Urine metabolite detection method taking trimetal oxide nanocage as matrix

Abstract

The invention discloses a urine metabolite detection method taking a trimetal oxide nanocage as a matrix, which comprises the following steps: dropwise adding a trimetal oxide nanocage solution on a target plate containing a sample, air-drying, and carrying out laser desorption ionization mass spectrometry to obtain a urine metabolite fingerprint spectrum, the method is applied to detection of bladder cancer, prostatic cancer and kidney cancer in combination with a Gradient Boosting machine learning algorithm, and has extremely high accuracy and precision. The trimetal oxide nanocage synthesized by the invention is obtained by one-step simple ion exchange and high-temperature calcination, has a rough surface, and has good crystal form, excellent ultraviolet absorption capacity, charge and heat transfer capacity and higher heat stability. The material is used as a matrix of laser desorption / ionization mass spectrometry, and has wide application prospects in the aspects of large-scale population screening and disease diagnosis and identification.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of nanomaterials and biological detection technology, and particularly relates to a urine metabolite detection method using a trimetal oxide nanocage as a matrix, especially to a urine metabolite detection method using a hollow dodecahedral trimetal oxide nanocage as an LDI-MS matrix, and application of the method in detection of bladder cancer, kidney cancer and prostate cancer. BACKGROUND

[0002] As a powerful omics tool, metabolomics can comprehensively analyze metabolites in biological fluids, tissues and cells, and has shown great potential in disease diagnosis and biomarker discovery, especially for the analysis of easily accessible samples such as urine, serum and plasma. Among commonly used biological fluids, urine has become an important carrier for clinical metabolomics research due to its non-invasive collection characteristics and ability to reflect integrated phenotypic information of multiple organs. This feature is particularly important in oncology, as cancer-related metabolic reprogramming is often reflected through changes in urine metabolite profiles. Therefore, urine-based metabolomics provides a new approach for the diagnosis and prognostic marker discovery of malignant tumors of the urinary system. Although a large number of studies have used urine metabolomics to distinguish urinary cancer patients from healthy people, studies directly comparing the metabolic characteristics of bladder cancer, kidney cancer and prostate cancer are still insufficient. Such comparative analysis is crucial for revealing common and cancer-specific metabolic changes, thereby achieving precise molecular typing through the "divide and conquer" strategy and promoting early detection of different types of urinary system cancer.

[0003] Mass spectrometry (MS) is widely recognized as a basic technology for metabolite analysis due to its broad molecular coverage, high sensitivity and structural resolution. Among various mass spectrometry methods, laser desorption / ionization mass spectrometry (LDI-MS) has increasingly demonstrated its value in metabolomics due to its high throughput, low sample consumption and rapid analysis. The efficiency of LDI-MS for small molecule detection mainly depends on the matrix used. Traditional organic matrices such as 2,5-dihydroxybenzoic acid (DHB) and α-cyano-4-hydroxycinnamic acid (CHCA) can produce significant background interference and coffee ring effect in the low molecular weight range (<1000 Da), limiting their application. To overcome these limitations, researchers have developed alternative matrices such as noble metals, metal oxides, carbon-based and silicon-based nanomaterials. Among them, metal oxides are favored due to their high light absorption rate, excellent stability, and ability to form uniform crystal structures with analytes, thereby improving ionization efficiency and sensitivity.

[0004] Metal-organic frameworks (MOFs) have been considered as ideal precursors for the preparation of nanostructured metal oxides through one-step thermal conversion due to their tunable structures. Among them, cobalt-based zeolitic imidazolate frameworks (ZIF-67) have attracted much attention due to their excellent structural tunability, rich active sites and efficient charge transport performance. Studies have shown that the introduction of nickel (Ni) and manganese (Mn) into the ZIF-67 framework can respectively enhance the electrical conductivity and redox activity of the material, and this synergistic effect can significantly promote ion / electron transfer, thereby improving the ionization efficiency. SUMMARY

[0005] The purpose of the present application is to provide a urine metabolite detection method based on a three-metal oxide nanocage matrix.

