Urine metabolism combined marker for diagnosing bladder cancer and screening method thereof

Abstract

The invention relates to a urine metabolism combined marker for diagnosing bladder cancer and a screening method of the urine metabolism combined marker. The urine metabolism combined marker is prepared from 2-methyl furan, dimethyl disulfide, octylene and acetophenone. The method for screening the urine metabolism combined marker comprises the following steps: carrying out chromatographic separation and mass spectrometry detection on a urine sample on the basis of a headspace automatic sampler in combination with a chromatographic column and a photochemical ionization time-of-flight mass spectrometry, preprocessing mass spectrometry data, and extracting spectrum peak data of metabolites in urine; the urine metabolism combination marker for diagnosing bladder cancer is obtained through further statistical analysis. The screening method of the urine metabolism combined marker for diagnosing the bladder cancer has the advantages of non-invasiveness, high-throughput detection, simplicity in operation and accurate result, and has a wide application prospect.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical diagnosis of bladder cancer, and particularly relates to a urine metabonomic marker for diagnosing bladder cancer and a screening method thereof. BACKGROUND

[0002] Bladder cancer is one of the most common malignant tumors of the urinary system worldwide, and its incidence is on the rise. Early and accurate diagnosis is crucial for improving patient survival and quality of life. At present, the "gold standard" for the diagnosis of bladder cancer is invasive cystoscopy combined with suspicious lesion tissue biopsy. Although this method is accurate, it has obvious limitations: strong invasiveness, which brings discomfort to patients (such as pain, bleeding, and the risk of urinary tract infection); high cost, especially for high-risk patients who need long-term monitoring; and complex operation, which requires professional urologists and equipment. Non-invasive or minimally invasive screening and auxiliary diagnostic methods mainly include urine cytology and urine-based molecular marker detection (such as NMP22, BTA, UroVysion FISH, etc.). However, urine cytology has high specificity but severely insufficient sensitivity, especially for low-grade tumors, resulting in a large number of false negative results. Existing molecular marker detection has improved sensitivity in some cases, but its specificity is often insufficient, can be disturbed by benign diseases (such as inflammation, stones), and can lead to false positives; the sensitivity and specificity vary greatly in different studies, and the stability needs to be improved; and the diagnostic performance of a single marker is limited.

[0003] Metabolomics is a discipline that systematically studies all small molecule metabolites in the body. These metabolites are the final downstream products of gene expression, protein activity, and cell physiological state. Cancer is a metabolic disease, and tumor cells undergo deep metabolic reprogramming to meet their rapid proliferation and invasion needs. This reprogramming leaves a unique metabolic fingerprint or profile in the tumor microenvironment and body fluids. Urine, as an easily accessible and completely non-invasive body fluid, collects the end products of whole-body metabolism and metabolic changes in the local environment of the kidney. Therefore, urinary metabolomics analysis provides a promising direction for exploring molecular diagnostic markers for bladder cancer. Urine samples can be repeatedly obtained without special treatment, making them very suitable for early screening, diagnosis, staging and grading, prognosis evaluation, and treatment response monitoring.

[0004] In summary, there is an urgent need and great clinical application value for developing an efficient, stable, standardized, and excellent diagnostic urine metabolic marker and its screening method to overcome the limitations of existing bladder cancer diagnosis methods (invasiveness, high cost, insufficient sensitivity / specificity, lack of universal standard markers), achieve non-invasive, precise, and rapid bladder cancer screening, auxiliary diagnosis, and recurrence monitoring. SUMMARY

[0005] In view of the defects of the existing bladder cancer diagnosis methods, such as invasiveness, high cost, insufficient sensitivity / specificity, and lack of universal standard markers, the purpose of the present application is to provide a urine metabolic combination marker for diagnosing bladder cancer and a screening method thereof, wherein the screening method is based on headspace autosampler combined with chromatographic column and photochemical ionization time-of-flight mass spectrometry for chromatographic separation and mass spectrometry detection of urine samples, and further obtaining the urine metabolic combination marker for diagnosing bladder cancer through statistical analysis.

