CN116754772A - Peripheral blood protein marker for early diagnosis of senile dementia, application and auxiliary diagnosis system - Google Patents

Abstract

The application provides a peripheral blood protein marker for early diagnosis of senile dementia, which comprises proteins P02753 (RBP 4), P22352 (GPX 3), P23560 (BDNF), P02765 (AHSG) or P00736 (C1R), and application of the protein marker in a diagnosis kit, and finally a model for early diagnosis of senile dementia is established, wherein the area under the overall prediction curve of the model is AUC=0.758, the sensitivity is 78%, and the specificity is 75%.

Description

Peripheral blood protein markers, applications and auxiliary diagnostic systems for early diagnosis of Alzheimer's disease

Cross-references to related applications

This application claims priority to the Chinese invention patent application with application number 202210245944.X submitted on March 14, 2022, the content of which is incorporated into this application by reference.

Technical field

The invention belongs to the field of biological detection technology, and specifically relates to a blood-based peripheral blood protein marker for diagnosing early Alzheimer's disease (mild cognitive impairment, MCI), its application and its medical auxiliary diagnosis system.

Background technique

(1) Existing technology

Alzheimer's disease (AD), commonly known as Alzheimer's disease, is a neurodegenerative disease mainly characterized by decline in memory, cognition, and daily living abilities. It has a high prevalence, disability rate, and disease among the elderly. The burden is heavy. Epidemiological surveys show that the prevalence rate in people over 60 years old is 3-5%. For every 5 years of age, the prevalence rate doubles. It can reach 20% over 80 years old and 50% over 90 years old. Our country is an aging society. According to data released by the National Bureau of Statistics, as of the end of 2020, our country's population aged 60 and above reached 260 million, accounting for 18.7% of the country's total population. Among them, there are 10 million dementia patients, and what is even more huge is that my country also has more than 30 million mild cognitive impairment (MCI) patients in the early stages of dementia (2019, The Lancet), [Longfei J, Meina Q, et al. Dementia in China: epidemiology, clinical management, and research advances. Lancet Neurology. 2019, 19 (1): p81-92], the medical and nursing burden is extremely heavy. There are currently no drugs that can slow or reverse the progression of the disease, so early diagnosis and intervention of Alzheimer's disease are crucial.

Alzheimer's disease can be roughly divided into asymptomatic stages, mild cognitive impairment (MCI) and dementia stages. Patients may have mild cognitive impairment for about 10 years before developing obvious clinical symptoms. Although patients at this stage have no impact on their daily living abilities, the pathological changes of dementia have already existed and developed in the brain until obvious clinical symptoms appear. Therefore, MCI is considered the optimal stage for early diagnosis and intervention.

