischemic stroke; risk assessment and early warning of individual ischemic stroke; objective grading of patient disease severity; dynamic monitoring and evaluation of clinical treatment effects; and prediction of patient prognosis.

[0007] Furthermore, this invention discloses the application of 24S-HC in constructing an algorithmic model for assessing the risk of ischemic stroke or for auxiliary diagnosis. This invention constructs an early diagnosis model for ischemic stroke (AIS) with 24S-HC as the core parameter using machine learning algorithms such as LASSO regression and random forest. The model demonstrates extremely high discriminative power (AUC value exceeding 0.92) in validation, significantly outperforming traditional clinical indicators, proving the core value of 24S-HC in constructing a high-precision intelligent diagnostic system. Specifically, the algorithmic model is a random forest model constructed using 10 feature variables as inputs: 24S-HC, 5β,6β-epoxycholesterol, 6α-hydroxy-5α-cholesterol, cholesterol-4,6-dien-3-one, history of atrial fibrillation, history of coronary heart disease, blood glucose level, weight, lipoprotein(a), and high-density lipoprotein cholesterol.

[0008] Secondly, this invention provides the application of 24S-HC as a drug target. The 24S-HC can be used to screen or develop drugs for the prevention or treatment of ischemic stroke. Intervention targeting its metabolic pathways (such as regulating the activity of its synthase CYP46A1) or signal transduction pathways can provide clear directions for new drug development.

[0009] (III) Beneficial Effects Compared with the prior art, the technical solution disclosed in this invention has the following significant effects and advantages: 1. High biomarker specificity and clear mechanism association: 24S-HC is a neuron-specific metabolite, and changes in its peripheral blood level directly reflect the pathological state of the central nervous system. It is deeply associated with the core mechanism of lipid metabolism disorder in ischemic stroke. The biomarker has a clear source and high specificity.

[0010] 2. Excellent early diagnostic performance with significant clinical value: This invention is the first to demonstrate in a case-control study that the level of 24S-HC in the peripheral blood of AIS patients is significantly upregulated in the early stages. The diagnostic model built based on this core feature shows extremely high discriminative power, providing a rapid and reliable blood testing aid in scenarios where imaging examinations are limited or cannot be performed in a timely manner (such as pre-hospital emergency care and primary care at the grassroots level).

[0011] 3. Broad application prospects, covering key aspects of disease: The applications proposed in this invention are not limited to early auxiliary diagnosis of diseases, but also extend to early warning of disease risk, assessment of disease severity, monitoring of treatment effects and prognosis. At the same time, it provides direction as a new target for drug development, realizing full-cycle coverage from prevention and diagnosis to treatment evaluation, and has clear clinical translation potential.

[0012] 4. Solid scientific foundation, providing multi-dimensional evidence: The conclusions of this invention are supported by multi-level research evidence. In addition to the findings of the clinical cohort, Mendelian randomization analysis provides preliminary evidence at the genetic level for a potential causal association between the oxidative sterol metabolic pathway and the risk of ischemic stroke, further enhancing the scientific persuasiveness and reliability of 24S-HC as a biomarker. Attached Figure Description

[0013] Figure 1 Classification and statistical chart of serum oxidized cholesterol metabolism profiles in AIS patients and HC population.

[0014] Figure 2 Principal component analysis score plot of quality control samples of serum oxidized cholesterol metabolomics data.

[0015] Figure 3 Correlation and distribution assessment of oxidized cholesterol detection signals among quality control samples.

[0016] Figure 4 Orthogonal partial least squares discriminant analysis (OPLS-DA) of serum oxidized cholesterol metabolic profiles in AIS patients and HC populations A. OPLS-DA score plot (showing between-group separation); B. Model interpretability, predictive evaluation, and permutation test results; C. Sample outlier detection graph; D. AIS discriminant ROC curve based on the OPLS-DA model.

[0017] Figure 5 Cluster heatmap of serum oxidized cholesterol metabolism profiles in AIS patients and HC populations.

[0018] Figure 6 Absolute quantitative statistical chart of representative oxidized cholesterol in serum of AIS patients and HC population A. Comparison of 24S-hydroxycholesterol levels between groups; B. Comparison of FF-MAS content between groups; C. Comparison of 27-hydroxycholesterol content between groups; D. Comparison of 7-ketocholesterol content between groups.

[0019] Figure 7 A flowchart illustrating the screening process for characteristic variables in the early diagnosis of ischemic stroke based on LASSO regression. A. Determine the optimal penalty parameter (λ) using 10-fold cross-validation; B. Characteristic variable coefficient path diagram.