[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:

[0007] A urine metabolic combination marker for diagnosing bladder cancer, which is composed of 2-methylfuran (m / z=82.05), dimethyl disulfide (m / z=93.95), octene (m / z=113), and acetophenone (m / z=120.9).

[0008] A method for screening a urine metabolic combination marker for diagnosing bladder cancer, comprising the following steps:

[0009] 1) Collect urine samples of bladder cancer patients and healthy volunteers respectively; and pretreat the urine samples;

[0010] 2) Use headspace autosampler combined with chromatographic column and photochemical ionization time-of-flight mass spectrometry to perform chromatographic separation and mass spectrometry detection on the urine samples, and obtain original mass spectrometry data;

[0011] 3) Pretreat the obtained original mass spectrometry data and extract metabolite spectrum peak data in the urine;

[0012] 4) Perform multivariate analysis and univariate analysis on the metabolite spectrum peak data to obtain potential differential metabolic markers; and further obtain the urine metabolic combination marker for diagnosing bladder cancer through binary logistic regression analysis.

[0013] All urine collectors fast overnight before collecting urine samples, and collect midstream urine of the first morning urine on the next day; after the urine samples are divided into sample bottles, they are stored in a-80℃ refrigerator; the urine collectors are bladder cancer patients and healthy volunteers; the healthy urine samples and the bladder cancer urine samples are taken from people aged 20-75 years old; the healthy urine samples are more than 20 cases, and the bladder cancer urine samples are more than 20 cases.

[0014] The 1.0-5.0 mL of the thawed urine sample is taken by a pipette and placed in a glass sample bottle with a volume of 10-20 mL, then 100-200 μL of HCl with a concentration of 4-8 mol / L is added to the glass sample bottle to adjust the pH of the urine to 1-2, then 3.0-6.0 g of NaCl with a purity of 99.8% is added, and finally 100-500 μL of chlorobenzene internal standard solution with a concentration of 10-50 ppm is added. After oscillation and dissolution, the sample is stored in the refrigerator for detection.

[0015] When the urine sample is subjected to chromatographic separation, the chromatographic column used is an FFAP-13-2 polar fused silica capillary column from Agilent.

[0016] When the urine sample is subjected to mass spectrometry data acquisition, photochemical ionization time-of-flight mass spectrometry is used for detection. The ionization source uses a dual-channel sampling ionization source based on a vacuum ultraviolet lamp. One channel is a water molecule sampling pipeline, which is connected to a helium purge gas source through a bubbling bottle filled with water. The helium gas is humidified by the water in the bubbling bottle and introduced into the ionization source in large quantities. The helium gas flow rate is 10-100 mL / min. One channel is connected to the sample gas after chromatographic separation. The sample gas flow rate is 1-20 mL / min. The ionization source pressure is 200-800 Pa. The ionization source repulsion voltage is 10-100 V. The spectrum acquisition time is 0.5-10 min. The mass-to-charge ratio range of the spectrum data acquisition is 0-500 amu.

[0017] The process of preprocessing the obtained raw mass spectrum and extracting the metabolite spectrum peak data in the urine is as follows: the water cluster peak (m / z = 37.02897), acetone (m / z = 59.04969) and chlorobenzene (m / z = 112.00798) are used for mass correction and alignment of the obtained raw mass spectrum. When extracting the metabolite spectrum peak data, the data processing software "PI_TOFMS_DATA_PROCESS.exe" is used to extract the metabolite spectrum peak data. The signal peaks with a signal value less than "100" are set to "0", and the metabolite spectrum peaks with a signal peak of "0" accounting for more than 80% in the sample data are removed.

[0018] The process of multivariate analysis and univariate analysis of the metabolite spectrum peak data is as follows: Simca-P software is used for multivariate analysis, and metabolite spectrum peak data with a VIP greater than 1 is screened out by partial least squares discriminant analysis. SPSS software is used for univariate analysis, and metabolite spectrum peak data with a p value less than 0.05 is screened out by T test. The intersection of the two sets of data is obtained, i.e. the potential differential metabolites.