However, early diagnosis of Alzheimer's disease is very difficult. In the past 10 years, the National Institute on Aging and Alzheimer's Association (NIA-AA) have introduced biomarker testing, allowing early diagnosis of AD patients before they develop obvious clinical symptoms. In 2011, NIA-AA recommended the use of fluorodeoxyglucose-positron tomography (FDG-PET), amyloid-positron tomography (Aβ-PET), and cerebrospinal fluid Aβ42, T-tau, and P-tau Detection as a biomarker for early diagnosis of AD. However, these examinations are either expensive (FDG-PET, Aβ-PET) or invasive (cerebrospinal fluid examination). They are still mainly used in the research field and cannot be promoted. Therefore, biomarkers that are easy to obtain and detect have become the focus of research. In recent years, studies have found that changes in the composition of some substances in the blood of patients with dementia are related to pathological changes in their brains, and can reflect pathological changes in the brain. At the same time, blood also has the advantages of easy acquisition and repeatable testing. Therefore, blood testing may become an effective way to diagnose early dementia. However, the components in blood are complex in composition, low in content, and easily affected by various factors. Therefore, there is a need to find sensitive and reliable biomarkers for early diagnosis related to dementia. Current research has discovered potential biomarkers of dementia in peripheral blood, mainly including MicroRNA, plasma proteins and other indicators in the blood, but they are still in the exploratory stage. The research results of MicroRNA in blood are not consistent enough, and there is still a lack of research evidence support from large-sample data [Miller, J.and J.Kauwe, Predicting Clinical Dementia Rating Using Blood RNA Levels.Genes, 2020.11:p.706.]; while plasma proteins are because Its content is low and there are many types. However, due to the sensitivity and stability of the detection method, the results are not consistent enough. As far as proteomics research is concerned, Kiddle et al. (2014) summarized and analyzed 21 studies and obtained 163 candidate proteins. After screening, 94 proteins were verified again. The research results were not consistent with the previous ones [Kiddle, S.J., et al. al., Candidate blood proteome markers of Alzheimer's disease onset and progression: a systematic review and replication study. Journal of Alzheimers Disease Jad, 2014.38(3):p.515., Xu Hua, Xiao Shifu, Peripheral blood protein biomarkers of Alzheimer's disease Progress in biological research. Chinese Journal of Clinicians: Electronic Edition, 2017(10): p.1821-1824.]. In 2015, another researcher used the same detection method as Kiddle et al. to screen out a group of proteins (including S100-A9 protein, CD226 antigen, transplantation inflammatory factor 1, endothelial cell selective adhesion molecule, and leukocyte differentiation antigen CD84 ), the sensitivity for diagnosing MCI is 96.7%, the specificity is 80.0%, and the accuracy is 92.5% [Zhao X, Lejnine S, Spond J, et al. A candidate plasma protein classifier to identify Alzheimer'sdisease[J].J Alzheimers Disease Jad,2015,43(2):549-563], but again, this result could not be repeated and verified by subsequent studies. It was not until Karikari et al. published the results of a large-sample study in "Lancet Neurology" in 2020 that the diagnostic value of proteins in peripheral blood as biomarkers for dementia diagnosis was recognized. This study verified the reliability and stability of Tau-181 protein in diagnosing dementia on more than 1,000 patients. The results found that Tau-181 has good sensitivity and stability in diagnosing patients in the stage of dementia [Karikari, T., et al., Bloodphosphorylated tau181 as a biomarker for Alzheimer's disease: a diagnostic performance and prediction modeling study using data from four prospective cohorts. The Lancet Neurology, 2020.19:p.422-433.]. Research in the past 2 years also includes the diagnostic value of plasma GFAP protein, p-tau217, p-tau231, etc. Overall, the accuracy of these proteins in the diagnosis of clinical dementia can reach more than 90%, but they are not effective in the early MCI stage of dementia. The diagnostic value is still not stable enough and needs further confirmation [Oeckl, P., et al., Glial Fibrillary Acidic Protein in Serum is Increased in Alzheimer'sDisease and Correlates with Cognitive Impairment.J Alzheimers Dis, 2019.67(2):p.481- 488. Chatterjee, P., et al., Diagnostic and prognostic plasma biomarkers for preclinical Alzheimer's disease. Alzheimers Dement, 2021. Ashton, N.J., et al., Plasma p-tau231: a new biomarker for incipient Alzheimer's disease pathology. Acta Neuropathol, 2021.141 (5): P.709-724; Thijssen, E.H., ETAL., Plasma Phosphorylated Tau 217And Phosphorylated Tau 181as Biomarker Inlzheimer's Di Sease and Frontotemporal Lobar Degeneration: A Retrospectivediagnostic Performance Study.lancet Neurol, 2021.20 (9): P. .739-752.].

In summary, early diagnosis of dementia has important social needs. Currently, there is still a lack of early diagnosis methods for AD that are convenient to detect and easy to promote. Peripheral blood is easy to obtain and can be tested repeatedly, so it has the potential to be used as an early diagnostic tool for dementia. However, the protein content in blood is low and diverse, which requires high sensitivity and stability of detection technology. Therefore, it is necessary to find reliable and sensitive biomarkers for early diagnosis of AD through effective technology.

(2) Problems or deficiencies in existing technology

With the advancement of proteomics technology, differential proteins in blood can be searched for on a large scale, and potential biomarkers can be found through screening of differential proteins. Currently, proteomics technologies used for high-throughput peripheral blood detection and screening mainly include the following categories:

(a) Detection methods based on immunocapture: This type of method is based on the specific binding of antigens and antibodies and combines a variety of specific technologies to achieve qualitative and quantitative detection of proteins, such as: protein liquid phase chip (Lumicaninex xMAP), Electrochemiluminescence technology MSD (mesoscale discovery), etc. [Xu Hua, Xiao Shifu, Research progress on peripheral blood protein biomarkers of Alzheimer's disease. Chinese Journal of Clinicians: Electronic Edition, 2017(10): p.1821-1824.] . A large number of target proteins can be detected in a small blood sample. Its high-throughput characteristics can meet the needs of clinical research. However, this type of method is based on the specific binding of antigens and antibodies and is not suitable for detection without hypothesis targets. It is also affected and limited by the selected antibodies. Larger, the results are not stable enough.

