KR20230173319A - Biomarker for determining major depressive disorder, polar disorder and zophrenia based on mass spectrometry and its use - Google Patents

Abstract

We disclose a biomarker panel that can detect by mass spectrometry whether a subject has major mental illness such as depressive disorder, bipolar disorder, or schizophrenia in a blood sample. The panel according to this institute can provide information using blood, which can slow the progression of the disease, contributing to improving the patient's quality of life as well as reducing social and economic costs that may be added as the disease worsens.

Description

Biomarker for determining major depressive disorder, polar disorder and zophrenia based on mass spectrometry and its use}

This institute is concerned with biomarkers that can distinguish major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia.

Major mental disorders include major depressive disorder (MDD), bipolar disorder (BD), and schizophrenia (SPR). Major depressive disorder is characterized by a constant depressive state without rapid cycling, while bipolar disorder displays BD-1, BD-2 states, or BD-not otherwise specified (NOS) depending on the change in mood swings. . BD-1 represents a state of depression and strong manic symptoms (hypermanic symptoms), and BD-2 represents a state of depression and mild manic symptoms (hypomanic symptoms). Additionally, BD-NOS represents an asymptomatic state. Schizophrenia is characterized by hallucinations and hallucinations as well as cognitive impairment.

In the 2016 epidemiological survey of mental disorders in Korea, the prevalence of major depressive disorder was 5%, the annual prevalence was 1.5%, and the prevalence of bipolar disorder was 0.1%. In addition, according to the Health Insurance Review and Assessment Service, the number of depressive disorders increased by an average of 7.3% per year over the past four years from approximately 600,000 in 2015 to approximately 790,000 in 2019, and has continued to increase recently. In addition, according to the Health Insurance Review and Assessment Service's statistics on diseases of national interest, the number of patients with schizophrenia has been increasing every year for the past five years, with the number of patients increasing by about 6% in 2018 compared to 2014.

Because the diagnosis of patients with depressive disorder, bipolar disorder, and schizophrenia is made subjectively by clinicians by observing the patient's symptoms and behavior, it is difficult to accurately differentiate between these three disorders. In particular, bipolar disorder shows symptoms similar to major depressive disorder when depressive symptoms appear, and symptoms similar to schizophrenia when bipolar disorder shows mania. Additionally, severe depressive symptoms in depressive disorder and bipolar disorder may be mistaken for symptoms of schizophrenia. Approximately 40% of bipolar disorder patients are misdiagnosed as depressive disorder. Additionally, approximately 30% of bipolar disorder patients are misdiagnosed as schizophrenia. This misdiagnosis leads to incorrect prescriptions, which further worsens the symptoms of the disease.

Therefore, in order to solve these problems, there is a need for a method that can accurately distinguish the three diseases, but diagnosis is mainly performed through methods based on symptom and behavioral observation results or through interviews between patients and clinicians. Because the types of questionnaires used are diverse and the criteria for diagnosing depressive disorder, bipolar disorder, and schizophrenia are not unified, there is a possibility that different diagnostic results may be derived depending on the questionnaire. Because depressive disorder, bipolar disorder, and schizophrenia must be diagnosed more accurately to reduce the misdiagnosis rate between the two disorders, it is necessary to develop technology based on objective values rather than the current subjective observation-based diagnosis method.

Prior art

KR Patent Application Publication Publication 10-2022-0034708 (2022.03.18.)

JP Patent Application Publication 2021-185791 A (2021.12.13)

This study relates to biomarkers that can distinguish major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia based on mass spectrometry. We aim to provide biomarkers and methods.

In one aspect, the disclosure relates to a combination of biomarkers to distinguish, diagnose, or determine major mental disorders, including depressive disorders, bipolar disorders, and schizophrenia.

In another aspect, the present application relates to a biomarker composition for diagnosing or determining major mental disorders, including a substance for measuring the expression level of the biomarker combination by mass spectrometry.

In one embodiment, the biomarker combination is as follows:

The combination of biomarkers for diagnosing or determining depressive disorder and bipolar disorder is ALDOC (Fructose-bisphosphate aldolase C), ANPEP (Aminopeptidase N), ARMH4 (Armadillo-like helical domain-containing protein 4), and C1RL (Complement C1r subcomponent). -like protein), CDH13 (Cadherin-13), CETP (Cholesteryl ester transfer protein), COL10A1 (Collagen alpha-1), CTNND1 (Catenin delta-1), DDR1 (Epithelial discoidin domain-containing receptor 1), DBH (Dopamine) beta-hydroxylase ), SERPING1 (Plasma protease C1 inhibitor), IL1RAP (Interleukin-1 receptor accessory protein), ITIH2 (Inter-alpha-trypsin inhibitor heavy chain H2), NPC2 (NPC intracellular cholesterol transporter 2), RAN (GTP-binding) nuclear protein Ran), SAA1 (Serum amyloid A-1 protein), and TF (Serotransferrin), and the combination of biomarkers for diagnosing or determining depressive disorder and schizophrenia is ALDOC (Fructose-bisphosphate aldolase C), IGFALS ( Insulin-like growth factor-binding protein complex acid labile subunit ), NAGLU (Alpha-N-acetylglucosaminidase), ATP1A1 (Sodium/potassium-transporting ATPase subunit alpha-1), CTSS (Cathepsin S), SERPINA6 (Corticosteroid-binding globulin) , CPB2 (Carboxypeptidase B2), COL10A1 (Collagen alpha-1), CRYM (Ketimine reductase mu-crystallin), GPX3 (Glutathione peroxidase 3), IGFBP3 (Insulin-like growth factor-binding protein 3), IGFBP5 (Insulin-like growth) factor-binding protein 5), ITIH2 (Inter-alpha-trypsin inhibitor heavy chain H2), PGC (Gastricsin), PROC (Vitamin K-dependent protein C), PROS1 (Vitamin K-dependent protein S), RIDA (2-iminobutanoate) /2-iminopropanoate deaminase), SAA1 (Serum amyloid A-1 protein), SAA4 (Serum amyloid A-4 protein), and TFPI (Tissue factor pathway inhibitor), and are used to diagnose or determine bipolar disorder and schizophrenia. The biomarker combination includes SERPINA3 (Alpha-1-antichymotrypsin), ANPEPM (Aminopeptidase N), BPIFB1 (BPI fold-containing family B member 1), C1RL (Complement C1r subcomponent-like protein), CFB (Complement factor B), and CLDN3 ( Claudin-3), DBH (Dopamine beta-hydroxylase), GPR37 (Prosaposin receptor GPR37), SERPIND1 (Heparin cofactor 2), IGFBP5 (Insulin-like growth factor-binding protein 5), SERPING1 (Plasma protease C1 inhibitor), MBL2 ( Mannose-binding protein C), NPC2 (NPC intracellular cholesterol transporter 2), PLXNC1 (Plexin-C1), PSMD1 (26S proteasome non-ATPase regulatory subunit 1), TFPI (Tissue factor pathway inhibitor), and UMOD (Uromodulin).

In one embodiment, the mass spectrometry method for quantification of each biomarker in each combination according to the present disclosure includes tandem mass spectrometry, ion trap mass spectrometry, triple quadrupole mass spectrometry, hybrid ion trap/quadrupol mass spectrometry, or time-of-flight mass spectrometry. .

In other embodiments, the mode used for mass spectrometry includes Selected Reaction Monitoring (SRM) or Multiple Reaction Monitoring (MRM).

In another embodiment, the mass spectrometry mode is MRM, and the peptides of each biomarker used in the MRM analysis are listed in Table 1.