[0019] Screening of the metabolic combination marker: first, the obtained potential differential metabolites are subjected to binary logistic regression analysis on the SPSS software, the grouping information of the healthy volunteers and the bladder cancer patients is set as the dependent variable, the relative intensity of the different metabolites is set as the covariate, and the method is "forward: formatted", the first ranked combination in all the obtained combinations is the optimal differential metabolite combination, and the differential metabolite combination is the urine metabolic combination marker for diagnosing bladder cancer.

[0020] The screening method of the present application can also establish a working characteristic curve discrimination model of the subject based on the metabolic combination marker, and obtain the discrimination formula and discrimination threshold of the bladder cancer and the healthy volunteers. Specifically, the metabolic marker combination is subjected to binary logistic regression analysis on the SPSS software, then the metabolites in the optimal combination are selected as the covariate, and binary logistic regression analysis is performed again to obtain the prediction probability of the combination;

[0021] Then, the obtained metabolic marker combination is subjected to receiver operating characteristic curve (ROC) analysis on the SPSS software, the prediction probability of the differential metabolite combination is set as the test variable, the grouping information is set as the state variable, and the state variable value is the disease grouping. Finally, the ROC curve is obtained, and the abscissa and ordinate are 1-specificity and sensitivity, respectively. When the sum of specificity and sensitivity is maximum, the corresponding value is the discrimination threshold of the healthy volunteers and the bladder cancer patients. Greater than the threshold value is a healthy person, and less than or equal to the threshold value is a bladder cancer patient.

[0022] Compared with the prior art, the advantages of the present application are that the urine metabolic combination marker for diagnosing bladder cancer in the present application has high sensitivity, strong specificity, high accuracy, and good prediction ability for diagnosis. Compared with the traditional GC-MS or LC-MS detection method, the analysis speed of the present application is faster, and compared with the direct injection detection of the urine sample, the present application can effectively reduce the interference of the complex matrix in the urine, and the analysis result is more accurate and reliable. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The multivariate analysis OPLS-DA analysis result graph and the 200 times permutation verification result graph in the embodiment, wherein (a) is the multivariate analysis OPLS-DA analysis result graph; (b) is the 200 times permutation verification result graph.

[0024] Figure 2 The receiver operating characteristic curve analysis result graph of the established bladder cancer diagnosis model. DETAILED DESCRIPTION

[0025] EMBODIMENT

[0026] 1. Subject

[0027] Methods: 69 patients with early bladder cancer (BC group) and 129 healthy controls (CTL group) were enrolled in this study. The age range was 40-70 years.

[0028] 2. Collect urine samples and pretreat them;

[0029] All subjects fasted overnight before urine sample collection. The midstream urine samples were collected from the first morning urine of all subjects. The collected urine samples were stored in sample bottles and kept in a -80℃ refrigerator.

[0030] When the urine samples were detected, they were thawed at 4℃. 1.0 mL of the thawed sample was taken with a pipette and placed in a 10 mL glass sample bottle. Then, 100 μL of 4 mol / L HCl was added to adjust the pH of the urine to 1-2. Then, 3.0 g of 99.8% pure NaCl was added. Finally, 100 μL of 50 ppm chlorobenzene solution was added. After shaking and dissolving, the sample was stored in the cold room of a 4℃ refrigerator for detection.

[0031] 3. The urine samples were separated and detected by headspace autosampler, chromatographic column, and photochemical ionization time-of-flight mass spectrometry.

[0032] The ionization source of the photochemical ionization time-of-flight mass spectrometer was a dual-channel ionization source based on a vacuum ultraviolet lamp (VUV lamp). One channel was a sample inlet line connected to the outlet of the chromatographic column, through which the sample gas separated by chromatography was introduced. The other channel was a water molecule inlet line connected to a helium purge gas source through a bubbling bottle filled with water. The helium gas was humidified by the water in the bubbling bottle and introduced a large amount of water molecules into the ionization source. The ionization source pressure was 500 Pa, the ionization source repelling voltage was 28 V, and the MCP voltage was 4200 V. The spectrum acquisition time was 10 min, and the mass-to-charge ratio range of the spectrum data acquisition was 0-300 amu.