(b) Aptamer-based detection methods: This type of method is based on the specific binding of aptamers and target proteins. Aptamers are single-stranded oligonucleotides that can effectively recognize and bind to target proteins with high affinity and specificity. This method also has the characteristics of high-throughput, and its results have better stability. It is also the most commonly used method at present, but its results may still fluctuate due to the influence of the selected aptamer [Gold, L., et al., Aptamer-based multiplexed proteomic technology for biomarker discovery. PlosOne, 2010.5(12):p.e15004.]. This reduces stability.

(c) Detection method based on mass spectrometry (MS): Mass spectrometry detection is more expensive than the first two methods. It is characterized by getting rid of the influence of antibodies on the results and has the best stability and accuracy. Moreover, in recent years, with the continuous advancement of mass spectrometry detection technology, its sensitivity has been continuously improved. It can not only perform non-target (non-hypothetical) detection of proteins in blood, but also directly quantitatively detect candidate proteins, combined with radionuclide labeling. Technologies such as this can achieve accurate quantification of low-abundance proteins in blood [Xu Hua, Xiao Shifu, Research Progress in Peripheral Blood Protein Biomarkers for Alzheimer's Disease. Chinese Journal of Clinicians: Electronic Edition, 2017(10):p .1821-1824.], the research results are stable and reliable, and it is the gold standard for low-abundance protein detection.

Contents of the invention

The main purpose of the present invention is to provide a stable, reliable and simple detection technology that can be used to diagnose the early stages of Alzheimer's disease (MCI) in order to solve the problem of difficulty in early diagnosis of Alzheimer's disease existing in the prior art.

In order to achieve the above object, on one hand, the present invention provides a peripheral blood protein marker for early diagnosis of Alzheimer's disease. The protein marker includes proteins P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765 (AHSG) or P00736(C1R).

A second aspect of the present invention provides the use of the above-mentioned reagent for detecting peripheral blood protein markers for early diagnosis of Alzheimer's disease in the preparation of an early diagnosis kit for Alzheimer's disease.

Preferably, the kit can measure the content of P02753 (RBP4) protein, P22352 (GPX3) protein, P23560 (BDNF) protein, P02765 (AHSG) protein or P00736 (C1R) protein in plasma.

Preferably, the kit can measure the content of any peptide segment of P02753 (RBP4) or P22352 (GPX3) or P23560 (BDNF) or P02765 (AHSG) or P00736 (C1R).

A third aspect of the present invention provides a medical auxiliary diagnostic system for early diagnosis of Alzheimer's disease, wherein the diagnostic system includes an MCI diagnostic model, and the MCI diagnostic model includes proteins P02753 (RBP4), P22352 (GPX3), P23560 ( BDNF), P02765(AHSG), P00736(C1R), age, and sex composition.

Using the blood-based peripheral blood protein markers and applications for early diagnosis of Alzheimer's disease and its medical auxiliary diagnosis system provided by the present invention has the following beneficial effects:

1. Use non-targeted proteomics technology (DIA) to conduct open exploration of differential proteins between groups, reducing human bias and achieving better objectivity.

2. Use the IPA (Ingenuity pathway analysis) bioanalysis software to analyze the functional pathways and related diseases of the proteins, screen them layer by layer, and select proteins that are related to neurological diseases and dementia; then verify them step by step, and finally determine the use. Biomarkers for early diagnosis of dementia have good stability and reliability.

3. Using an extremely high-precision mass spectrometer and combining isotope standard peptides to accurately quantify low-abundance proteins in blood samples, the detection results are stable and reliable. On this basis, a stable prediction model was established, and finally a prediction model consisting of five proteins, P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765 (AHSG), and P00736 (C1R), as well as gender and age factors was obtained. A set of prediction models, the area under the curve (AUC) of the overall model predicts MCI = 0.758, the sensitivity is 78%, and the specificity is 75%, which has good stability and good sensitivity.

Description of the drawings

Figure 1 is a diagram showing the pathway analysis results of differential proteins between MCI and NC groups of the present invention.

Figure 2 is a schematic diagram of the PRM experimental process and detection of the present invention.

Figures 3a-1 to 3e-2 respectively show the concentrations of the five proteins P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765 (AHSG), and P00736 (C1R) and the corresponding results for diagnosing MCI.

Figure 4 is a diagram showing the results of diagnosing MCI using the prediction model of the present invention.

Detailed ways

In order to make it easy to understand the technical means, creative features, objectives and effects of the present invention, the present invention will be further elaborated below in conjunction with specific embodiments.