[0006] The urine metabolite detection method based on a three-metal oxide nanocage matrix proposed by the present application drops urine on a stainless steel target plate, air dries it, drops a three-metal oxide nanocage solution on the target plate containing the sample, air dries it, and performs laser desorption ionization mass spectrometry analysis; the laser desorption ionization mass spectrometry uses a 355 nm Nd:YAG laser light source, the laser frequency is 2000 Hz, and the acceleration voltage is 20 kV; the acquisition mode is the reflection positive ion mode, and the mass-to-charge ratio range for acquisition is 100-1000 Da; the mass spectrometry data is obtained from flexControl 3.4, and the data is exported in flexAnalysis 3.4, and after processing by RStudio, the urine metabolite fingerprint spectrum is obtained, and the urine metabolite fingerprint spectrum uses the Gradient Boosting machine learning algorithm; Among them, the three-metal oxide nanocage is a hollow dodecahedron three-metal oxide nanocage (TONC) (CoNiMnO) derived from a metal-organic framework derived from a cobalt-based zeolitic imidazolate framework-67 in which nickel and manganese are introduced; In the present application, the synthesis method of the three-metal oxide nanocage comprises the following steps: (1) Dissolve cobalt nitrate hexahydrate (Co(NO3)2·6H2O) in methanol to form solution A, wherein the amount ratio of cobalt nitrate hexahydrate to methanol is 1.164 g:100 mL; (2) Dissolve 2-methylimidazole in methanol to form solution B, wherein the amount ratio of 2-methylimidazole to methanol is 1.314 g:100 mL; (3) Pour solution A into solution B and stir vigorously to form solution C; (4) Age solution C at room temperature for 20-24 hours, centrifuge, wash with methanol, and dry to obtain zeolitic imidazolate framework-67; (5) The zeolitic imidazolate framework-67 is dissolved with ethanol, and then nickel chloride hexahydrate (NiCl2·6H2O) and manganese chloride tetrahydrate (MnCl2·4H2O) are added, and then ultrasonic treatment is performed, and the stirring temperature is 50 DEG C, and then centrifugal separation, washing and drying are performed to obtain a TONCs precursor; wherein the amount ratio of the zeolitic imidazolate framework-67, ethanol, nickel chloride hexahydrate and manganese chloride tetrahydrate is 0.1 g:40 mL:0.048 g:0.040 g; (6) The TONCs precursor is placed in a muffle furnace for high-temperature calcination, the calcination time is 2 hours, and the calcination temperature is 350 DEG C to obtain a hollow dodecahedron TONCs.

[0007] The second object of the application is to propose the application of the urine metabolite detection method based on the trimetal oxide nanocage in the detection of bladder cancer, prostate cancer and kidney cancer, and the MALDIquant software package is used to process the mass spectrum raw data in RStudio processing, including intensity conversion, smoothing and calibration, baseline repositioning and spectrum averaging, etc., and the mass-to-charge ratio characteristics are extracted.

[0008] In the application, the specific steps of the Gradient Boosting machine learning algorithm are as follows: (1) 120 bladder cancer samples with mass-to-charge ratio characteristics ranging from 100 to 1000, 125 prostate cancer samples with mass-to-charge ratio characteristics ranging from 100 to 1000 and 138 kidney cancer samples with mass-to-charge ratio characteristics ranging from 100 to 1000 are randomly divided into a training set and a test set in a ratio of 7:3. (2) The Gradient Boosting model is trained and optimized using the training set data, and the effect of the model is further verified using the test set data.

