Biomarkers for migraine diagnosis, kits and uses thereof

Abstract

The present application relates to the field of biological medicine, and particularly relates to biomarkers for migraine diagnosis, kits and application thereof. The biomarkers related to migraine disease according to the present application include COL4A2, and also include any one or a combination of multiple of MMP-14, LCAT, ADAMTS13 and CPXM2. The present application is based on proteomic data of clinical healthy controls and migraine patients, and uses bioinformatics analysis and machine learning methods to deeply mine protein combinations with the most early warning value for migraine disease diagnosis, and to verify in a clinical cohort with expanded samples, to provide a good prospect for clinical transformation of newly discovered biomarkers, and also to lay a foundation for subsequent mechanism research.

Description

Technical Field

[0001] This invention relates to the field of biomedicine, specifically to biomarkers, reagent kits, and applications for the diagnosis of migraine. Background Technology

[0002] Migraine is one of the leading causes of disability worldwide. According to the 2021 Global Burden of Disease (GBD) data, migraine accounts for 4.9% of global years lived with disability (YLDs). The disease is characterized by recurrent attacks of moderate to severe throbbing headaches, often accompanied by nausea, vomiting, photophobia, and phonophobia, severely impairing patients' quality of life and social function. Crucially, migraine has a clear chronic progressive nature, with an annual disability progression rate of 21.7%. Patients experiencing disease progression are prone to acute drug overdose, treatment resistance, and systemic degeneration of the brain's macroscopic structures, significantly exacerbating the disease's harm. Therefore, international headache experts strongly urge a shift in migraine prevention strategies from "passively responding to attacks" to "actively preventing disease progression" to halt the disease's deterioration.

[0003] Current clinical diagnosis and treatment face two major bottlenecks: firstly, the key driving mechanisms of chronic migraine progression are not fully understood; secondly, the lack of peripheral blood-specific biomarkers that can accurately assess disease progression hinders early identification and timely intervention of high-risk individuals. Previous studies have largely focused on central sensitization mediating migraine chronicity, but have struggled to explain the effectiveness of peripheral nerve blocks, and clinically effective monoclonal antibody drugs have difficulty crossing the blood-brain barrier. Our previous research and clinical practice have revealed that the positive rate of pericranial nerve tenderness in patients with frequent migraine attacks is as high as 44.8%, peripheral cold stimulation easily induces headaches, and after effective intervention to alleviate symptoms, pericranial nerve tenderness disappears. These results suggest that peripheral sensitization may be a key driving factor promoting the chronicity of migraine.

[0004] Based on proteomics data from clinical healthy controls and migraine patients (frequent migraine and chronic migraine), this study uses bioinformatics analysis and machine learning to deeply explore protein combinations that have the most predictive value for the diagnosis or progression of migraine disease. The results were validated in an expanded clinical cohort, providing a promising prospect for the clinical translation of newly discovered biomarkers and laying the foundation for subsequent mechanistic research. Summary of the Invention

[0005] Purpose of the invention: The technical problem to be solved by the present invention is to provide a biomarker, reagent kit and application for the diagnosis of migraine, which are in addition to the shortcomings of the prior art.

[0006] To address the aforementioned technical problems, this invention discloses a biomarker, reagent kit, and their applications for migraine diagnosis. The specific technical solution is as follows: In a first aspect, the present invention provides biomarkers associated with migraine disease, said biomarkers including COL4A2.

[0007] The biomarkers mentioned also include any one or more combinations of MMP-14, LCAT, ADAMTS13 and CPXM2.

[0008] In some embodiments of the present invention, the biomarkers include a combination of COL4A2, LCAT, ADAMTS13 and CPXM2.

[0009] In other embodiments of the invention, the biomarkers include a combination of COL4A2, MMP-14, LCAT, ADAMTS13, and CPXM2.

[0010] The specific information about the biomarkers is as follows: COL4A2: Collagen type IV alpha 2 chain, Uniprot serial number P08572; MMP-14: Matrix metalloproteinase-14, Uniprot sequence number is P50281; LCAT: Phosphatidylcholine-sterol acyltransferase, Uniprot sequence number P04180; ADAMTS13: A disintegrin and metalloproteinase with thrombospondinmotifs 13, Uniprot sequence number Q76LX8. CPXM2: Inactive carboxypeptidase-like protein X2, Uniprot sequence number Q8N436.

[0011] In a second aspect, the present invention provides a detection kit comprising reagents for detecting the expression level of the biomarkers described in the first aspect.