The present invention adopts mass spectrometry proteomics detection technology. First, it detects and compares the plasma proteins in the peripheral blood of aMCI (amnestic mild cognitive impairment) patients in the early stages of Alzheimer's disease and the elderly (NC) with normal cognition. Through layer-by-layer screening and step-by-step verification of differential proteins, combined with isotope labeling methods, the obtained target proteins can be accurately quantified and verified to find biomarkers that can be used for early diagnosis of AD. The entire research process is divided into four stages: the first stage uses data independent acquisition (DIA) to conduct non-targeted and open exploration of the differential proteins in the peripheral blood of aMCI and normal elderly people (NC); the second stage In this stage, bioinformatics technology is used to gradually screen differential proteins between groups. Through layer-by-layer screening of molecular pathways, functional mechanisms, etc., irrelevant proteins are eliminated and the scope of candidate proteins is established. Screening strategies include: (1) Proteins with statistical differences between groups: Student's t Test P < 0.05 and difference multiples of 1.5 and above; (2) IPA (Ingenuity pathway analysis) bioanalysis software, which analyzes the functional pathways to which the proteins belong, Related diseases were analyzed and screened layer by layer to select proteins related to the pathogenesis of neurological diseases and dementia; in the third stage, candidate proteins were obtained for preliminary verification in 38 elderly people, and those with poor detection and prediction effects were eliminated. Proteins were further screened to obtain 20 candidate biomarkers; in the fourth stage, the diagnostic efficacy of the candidate biomarkers was verified in 155 elderly people. Parallel reaction monitoring (PRM) mass spectrometry technology was used, combined with isotope-labeled synthetic peptides, to accurately quantify 20 target proteins. After verification in 155 elderly people, a group of proteins (RBP4, GPX3, BDNF, AHSG, C1R) was finally obtained. Combined with age and gender, the sensitivity for early diagnosis of aMCI reached 78%, and the specificity reached 75%.

As shown in Figure 1, IPA software analyzes the differential proteins obtained from the comparison between aMCI and NC groups, and the functional pathways to which they belong. There are a total of 17 pathways with statistically significant differences when the -log value of the P value is above 1.3 (equivalent to the p<0.05 level). According to the significance of the statistical differences, they are sorted from large to small: acute response signaling pathway, LXR/RXR Activation, coagulation system, NADH repair, role of tissue factor in cancer, FXR/RXR activation, GP6 signaling pathway, IL-8 signaling, choline biosynthesis III, extrinsic prothrombin activation pathway, early-onset diabetes in youth (MODY) signaling, glycolysis I, gluconeogenesis I, intrinsic prothrombin activation pathway, phospholipases, agrin interactions at the neuromuscular junction, etc.

As shown in Figure 2, the SpectroDive8 software was used to select targeting peptides for more than 110 target proteins. Pre-experiments were conducted based on the selected targeting peptides to analyze the detection effect of the target peptides and eliminate poor signal intensity and detection. Proteins with unsatisfactory detection results will be retained for formal PRM quantitative testing and verification.

As shown in Figure 3, the model consists of five proteins: P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765 (AHSG), and P00736 (C1R), as well as gender and age factors. The overall prediction of MCI by the model The AUC=0.758, the sensitivity was 78%, and the specificity was 75%.

1.DIA mass spectrometry detection

In this example, the number of cases that underwent DIA detection was: 6 cases in the aMCI group and 6 cases in the NC group. The average age of the aMCI group was 68.17±6.369 years old, and the average age of the NC group was 70.50±10.585 years old. There was no statistical difference in age between the two groups (F=1.775, P=0.412>0.05). There was no statistical difference in gender composition between aMCI group and NC group (X ² =0.670, P=0.733>0.05). The test sample is human peripheral blood plasma. Sample collection and processing are completed strictly in accordance with unified specifications. The conditions for blood collection are as follows: the subjects fasted for 12 hours, and the cubital venous blood was taken early the next morning and placed in a red anticoagulant tube. Complete blood centrifugation within 2 hours after collection to separate plasma and blood cells.

Table 1. Reagents and instrument specifications required for DIA mass spectrometry experiments

The DIA mass spectrometry detection process includes sample pretreatment, reverse liquid chromatography separation (RPLC), DDA detection and library construction, DIA individual sample detection, etc. The proteins obtained were detected using the professional analysis software Spectronaut Pulsar of DIA mass spectrometry to extract, correct and integrate the peak information.