[0009] In the application, all mass-to-charge ratio characteristics are filtered using gain ratio, Gini index and GB importance, and the filtering conditions are gain ratio > 0.19, Gini index < 0.135 and GB importance > 0.007; 9 mass-to-charge ratio values, i.e., m / z 327.06, m / z 152.05, m / z 353.10, m / z 260.89, m / z 505.83, m / z 184.06, m / z 389.02, m / z 232.95 and m / z 126.92, are screened out as key characteristics to construct a biomarker panel.

[0010] The application has the following beneficial effects: The urine metabolite detection method provided by the application uses hollow dodecahedron trimetal oxide nanocage as a substrate, and uses LDI-MS for high-sensitivity detection of urine metabolites. The method has the following advantages. (1) The hollow dodecahedron trimetal oxide nanocage provided by the application is obtained by one-step simple ion exchange and high-temperature calcination, does not involve the use of noble metal materials and secondary modification, is simple to make, low in cost, and suitable for large-scale production and application. (2) The hollow dodecahedron trimetal oxide nanocage provided by the application has strong ultraviolet absorption, can promote energy transfer and metabolite desorption ionization, and the unique hollow dodecahedron structure further enhances the light capture ability and light-heat conversion efficiency of the material, and has excellent heat-driven ion desorption performance, can effectively protect the metabolites from excessive fragmentation caused by direct laser irradiation, and is suitable for being used as a substrate of LDI-MS. (3) The application identifies the urine metabolic differences of bladder cancer, prostate cancer and kidney cancer patients through a machine learning algorithm, thereby realizing the direct differentiation and paired discrimination of the three types of urinary system cancers, and has high accuracy and precision.

[0011] The application of the urine metabolite detection method provided by the application to the detection of bladder cancer, prostate cancer and kidney cancer uses the hollow dodecahedron trimetal oxide nanocage (TONC) derived from the metal organic framework, which exhibits strong light response and heat transfer performance due to the existence of synergistic effect in composition and structure. The urine metabolic fingerprint of the patient with urinary system cancer is successfully extracted by using the TONC-assisted LDI-MS. The method can complete the direct urine detection of metabolites in a few seconds, avoiding complex pretreatment steps. Combined with a machine learning algorithm, the method can realize rapid diagnosis and identification of cancer. The application promotes the application of substrate-assisted LDI-MS metabolomics in clinical practice. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 It is a transmission electron microscope photo of the trimetal oxide nanocage of Example 1. Figure 2 It is a scanning electron microscope photo of the trimetal oxide nanocage of Example 1. Figure 3 It is an X-ray powder diffraction spectrum of the trimetal oxide nanocage of Example 1. Figure 4 It is an ultraviolet-visible absorption spectrum of the trimetal oxide nanocage of Example 1 compared with the bimetal oxide CoNi and CoMn. Figure 5 It is a typical mass spectrum of the urine metabolites of the bladder cancer patient, the prostate cancer patient and the kidney cancer patient of Example 2. Figure 6 Confusion matrix plot for direct discrimination of bladder cancer, prostate cancer and kidney cancer for Example 3; Figure 7 Radar plot for pairwise discrimination between the three groups for Example 3, including area under the curve (AUC), accuracy, F1 score, precision and recall; Figure 8 Score scatter plot for mutual discrimination of bladder cancer and kidney cancer based on key features for Example 4; Figure 9 Score scatter plot for mutual discrimination of prostate cancer and kidney cancer based on key features for Example 4; Figure 10 Score scatter plot for mutual discrimination of bladder cancer and prostate cancer based on key features for Example 4. DETAILED DESCRIPTION

[0013] The present application utilizes trimetallic oxide nanocages as a matrix for assisting LDI-MS, realizing the detection and analysis of urine metabolites of patients with urinary system cancer.