In another aspect, the hospital provides a method of providing information necessary for judgment or diagnosis by distinguishing between major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia, or distinguishes depressive disorder, bipolar disorder, and schizophrenia in vitro to provide the above information. This relates to a method of detecting biomarkers for each combination to determine this.

In the method according to the present application, more than one combination of biomarkers may be tested as needed.

In one embodiment, the method includes measuring or quantifying the expression level of each biomarker of each biomarker combination according to Table 1 from blood isolated from a subject by mass spectrometry; and correlating the measurement or quantitative result with depressive disorder, bipolar disorder, or schizophrenia.

In one embodiment, the mass spectrometry method includes tandem mass spectrometry, ion trap mass spectrometry, triple quadrupole mass spectrometry, hybrid ion trap/quadrupole mass spectrometry, or time-of-flight mass spectrometry.

In another embodiment, the mode used for mass spectrometry is Selected Reaction Monitoring (SRM) or Multiple Reaction Monitoring (MRM).

In another embodiment, the mass spectrometry mode is MRM, and the peptides for each protein constituting each biomarker combination used for analysis in the MRM are shown in Table 1.

In another embodiment, the method according to the present application includes score data of the major mental disorder symptom evaluation index [Symptom Checklist 90-revised (SCL-90R)] in addition to the biomarker detection or quantitative results for depressive disorder, bipolar disorder, and schizophrenia. Used for judging, diagnosing, or classifying diseases.

In one embodiment, the method herein further includes the step of obtaining score data of the subject's major mental disorder symptom evaluation index [Symptom Checklist 90-revised (SCL-90R)], and the biomarker detection results in the linking step. In addition, the above score data is used to determine depressive disorder, bipolar disorder, and schizophrenia.

In another embodiment, the linking step of the method herein involves substituting the biomarker detection results and major mental disorder symptom evaluation index score data into Equation I described below to obtain a result, and dividing the result into a threshold value determined for each biomarker combination. Compare.

The biomarker according to this institute diagnoses based on quantitative levels of protein compared to the current method of diagnosing major mental disorders, including depressive disorder, bipolar disorder, and schizophrenia, which is based on symptoms and behavioral observation results. Since the two diseases can be distinguished more objectively than the diagnostic method, the misdiagnosis rate between the two diseases can be reduced, enabling accurate diagnosis and prescription accordingly. In addition, the model developed at our center was able to distinguish disease subtypes of depressive disorder, schizophrenia, and bipolar disorder in addition to depressive disorder, bipolar disorder, and schizophrenia. In addition, an ensemble model was developed combining a proteomic multi-marker model and self-reported major mental illness symptom evaluation index clinical data variables, enabling differentiation and diagnosis of depressive disorder, bipolar disorder, and schizophrenia with improved accuracy.

Figure 1 is a schematic diagram of the core technical means of the present invention.
Figure 2 is a schematic diagram of the process of constructing a list of protein biomarker candidates for major mental disorders.
Figure 3 is a schematic diagram of blood (plasma) sample composition and pretreatment.
Figure 4 is a schematic diagram of 6490 Triple Quadrupole LC-MRM-MS analysis used in one embodiment of the present application.
Figure 5 is a schematic diagram of the development of the multiprotein-marker model and ensemble model used in the discovery of marker combinations at our institute.
Figures 6a to 6c show the discrimination performance results of the proteomic multi-marker model developed herein to distinguish MDD vs. BD, MDD vs. SPR, and BD vs. SPR, respectively. The proteomic multi-marker model that differentiates MDD from BD combined 17 protein markers. Among the 17 protein markers, 7 protein markers were upregulated in MDD and 10 protein markers were upregulated in BD. With 17 protein markers, the Training set discriminates between MDD and BD with a probability of 84%, and the Validation set, Independent test set, and Total set discriminate between MDD and BD with a probability of 73%, 74%, and 80%, respectively. The proteomic multi-marker model that differentiates MDD from SPR combined 20 protein markers. Among the 20 protein markers, 13 protein markers were up-regulated in MDD and 7 protein markers were up-regulated in SPR. With 20 protein markers, MDD and SPR are distinguished with a probability of 87% in the training set, and MDD and SPR are distinguished with a probability of 74%, 82%, and 84% in the validation set, independent test set, and total set, respectively. The proteomic multi-marker model for differentiating BD and SPR combined 17 protein markers. Among the 17 protein markers, 10 protein markers were up-regulated in BD and 7 protein markers were up-regulated in SPR. With 17 protein markers, MDD and SPR are distinguished with a probability of 88% in the training set, and BD and SPR are distinguished with a probability of 72%, 78%, and 83% in the validation set, independent test set, and total set, respectively. The proteomic multi-marker models developed at our institute showed good discrimination performance of AUC > 0.8 (80%) in the total data (total set), so it is believed that they will be helpful in objective discrimination and diagnosis of mental disorders.
Figure 7 shows the results of the discrimination performance between depressive disorder, schizophrenia, and bipolar disorder sub-subtype groups of the proteomic multi-marker model developed at our institute. Figure 7a shows the discrimination performance between BD II and BD NOS, which are subtypes of MDD and BD, and discriminates between MDD and BD-II+BD NOS with a probability of 78%. Figure 7b shows the discrimination performance between SPR and BD-I, a subtype of BD, and discriminates between SPR and BD-I with a probability of 82%. Figure 7c shows the discrimination performance between MDD and BD without current hypomanic/manic/mixed symptoms, a subtype of BD, and discriminates between SPR and BD without current hypomanic/manic/mixed symptoms with a probability of 80%. The multi-marker model developed at our center showed a good discrimination performance of AUC > 0.75 (75%) in the comparative analysis between sub-subtype groups of BD, so it can be used for objective discrimination and diagnosis between sub-subtypes in mental disorders that cannot be differentiated due to similar symptoms. I think it will help.
Figure 8 shows the discrimination and diagnostic performance results of the ensemble model developed in this institution. Figure 8a shows the discrimination and diagnostic performance results of the ensemble model developed by combining clinical data with a proteomic multi-marker model to differentiate MDD and BD. In terms of discrimination performance, MDD and BD are discriminated with probabilities of 84%, 82%, 77%, and 80% in Training, Validation, Independent test, and Total set, respectively. In particular, the diagnostic performance in the independent test set showed accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of 76%, 71%, 82%, 77%, and 75%, respectively. Figure 8b shows the discrimination and diagnostic performance results of the ensemble model developed by combining clinical data with a proteomic multi-marker model to differentiate MDD and SPR. In case of discrimination performance, MDD and SPR are discriminated with a probability of 91%, 83%, 90%, and 88% in Training, Validation, Independent test, and Total set, respectively. In particular, the diagnostic performance in the independent test set showed accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of 89%, 86%, 92%, 87%, and 91%, respectively. Figure 8c shows the discrimination and diagnostic performance results of the ensemble model developed by combining clinical data with a proteomic multi-marker model to differentiate BD and SPR. In case of discrimination performance, MDD and SPR are discriminated with a probability of 89%, 73%, 88%, and 88% in Training, Validation, Independent test, and Total set, respectively. In particular, the diagnostic performance in the independent test set showed accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of 83%, 89%, 77%, 80%, and 87%, respectively. In conclusion, our ensemble models were developed by additionally combining clinical data with the established proteomic multi-marker model and showed improved discrimination and diagnostic performance than the proteomic public marker model.
Figure 9 shows the comparison results of the discrimination and diagnostic performance of the ensemble model developed in this institution and the CRSB model. Figure 9a shows the results of comparison of discrimination and diagnosis performance between the ensemble model and CRSB model that distinguish MDD and BD. The ensemble model showed overall higher discrimination performance than the CRSB model in training, validation, and independent tests. In particular, in the independent test, the ensemble model overall showed higher diagnostic performance in accuracy, sensitivity, and negative predictive value than the CRSB model. Figure 9b shows the comparison results of discrimination and diagnosis performance between the ensemble model and CRSB model that distinguish MDD and SPR. The ensemble model showed overall lower discrimination performance than the CRSB model, but in the independent test, the ensemble model showed overall similar diagnostic performance to the CRSB model. Figure 9c shows the comparison results of discrimination and diagnosis performance between the ensemble model and CRSB model that distinguish BD and SPR. The ensemble model showed higher performance in training, independent test, and total set than the CRSB model. In particular, in the independent test, the ensemble model overall showed higher diagnostic performance than the CRSB model. In conclusion, our ensemble models showed overall similar discrimination and diagnostic performance to the CRSB models developed by combining clinical diagnostic evaluation data. Therefore, the possibility of an ensemble model in the differentiation and diagnosis of mental illness was shown.