1.1 Experimental steps

1.1.1 Sample preparation

The protein extraction requirements required for detection must be free of contamination from environmental proteins, PEG, Triton, glycerol, etc. The peptide preparation requirements must be less than 30% non-specific reductive alkylation, less than 5% non-specific enzyme cleavage rate, and no obvious non-specific Heterosexual grooming. The pre-processing process is as follows:

(1) Plasma is diluted 100 times, and protein quantitative analysis is performed using the BCA method;

(2) According to the measured protein concentration, take 100ug of protein and dilute it to 100ul with 100mM TEAB;

(3) Add 1uL of 1M reducing agent DTT to a final concentration of 10mM, and incubate at 37°C for 1 hour;

(4) Add 2uL 1M iodoacetamide to a final concentration of 20mM, incubate at room temperature, protected from light, for 1 hour;

(5) For precipitation, add 5 times the volume of acetone to the sample and let it stand overnight (12 hours) at -20°C;

(6) Centrifuge at 12000g for 20 minutes at 4°C;

(7) Remove the supernatant, add 1 mL of -20°C ethanol and acetone (ratio 1:1) to the protein precipitate, mix, shake and wash for 5 minutes, repeat twice;

(8) Centrifuge at 12000g for 20 minutes at 4°C;

(9) Remove the supernatant and let the protein precipitate dry naturally at room temperature until the precipitate becomes transparent;

(10) Add 100ul of 50mM ammonium bicarbonate to the precipitate, add 2.5ug of sequencing grade trypsin at a ratio of 1:40 (mass ratio w:w), and enzymatically digest at 37°C overnight.

1.1.2 High performance liquid chromatography separation

The samples were separated using a high pH reversed phase separation method. Take equal amounts of peptide fragments from all samples, mix and freeze-dry, add buffer A (buffer A: 20mM ammonium formate aqueous solution, ammonia water adjusted to pH 10.0), and perform fractionation. The high-performance liquid chromatograph Ultimate3000 system is connected to the XBridge C18coLumicann reversed-phase column and uses a linear gradient for high pH separation. Within 30 minutes, adjust the concentration of mobile phase B (add 20mM ammonium formate to 80% acetonitrile, and adjust the pH value of ammonia water to pH10.0) from 5% to 45%, and balance the column under the initial conditions for 15 minutes. The column maintained a flow rate of 1 ml per minute and a stable temperature of 30°C. A total of 12 fractions were collected under these conditions.

1.1.3 Mass spectrometry data-dependent collection and library construction

After adding 30 μl of solvent C (0.1% formic acid aqueous solution) to each fraction to make a suspension, iRT peptide was added for correction. After separation by nano-liquid phase (nano-LC), tandem mass spectrometry was analyzed by online electrospray. The experiment was performed on an Easy-nLC1200 system connected to a Q Exatives Plus mass spectrometer equipped with an electrospray ion source. After 3 μl of peptide sample was loaded onto the trapping column at a flow rate of 10 μl per minute, it was separated in the analytical column with a linear gradient. Adjust 5% solution D (D: 0.1% formic acid acetonitrile solution) to 40% D within 150 minutes. The column was equilibrated under initial conditions for 10 min, the flow rate was controlled at 300 nL per minute, and the voltage of the electrospray ion source was set to 2 kV. The mass spectrometer works in data-dependent acquisition mode, automatically switching between primary mass spectrometry (MS) and secondary mass spectrometry (MS/MS). A full scan spectrum was obtained at 70K mass resolution, with a mass-to-charge ratio (m/z) range of 350-1550. Automatic gain control (AGC) was used to control the number of ions entering the ion trap, and the target was set to 3e6 , followed by subsequent high-energy collisional dissociation MS/MS scans at 17.5K resolution. Parameter settings: The isolation window is set to 1.6Da, the AGC target is set to 1e5, the MS/MS Fixedfirst mass is set to 200, the microscan is set to 1, and the dynamic exclusion time is 30 seconds.

The original data were searched through the Swiss-Prot homo sapiens (2017-3-27) database. When searching the database, trypsin was set as the digestive enzyme. The specific parameters were set as: (1) The fragment ion mass allowed during the search The error range is 0.050 Daltons; (2) The allowed error range for the precursor ion mass during search is 10.0PPM; (3) Allowed fixed modifications: carbamidomethyl (C); (4) Allowed variables Modifications: asparagine and glutamine deamidation, methionine oxidation, and protein N-terminal acetylation. Establish a spectral database through the above search information and establish comparison standards for subsequent DIA testing.