[0020] Figure 8 Construction and evaluation of machine learning models for early diagnosis of ischemic stroke A. Key feature variables identified by LASSO regression; B. Comparison of ROC curves for the discriminative performance of Random Forest and Support Vector Machine models.

[0021] Figure 9 Mendelian randomized causal inference analysis diagram of oxidative cholesterol metabolism and ischemic stroke risk A. Effect values ​​(OR and 95% CI) of oxidized sterols on the risk of AIS estimated by different MR methods. B. Scatter plot of effect sizes for the instrumental variable SNP; C. Forest plot showing the direction of effect of each SNP site. Detailed Implementation

[0022] The embodiments of the present invention are further described below with reference to the accompanying drawings. This detailed description should not be considered as a limitation of the present invention, but should be understood as a more detailed description of certain aspects, features and embodiments of the present invention.

[0023] It should be understood that the terminology used in this invention is merely for describing particular embodiments and is not intended to limit the invention. Furthermore, with respect to numerical ranges in this invention, it should be understood that each intermediate value between the upper and lower limits of the range is also specifically disclosed. Every smaller range between any stated value or intermediate value within a stated range, and any other stated value or intermediate value within said range, is also included in this invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0024] Unless otherwise stated, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. While only preferred methods and materials have been described herein, any methods and materials similar or equivalent to those described herein may be used in the implementation or testing of this invention. All references to this specification are incorporated by way of citation to disclose and describe methods and / or materials associated with those references. In the event of any conflict with any incorporated reference, the content of this specification shall prevail.

[0025] Various modifications and variations can be made to the specific embodiments described in this specification without departing from the scope or spirit of the invention, as will be apparent to those skilled in the art. Other embodiments derived from this specification will also be readily apparent to those skilled in the art. This application specification and embodiments are merely exemplary.

[0026] The terms “include,” “including,” “have,” “contain,” etc., used in this article are all open-ended terms, meaning that they include but are not limited to.

[0027] Example 1: Detection of 24S-HC in the serum of AIS patients and its validation as a diagnostic marker 1. Research Subjects and Sample Collection This study was reviewed and approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University (Approval No.: 2021-SR-187), and all participants signed informed consent forms.

[0028] (1) AIS patient group: Patients with acute ischemic stroke confirmed by imaging (CTA / MRA) of occlusion of the middle cerebral artery (M1 or M2 segment) were recruited. Inclusion criteria: age > 18 years, onset time < 24 hours. Exclusion criteria: hyperlipidemia, type 2 diabetes, cerebral hemorrhage, active malignancy, hematologic disorders, chronic inflammatory diseases or recent infection history; previous use of lipid-lowering drugs.

[0029] (2) Healthy control group: recruit healthy volunteers of the same age and sex, with no history of neurological diseases or diseases listed in the above exclusion criteria.

[0030] All subjects received 3-5 mL of fasting venous blood immediately upon admission (patient group) or during physical examination (control group) using a separation gel coagulation tube. After the blood samples were allowed to stand at room temperature for 30 min, they were centrifuged at 4°C and 3000 r / min for 20 min, and the serum was carefully separated. The obtained serum was immediately flash-frozen in liquid nitrogen and stored in an ultra-low temperature freezer at -80°C until analysis.

[0031] 2. Sample pretreatment and detection for targeted lipidomics Lipids were extracted from serum using a modified Bligh-Dyer method. The simplified steps are as follows: (1) Take 100 μL of serum and add a mixed solution containing deuterated internal standards (including d6-lanosterol, d5-enzymes, d7-lathosterol, d7-7-dehydrocholesterol, d7-sitosterol, d6-cholesterol, d7-24-hydroxycholesterol, d7-7β-hydroxycholesterol, d6-25-hydroxycholesterol, d6-27-hydroxycholesterol, d7-7-ketocholesterol, d7-7α-hydroxycholesterolone, d6-TMAS, d7-4β-hydroxycholesterol, d6-24,25-epoxycholesterol, d7-demethylcholesterol, and d3-3β-7α-dihydroxycholesterol-5-enoic acid).

[0032] (2) Add chloroform:methanol (2:1, v / v) mixture, vortex and centrifuge to separate the layers.

[0033] (3) Collect the lower organic phase and dry it with nitrogen.