We compared the current diagnostic methods for major depressive disorder (MDD), bipolar disorder (BD), and schizophrenia (SPR) based on symptoms and behavioral observations. It is based on the discovery of a blood biomarker panel that can distinguish and diagnose patients with major mental disorders based on objective values based on quantitative biomarker proteins.

Major depressive disorder is a disease characterized by a constant depressive state without rapid cycling, with decreased motivation and depression as the main symptoms, causing various cognitive and psychosomatic symptoms, leading to a decline in daily functioning. Diagnosis is made through a clinician's interview and survey according to the standards of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR).

Bipolar disorder is a disease with mood swings, and depending on the change in mood swings, it can be classified into BD-1, BD-2, BD-not otherwise specified (NOS), or BD without current hypomanic/manic/mixed symptoms. BD-1 represents a state of depression and strong mania (hypomanic symptoms), BD-2 represents a state of depression and mild mania (hypomanic symptoms), and BD-NOS represents an asymptomatic state.

Schizophrenia is characterized by hallucinations and hallucinations as well as cognitive impairment.

The diagnosis of patients with bipolar disorder, major depressive disorder, and schizophrenia is based on the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) (Regier, D. A., Kuhl, E. A., & Kupfer, D. J. (2013). The DSM-5: Classification and criteria changes. World psychiatry, 12(2), 92-98; American Psychiatric Association, & American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders: DSM-5. Arlington, VA) and the final diagnosis is determined through the Mini-International Neuropsychiatric Interview (MINI) (Lecrubier, Y., Sheehan, D.V., Weiller, E., Amorim) , P., Bonora, I., Sheehan, K. H., ... & Dunbar, G. C. (1997). The Mini International Neuropsychiatric Interview (MINI). A short diagnostic structured interview: reliability and validity according to the CIDI. European psychiatry, 12(5), 224-231;Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., ... & Dunbar, G. C. (1998).The Mini -International Neuropsychiatric Interview (MINI): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of clinical psychiatry, 59(20), 22-33.).

Additionally, the diagnosis of patients with bipolar disorder, major depressive disorder, and schizophrenia can be made using the patient self-report brief psychiatric diagnostic test [Symptom Checklist 90-revised (SCL-90R)] (Derogatis LR. SCL-90-R: Administration, scoring & procedures) manual-II for the R (evised) version and other instruments of the psychopathology rating scale series. Clinical Psychometric Research . 1992 1992:1-16.)'s 9 symptom dimensions, Somatization (SOM), 12 items, Obsessive - Compulsive (OC) 10 questions, Interpersonal Sensitivity (IS) 9 questions, Depression (DEP) 13 questions, Anxiety (ANX) 10 questions, Hostility (HOS) 6 questions, Fear Anxiety ( The average value (sum of scores/number of items) of each symptom dimension item corresponding to 7 items for Phobic Anxiety (PHOB), 6 items for Paranoid Ideation (PAR), and 10 items for Psychoticism (PSY) is (mean value) 1~2 points: No symptoms at all, average value 2~3 points: Some symptoms, average value 3~4 points: Some symptoms, average value 4~5 points Symptoms are quite severe, average value 5 points or more : the symptoms are very severe), the clinician determines the final diagnosis.

At our center, we developed a blood proteome-based multiple biomarker model that can distinguish patients with major mental disorders such as major depressive disorder, bipolar disorder, and schizophrenia.

In addition, by combining the score data of the patient self-reported major mental illness symptom evaluation index [Symptom Checklist 90-revised (SCL-90R)] with the developed blood proteome multi-biomarker model, a highly accurate ensemble for the differentiation and diagnosis of major mental illness ( Ensemble model was developed. The ensemble model is a machine learning technique called the stacking ensemble strategy (Dzeroski, S., & Zenko, B. (2004). Is combining classifiers with stacking better than selecting the best one?. Machine learning, 54(3), 255-273) It was developed through and the model developed here is represented by the following regression equation 1.

In the above equation, the constant represents the intercept value of the regression equation of the model developed here. [Protein number] refers to the concentration of each biomarker determined by the method according to the present application in the regression equation and is an independent variable. [Dimension number] refers to the previously mentioned SCL-90R symptom dimension combined in the regression equation and is an independent variable. B is a coefficient that represents the relative contribution of each combined protein to distinguishing mental disorders in the model. C is a coefficient that represents the relative contribution of each combined SCL-90R symptom dimension to distinguishing diseases in the model. B and C have + or - values. For example, in the case of an ensemble model that discriminates between depressive disorder and bipolar disorder, the two groups were designated as dichotomous numeric values of 0 for depressive disorder and 1 for bipolar disorder, so if the coefficient value (B or C) is a + value, bipolar disorder is considered more important than depressive disorder. It contributes more to the differentiation, and if it is a -value, it contributes more to the differentiation of depressive disorder than bipolar disorder. p is the probability of discrimination between the two mental disorders finally calculated from the model. The developed model is used to differentiate between two mental disorders through the final calculated p.

In addition, Youden's index [J=max(sensitivity+specificity-1)] (Youden, W. J. (1950). Index for rating diagnostic tests. Cancer, 3(1)) was used to differentiate the two diseases based on specific criteria in each ensemble model developed. , 32-35.) value was set as the optimal cutoff value. Therefore, the two mental disorders can be distinguished based on the optimal threshold (cutoff) calculated from each ensemble model.

Specifically, the proteins included in the ensemble model developed from the patients' blood samples were quantified using MRM-MS, our mass spectrometry equipment, and the protein quantitative value data was used in the model developed along with the SCL-90R clinical data of the patients. By applying this, discrimination performance can be calculated through AUROC analysis. In the AUROC curve calculated in this way, there are sensitivity and specificity values calculated for each patient, and therefore there is a distribution of sensitivity and specificity values for all patients. The Youden index value for each patient can be obtained through sensitivity+specificity-1, so there is a distribution of Youden index values for all patients. The largest Youden index value among the Youden index values is determined as the optimal cutoff. For the threshold value (cutoff) according to an implementation of the present application, refer to Example 5, etc.