1.1.4 Mass spectrometry data independent acquisition and data analysis

The DIA detection process is as follows: Take 9ul of each sample, add iRT peptide, and perform it on the Easy-nLC 1200 system. The system is connected to a Q Exactives Plus mass spectrometer equipped with an online electrospray ion source. 3 μl of peptide sample was loaded onto the capture column at a flow rate of 10 μl per minute, and then separated in a linear gradient in the analytical column. The concentration of D solvent was adjusted from 5% to 40% within 150 minutes. The column was equilibrated under initial conditions for 10 min, the flow rate was controlled at 300 nL per minute, and the voltage of the electrospray ion source was set to 2 kV. The mass spectrometer operates in data-independent acquisition mode, automatically switching between primary and secondary mass spectrometry acquisitions. A full scan spectrum was obtained at 70K mass resolution with a mass-to-charge ratio range of 350-1550 and an AGC target of 3e6, followed by a subsequent high-energy collision dissociation MS/MS scan at 17.5K resolution with an AGC target of 3e6 . 39 windows were used to fragment and scan all peptides within this range, and the window coverage is as follows (Table 2).

Table 2. DIA scan window coverage

Spectronaut Pulsar software was used to optimize data acquisition and analysis, which mainly included: setting the cycle time (to obtain optimal extraction of elution peaks), adjusting the precursor ion window according to the chromatographic peak width, and adjusting the cycle time. Extract 6-8 data points for each peak for analysis, extract, correct, and integrate the peak information, and analyze the DIA mass spectrometry data with the default parameter settings (BGS FactorySettings-default). The protein qualitative standard is the precursor ion threshold. (precursorthreshold) is set to 1.0% FDR (false discovery rate). At the same time, the software was used to conduct preliminary analysis of differential proteins among the three groups. The standard for differential proteins was pairwise comparison between the three groups: Student’s t Test analyzed proteins with P < 0.05 and a difference fold of 1.5 or above.

2. Analysis of differential proteins between groups and screening of candidate biomarkers

Since PRM targeted mass spectrometry detection has certain limitations on the number of target proteins, in order to ensure the accuracy of quantification of low-abundance proteins in subsequent studies, hundreds of proteins obtained by DIA detection need to be analyzed and gradually screened to identify candidate organisms. markers for subsequent validation and diagnostic analysis. The differential proteins are imported into the IPA (Ingenuity pathway analysis) bioanalysis software, and the functional pathways and related diseases of the proteins are analyzed (Figure 1). Screening is carried out layer by layer to select proteins related to the pathogenesis of neurological diseases and dementia; identify the most Target proteins that may be related to AD enter the next step of verification;

3.PRM quantitative analysis (preliminary verification)

The candidate proteins were obtained for preliminary verification in 38 elderly people, and proteins with poor detection results were eliminated for further screening.

3.1 Experimental materials, reagents and equipment

The collected biological samples were subjected to protein extraction, reductive alkylation and enzymatic hydrolysis using Biognosys Preparation KitPro (Biognosys AG), and then an appropriate amount of enzymatically digested peptides from each sample was mixed into PQ500 at a certain ratio. Peptide labeling reagent. The prepared samples first use the Unscheduled PRM method to collect target peptides to optimize the retention time window. After optimizing the method, formal experiments can be carried out. The collected data were used for identification and quantitative analysis of target peptides using SpectroDive (Biognosys AG) software.

Table 3. Reagents and instrument specifications required for PRM detection

Experimental steps

3.2PRM pre-experiment

The criteria for selecting targeted peptides are: ① Characteristic peptide (unique peptide); ② The length of the peptide sequence is greater than 7 amino acids and less than 20 amino acids; ③ The peptide is completely digested with no missed cleavage sites, and the peptide sequence is Does not contain proline; ④ The peptide segment may contain fixed modifications, but does not contain variable modifications.

Based on the selected peptide information, conduct a preliminary experiment and use 2 prepared peptide samples for detection. Take an equal amount of peptide fragments from the sample, mix and freeze-dry, redissolve in buffer A (buffer A: 20mM ammonium formate aqueous solution, ammonia water adjusted to pH 10.0), connect the reverse column to the Ultimate3000 system, and perform high pH reverse separation. Use a linear gradient, rising from 5% B to 45% B within 30 minutes (B: 20mM ammonium formate was added to 80% acetonitrile, and ammonia was adjusted to a pH of 10.0). The column was equilibrated under initial conditions for 15 minutes. The column maintained a flow rate of 1 ml per minute and the temperature was maintained at a stable temperature of 30°C. Six fractions were collected under these conditions for analysis.