[0014] In order to make the present application more apparent and easy to understand, the preferred embodiments are described in detail below with reference to the accompanying drawings. It should be understood by those skilled in the art that the following embodiments are only used to illustrate the present application, and are not intended to limit the present application in any form and in substance. Example 1

[0015] Preparation of metal-organic framework derived hollow dodecahedral trimetallic oxide nanocage: (1) Dissolve 1.164 g of cobalt nitrate hexahydrate in 100 mL of methanol to form solution A; (2) Dissolve 1.314 g of 2-methylimidazole in 100 mL of methanol to form solution B; (3) Pour solution A into solution B and stir thoroughly to form purple solution C; (4) Age solution C obtained in step (3) at room temperature for 20 hours, then centrifuge to obtain a solid product, wash with methanol three times, and vacuum dry at 50°C overnight; (5) Take 100 mg of the product obtained in step (4), dissolve in 40 mL of ethanol, then add 0.048 g of nickel chloride hexahydrate and 0.040 g of manganese chloride tetrahydrate, ultrasonic for 10 minutes, 50°C vigorous stirring for 30 minutes, then centrifuge, wash with ethanol three times, and vacuum dry at 50°C overnight; (6) Place the product obtained in step (5) in a muffle furnace and calcine at 350°C in air atmosphere for 2 hours, with a heating rate of 2°C / min, to obtain the final product, trimetallic oxide nanocage.

[0016] Material characterization: Transmission electron microscope image of the trimetallic oxide nanocage is shown in Figure 1 ; scanning electron microscope image of the trimetallic oxide nanocage is shown in Figure 2 ; X-ray powder diffraction pattern of the trimetallic oxide nanocage is shown in Figure 3 ; ultraviolet-visible absorption spectrum of the trimetallic oxide nanocage compared with bimetallic oxide CoNi and CoMn is shown in Figure 4 .

[0017] Analysis results: From Figure 1 and Figure 2 , it can be seen that the trimetallic oxide nanocage has a hollow dodecahedron structure, a rough surface and sharp corners, and a diameter of about 500 nm; from Figure 3 , it can be seen that the trimetallic oxide nanocage has a highly crystalline structure; from Figure 4 , it can be concluded that the trimetallic oxide nanocage has excellent ultraviolet absorption ability at the wavelength of 355 nm of the laser light source of the mass spectrometer. Example 2

[0018] The trimetallic oxide nanocage obtained in Example 1 was used as a matrix, and the urine of 120 bladder cancer samples, 125 prostate cancer samples and 138 kidney cancer samples taken from Zhongshan Hospital Affiliated to Fudan University was subjected to metabolite mass spectrometry detection.

[0019] (1) 1 mg of the trimetallic oxide nanocage material synthesized in Example 1 was uniformly dispersed in 1 mL of deionized water to obtain a trimetallic oxide nanocage suspension of 1 mg / mL.

[0020] (2) The uniformly dispersed urine sample was directly added to a stainless steel target plate, and after the sample point was dried, the trimetallic oxide nanocage suspension obtained in step (1) was added to the corresponding sample point on the stainless steel target plate, and was naturally dried.

[0021] (3) The stainless steel target plate in step (2) was directly subjected to laser desorption ionization mass spectrometry analysis, using a Bruker ultrafleXtreme MALDI-TOF / TOF mass spectrometer, using a 355 nm Nd:YAG laser light source, a laser frequency of 2000 Hz, and an acceleration voltage of 20 kV; the acquisition mode was a positive ion reflector mode, and the acquisition mass-to-charge ratio range was 100-1000 Da; the mass spectrometry data was obtained from flexControl 3.4, and the data was exported in flexAnalysis 3.4 to obtain the urine metabolite mass spectrum.

[0022] Typical mass spectra of urine metabolites of bladder cancer patients, prostate cancer patients and kidney cancer patients are shown in Figure 5shown.

[0023] Analysis results: From Figure 5 It can be observed that there are multiple distinct peaks between the three groups of samples, which demonstrates the feasibility of using TONCs assisted LDI-MS to distinguish urinary system cancers. Example 3

[0024] The urine metabolic fingerprint obtained in Example 2 was used to directly distinguish and pairwise discriminate bladder cancer, prostate cancer and kidney cancer using the Gradient Boosting algorithm.