Specifically, the subjects registered in the present invention were 174 people with major depressive disorder, 170 people with bipolar disorder, 171 people with schizophrenia, and 160 normal controls. In particular, of the 170 patients with bipolar disorder, 75 had BD-1, 84 had BD-2, 11 had BD-NOS, and 143 had BD without current hypomanic/manic/mixed symptoms. Candidates for depressive disorder, bipolar disorder, and schizophrenia biomarkers (depressive disorder, bipolar disorder) established using mass spectrometry-based multiple reaction detection (MRM-MS or MRM) in 675 blood (plasma) samples from a total of 675 subjects. Proteomic data was collected by quantifying the disorder and schizophrenia-related genome, transcriptome, and proteome database and establishing a list of biomarker candidates through literature review. A proteome-based multi-marker model [multiprotein-marker (MPM) model] was constructed and developed by applying machine learning techniques using the least absolute shrinkage and selection operator (LASSO) and crossvalidation to the collected proteome data. The model's discriminatory performance of major mental disorders was measured through area under receiver operating characteristic (AUROC) analysis. Clinical data on patient self-reported major mental disorder symptom evaluation index [symptom checklist 90-revised (SCL-90R)] and clinical major mental disorder symptom evaluation index [clinician rater score] were collected for a total of 675 subjects. Using the machine learning-based Stacking Ensemble method, we developed an ensemble model that combines a previously constructed proteomic multi-marker model with variables from clinical data of patient self-reported major mental illness symptom evaluation indicators. In addition, a Clinician rater score-based (CRSB) model was developed using only the clinical data variables of the clinical major mental disorder symptom evaluation index. By comparing the discrimination and diagnostic performance of the Ensemble model developed at our center and the Clinician rater score-based (CRSB) model, the main model is a combination of proteomic data without clinician intervention and patient self-reported diagnostic evaluation clinical data variables. It presented the possibility of objective differentiation and diagnosis of mental illness. Discrimination performance was measured through Area under receiver operating characteristic (AUROC) analysis, and diagnostic performance was measured through Accuracy, Sensitivity, Specificity, Positive predictive value (PPV), and Negative predictive value (NPV) diagnostic parameter analysis.

This technology uses a multiple reaction detection method to quantify up to 300 biomarkers during an analysis time of 70 minutes with a single small amount of sample injection. Compared to quantitative methods using antibodies, the variability of quantitative results is relatively small and quantification can be performed more economically. In hospitals where mass spectrometry-based multiple reaction detection methods are established, it is possible to diagnose major mental disorders with improved accuracy by quantifying protein biomarkers in the blood with low variability, high reproducibility, and accuracy.

Accordingly, in one aspect, our research relates to a combination of biomarkers that can distinguish between major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia.

In one embodiment, the combination of biomarkers for distinguishing between depressive disorder and bipolar disorder includes ALDOC (Fructose-bisphosphate aldolase C), ANPEP (Aminopeptidase N), ARMH4 (Armadillo-like helical domain-containing protein 4), and C1RL (Complement C1r subcomponent- like protein), CDH13 (Cadherin-13), CETP (Cholesteryl ester transfer protein), COL10A1 (Collagen alpha-1), CTNND1 (Catenin delta-1), DDR1 (Epithelial discoidin domain-containing receptor 1), DBH (Dopamine beta) -hydroxylase ), SERPING1 (Plasma protease C1 inhibitor), IL1RAP (Interleukin-1 receptor accessory protein), ITIH2 (Inter-alpha-trypsin inhibitor heavy chain H2), NPC2 (NPC intracellular cholesterol transporter 2), RAN (GTP-binding nuclear) protein Ran), SAA1 (Serum amyloid A-1 protein), and TF (Serotransferrin).

In another embodiment, the combination of biomarkers for distinguishing between depressive disorder and schizophrenia includes ALDOC (Fructose-bisphosphate aldolase C), IGFALS (Insulin-like growth factor-binding protein complex acid labile subunit), and NAGLU (Alpha-N-acetylglucosaminidase). , ATP1A1 (Sodium/potassium-transporting ATPase subunit alpha-1), CTSS (Cathepsin S), SERPINA6 (Corticosteroid-binding globulin), CPB2 (Carboxypeptidase B2), COL10A1 (Collagen alpha-1), CRYM (Ketimine reductase mu-crystallin) ), GPX3 (Glutathione peroxidase 3), IGFBP3 (Insulin-like growth factor-binding protein 3), IGFBP5 (Insulin-like growth factor-binding protein 5), ITIH2 (Inter-alpha-trypsin inhibitor heavy chain H2), PGC ( Gastricsin), PROC (Vitamin K-dependent protein C), PROS1 (Vitamin K-dependent protein S), RIDA (2-iminobutanoate/2-iminopropanoate deaminase), SAA1 (Serum amyloid A-1 protein), SAA4 (Serum amyloid A) -4 protein), and TFPI (Tissue factor pathway inhibitor).

In another embodiment, the combination of biomarkers for distinguishing bipolar disorder and schizophrenia includes SERPINA3 (Alpha-1-antichymotrypsin), ANPEPM (Aminopeptidase N), BPIFB1 (BPI fold-containing family B member 1), and C1RL (Complement C1r subcomponent- like protein), CFB (Complement factor B), CLDN3 (Claudin-3), DBH (Dopamine beta-hydroxylase), GPR37 (Prosaposin receptor GPR37), SERPIND1 (Heparin cofactor 2), IGFBP5 (Insulin-like growth factor-binding protein) 5), SERPING1 (Plasma protease C1 inhibitor), MBL2 (Mannose-binding protein C), NPC2 (NPC intracellular cholesterol transporter 2), PLXNC1 (Plexin-C1), PSMD1 (26S proteasome non-ATPase regulatory subunit 1), TFPI ( Tissue factor pathway inhibitor), and UMOD (Uromodulin).

In another aspect, the present application also relates to a biomarker composition for distinguishing major mental disorders, including a substance for measuring the biomarker expression level of each biomarker or combination of markers by mass spectrometry.

Mass spectrometry methods for biomarker analysis according to the present disclosure include tandem mass spectrometry, ion trap mass spectrometry, triple quadrupole mass spectrometry, hybrid ion trap/quadrupole mass spectrometry, or time-of-flight mass spectrometry.

In one embodiment, the mode used in mass spectrometry according to the present disclosure is Selected Reaction Monitoring (SRM) or Multiple Reaction Monitoring (MRM).

Mass spectrometry methods used in the methods according to the present disclosure may include tandem mass spectrometry, ion trap mass spectrometry, triple quadrupole mass spectrometry, hybrid ion trap/quadrupole mass spectrometry, and/or time-of-flight mass spectrometry. The mass spectrometry mode used at this time may be, for example, Selected Reaction Monitoring (SRM) or Multiple Reaction Monitoring (MRM).

In one implementation, in particular MRM mode is used. The principle of MRM mass spectrometry is to hydrolyze all selected target proteins into peptides and then select a peptide (precursor ion, MS1) with a mass to charge (m/z) specific to each target protein. When this specific peptide collides (Quadruple 2, Q2), a fragment (fragmentation ion, MS2) with a specific mass and a characteristic m/z is selected from the generated fragments. The pair of precursor ion/fragment ions obtained from MS1/MS2 is called the specific transition of the target protein (specific mass fingerprint of the target protein), and when these transitions are measured for all target proteins (more than 300 proteins), the sample The amount of all target proteins can be quantified simultaneously relative or absolute. For relative or absolute quantification, SIS (isotopically substituted identical amino acid sequence) peptide is used as a standard material. Since the amount of input standard material (SIS peptide) of the measurement sample is known, the amount of target peptide can be calculated proportionally. It is a principle. The transition that passes through MS2 is converted into a digital signal at the detector and converted into a peak chromatogram, and the peak area can be calculated to perform relative and absolute quantitative analysis.

When the biomarkers according to the present application are performed by mass spectrometry, the peptide sequences detected in each biomarker are as follows.

[Table 1]

The specimen or sample that can detect the biomarker according to the present application is blood, whole blood, serum, or plasma. In one embodiment of the present application, the sample is blood, particularly plasma sample. When using blood as a sample, it can be used by depleting and denaturing the protein and then digesting it with a proteolytic enzyme (eg, trypsin and/or chymotrypsin).