[0033] Mass spectrometry data acquisition on urine samples: The prepared urine samples were placed in the sample disc of the headspace autosampler, and the mechanical arm would move the samples to the heated disc according to the set program. After heating and equilibrating at 70℃ for 4 min, the pressurization and sampling instructions were executed, and the headspace gas of the sample bottle was temporarily stored in a 3 mL quantification ring using nitrogen purge. Finally, the sample gas in the quantification ring was blown into the chromatographic column for volatile organic compound (VOC) separation by switching the six-way valve. The chromatographic column used was Agilent's FFAP-13-2 polar fused silica capillary column (20 m x 0.53 mm x 1 μm), and nitrogen was used as the purge elution gas. The samples separated by the chromatographic column entered the photochemical ionization time-of-flight mass spectrometer in turn through the sample inlet line at a flow rate of 12 mL / min. The water molecule inlet line of the mass spectrometry ionization source was connected to a helium gas flow of 50 mL / min, which was humidified by a water-filled bubbling bottle to introduce a large amount of water molecules into the ionization source. The water molecules were first ionized by VUV lamp irradiation to H3O+, and then underwent chemical ionization reaction: H3O + + M → M + H + + H2O. The ions were continuously transmitted and shaped in the mass spectrometer, finally reached the field-free flight region, and then finally reached the MCP detector. After signal amplification, the time-of-flight and signal intensity mass spectrum was obtained, and finally converted to the mass spectrum of mass-to-charge ratio and intensity. During the entire analysis process, the sample transmission line and mass spectrometry ionization zone temperature were maintained at 150℃, the chromatographic carrier gas was nitrogen, the flow rate was 14 ml / min, and the single urine sample test time was 600 s.

[0034] 4. Data preprocessing

[0035] The obtained raw mass spectrum was corrected and aligned using water cluster peaks (m / z = 37.02897, water was introduced as a reaction reagent ion), acetone (m / z = 59.04969, human urine contains a high concentration of acetone), and chlorobenzene (m / z = 112.00798, internal standard, 100 μL was added during sample pretreatment). The mass spectrometry data processing software "PI_TOFMS_DATA_PROCESS.exe" was used to extract the metabolite spectrum peak data, and the signal peaks with signal values less than "200" were set to "0". Metabolite spectrum peaks with a signal peak of "0" accounting for more than 80% in the sample data were removed. Finally, an Excel table containing metabolite relative intensity information was obtained, and part of the results are shown in Table 1. Each row represents a sample information, and each column represents a metabolite information.

[0036] Table 1 Partial results table containing metabolite relative intensity information

[0037]

[0038] 5. Screening process for differentially metabolites

[0039] Import the Excel file containing metabolite information into Simca-P software for multivariate analysis. Open the software settings interface, select "Par" as the Scaling mode, and then set the grouping information: one group is bladder cancer patients ("BC"), and the other group is healthy controls ("CTL"). Select "OPLS-DA" (Partial Least Squares Discriminant Analysis) as the model type, and close the settings window after confirmation. Click "Fit" to start fitting the data. Then click "Scores" to plot the scores (e.g., ...). Figure 1 (a) shows the separation of the two groups, where R 2 =0.80Q 2 =0.64, indicating that the model is robust and has good predictive ability. The model was then validated with 200 permutations, where R² = 0.296 and Q² = -0.463, indicating that the model is not overfitting. Figure 1 (b) shows the process. Next, click "VIP" to obtain the metabolite importance ranking table, and export the data for metabolites with a VIP greater than 1. Then, import the Excel file containing metabolite information into SPSS software for univariate analysis. Select "Independent Samples T-Test" in the "Analysis" menu to obtain a table containing significance (p-value). After exporting the data, filter out the metabolite data with p-values ​​less than 0.05. Finally, compare the data obtained after processing by the two methods, and take the intersection of the two sets of data as the potential differential metabolic biomarkers, as shown in Table 2.

[0040] Table 2 Potential metabolic biomarkers obtained from univariate and multivariate analyses.