[0025] (1) The MALDIquant software package of RStudio was used to process the mass spectrum raw data, including intensity conversion, smoothing and calibration, baseline re- moving and average value of the spectrum, etc. 452 mass-to-charge ratio features were extracted.

[0026] (2) 120 bladder cancer samples, 125 prostate cancer samples and 138 kidney cancer samples were randomly divided into training set and test set in the ratio of 7:3, among which 120 bladder cancer samples with mass-to-charge ratio features ranging from 100 to 1000, 125 prostate cancer samples with mass-to-charge ratio features ranging from 100 to 1000 and 138 kidney cancer samples with mass-to-charge ratio features ranging from 100 to 1000.

[0027] (3) The Gradient Boosting model was trained and optimized using the training set data, and the test set data was used to further verify the effect of the model.

[0028] The confusion matrix of using the Gradient Boosting model to directly distinguish bladder cancer, prostate cancer and kidney cancer in the training set is shown in Figure 6 The radar chart of pairwise discrimination between the three groups using the Gradient Boosting model is shown in Figure 7 , which includes the area under the curve (AUC), accuracy, F1 score, precision and recall.

[0029] Analysis results: From Figure 6 It can be found that the Gradient Boosting algorithm can well distinguish the three cancers with high accuracy, among which the discrimination effect of kidney cancer from the other two cancers is the best. Figure 7 It can be concluded that the pairwise discrimination of the training set has achieved excellent results, especially the mutual discrimination between bladder cancer and kidney cancer. Figure 6 and Figure 7 It is proved that the Gradient Boosting machine learning algorithm has high reliability and great application potential in both direct discrimination and pairwise discrimination of the three groups. Example 4

[0030] Based on the 452 mass-to-charge ratio characteristics obtained from Example 3, key characteristics were screened as biomarkers to construct a biomarker panel for the diagnosis and identification of urinary system cancer.

[0031] (1) The important indicators closely related to the Gradient Boosting algorithm, i.e., gain ratio, Gini index, and GB importance, were used for filtering on all mass-to-charge ratio characteristics. The filtering conditions were gain ratio > 0.19, Gini index < 0.135, and GB importance > 0.007.

[0032] (2) Comparison with the human metabolome database screened 9 key characteristics, and their mass-to-charge ratio values were m / z 327.06, m / z 152.05, m / z 353.10, m / z 260.89, m / z 505.83, m / z 184.06, m / z 389.02, m / z 232.95, and m / z 126.92.

[0033] (3) Based on the 9 key characteristics screened in step (2), a biomarker panel was constructed, a Gradient Boosting model was trained using the training set data, and the results of the test set data were predicted using this model.

[0034] The score scatter plot of the mutual differentiation of bladder cancer and kidney cancer based on the key characteristics is shown in FIG. 2. Figure 8 The score scatter plot of the mutual differentiation of prostate cancer and kidney cancer is shown in FIG. 3. Figure 9 The score scatter plot of the mutual differentiation of bladder cancer and prostate cancer is shown in FIG. 4. Figure 10

[0035] The analysis results are shown in FIG. 5. Figures 8 to 10 It can be seen that the biomarker panel constructed based on the key characteristics can effectively distinguish bladder cancer, prostate cancer, and kidney cancer, has high accuracy and reliability, and further confirms the feasibility of the advanced metabolic analysis method based on TONCs in clinical diagnosis.