Specimens or samples used for biomarker detection according to the present application are obtained from mammals, particularly humans. The human subject is a patient with a mood disorder who needs to distinguish between major depressive disorder and bipolar disorder.

The term diagnosis or judgment herein refers to determining the susceptibility of a test subject to a specific disease or condition, determining whether the subject currently has a specific disease or condition, or therametrics (e.g., and monitoring the condition of the subject to provide information about treatment efficacy.

In another aspect, our hospital provides a method of providing information necessary for judgment or diagnosis by distinguishing between major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia, or distinguishing depressive disorder, bipolar disorder, and schizophrenia in vitro to provide the above information. This relates to a method of detecting biomarkers for diagnosis.

The method includes measuring the expression level of each biomarker of the biomarker combination according to Table 1 from blood isolated from the subject by mass spectrometry; and correlating the measurement result with depressive disorder, bipolar disorder, or schizophrenia.

The peptide sequences of mass spectrometry, samples, and each biomarker used in the method according to the present application can be referred to as mentioned above.

In one embodiment, the mass spectrometry mode is MRM, and the peptides for each protein constituting the panel used for MRM analysis are shown in Table 1.

In another embodiment, the method according to the present application uses score data of the major mental disorder symptom evaluation index [Symptom Checklist 90-revised (SCL-90R)] in addition to the biomarker detection results, i.e., in combination with depressive disorder, bipolar disorder, and schizophrenia. It can be used to diagnose depressive disorder, bipolar disorder, and schizophrenia, which is used to classify diseases. Regarding this, please refer to what was mentioned above.

In this case, the linking step of the method according to the present application obtains a result by substituting the biomarker detection result and major mental disorder symptom evaluation index score data into the following equation 1, and the result is adjusted to the threshold value determined for each biomarker combination. Compare with

Below, examples are presented to aid understanding of the present invention. However, the following examples are provided only for easier understanding of the present invention, and the present invention is not limited to the following examples.

Example 1. Collection of blood samples for major psychiatric disorders and analysis of demographic and clinical characteristics

To determine whether candidate protein biomarkers related to major mental disorders were detected, blood (plasma) samples from 160 normal controls, 174 cases of depressive disorder, 170 cases of bipolar disorder, and 171 cases of schizophrenia were collected from a multicenter/large-scale cohort. In particular, the bipolar disorder sample consisted of three disease subtypes (75 cases of BD-1, 84 cases of BD-2, and 11 cases of BD-NOS), allowing the results of the present invention to be applied to the diagnosis of these subtypes. In addition, demographic and clinical data were collected for the patients and normal controls, and statistical analysis (univariate analysis) was performed to analyze demographic and clinical characteristics. Our hospital conducted the study based on the latest version of the Declaration of Helsinki and was reviewed by the Institutional Review Board of Seoul National University Hospital (IRB no. 1806-106-951) and all participating hospitals. Our hospital obtained informed consent from patients for participation and use of patient samples.

The demographic and clinical characteristics of the patient group and normal control group participating in this outbreak are shown in Table 2 below.

[Table 2]

Abbreviations

BPRS: Brief Psychiatric Rating Scale

YMRS: Young Mania Rating Scale

MADRS: Montgomery-Asberg Depression Rating Scale

HAM-A: Hamilton Anxiety Scale

MDD: major depressive disorder

BD: bipolar disorder

SPR: schizophrenia

HC: healthy control

BMI: body mass index

Example 2. Construction of a list of protein biomarker candidates

A list of candidate protein biomarkers related to depressive disorder, bipolar disorder, and schizophrenia was constructed through the inventor's established list of disease biomarkers, prior literature search, and data mining. A total of 5 databases related to depressive disorder, bipolar disorder, and schizophrenia [PsyGeNET (http://www.psygenet.org), Schizophrenia Gene Resource 2 (SZGR2) (https://bioinfo.uth.edu/SZGR/) , Laboratory of Neurophenomics (http://www.neurophenomics.info/), Comprehensive Database for Schizophrenia (SZDB2) (www.SPRdb.org), and The Stanley Neuropathology Consortium Integrative Database (SNCID) (http://sncid.stanleyresearch .org)] to construct a list of 8081 genes related to depressive disorder, bipolar disorder, and schizophrenia. Afterwards, 1462 proteins with high detectability in blood were selected through the blood proteome databases The Human Blood Protein Atlas and Plasma Proteome Database (PPD). Afterwards, 407 proteins with a high probability of being detected by mass spectrometry were selected using 8 mass spectrometry-based spectral libraries. Constructed from previous research (Shin, D. et al. (2021). Quantitative Proteomic Approach for Discriminating Major Depressive Disorder and Bipolar Disorder by Multiple Reaction Monitoring-Mass Spectrometry. Journal of Proteome Research, 20(6), 3188-3203) A total of 1,667 protein target lists were constructed by adding 210 proteins related to depressive disorder and bipolar disorder and 1,048 disease-related proteins uniquely constructed by the laboratory. Using multiple reaction detection method (MRM-MS or MRM), a final list of 642 protein biomarker candidates showing detection and quantification in blood (plasma pooled samples) of 50 cases of depressive disorder, bipolar disorder, and schizophrenia each was created. established (see Figure 2). Then, it was analyzed as follows.

Example 3. Protein biomarker analysis using MRM-MS

01. Protein depletion and concentration

Six high-abundant proteins (albumin, IgG, IgA, haptoglobin, transferrin, alpha-1-antitrypsin), which make up about 90% of blood samples, were analyzed using a high performance liquid chromatogram (HPLC) combined with a MARS-6 column. It was removed through this method. Since most blood disease biomarkers exist at very low concentrations in the blood, high-abundant proteins were removed to increase the efficiency of quantifying low-abundant proteins in the blood.

02. Protein concentration measurement and Rapigest In-solution digestion

BCA assay was performed to measure blood protein concentration. Afterwards, 100ug of blood proteins were subjected to in-solution digestion using Rapigest-SF (Waters) and Trypsin according to the manufacturer's method.

03. SIS peptide spiking

For relative quantification of protein biomarkers, stable isotope standard (SIS) peptide was injected into the sample and quantification was performed. SIS peptides have the same chemical properties as protein biomarkers in the blood, so two types of peptides are analyzed at the same retention time, but there is a difference in mass values, so when performing peak integration, the peaks of the two peptides can be obtained separately. By performing relative quantification by calculating the ratio of the peak intensities of the two peptides as (peptide intensity of blood protein)/(SIS peptide intensity), differences in sensitivity that may occur during protein quantification can be corrected.

The analysis method may refer to FIG. 3.

04. 6490 Triple Quadrupole LC-MS/MS analysis (multiple reaction monitoring-mass spectrometry, MRM-MS or MRM)

Multiple reaction monitoring-mass spectrometry (MRM-MS or MRM) is a technology that can quantify up to 800 proteins during an analysis time of 70 minutes by injecting a small amount of sample once. Through the sample pretreatment process, the protein is fragmented into peptide units, ionized, and then injected into the mass spectrometer. When the ionized peptide is injected into triple quadrupole mass spectrometry, ions with a specific mass value are quantified as they pass through Q1, Q2, and Q3. In Q1, peptide proteins with a specific mass value are selected by the mass filter function and delivered to Q2. In Q2, collision energy is applied to the delivered peptide ion to fragment it and create a product ion. When this is passed to Q3, Q3 selects product ions with a specific mass value using the mass filter function. Since there are many types of proteins that make up the multi-marker model and the number of samples to be analyzed is relatively large (675 in total), we used the MRM-MS technique, which can quantify a large amount of proteins in one analysis (see Figure 4) .