[0041]

[0042]

[0043] 6. Screening of differential metabolite combinations and establishment of differential metabolite combination models

[0044] The dataset of potential metabolic markers was imported into SPSS, the binary Logistic function in the method >> regression was selected, all metabolites were imported into the analysis box, the grouping information of healthy volunteers and bladder cancer patients was set as the dependent variable, and then the function was selected: forward, click OK. In the "variables in the equation" in the analysis results, find the differential metabolites, then go back to the binary Logistic analysis, import the differential metabolites into the analysis box, and select the function: conditional, click OK. Further screen and combine the potential differential metabolites, and the first ranked combination in all combinations is the optimal differential metabolite combination. The metabolic marker combination consists of four compounds, and the final determined differential metabolite combination is m / z = 82.05, 93.95, 113 and 120.9, which correspond to 2-methyl furan, dimethyl disulfide, octene and acetophenone respectively.

[0045] Then go back to the binary Logistic analysis, import the differential metabolite combination into the analysis box, select the function "conditional", click "OK", and get the data for drawing the receiver operating characteristic curve (ROC). The ROC curve is drawn using the "ROC curve" function in SPSS "analysis", setting the prediction probability of the differential metabolite combination as the test variable, the grouping information as the state variable, and the state variable value as the disease grouping. The results are shown in Figure 2 The area under the ROC curve (AUC) is 0.927, the sensitivity is 81.0%, the specificity is 89%, the accuracy is 86.2%, and the model has good prediction ability. When the sensitivity and specificity add up to the maximum, the corresponding threshold value is 0.6917, that is, when the threshold value is higher than 0.6917, it is a healthy person, and lower than 0.6917, it is a bladder cancer patient. Finally, the discriminant formula for diagnosing bladder cancer and non-bladder cancer is obtained as

[0046] y = e x / (x+1)

[0047] where x = 0.002112 x I 2-甲基呋喃 -0.000049 x I 二甲基二硫醚 +0.000229 x I 辛烯 +0.000637 x I 苯乙酮 -7.511, I represents the signal intensity of the differential metabolite, and y represents the discriminant threshold.

Claims

1. A urinary metabonomic marker for the diagnosis of bladder cancer, characterized in that, The urine metabolome combined marker consists of 2-methylfuran (m / z=82.05), dimethyl disulfide (m / z=93.95), octene (m / z=113) and acetophenone (m / z=120.9).

2. A method of screening the urinary metabonomic markers for diagnosing bladder cancer according to claim 1, characterized by, The method comprises the following steps: 1) Collecting urine samples of bladder cancer patients and healthy volunteers respectively; and pre-treating the urine samples; 2) Carrying out chromatographic separation and mass spectrometric detection on the urine samples by using a headspace autosampler in combination with a chromatographic column and a photochemical ionization time-of-flight mass spectrometer to obtain original mass spectrometric data; 3) Pre-treating the obtained original mass spectrometric data and extracting metabolite spectrum peak data in the urine; 4) Carrying out multivariate analysis and univariate analysis on the metabolite spectrum peak data to obtain potential differential metabolite markers; and further obtaining a urine metabolome combined marker for diagnosing bladder cancer by binary logistic regression analysis.

3. The method of claim 2, wherein, All urine collectors fast overnight before collecting the urine samples, and collect midstream urine of the first morning urine on the next day; the collected urine samples are divided into sample bottles and stored in a-80℃ refrigerator; the urine collectors are bladder cancer patients and healthy volunteers; the healthy urine samples and the bladder cancer urine samples are taken from people aged 20-75; the healthy urine samples are more than 20, and the bladder cancer urine samples are more than 20.

4. The method of claim 2, wherein, The pre-treatment process of the urine samples is as follows: 1.0-5.0 mL of the thawed urine sample is taken out by a pipette and placed in a glass sample bottle with a volume of 10-20 mL, then 100-200 μL of HCl with a concentration of 4-8 mol / L is added to the glass sample bottle to adjust the pH of the urine to 1-2, then 3.0-6.0 g of NaCl with a purity of 99.8% is added, and finally 100-500 μL of chlorobenzene internal standard solution with a concentration of 10-50 ppm is added, and after oscillation and dissolution, the sample is stored in a refrigerator for detection.