[0034] (4) Derivatize with 50 μL pyridine and 50 μL acetic anhydride at 60 °C for 1 hour, dry with nitrogen, and then reconstitute with 100 μL acetonitrile:isopropanol (1:1, v / v).

[0035] Analysis was performed using an ultra-high performance liquid chromatography-tandem mass spectrometry (UHPLC-MS / MS) system. Chromatographic conditions: Waters ACQUITY UPLC HSS T3 column (2.1 × 100 mm, 1.8 μm); mobile phase A was an aqueous solution containing 5 mM ammonium formate, and mobile phase B was an acetonitrile:isopropanol (9:1, v / v) solution containing 5 mM ammonium formate; gradient elution. Mass spectrometric conditions: electrospray ionization source, positive ion mode, multiple reaction monitoring (MRM).

[0036] 3. Data Processing and Biomarker Screening The raw data were processed using Analyst 1.6.3 software, and quantification was performed using the internal standard method.

[0037] (1) First, the data is preprocessed, including missing value imputation (replacing with 1 / 2 of the minimum value) and standardization (Pareto scaling).

[0038] (2) Principal component analysis was used to preliminarily observe the overall metabolic differences between the AIS group and the healthy control group.

[0039] (3) Further orthogonal partial least squares discriminant analysis was used to screen metabolites with a variable projection importance value greater than 1.0 and statistically significant differences between groups as candidate biomarkers.

[0040] 4. Results (1) Metabolic profiling and discovery of key biomarkers Targeted lipidomics assays and data analysis revealed overall differences in serum oxidative sterol metabolic profiles between the AIS and HC patient groups. Principal component analysis showed a trend of separation in the spatial distribution of metabolic profiles between the two groups (see [link to study]). Figure 2 To reinforce this finding and screen key biomarkers, a supervised learning-based orthogonal partial least squares discriminant analysis model was further constructed. This model demonstrated excellent discriminant ability, achieving an area under the receiver operating characteristic curve (AUC) of 0.925 (see [link to model].) Figure 4 D). Variable importance projection analysis identified 24S-HC as one of the core differentially expressed metabolites between the two groups (VIP>1). Quantitative comparison confirmed that the serum 24S-HC level in AIS patients was significantly higher than that in the healthy control group (P<0.05) (see [link to relevant documentation]). Figure 6 A).

[0041] (2) Diagnostic model construction and internal validation To evaluate the diagnostic value of 24S-HC and build a better classification model, this invention employs machine learning methods for modeling and internal validation: ① Feature Variable Screening: The LASSO regression algorithm was used to screen all detected oxidized sterol metabolites and collected clinical variables. The optimal penalty parameter λ was determined through 10-fold cross-validation, ultimately resulting in 10 stable key feature variables. These variables specifically include: Oxidized sterol metabolites (4 items): 24S-HC, 5β,6β-epoxycholesterol, 6α-hydroxy-5α-cholesterol, and cholesterol-4,6-dien-3-one. Clinical indicators (6 items): history of atrial fibrillation, history of coronary artery disease, blood glucose level, weight, lipoprotein(a), and high-density lipoprotein cholesterol. Among these, 24S-HC had the largest absolute value of its regression coefficient, indicating its highest contribution to the model's judgment (see the screening process for details). Figure 7 and Figure 8 A).

[0042] ② Model Construction and Performance Evaluation: Using the 10 selected feature variables as inputs, support vector machines and random forest classifiers were constructed in the same study cohort, respectively. A 10-fold cross-validation strategy was employed to evaluate model performance, avoiding overfitting and providing robust performance estimates. Performance evaluation showed that the random forest model performed best, achieving an AUC of 0.92 for distinguishing AIS, significantly outperforming the support vector machine model (AUC = 0.81) and any single clinical indicator (see [link to relevant documentation]). Figure 8 B). This result demonstrates that the multivariate combination model constructed based on features such as 24S-HC has excellent differential diagnostic capabilities.

[0043] 5. Conclusion This embodiment demonstrates through a systematic clinical cohort study that 24S-HC levels are specifically and significantly upregulated in the peripheral blood of AIS patients. More importantly, the random forest machine learning model constructed using the 10 feature variables (including 4 oxidized sterol metabolites and 6 clinical indicators, with 24S-HC as the core) selected by the aforementioned LASSO regression showed high discrimination accuracy (AUC = 0.92) in cross-validation. This comprehensively validates the effectiveness and application potential of 24S-HC and its multivariate combination model with specific clinical indicators and metabolites as a novel diagnostic tool for ischemic stroke, from both the "discovery of key biomarkers" and the "construction of a high-precision diagnostic model," providing ample support for the related applications of this invention.

[0044] Example 2: A study on the genetic causal relationship between oxidative sterol metabolism and AIS risk based on Mendelian randomization analysis 1. Research Objectives To explore, from a genetic epidemiological perspective, whether there is a potential causal relationship between oxidative sterol metabolism (including the 24S-HC pathway) and the risk of AIS, this embodiment uses publicly available genetic data to conduct a two-sample Mendelian randomization analysis.

[0045] 2. Analytical Methods and Data Sources This study employs a two-sample Mendelian randomization analysis framework.

[0046] (1) Data source: Exposure data (peripheral blood oxidized cholesterol levels) were obtained from a genome-wide association study (GWAS) involving 13,814 European individuals. Outcome data (risk of developing AIS) were obtained from the public GWAS catalog (dataset ID: GCST90038613, which included 6,925 AIS cases and 477,673 controls).

[0047] (2) Instrumental variable screening: Screening for GWAS exposures that were significantly associated with oxidative sterol levels at the genome-wide level (P<5×10⁻).8 The single nucleotide polymorphisms (SNPs) were used as instrumental variables, and their compliance with the three core assumptions of MR analysis was ensured.

[0048] (3) Statistical methods: The inverse variance weighted method was mainly used to estimate causal effects, supplemented and validated by MR-Egger regression and weighted median method. Sensitivity analysis was performed using the MR-PRESSO method to identify and remove outliers and assess the robustness of the results.

[0049] 3. Results Mendelian randomization analysis showed a statistically significant positive association between elevated peripheral blood oxidative cholesterol levels predicted by genetic tools and an increased risk of AIS. The odds ratio estimated by the weighted median method was 1.01 (95% confidence interval: 1.00–1.01, P = 0.003). The inverse variance weighted method and MR-Egger regression also showed consistent effect directions. Scatter plots of the instrumental variables showed good clustering of effect points for each SNP. Figure 9 B), the forest plot further shows that the effect orientation of key SNP sites (such as rs2277119) is stable in different models (B). Figure 9 C). Sensitivity analysis did not detect significant level pleiotropy, supporting the reliability of the primary analytical conclusions (see details). Figure 9 AC).

[0050] 4. Conclusion This embodiment, through Mendelian randomization analysis, provides preliminary inferential evidence from a genetic perspective that "abnormal oxidative sterol metabolism is a potential causal factor increasing the risk of AIS." This finding, from the perspective of upstream genetic susceptibility, supports the hypothesis that the oxidative sterol metabolic pathway is involved in the pathogenesis of AIS, thereby further strengthening the biological rationale and scientific basis of 24S-HC as an AIS-related biomarker.

Claims

1. Use of 24S-hydroxycholesterol as a biomarker in the preparation of an in vitro diagnostic preparation for assisting in the diagnosis of ischemic stroke.

2. Use according to claim 1, characterized in that, The use is based on the significantly up-regulated expression level of 24S-hydroxycholesterol in the peripheral blood of the patient with ischemic stroke compared with healthy controls.

3. Use according to claim 1 or 2, characterized in that, The in vitro diagnostic preparation is used for any one or more of the following purposes: (1) early assisting in the diagnosis of ischemic stroke; (2) assessing the risk of an individual suffering from ischemic stroke; (3) grading the severity of the disease in patients with ischemic stroke; (4) monitoring or evaluating the therapeutic effect on ischemic stroke; (5) judging the prognosis of patients with ischemic stroke.

4. Use according to claim 1, characterized in that, The use comprises inputting the expression amount of 24S-hydroxycholesterol obtained by detection into an algorithm model for assessing the risk of ischemic stroke or assisting in the diagnosis trained by machine learning method for analysis.

5. Use of 24S-hydroxycholesterol as a drug action target in screening or developing a drug for preventing or treating ischemic stroke.

6. Use of 24S-hydroxycholesterol in constructing an algorithm model for assessing the risk of ischemic stroke or assisting in the diagnosis.

7. Use according to claim 6, characterized in that, The algorithm model is a random forest model constructed with 24S-HC, 5β, 6β-epoxycholesterol, 6ɑ-hydroxy-5ɑ-cholestanol, cholesta-4, 6-dien-3-one, history of atrial fibrillation, history of coronary heart disease, blood glucose level, body weight, lipoprotein (a), high-density lipoprotein cholesterol as input.

Images