After preprocessing a total of 515 blood (plasma) samples, multiple reaction monitoring-mass spectrometry was performed using a 6490 Triple Quadrupole mass spectrometer. The sample injected into the mass spectrometry equipment was fractionated for 52 minutes while increasing the ACN gradient from 3% to 40%, and peptide ions with a specific mass value (m/z) were selected in Q1 and fragmented into product ions in Q2. Afterwards, in Q3, product ions with specific mass values were selected. The raw data obtained after analysis was subjected to peak integration using Skyline software, and through this, the relative quantitative value (peak area ratio) of each protein biomarker candidate was calculated.

Example 4. Production and performance evaluation of a multi-marker model for distinguishing depressive disorder and bipolar disorder using machine learning techniques

LASSO (Friedman, J., Hastie, T., & Tibshirani, R. (2010). Regularization paths for generalized linear models via coordinate descent. Journal of statistical software, one of the machine learning techniques for the 642 proteins quantified above. , 33(1), 1.) and cross-validation (5-fold crossvalidation) was used to construct a multi-marker model. The developed multi-marker model differentiates between depressive disorder and bipolar disorder, depressive disorder and schizophrenia, and bipolar disorder and schizophrenia, and subtypes of depressive disorder and bipolar disorder (BD-2 and BD-NOS), depressive disorder and bipolar disorder. It plays a role in differentiating subtypes of schizophrenia and bipolar disorder (BD-1) of subtype (BD without current hypomanic/manic/mixed symptoms). In the case of depressive disorder, bipolar disorder, and schizophrenia, because the causative factors are complex, including family history, living environment, and lifestyle habits, it was judged that it was not appropriate to differentiate and diagnose them using a single biomarker, so multiple proteins were combined to determine the diagnosis. A marker diagnostic model was created.

Specifically, MRM-MS data for 515 patients (174 cases of depressive disorder, 170 cases of bipolar disorder, and 171 cases of schizophrenia) were used in the training set (104 cases of depressive disorder, 102 cases of bipolar disorder, and 102 cases of schizophrenia). It was divided into a validation set (35 cases of depressive disorder, 34 cases of bipolar disorder, and 34 cases of schizophrenia) and an independent test set (35 cases of depressive disorder, 34 cases of bipolar disorder, and 35 cases of schizophrenia). As a result of performing machine learning-based (LASSO regression and cross-validation) model development on the training set, a total of three proteomic multi-marker models were constructed. That is, in the proteomic multi-marker model (MDD vs BD) for differentiating depressive disorder and bipolar disorder, a total of 17 proteins (ALDOC, ANPEP, ARMH4, C1RL, CDH13, CETP, COL10A1, CTNND1, DDR1, DBH, SERPING1, IL1RAP, ITIH2 , NPC2, RAN, SAA1, and TF) were combined, and a total of 20 proteins (ALDOC, IGFALS, NAGLU, ATP1A1, CTSS, SERPINA6, CPB2) were combined in the proteomic multi-marker model (MDD vs SPR) for differentiating depressive disorder and schizophrenia. , COL10A1, CRYM, GPX3, IGFBP3, IGFBP5, ITIH2, PGC, PROC, PROS1, RIDA, SAA1, SAA4, and TFPI) were combined, and in the proteomic multi-marker model (BD vs SPR) for differentiating bipolar disorder and schizophrenia, A total of 17 proteins (SERPINA3, ANPEPM BPIFB1, C1RL, CFB, CLDN3, DBH, GPR37, SERPIND1, IGFBP5, SERPING1, MBL2, NPC2, PLXNC1, PSMD1, TFPI, and UMOD) were combined.

The proteomic multi-marker model for differentiating depressive disorder and bipolar disorder showed discrimination performance of AUROC=0.84, 0.73, and 0.74 in training set/validation set/independent test set, respectively. The proteomic multi-marker model for differentiating depressive disorder and schizophrenia showed discrimination performance of AUROC=0.87, 0.74, and 0.72 in training set/validation set/independent test set, respectively. The proteomic multi-marker model for distinguishing bipolar disorder and schizophrenia showed discrimination performance of AUROC=0.88, 0.72, and 0.78 in the training set/validation set/independent test set, respectively (see Figure 6).

In addition, models developed by applying the proteomic multi-marker model to sub-subtype groups (BD-1, BD-2, BD-NOS, and BD without current hypomanic/manic/mixed symptoms) of bipolar disorder patients showed depressive disorder, schizophrenia, In addition, the discrimination performance between sub-subtype groups of bipolar disorder patients was verified. Using a proteomic multi-marker model to differentiate between depressive disorder and bipolar disorder, the discrimination performance between 174 depressive disorder patients and a total of 95 BD-2 and BD-NOS patients was UROC=0.78. Additionally, the discrimination performance between 174 patients with depressive disorder and a total of 143 patients with BD without current hypomanic/manic/mixed symptoms was AUROC=0.80. Using the proteomic multi-marker model for differentiating bipolar disorder and schizophrenia, the discrimination performance between 171 schizophrenia patients and a total of 75 BD-1 patients was AUROC = 0.82 (see Figure 7).

Example 5. Production and performance evaluation of an ensemble model [Ensemble (ES) model] using machine learning techniques

In addition, using a machine learning-based stacking ensemble method, we developed an ensemble model that combined the clinical variables of the patient self-reported major mental illness symptom evaluation index [symptom checklist 90-revised (SCL-90R)] with the previously constructed proteomic multi-marker model. (see Figure 8). It was determined that combining additional clinical variables with the previously constructed proteomic multi-marker model could provide higher accuracy in discriminating and diagnosing major mental disorders. In addition, a Clinician rater score-based (CRSB) model was developed by combining only the clinical major mental disorder symptom evaluation index [clinician rater score], which was created to compare and analyze the discrimination and diagnostic performance with the ensemble model (see Figure 9). )

The ensemble model specifically combines the clinical variables of the patient self-reported major mental illness symptom evaluation index [symptom checklist 90-revised (SCL-90R)] with the proteomic multi-marker model developed at our institute using a machine learning-based stacking ensemble technique. developed. An ensemble model (MDD vs BD) by combining the five SCL-90R items Somatization dimension, Psychoticism dimension, Depression dimension, Overeating item, and Interpersonal sensitivity dimension with the proteomic multi-marker model (MDD vs BD) for differentiating depressive disorder and bipolar disorder. was developed. An ensemble model (MDD vs SPR) was created by combining the 5 SCL-90R items Somatization dimension, Interpersonal sensitivity dimension, Psychoticism dimension, Paranoid ideation dimension, and Depression dimension with the proteomic multi-marker model (MDD vs SPR) for differentiating depressive disorder and schizophrenia. ) was developed. The proteomic multi-marker model (BD vs SPR) for differentiating bipolar disorder and schizophrenia was created by combining the six items of SCL-90R: Obsessive-compulsive dimension, Interpersonal sensitivity dimension, Psychoticism dimension, Phobic anxiety dimension, Hostility dimension, and Depression dimension. A model (BD vs SPR) was developed. The discriminatory power of the ensemble model (MDD vs BD) showed AUROC=0.84, 0.82, and 0.77 in the training/validation/independent test set, and the optimal cutoff value (J) in the independent test set was 0.48. If it is greater than this value, it is classified as bipolar disorder, and if it is smaller than this value, it is classified as depressive disorder. Therefore, based on J=0.48, the diagnostic performance in the independent test set showed Accuracy=0.76, Sensitivity=0.71, Specificity=0.82, PPV=0.77, and NPV=0.75. The discriminatory power of the ensemble model (MDD vs SPR) showed AUROC=0.91, 0.83, and 0.90 in the training/validation/independent test set, and the optimal cutoff value (J) in the independent test set was 0.50. If it is greater than this value, it is classified as schizophrenia, and if it is smaller than this value, it is classified as depressive disorder. Therefore, based on J=0.50, the diagnostic performance in the independent test set showed Accuracy=0.89, Sensitivity=0.86, Specificity=0.92, PPV=0.87, and NPV=0.91. The discriminatory power of the ensemble model (BD vs SPR) showed AUROC=0.89, 0.73, and 0.88 in the training/validation/independent test set, and the optimal cutoff value (J) in the independent test set was 0.62. If it is greater than this value, it is classified as schizophrenia, and if it is less than this value, it is classified as bipolar disorder. Therefore, based on J = 0.62, the diagnostic performance in the independent test set showed Accuracy = 0.83, Sensitivity = 0.89, Specificity = 0.77, PPV = 0.80, and NPV = 0.87 (see Figure 8).