Add 30uL solvent C to each fraction (solvent C is 0.1% formic acid aqueous solution; solvent D is 0.1% formic acid acetonitrile solution) to form a suspension, add iRT peptide, use reverse liquid phase separation, and use online electrospray ion source for tandem mass spectrometry. Detection and analysis. The experiment was completed using Easy-nLC 1000system. At the same time, the system was connected to the Orbitrap Fusion Tribrid mass spectrometry detection instrument equipped with an electrospray ion source. 10uL peptide sample was loaded onto the trap column at a rate of 10uL per minute, and then separated in a linear gradient in the analytical column, increasing the 3% D concentration to 32% D concentration within 120 minutes, and the column was equilibrated under the initial conditions. For 10 minutes, control the flow rate of 300nL per minute, and the voltage of the electrospray ion source is 2 kV.

The mass spectrometer first works in data-dependent acquisition mode, and the instrument automatically switches between primary mass spectrometry (MS) and secondary mass spectrometry (MS/MS) acquisition. Scanning at 60K mass resolution, the mass-to-charge ratio (m/z) range of the full scan spectrum is between 350-1550, and then completing the subsequent high-energy collision (HCD) secondary mass spectrometry (MS/MS) at 15K resolution )scanning. Specific parameter settings: the isolation window is set to 1.6Da, the AGC target is set to 400000, the MS/MS Fixedfirst mass is set to 110, the Microscan is set to 1, and the dynamic exclusion time is 45 seconds.

The pre-experimental data was imported into Spectronaut Pulsar, and the database was searched through the embedded Mascot Distiller version 2.6. The database was set to Swissprot homo sapiens (2017-3-27), and trypsin was set as the enzymatic enzyme. Specific parameters used in the database search: the allowed fragment ion mass error range is 0.050Da; the allowed precursor ion mass error range is 10.0PPM; the allowed fixed modification is Carbamidomethyl (C); the allowed variable modification is Asparagine and Glutamine deamidation, Methionine oxidation , Protein N-terminal acetylation. The library search results were imported into SpectroDive 8 software for analysis, and proteins that could not be detected or whose target peptide signal intensity was not ideal were removed. Precursor ions and analysis methods were set based on the pre-experiment results, and then formal PRM detection was performed.

3.3PRM formal testing and data collection

Take 9uL sample, add iRT peptide, and detect it on the same mass spectrometer after passing through the online electrospray ion source. Set the mass spectrometry detection gradient to 120 minutes. 2uL peptide sample was loaded onto the trapping column at a flow rate of 10uL per minute, and separated in the analytical column using a linear gradient: the concentration of solvent D was increased from 3% to 32% within 120 minutes. The analytical column was equilibrated under initial conditions for 10 minutes, the flow rate was controlled at 300 nL per minute, and the electrospray ion source voltage was 2 kV. The mass spectrometer first works in data-dependent acquisition mode, and the instrument automatically switches between MS (first-level mass spectrometry) and MS/MS (secondary-level mass spectrometry) acquisition. Scan at 60K mass resolution to obtain a full scan spectrum with a mass-to-charge ratio range of 350-1550. Then perform a subsequent high-energy collision MS/MS scan at 15K resolution, and select the PRM module for mass spectrometry. Each sample was tested in duplicate based on the initial test to increase the accuracy of protein quantification. The collected data was imported into SpectroDive 8 for analysis, and the signal intensity of the sample was corrected to eliminate human errors caused by sample preparation and instrument detection. The best six fragment ion (product ion) signals are collected for each peptide, and the sum of the peak areas of the fragment ion spectrum represents the signal intensity of the peptide. Select 3-5 peptides for each protein, and the average signal intensity of the peptides represents the expression level of the protein.

4. Isotope-labeled synthetic peptides are combined with PRM technology for accurate quantification and verification (expanded sample verification).

The 20 candidate biomarkers obtained from the above screening were combined with isotope-labeled synthetic peptide PRM technology for precise quantification and expanded sample verification. The verification subjects were 155 elderly people, and their general information is shown in Table 4.

Table 4. Comparison of general information between NC and aMCI groups

Note: NC=normal cognitive function; aMCI=amnestic mild cognitive impairment

Quantification was performed using an Orbitrap Fusion ^™ Lumos ^™ Tribrid ^™ mass spectrometer (Thermo Fisher Scientific, MA, USA) with a high-precision tandem EASY-nanoLC1200. First, by selecting representative peptides of the target protein and collecting mass spectrometry signals on the peptides, quantitative information of the target protein is obtained. At the same time, by synthesizing isotope standard peptides and drawing a standard curve, absolute quantification of the protein is achieved.

First, the precursor ion information of the target peptide is selectively detected in the primary mass spectrometer (Q1); the precursor ions are fragmented; finally, a high-resolution, high-mass-accuracy analyzer is used to detect the selected precursor ions in the secondary mass spectrometer. Information about all fragments within the window. The experiment contains 2 sets of samples. During sample preparation, PQ500 (Ki-3019-96, Biognosys AG, Switzerland) isotope-labeled peptides were added to all samples to be detected for identification and quantitative analysis of target peptides. All samples were collected randomly. The collected PRM data were normalized and quantitatively analyzed using SpectronDive 9.10 (Biognosys AG, Switzerland) software. The p value was calculated using the Students ttest (unpaired, two-tailed) algorithm. The average value of each group was used to calculate the fold difference. Sample preparation was performed according to the instructions of BiognosysSample Preparation Kit Pro (Ki-3013, Biognosys AG, Switzerland). The prepared peptides were analyzed by PRM via an LC-MS/MS system equipped with an online nanospray ion source.

5Data Analysis and Modeling

5.1 Perform statistical analysis on the obtained quantitative results of the target protein. First, we performed an analysis of the efficacy of a single protein in diagnosing MCI. The obtained results were screened according to the standards of P<0.05 and AUC>0.58. The five proteins finally obtained were P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765(AHSG) and P00736(C1R) have good diagnostic value. The AUCs of predicted MCI were 0.61, 0.62, 0.60, 0.60, and 0.58 respectively. The concentration and AUC of the target protein are shown in Figure 3.

5.2 In order to further improve the diagnostic value, the obtained protein and other data are then modeled. The purpose of modeling is to find an optimal model to more effectively distinguish aMCI from healthy people. Adjust the ratio of training set and test set to 75%:25%; perform Bagging Tree missing value interpolation on all data. First, use the data with missing values filled in to perform peptide screening, and perform a t-test on each peptide. Secondly, the proteins with poor potency are eliminated, that is, only the peptide with the smallest pvalue is retained to represent the protein content. Carry out model training on the proteins screened above, using the Bootstrap 1000-time resampling method. Take 75% of all samples as training set samples, and the remaining 25% as prediction set samples. Logistic regression method was used to build the model. The model is composed of five proteins: P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765 (AHSG), and P00736 (C1R), as well as gender and age factors. The overall predicted MCI of the model is AUC=0.758, which is sensitive. The accuracy was 78% and the specificity was 75% (see Figure 4).

Fitting formula:

In addition to the technical solutions mentioned above, mass spectrometry can also be used to detect the five proteins P02753 (RBP4), P22352 (GPX3), P23560 (BDNF), P02765 (AHSG), P00736 (C1R) and their associated peptides, as well as using Antigen-antibody binding technology and aptamer binding technology were used to detect these five proteins.

In this specification, the invention has been described with reference to specific embodiments thereof. However, it is apparent that various modifications and changes can be made without departing from the spirit and scope of the invention. Accordingly, the specification and drawings are to be regarded as illustrative rather than restrictive.

Claims (5)

1. A peripheral blood protein marker for early diagnosis of senile dementia, wherein the protein marker comprises protein P02753 (RBP 4), P22352 (GPX 3), P23560 (BDNF), P02765 (AHSG) or P00736 (C1R).

2. Use of a reagent for detecting the content of the peripheral blood protein marker according to claim 1 in the preparation of a kit for early diagnosis of senile dementia.

3. The use according to claim 2, wherein the kit comprises reagents for determining the content of the P02753 (RBP 4) protein or the P22352 (GPX 3) protein or the P23560 (BDNF) protein or the P02765 (AHSG) protein or the P00736 (C1R) protein in the plasma.

4. The use according to claim 3, wherein the kit is used for determining any peptide fragment of P02753 (RBP 4) or P22352 (GPX 3) or P23560 (BDNF) or P02765 (AHSG) or P00736 (C1R).

5. The diagnosis system is characterized by comprising an early diagnosis model of senile dementia, wherein the model comprises proteins P02753 (RBP 4), P22352 (GPX 3), P23560 (BDNF), P02765 (AHSG), P00736 (C1R), age and gender.