5. The method of claim 2, wherein, When the urine samples are subjected to chromatographic separation, the chromatographic column used is an FFAP-13-2 polar fused silica capillary column of Agilent, When the urine samples are subjected to mass spectrometric data acquisition, photochemical ionization time-of-flight mass spectrometry is used for detection, the ionization source uses a double-path sampling ionization source based on a vacuum ultraviolet lamp, one path is a water molecule sampling pipeline, the water molecule sampling pipeline is connected with a helium gas blowing source through a bubbling bottle, the bubbling bottle is filled with water, the helium gas is humidified by the water in the bubbling bottle and introduced into the ionization source with a large amount of water molecules, the flow rate of the helium gas is 10-100 mL / min, one path is used to introduce the sample gas after chromatographic separation, the flow rate of the sample gas is 1-20 mL / min, the ionization source gas pressure is 200-800 Pa, the ionization source repulsion voltage is 10-100 V, the MCP voltage is 4000-4200 V, the spectrum acquisition time is 0.5-10 min, and the mass-to-charge ratio range of the spectrum data acquisition is 0-500 amu.

6. The method of claim 2, wherein, The pre-treatment process of the obtained original mass spectrum is as follows: the obtained original mass spectrum is subjected to mass correction and alignment by using water cluster peaks (m / z=37.02897), acetone (m / z=59.04969) and chlorobenzene (m / z=112.00798). When extracting the metabolite spectrum peak data in urine, the data processing software "PI_TOFMS_DATA_PROCESS.exe" is used to extract the metabolite spectrum peak data, the signal peak with a signal value lower than "100" is set to "0", and the metabolite spectrum peak with a signal peak of "0" accounting for more than 80% in the sample data is removed.

7. The method of claim 2, wherein, The multivariate analysis and univariate analysis process of the metabolite spectrum peak data is as follows: the Simca-P software is used for multivariate analysis, and the metabolite spectrum peak data with a VIP greater than 1 is screened out by the partial least squares discriminant analysis method; the SPSS software is used for univariate analysis, and the metabolite spectrum peak data with a p value less than 0.05 is screened out by T test, and the intersection of the two groups of data is obtained, that is, the potential difference metabolites.

8. The method of claim 2, wherein, Screening of metabonomic markers: first, binary logistic regression analysis is performed on the potential difference metabolites obtained on the SPSS software, the grouping information of the health volunteers and the bladder cancer patients is set as the dependent variable, the relative intensity of different metabolites is set as the covariate, and the method is "forward: formatted", and the combination ranked first in all combinations is the optimal difference metabolite combination, which is the urine metabonomic marker for diagnosing bladder cancer.

9. The method of claim 2, wherein, The method can also establish a receiver operating characteristic curve discriminant model of the subject based on the metabonomic marker combination, and obtain the discriminant formula and discriminant threshold of the bladder cancer and the health volunteers.

10. The method according to claim 2 or 9, characterized in that, The receiver operating characteristic curve discriminant model of the subject based on the metabonomic marker combination is established, specifically: binary logistic regression analysis is performed on the metabonomic marker combination on the SPSS software, Then, the method is set to "input", the metabolites in the optimal combination are selected as the covariate, and binary logistic regression analysis is performed again to obtain the prediction probability of the combination; Then, the metabonomic marker combination obtained is analyzed by receiver operating characteristic curve (ROC) in the SPSS software, the prediction probability of the difference metabolite combination is set as the test variable, the grouping information is set as the state variable, and the state variable value is the disease grouping, and finally the ROC curve is obtained, the horizontal and vertical coordinates are 1-specificity and sensitivity, respectively, and when the sum of specificity and sensitivity is maximum, the corresponding value is the discriminant threshold of the health volunteers and the bladder cancer patients, and the value greater than the threshold is the healthy person, and the value less than or equal to the threshold is the bladder cancer patient.

Images