Furthermore, the performance of the ensemble model developed at our hospital and clinical variables for major mental illness symptom evaluation indices [clinician rater score] [Brief Psychiatric Rating Scale (BPRS), Young Mania Rating Scale (YMRS), Montgomery-Asberg Depression Rating Scale] (MADRS), Hamilton Anxiety Scale (HAM-A)] were combined to develop a Clinician's rater score-based (CRSB) model and compare its performance. Overall, the differential and diagnostic performance of the ensemble model combining proteomic data and patient self-reported symptom assessment clinical data and the performance of the Clinician’s rater score-based (CRSB) model were found to be similar (see Figure 9).

These results indicate that the ensemble model developed at our center is capable of differentiating and objectively diagnosing major mental disorders with high accuracy even in situations where the active intervention and subjectivity of clinicians are excluded in the diagnosis of major mental disorders.

Although the exemplary embodiments of the present application have been described in detail above, the scope of the rights of the present application is not limited thereto, and various modifications and improvements made by those skilled in the art using the basic concept of the present application defined in the following claims are also included in the scope of the rights of the present application. belongs to

All technical terms used in the present invention, unless otherwise defined, are used with the same meaning as commonly understood by a person skilled in the art in the field related to the present invention. The contents of all publications incorporated by reference herein are hereby incorporated by reference.

<110> Seoul National University R&DB Foundation <120> Biomarker for determining major depressive disorder, polar disorder and zophrenia based on mass spectrometry and its use <130> DP202203004P <160> 43 <170> KoPatentIn 3.0 <210> 1 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223> ALDOC <400> 1 Ala Leu Gln Ala Ser Ala Leu Asn Ala Trp Arg 1 5 10 <210> 2 <211> 15 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(15) <223> AMPN <400> 2 Ala Gln Ile Ile Asn Asp Ala Phe Asn Leu Ala Ser Ala His Lys 1 5 10 15 <210> 3 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> ARMD4 <400> 3 Thr Val Val Pro Ser Ile Thr Arg 1 5 <210> 4 <211> 14 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(14) <223>C1RL <400> 4 Gly Ser Glu Ala Ile Asn Ala Pro Gly Asp Asn Pro Ala Lys 1 5 10 <210> 5 <211> 16 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(16) <223>CAD13 <400> 5 Ile Asn Asn Thr His Ala Leu Val Ser Leu Leu Gln Asn Leu Asn Lys 1 5 10 15 <210> 6 <211> 10 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(10) <223> CETP <400> 6 Ala Ser Tyr Pro Asp Ile Thr Gly Glu Lys 1 5 10 <210> 7 <211> 10 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(10) <223> COAA1 <400> 7 Gly Thr His Val Trp Val Gly Leu Tyr Lys 1 5 10 <210> 8 <211> 12 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(12) <223>CTND1 <400> 8 Gly Tyr Glu Leu Leu Phe Gln Pro Glu Val Val Arg 1 5 10 <210> 9 <211> 12 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(12) <223>DDR1 <400> 9 Leu His Leu Val Ala Leu Val Gly Thr Gln Gly Arg 1 5 10 <210> 10 <211> 10 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(10) <223> DOPO <400> 10 Thr Pro Glu Gly Leu Thr Leu Leu Phe Lys 1 5 10 <210> 11 <211> 6 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(6) <223>IC1 <400> 11 Thr Thr Phe Asp Pro Lys 1 5 <210> 12 <211> 10 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(10) <223> IL1AP <400> 12 Asn Glu Val Trp Trp Thr Ile Asp Gly Lys 1 5 10 <210> 13 <211> 15 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(15) <223> ITIH2 <400> 13 Ile Gln Pro Ser Gly Gly Thr Asn Ile Asn Glu Ala Leu Leu Arg 1 5 10 15 <210> 14 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223> NPC2 <400> 14 Leu Val Val Glu Trp Gln Leu Gln Asp Asp Lys 1 5 10 <210> 15 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223> RAN <400> 15 Phe Asn Val Trp Asp Thr Ala Gly Gln Glu Lys 1 5 10 <210> 16 <211> 20 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(20) <223>SAA1 <400> 16 Phe Phe Gly His Gly Ala Glu Asp Ser Leu Ala Asp Gln Ala Ala Asn 1 5 10 15 Glu Trp Gly Arg 20 <210> 17 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> TRFE <400> 17 Ala Ser Tyr Leu Asp Cys Ile Arg 1 5 <210> 18 <211> 12 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(12) <223> ALS <400> 18 Asp Phe Ala Leu Gln Asn Pro Ser Ala Val Pro Arg 1 5 10 <210> 19 <211> 15 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(15) <223> ANAG <400> 19 Asp Phe Cys Gly Cys His Val Ala Trp Ser Gly Ser Gln Leu Arg 1 5 10 15 <210> 20 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223>AT1A1 <400> 20 Ile Val Glu Ile Pro Phe Asn Ser Thr Asn Lys 1 5 10 <210> 21 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> CATS <400> 21 Tyr Thr Glu Leu Pro Tyr Gly Arg 1 5 <210> 22 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> CBG <400> 22 His Leu Val Ala Leu Ser Pro Lys 1 5 <210> 23 <211> 12 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(12) <223> CBPB2 <400> 23 Asp Thr Gly Thr Tyr Gly Phe Leu Leu Pro Glu Arg 1 5 10 <210> 24 <211> 7 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(7) <223> CRYM <400> 24 Thr Val Val Pro Val Thr Lys 1 5 <210> 25 <211> 6 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(6) <223> GPX3 <400> 25 Phe Tyr Thr Phe Leu Lys 1 5 <210> 26 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223>IBP3 <400> 26 Tyr Gly Gln Pro Leu Pro Gly Tyr Thr Thr Lys 1 5 10 <210> 27 <211> 9 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(9) <223> IBP5 <400> 27 Ala Val Tyr Leu Pro Asn Cys Asp Arg 1 5 <210> 28 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223> PEPC <400> 28 Ala Glu Cys Gly Leu Gly Val Pro Thr Thr Arg 1 5 10 <210> 29 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> PROC <400> 29 Thr Phe Val Leu Asn Phe Ile Lys 1 5 <210> 30 <211> 10 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(10) <223> PROS <400>30 Asn Asn Leu Glu Leu Ser Thr Pro Leu Lys 1 5 10 <210> 31 <211> 10 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(10) <223> RIDA <400> 31 Ala Ala Tyr Gln Val Ala Ala Leu Pro Lys 1 5 10 <210> 32 <211> 9 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(9) <223> SAA4 <400> 32 Gly Pro Gly Gly Val Trp Ala Ala Lys 1 5 <210> 33 <211> 9 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(9) <223>TFPI1 <400> 33 Ile Ala Tyr Glu Glu Ile Phe Val Lys 1 5 <210> 34 <211> 11 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(11) <223> AACT <400> 34 Ala Leu Glu Gln Asp Leu Pro Val Asn Ile Lys 1 5 10 <210> 35 <211> 12 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(12) <223> BPIB1 <400> 35 Asp Thr Gly Thr Tyr Gly Phe Leu Leu Pro Glu Arg 1 5 10 <210> 36 <211> 16 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(16) <223> CFAB <400> 36 Leu Asn Thr Ile Gly His Tyr Glu Ile Ser Asn Gly Ser Thr Ile Lys 1 5 10 15 <210> 37 <211> 12 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(12) <223> CLD3 <400> 37 Tyr Leu Ser Tyr Thr Leu Asn Pro Asp Leu Ile Arg 1 5 10 <210> 38 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> GPR37 <400> 38 Ile Pro Leu Asn Asp Leu Phe Arg 1 5 <210> 39 <211> 17 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(17) <223>HEP2 <400> 39 Ser Leu Glu Ala Gln Gly Asn Ser Ser His Leu Asp Ala Asp Thr Val 1 5 10 15 Arg <210> 40 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> MBL2 <400> 40 Leu Leu Glu Leu Thr Gly Pro Lys 1 5 <210> 41 <211> 9 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(9) <223>PLXC1 <400> 41 Asp Ile Ser Glu Val Val Thr Pro Arg 1 5 <210> 42 <211> 8 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(8) <223> PSMD1 <400> 42 Ser Leu Glu Ser Ile Asn Ser Arg 1 5 <210> 43 <211> 14 <212> PRT <213> Homo sapiens <220> <221> PEPTIDE <222> (1)..(14) <223> UROM <400> 43 Leu Leu Asn Leu Glu Gly Phe Pro Ser Gly Ser Gln Ser Arg 1 5 10