[0036] The above examples are only preferred embodiments of the present application. It should be noted that any modifications, equivalent replacements, or improvements made by those skilled in the art without departing from the design principles of the present application shall be included in the protection scope of the present application.​

Claims

1. A method for detecting urine metabolites using trimetallic oxide nanocages as a matrix, characterized in that: The trimetallic oxide nanocage solution was dropped onto a target plate with air-dried urine samples, air-dried, and then subjected to laser desorption ionization mass spectrometry analysis; Laser desorption ionization mass spectrometry was performed using a 355 nm Nd:YAG laser source with a laser frequency of 2000 Hz and an accelerating voltage of 20 kV. The acquisition mode was reflectron cation mode, with a mass-to-charge ratio range of 100–1000 Da. Mass spectrometric data were acquired using flexControl 3.4 and exported in flexAnalysis 3.

4. After processing in RStudio, urine metabolite fingerprints were generated using the Gradient Boosting machine learning algorithm. The trimetallic oxide nanocage is a hollow dodecahedral trimetallic oxide nanocage derived from a metal organic framework in which nickel and manganese are introduced into a zeolite imidazolate framework material-67.

2. The method for detecting urine metabolites based on trimetallic oxide nanocages according to claim 1, wherein: The synthesis method of the trimetallic oxide nanocage comprises the following steps: (1) Dissolve cobalt nitrate hexahydrate in methanol to form solution A, wherein the ratio of cobalt nitrate hexahydrate to methanol is 1.164 g:100 mL; (2) Dissolve 2-methylimidazole in methanol to form solution B, wherein the ratio of 2-methylimidazole to methanol is 1.314 g:100 mL; (3) Pour solution A into solution B and stir vigorously to form solution C; (4) Solution C was aged at room temperature for 20-24 hours, centrifuged, washed with methanol, and dried to obtain zeolite imidazolate framework material-67; (5) Dissolve the zeolite imidazolate framework material-67 in ethanol, then add nickel chloride hexahydrate and manganese chloride tetrahydrate, ultrasonicate, fully stir at a stirring temperature of 50°C, centrifuge, wash, and dry to obtain a trimetallic oxide nanocage precursor; wherein the amount ratio of zeolite imidazolate framework material-67, ethanol, nickel chloride hexahydrate and manganese chloride tetrahydrate is 0.1 g:40 mL:0.048 g:0.040 g; (6) The trimetallic oxide nanocage precursor is placed in a muffle furnace and calcined at a high temperature for 2 hours at a temperature of 350° C., thereby obtaining the trimetallic oxide nanocage.

3. An application of the urine metabolite detection method based on trimetallic oxide nanocages as claimed in claim 1 in the detection of bladder cancer, prostate cancer and kidney cancer, characterized in that: The MALDIquant software package was used in RStudio to process the raw mass spectrometry data, including intensity conversion, smoothing and calibration, baseline re-shifting, and spectral averaging to extract the mass-to-charge ratio features.

4. The use of the urine metabolite detection method based on trimetallic oxide nanocages as a matrix in the detection of bladder cancer, prostate cancer and kidney cancer according to claim 3, characterized in that: The specific steps of the Gradient Boosting machine learning algorithm are as follows: (1) 120 bladder cancer samples with mass-to-charge ratios ranging from 100 to 1000, 125 prostate cancer samples with mass-to-charge ratios ranging from 100 to 1000, and 138 kidney cancer samples with mass-to-charge ratios ranging from 100 to 1000 were randomly divided into training and test sets in a ratio of 7:

3. (2) Use the training set data to train and optimize the Gradient Boosting model, and use the test set data to further verify the effect of the model.

5. The use of the urine metabolite detection method based on trimetallic oxide nanocages as a matrix in the detection of bladder cancer, prostate cancer and kidney cancer according to claim 3, characterized in that: All mass-to-charge ratio features were filtered using gain ratio, Giniindex, and GB importance, with the filtering conditions being gain ratio > 0.19, Gini index <0.135, and GB importance >0.

007. Nine key features with mass-to-charge ratio values ​​of m / z 327.06, m / z 152.05, m / z 353.10, m / z 260.89, m / z 505.83, m / z 184.06, m / z 389.02, m / z 232.95, and m / z 126.92 were selected to construct a biomarker panel.

Images