[0012] In some embodiments of the present invention, when the biomarker is COL4A2, the detection kit is a single-marker kit; when the biomarker is a combination of COL4A2, LCAT, ADAMTS13 and CPXM2, or a combination of COL4A2, MMP-14, LCAT, ADAMTS13 and CPXM2, the detection kit is a combination kit.

[0013] The detection kit includes specific amplification gene primers, specific recognition gene probes, or specific binding protein antibodies or ligands for detecting the expression level of the biomarkers described in the first aspect.

[0014] Thirdly, the present invention provides the application of the biomarkers described in the first aspect, or the detection kits described in the second aspect, in the preparation of diagnostic products for migraine diseases, wherein the products include reagents or kits.

[0015] The product described herein is used to detect the expression levels of biomarkers in test samples for non-diagnostic purposes.

[0016] The test samples include serum.

[0017] The migraines mentioned above include episodic migraines and / or chronic migraines.

[0018] Beneficial effects: Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention establishes a diagnostic method for migraine based on serum protein biomarkers, identifies highly specific predictive biomarkers, effectively fills research gaps both domestically and internationally, provides a powerful tool for diagnosing migraines in clinical practice, and offers important technical support for early intervention and mitigation of migraine progression; furthermore, the serum testing protocol provided in this application is easy to standardize and implement with quality control, and has better clinical scalability. 2. This invention employs a multi-marker joint detection strategy combining COL4A2, LCAT, ADAMTS13, and CPXM2, or COL4A2, LCAT, ADAMTS13, CPXM2, and MMP-14, significantly improving the diagnostic efficacy for migraine. Compared to the limited diagnostic efficacy and significant population heterogeneity of existing diagnostic markers (such as CGRP), this application focuses on the core aspects of disease progression. Through multi-dimensional protein marker combination modeling, it achieves high receiver operating characteristic (AUC), sensitivity, and specificity in an external independent validation cohort, demonstrating stronger discriminative ability and lower false positive risk for disease diagnosis. This provides a reliable basis for early identification of migraine progression risk and timely development and implementation of intervention programs. 3. This application constructs a complete technical route of "discovery and screening - feature optimization - targeted quantitative validation," and uses parallel response monitoring (PRM) for validation in an independent external cohort. Compared with existing technologies that only focus on non-targeted proteomics screening, this significantly reduces batch effects and the risk of model overfitting, and improves the reliability and clinical translation feasibility of biomarker combinations. Furthermore, multi-stage validation of candidate biomarkers shows consistent expression trends in both population groups, and the detection results have good reproducibility. 4. The biomarker combination screened in this application has mechanistic complementarity, involving multiple biological dimensions such as vascular basement membrane / extracellular matrix stability, lipid metabolism homeostasis, and endothelial-thrombosis-related processes. Compared with single-pathway biomarkers, which are easily affected by individual differences, the multi-dimensional combination of this application can maintain good robustness in patients with different pathological backgrounds, thereby improving the generalization ability of the detection system; 5. In practical applications, this method only requires detecting the expression levels of relevant proteins in serum and comparing them with preset thresholds to determine high-risk disease status and identify high-risk individuals at an early stage. This method does not rely on complex computational models or high-performance computing equipment and can be directly applied to routine clinical testing platforms (such as chemiluminescence systems), offering advantages such as ease of operation and widespread adoption. In summary, this application demonstrates significant improvements over existing technologies in terms of diagnostic objectivity, detection performance, technical reliability, and the breadth of mechanism coverage, and has promising prospects for clinical application. Attached Figure Description

[0019] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0020] Figure 1 The overall distribution of plasma protein expression characteristics in migraine patients and healthy controls (HC) is shown. In this diagram, A is a two-dimensional scatter plot based on principal component analysis, and B is a volcano plot of differentially expressed proteins.

[0021] Figure 2 Results of KEGG pathway enrichment analysis of differentially expressed plasma proteins in migraine patients and healthy controls.

[0022] Figure 3 The results of GO functional enrichment analysis of plasma differentially expressed proteins in migraine patients and healthy controls are shown. In this analysis, A represents biological processes, B represents molecular functions, and C represents cellular components.

[0023] Figure 4 This is a schematic diagram of a machine learning process used to screen candidate biomarkers.

[0024] Figure 5 This is the UMAP distribution map obtained from dimensionality reduction analysis based on differential protein expression data.

[0025] Figure 6 The figure shows the receiver operating characteristic (ROC) curves after the model was built based on candidate proteins, where A represents the training set result and B represents the test set result.

[0026] Figure 7 This represents the confusion matrix results of the model in the training and test sets, where A is the training set and B is the test set.

[0027] Figure 8 This is a scatter plot of the model's probability predictions on the training and test sets, where A is the training set and B is the test set.

[0028] Figure 9 This graph shows the trend of expression levels of eight candidate proteins in the training set of the discovery cohort, the test set of the discovery cohort, and the targeted mass spectrometry validation of the external validation cohort.

[0029] Figure 10 The ROC curves are for individual proteins and their combinations in the external validation cohort, where the predicted probabilities represent the ROC curves for four protein combinations.

[0030] Figure 11 To discover the ROC curves of individual proteins and their combinations in the cohort, where the predicted probabilities represent the ROC curves of four protein combinations.

[0031] Figure 12 The expression of matrix metalloproteinase family members in the peripheral blood of migraine patients is shown in Figure A, where A represents the expression in proteomics and B represents the expression in the peripheral blood of the external validation cohort detected by ELISA.

[0032] Figure 13 To discover the ROC curves of the five proteins individually and in combination in the cohort. Detailed Implementation

[0033] Unless otherwise specified, the experimental methods described in the following examples are conventional methods; the reagents and materials described are commercially available unless otherwise specified.

[0034] Example 1 1. Clinical data of migraine sufferers and healthy control group subjects This study included patients who visited the dizziness and headache clinic of the Department of Neurology at Nanjing Drum Tower Hospital from May 2024 to July 2025. Inclusion criteria for migraine patients were: age 18–65 years, regardless of gender; meeting the diagnostic criteria for migraine in the International Classification of Headache Disorders, Third Edition (ICHD-3), including the diagnostic criteria for episodic migraine and chronic migraine; able to independently complete and maintain a standardized headache diary; and having episodic migraine with 2–8 attacks per month and peripheral blood discharge.

[0035] The healthy control group (HC) consisted of non-migraine sufferers recruited during the same period who were willing to participate in this study and donated peripheral blood.

[0036] Ultimately, 30 healthy controls, 62 cases of episodic migraine, and 21 cases of chronic migraine were included as the discovery cohort.

[0037] This study was reviewed and approved by the Ethics Committee of Gulou Hospital Affiliated to Nanjing University School of Medicine (Ethics Review Approval Numbers: 2025-0143-01 and 2025-0558-01), and all participants signed written informed consent forms. The research process complied with the Declaration of Helsinki.

[0038] Hematological processing procedure: 3-4 mL of peripheral blood was drawn from all migraine patients during the migraine attack or remission period. The blood was allowed to stand at room temperature for 30 min to coagulate, then centrifuged at 3000 r / min for 10 min. The supernatant was collected into 1.5 mL centrifuge tubes, with each tube containing approximately 500 μL for later use. The serum was stored in a -80 ℃ refrigerator.

[0039] 2. Differential abundance analysis of plasma proteins between migraine patients and healthy controls After collecting serum, differential abundance analysis of plasma proteins was first performed between migraine patients and healthy controls (HC). Figure 1 Figure A shows a two-dimensional scatter plot based on principal component analysis, indicating differences in the distribution of plasma proteins between the two groups. Subsequent differential analysis of plasma proteins was performed. When the p-value was <0.05, a change in expression exceeding 1.5 was used as the threshold for significant upregulation, and a change less than 1 / 1.5 was used as the threshold for significant downregulation. This identified 344 upregulated proteins and 112 downregulated proteins. The volcano plot marks the five proteins with the most significant differences in upregulation and downregulation. Figure 1 (B in the middle).

[0040] 3. Functional enrichment analysis of differentially abundant proteins To systematically explore the potential functions of differentially expressed proteins, KEGG pathway enrichment analysis and GO functional enrichment analysis were performed on the differentially expressed proteins. The results showed that the occurrence and development of migraine are significantly associated with multiple biological processes. KEGG analysis showed that the differentially expressed proteins were significantly enriched in pathways associated with neurodegenerative diseases such as Parkinson's disease, Alzheimer's disease, Huntington's disease, and amyotrophic lateral sclerosis (ALS), while oxidative phosphorylation and the TCA cycle pathway were also significantly enriched (see [link to KEGG analysis]). Figure 2 GO enrichment analysis showed that, compared with healthy controls, the main biological functions of differentially enriched proteins in migraine patients were concentrated in mitochondrial function, particularly those related to the electron transport chain (see [link to analysis]). Figure 3 ).

[0041] Example 2 This embodiment utilizes machine learning to discover potential biomarkers. To screen for differentially expressed proteins with research value, multiple machine learning algorithms are employed. A discovery queue (with a training set to test set ratio of 7:3) is used to select proteins with higher weight values ​​for subsequent validation. The specific machine learning process is detailed below. Figure 4 UMAP analysis showed significant aggregation differences in protein expression between the migraine group and the healthy control group, and the two groups were effectively distinguishable (see...). Figure 5 The top 30 protein molecules selected through feature screening were analyzed using different machine learning algorithms, and ROC curves were used to evaluate model performance. The results show the ROC curves of the top 30 protein combinations for nine different model constructions on the training and test sets. Figure 6 In the training set, the AUC values ​​of each model are close to 1.00, demonstrating good discriminative ability. In the test set, the AUC values ​​range from 0.90 to 0.99, with RandomForestClassifier, GradientBoostingClassifier, and ExtraTreesClassifier performing the best (AUC≈0.99), indicating that the models have strong generalization ability.

[0042] Considering that the ExtraTreesClassifier model showed the best performance on the test set, with an AUC of 0.991 (RandomForestClassifier and GradientBoostingClassifier had AUCs of 0.987), the ExtraTreesClassifier model was subsequently constructed using the TOP30 proteins, and confusion matrix analysis was performed on both the training and test sets. Figure 7 ) and probability prediction analysis ( Figure 8To further evaluate the model's classification ability, the confusion matrix results showed that in the training set, there were 79 samples (21 negative and 58 positive). The model correctly classified all samples, correctly predicting 21 negative samples and 58 positive samples, with no misclassifications, indicating that the model has a very high fitting ability on the training data. In the independent test set, there were 34 samples (9 negative and 25 positive). Among them, 8 negative samples were correctly predicted and 1 was misclassified; 24 positive samples were correctly predicted and 1 was misclassified. The model's overall accuracy on the test set was 94.1% (32 / 34), sensitivity was 96.0% (24 / 25), and specificity was 88.9% (8 / 9). The results show that the model maintains high discriminative ability and stable generalization performance on independent data. Figure 8 The predicted probability distribution plot shown indicates that the two groups of samples in the training set, namely migraine patients and healthy controls, are completely separated; in the test set, most samples are distributed around the classification threshold (i.e., 0.5), with only a small number of boundary samples, which further verifies the stability and good generalization ability of the model.

[0043] Example 3: External Verification Queue PRM Verification This embodiment also included 62 healthy controls and 184 migraine patients (153 with episodic migraine and 31 with chronic migraine) as an external validation cohort, with the same ethical review approval number as in Embodiment 1. The inclusion criteria for the external validation cohort were the same as those for the discovery cohort in Embodiment 1.

[0044] Given that mitochondrial-associated proteins are affected by various factors, resulting in poor stability and a lack of specificity for migraine diagnosis, 13 mitochondrial-associated proteins were removed from the top 30 proteins. From the remaining 17 proteins, eight important candidate proteins (FBN1, IVL, COL4A2, CFHR2, CPXM2, LCAT, ADAMTS13, and PCDH18) related to migraine pathogenesis were validated using PRM-MS. PRM validation results showed that only four of the eight proteins (COL4A2, LCAT, ADAMTS13, and CPXM2) showed statistically significant differences in expression between migraine patients and healthy controls, and the expression changes were consistent with the trends observed in the cohort. Therefore, these four proteins are considered potential biomarkers for migraine. Figure 9 ).

[0045] The information for the eight candidate proteins is as follows: FBN1: Fibrillin-1, Uniprot sequence number P35555; IVL: Involucrin, an inner lining protein, Uniprot sequence number P07476; COL4A2: Collagen type IV alpha 2 chain, Uniprot serial number P08572; CFHR2: Complement factor H-related protein 2, Uniprot sequence number P36980; CPXM2: Inactive carboxypeptidase-like protein X2, Uniprot sequence number Q8N436; LCAT: Phosphatidylcholine-sterol acyltransferase, Uniprot sequence number P04180; ADAMTS13: A disintegrin and metalloproteinase with thrombospondinmotifs 13, Uniprot sequence number Q76LX8. PCDH18: Protocadherin-18, Uniprotocadherin 18, Uniprotocode number Q9HCL0.

[0046] ROC curves were then plotted for the relative quantification results of the four proteins (COL4A2, LCAT, ADAMTS13, and CPXM2) individually and in combination, as provided by PRM validation. The results showed that the combination of the four proteins exhibited good performance in the diagnosis of migraine, with an AUC of 0.975, a sensitivity of 0.896, and a specificity of 0.934. Figure 10 (See Table 1). In the discovery cohort, the combined diagnostic performance of these four proteins was also excellent ( ). Figure 11 (See Table 2). COL4A2 demonstrated excellent diagnostic efficacy, with an AUC of 0.946, a sensitivity of 0.918, and a specificity of 0.869, demonstrating high diagnostic and predictive value even as a single protein. In contrast, LCAT and ADAMTS13 exhibited relatively lower diagnostic efficacy (see Table 2). Figure 10 ).

[0047] Table 1. AUC area, sensitivity, and specificity data for individual proteins and their combinations in the external validation cohort.

[0048] Table 2. AUC area, sensitivity, and specificity data for individual proteins and their combinations in the discovery cohort.

[0049] Example 4 Previous studies have shown that COL4A2 is a core structural protein of the vascular basement membrane (type IV collagen α2 chain), and its core function is to maintain the integrity of the basement membrane. Matrix metalloproteinases (MMPs) are COL4A2 degrading enzymes. Considering the upregulation of COL4A2 expression in the peripheral blood of migraine patients, this application further performed differential analysis on MMPs in the cohort, finding that MMP-14 and MMP-19 expression were significantly increased, while MMP-9 was significantly decreased. Figure 12 As shown in A in the diagram.

[0050] The specific information about MMPs is as follows: MMP-14: Matrix metalloproteinase-14, Uniprot sequence number is P50281; MMP-19: Matrix metalloproteinase-19, Uniprot sequence number Q99542; MMP-9: Matrix metalloproteinase-9, Uniprot sequence number P14780.

[0051] ELISA validation using an external validation cohort showed that only MMP-14 levels were significantly higher in the peripheral blood of the migraine group than in the healthy control group, while there were no differences in MMP-9 and MMP-19 between groups (see [link to ELISA results]). Figure 12 (As shown in B in the figure), therefore, MMP-14 is also considered a potential biomarker for migraine. Its combination with four other biomarkers showed excellent efficacy in diagnosing migraine, with an AUC of 0.977, a sensitivity of 0.901, and a specificity of 1 (see Figure B). Figure 13 (and Table 3).

[0052] Table 3 shows the ROC curves of the five proteins and their combinations in the cohort, along with their corresponding AUC, sensitivity, and specificity.

[0053] In actual clinical applications, this invention detects the expression levels of COL4A2, LCAT, ADAMTS13 and CPXM2 in the peripheral serum of subjects, or detects the expression levels of COL4A2, LCAT, ADAMTS13, CPXM2 and MMP-14 in the peripheral serum of subjects, and compares the detection results with preset thresholds to determine whether the subjects are in a high-risk state for migraine.

[0054] The optimal cutoff value was determined using ROC curve analysis and the Youden index (as shown in Tables 1 and 3). When P ≥ cutoff, the subject was identified as high-risk for migraine; when P < cutoff, the subject was identified as low-risk. The detection method was chemiluminescence immunoassay or ELISA, both of which have good clinical operability and reproducibility. Unlike methods relying on complex machine learning models, the determination method provided by this invention can directly determine based on the expression levels of one or more proteins, making it easy to promote and apply in routine clinical testing environments.

[0055] This invention provides a biomarker, reagent kit, and application method for migraine diagnosis. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. Biomarkers associated with migraine disease, characterized in that, The biomarkers mentioned include COL4A2.

2. The biomarker according to claim 1, characterized in that, The biomarkers include any one or more combinations of MMP-14, LCAT, ADAMTS13, and CPXM2.

3. The biomarker according to claim 2, characterized in that, The biomarkers include a combination of COL4A2, LCAT, ADAMTS13, and CPXM2.

4. The biomarker according to claim 2, characterized in that, The biomarkers include a combination of COL4A2, MMP-14, LCAT, ADAMTS13, and CPXM2.

5. A test kit, characterized in that, It includes reagents for detecting the expression level of the biomarker as described in any one of claims 1 to 4.

6. The detection kit according to claim 5, characterized in that, It includes specific amplification gene primers, specific recognition gene probes, or specific binding proteins antibodies or ligands for detecting the expression level of the biomarkers described in any one of claims 1 to 4.

7. The use of the biomarker according to any one of claims 1 to 4, or the detection kit according to claim 5 or 6, in the preparation of a diagnostic product for migraine, wherein the product comprises a reagent or a kit.

8. The application according to claim 7, characterized in that, The product described herein is used to detect the expression levels of biomarkers in test samples for non-diagnostic purposes.

9. The application according to claim 8, characterized in that, The test samples include serum.

10. The application according to claim 8, characterized in that, The migraines mentioned include episodic migraines and / or chronic migraines.

Images