Claims (10)

A biomarker composition for diagnosing major mental disorders containing a substance for measuring the expression level of a combination of biomarkers that distinguishes major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia by mass spectrometry,
The combination of biomarkers for diagnosing depressive disorder and bipolar disorder is ALDOC (Fructose-bisphosphate aldolase C), ANPEP (Aminopeptidase N), ARMH4 (Armadillo-like helical domain-containing protein 4), and C1RL (Complement C1r subcomponent-like). protein), CDH13 (Cadherin-13), CETP (Cholesteryl ester transfer protein), COL10A1 (Collagen alpha-1), CTNND1 (Catenin delta-1), DDR1 (Epithelial discoidin domain-containing receptor 1), DBH (Dopamine beta-1) hydroxylase ), SERPING1 (Plasma protease C1 inhibitor), IL1RAP (Interleukin-1 receptor accessory protein), ITIH2 (Inter-alpha-trypsin inhibitor heavy chain H2), NPC2 (NPC intracellular cholesterol transporter 2), RAN (GTP-binding nuclear protein) Ran), SAA1 (Serum amyloid A-1 protein), and TF (Serotransferrin),
The combination of biomarkers used to distinguish and diagnose depressive disorder and schizophrenia is ALDOC (Fructose-bisphosphate aldolase C), IGFALS (Insulin-like growth factor-binding protein complex acid labile subunit), NAGLU (Alpha-N-acetylglucosaminidase), and ATP1A1. (Sodium/potassium-transporting ATPase subunit alpha-1), CTSS (Cathepsin S), SERPINA6 (Corticosteroid-binding globulin), CPB2 (Carboxypeptidase B2), COL10A1 (Collagen alpha-1), CRYM (Ketimine reductase mu-crystallin), GPX3 (Glutathione peroxidase 3), IGFBP3 (Insulin-like growth factor-binding protein 3), IGFBP5 (Insulin-like growth factor-binding protein 5), ITIH2 (Inter-alpha-trypsin inhibitor heavy chain H2), PGC (Gastricsin) , PROC (Vitamin K-dependent protein C), PROS1 (Vitamin K-dependent protein S), RIDA (2-iminobutanoate/2-iminopropanoate deaminase), SAA1 (Serum amyloid A-1 protein), SAA4 (Serum amyloid A-4) protein), and TFPI (Tissue factor pathway inhibitor),
The combination of biomarkers for diagnosing bipolar disorder and schizophrenia is SERPINA3 (Alpha-1-antichymotrypsin), ANPEPM (Aminopeptidase N), BPIFB1 (BPI fold-containing family B member 1), and C1RL (Complement C1r subcomponent-like protein). ), CFB (Complement factor B), CLDN3 (Claudin-3), DBH (Dopamine beta-hydroxylase), GPR37 (Prosaposin receptor GPR37), SERPIND1 (Heparin cofactor 2), IGFBP5 (Insulin-like growth factor-binding protein 5) , SERPING1 (Plasma protease C1 inhibitor), MBL2 (Mannose-binding protein C), NPC2 (NPC intracellular cholesterol transporter 2), PLXNC1 (Plexin-C1), PSMD1 (26S proteasome non-ATPase regulatory subunit 1), TFPI (Tissue factor) pathway inhibitor), and UMOD (Uromodulin), a biomarker composition for distinguishing major mental disorders.
According to claim 1,
A biomarker composition for distinguishing major mental disorders, wherein the mass spectrometry method includes tandem mass spectrometry, ion trap mass spectrometry, triple quadrupole mass spectrometry, hybrid ion trap/quadrupol mass spectrometry, or time-of-flight mass spectrometry.
According to claim 2,
A biomarker composition for distinguishing major mental disorders, wherein the mode used in the mass spectrometry is Selected Reaction Monitoring (SRM) or Multiple Reaction Monitoring (MRM).
According to claim 3,
The mass spectrometry mode is MRM, and the peptides of each biomarker used for analysis in the MRM are as shown in the following table. Biomarker composition for distinguishing major mental disorders:


As a method for providing information necessary for the diagnosis of major mental disorders such as depressive disorder, bipolar disorder, and schizophrenia, the method includes,
Measuring the expression level of each biomarker of each biomarker combination according to claim 1 from blood isolated from the subject using mass spectrometry; and
A method of providing information necessary for the diagnosis of depressive disorder, bipolar disorder, and schizophrenia, comprising correlating the measurement results with depressive disorder, bipolar disorder, or schizophrenia.
According to claim 5,
For the diagnosis of depressive disorder, bipolar disorder, and schizophrenia, wherein the mass spectrometry method includes tandem mass spectrometry, ion trap mass spectrometry, triple quadrupole mass spectrometry, hybrid ion trap/quadrupole mass spectrometry, or time-of-flight mass spectrometry. How to provide the information you need.
According to claim 6,
A method of providing information necessary for the diagnosis of depressive disorder, bipolar disorder, and schizophrenia, wherein the mode used in the mass spectrometry is Selected Reaction Monitoring (SRM) or Multiple Reaction Monitoring (MRM).
According to claim 7,
The mass spectrometry mode is MRM, and the peptides for each protein constituting the panel used for analysis in the MRM are as shown in the following table. A method of providing information necessary for the diagnosis of depressive disorder, bipolar disorder, and schizophrenia.


The method according to any one of claims 5 to 8,
The method further includes the step of obtaining score data on the subject's major mental illness symptom evaluation index [symptom checklist 90-revised (SCL-90R)],
A method of providing information necessary for the diagnosis of depressive disorder, bipolar disorder, and schizophrenia, wherein the score data in addition to the biomarker detection result in the linking step is used to determine depressive disorder, bipolar disorder, and schizophrenia.
According to clause 9,
The linking step involves substituting the biomarker detection results and major mental disorder symptom evaluation index score data into the following equation to obtain a result, and comparing the result with the threshold value determined for each biomarker combination. Depressive disorder, How to provide information needed for a diagnosis of bipolar disorder and schizophrenia: