A method to accurately detect differential metabolites in brown adipose tissue exposed to arsenic

Abstract

The present invention discloses a method for accurately detecting differential metabolites of arsenic-exposed brown adipose tissue, comprising the following steps: S1: extracting arsenic-exposed brown fat metabolites; S2: performing non-targeted metabolomics and non-targeted lipidomics analysis on metabolite samples; S3: performing multivariate analysis on raw data and searching a database; S4: constructing a differential metabolite ion pair database for differential metabolites; S5: performing quasi-targeted metabolomics and quasi-targeted lipidomics detection to obtain secondary spectra and MRM information; S6: using SCIEX OS software for peak extraction, performing multivariate statistical analysis, and screening to obtain accurate differential metabolites. The method of the present invention effectively reduces the occurrence of "false positives" in traditional non-targeted metabolomics and lipidomics, and has high accuracy. Compared with other quasi-targeted methods, it is more efficient, greatly shortens the detection time, and can obtain clear secondary spectra.

Description

Technical Field

[0001] The present invention relates to the field of biological analysis, and in particular to a method for accurately detecting differential metabolites of brown adipose tissue exposed to arsenic. Background Art

[0002] Arsenic, as a metalloid, is widely present in soil, water, and crops. Epidemiological studies have shown that arsenic exposure is associated with a variety of diseases such as intellectual and cognitive impairment, mammalian diabetes, obesity, kidney cancer, and cardiovascular disease. Brown fat can directly convert chemical energy into heat energy using glucose and free fatty acids in peripheral tissues, improving the body's glucose and lipid metabolism, and has been identified as one of the most promising therapeutic targets for preventing diet-induced obesity and diabetes. Therefore, it is very important to use metabolomics and lipidomics to understand the changes in brown adipose tissue metabolites after arsenic exposure, especially the mechanisms of differential metabolite changes.

[0003] Metabolomics is an emerging discipline that systematically studies metabolites and metabolic pathways in organisms. Metabolites reflect both the downstream output of genes and the effects of the environment on organisms. Genomics and proteomics can predict changes that will occur in organisms, while metabolomics, as the terminal of systems biology, reflects changes that have already occurred in organisms. Metabolomics has a wide range of applications in disease diagnosis, etiology research, exploration of new drug targets, precision medicine, and treatment monitoring. It can profile the overall and changes of metabolites in a specific organism, and use multivariate statistical methods to understand the level changes of overall metabolism, screen and identify differential metabolites with significant effects. Metabolomics can also explore the mechanisms of disease occurrence and drug treatment through metabolic pathway analysis. Lipidomics is a branch of metabolomics that conducts comprehensive and systematic analysis and identification of lipid metabolites and molecules that interact with them in organisms, tissues or cells, understands the structure and function of lipids, and then reveals the relationship between lipid metabolism and the physiological and pathological processes of cells, organs and even the body.

[0004] Common detection methods for metabolomics and lipidomics include targeted detection, non-targeted detection, and quasi-targeted detection. Non-targeted metabolomics and lipidomics can perform relatively comprehensive detection of metabolites. In the identification process, they mainly rely on public databases to compare the detection results. Their identification results are prone to "false positives" for the following reasons: (1) Compared with the data in the public database, the experimental mass spectrometry data may be collected under different experimental conditions; (2) A large part of the data in some public databases is not experimental, but predicted using computer tools. When processing data, the default parameters of the software will also compare these predicted data; (3) Metabolites of different species or even the entire known chemical space are used to identify metabolites of species of interest; (4) Because biological samples are full of chemicals from food, drugs and their metabolites, intestinal microbial flora, environmental pollutants, etc., the analysis of metabolites in these samples becomes complicated; (5) The recognition ability is weak for substances with low abundance.

[0005] Targeted metabolomics and lipidomics mainly use triple quadrupole mass spectrometry to detect specific metabolites, and generally use standards to obtain metabolite information for subsequent detection. It has high sensitivity and good repeatability, but it is difficult to achieve full signal coverage, low efficiency, and long time consumption.

[0006] Quasi-targeted metabolomics was proposed by Professor Xu Guowang's team at the Dalian Institute of Chemical Physics in 2012. This method first uses high-resolution mass spectrometry to comprehensively obtain the primary and secondary mass spectrometric information of metabolites, selects characteristic ion pairs from the secondary mass spectrometry, and then uses MRM mode (LC-MS / MS) or SIM mode (GC-MS) for targeted acquisition. Compared with targeted analysis, the quasi-targeted method can include both qualitative and unqualified metabolites by collecting the secondary mass spectrometric metabolic characteristics of mixed biological samples (QC), achieving a wide coverage. Compared with non-targeted analysis, the MRM mode has high data stability and a wide linear range, and can obtain high-quality data information, which can focus on more low-abundance metabolites. At present, quasi-targeted metabolomics has been widely used in the accurate identification of metabolites in samples such as drugs, food, urine and blood. However, due to the large number of isomers and the difficulty in fine structure identification of lipids, it is relatively complicated to apply quasi-targeting to lipidomics. In addition, common quasi-targeted metabolomics methods often require verification of all secondary mass spectrometric information obtained by high-resolution mass spectrometry, which is labor-intensive and time-consuming. Therefore, there is an urgent need for a method that can remove "false positive" metabolites, improve the accuracy of metabolite structure identification and have a fast detection speed, so as to improve the accuracy of identifying differential metabolites in arsenic-exposed brown adipose tissue. Summary of the invention

[0007] In view of the defects in the prior art, the present invention proposes a method for accurately screening differential metabolites of arsenic-exposed brown adipose tissue. The quasi-targeted metabolomics and lipidomics methods proposed in the present invention first comprehensively obtain the primary and secondary mass spectrometry information of metabolites through ultra-high performance liquid chromatography-electrostatic field orbital trap mass spectrometry (UHPLC-Q-Exactive Orbitrap MS), screen differential ion pairs through statistical methods, and then use ultra-high performance liquid chromatography-triple quadrupole / linear ion trap mass spectrometry (UHPLC-QTRAP-MS / MS) Multiple Reaction Monitoring-Information Dependence-Enhanced Mass Spectrometry Scanning (MRM-IDA-EPI) mode to perform targeted verification of differential ion pairs. This method greatly improves the accuracy of metabolite identification, and has a wide coverage, short time consumption and high efficiency. Compared with the traditional quasi-targeted method, this method first performs statistical analysis on the ion pair information obtained by high-resolution mass spectrometry analysis, screens out differential ion pairs, and then uses the MRM-IDA-EPI mode of UHPLC-QTRAP-MS / MS for verification. This can reduce the number of ion pairs to increase the detection speed, and at the same time obtain clearer secondary spectra of metabolites.

[0008] Accurate screening of differential metabolites in brown fat after arsenic exposure will help to gain a deeper understanding of the impact of arsenic toxicity on the body's metabolism, laying the foundation for the formulation of subsequent prevention and intervention measures and the development of potential biomarker detection kits.

[0009] The present invention provides a method for accurately detecting differential metabolites of brown adipose tissue exposed to arsenic, comprising the following steps:

[0010] S1: The Matyash method was used to extract arsenic-exposed brown fat metabolites to obtain metabolite samples;

[0011] S2: Perform non-targeted metabolomics and non-targeted lipidomics analysis on the metabolite samples described in S1 using ultra-high performance liquid chromatography-electrostatic field orbitrap mass spectrometry to obtain raw data;

[0012] S3: The original data described in S2 were searched in the database and multivariate analysis was performed to screen out the differential metabolites of arsenic-exposed brown adipose tissue;

[0013] S4: Use MRM-Ion Pair Finder to construct a differential metabolite ion pair database for the differential metabolites described in S3;

[0014] S5: Ultra-high performance liquid chromatography-quadrupole / linear ion trap mass spectrometry was used to perform pseudo-targeted lipidomics and pseudo-targeted metabolomics detection on the differential metabolite ions described in S4, and multiple reaction monitoring-information-dependent-enhanced mass spectrometry scanning mode (MRM-IDA-EPI) was used in positive and negative ion modes to obtain secondary spectra and MRM relative quantitative information;

[0015] S6: Analyze the secondary spectra obtained in S5, use SCIEX OS software for peak extraction, obtain relative quantitative data, perform multivariate statistical analysis, and screen for accurate differential metabolites.

[0016] Furthermore, the accurate differential metabolites described in S6 include low-polarity differential metabolites obtained by the quasi-targeted lipidomics analysis and high-polarity differential metabolites obtained by the quasi-targeted metabolomics analysis.

[0017] Furthermore, the low polarity difference metabolites are shown in Table 1:

[0018] Table 1

[0019]

[0020]

[0021]

[0022] The high polarity differential metabolites and mass spectrum information are shown in Table 2 below:

[0023] Table 2

[0024]

[0025] Furthermore, the conditions for collecting non-targeted metabolomics data using ultra-high performance liquid chromatography include:

[0026] In the positive ion mode, mobile phase A was 0.1% formic acid in water, and mobile phase B was acetonitrile solution containing 0.1% formic acid; in the negative ion mode, mobile phase A was 0.5 mM ammonium fluoride solution, and mobile phase B was acetonitrile;

[0027] The flow rate was 0.4 mL / min.

[0028] The gradient of mobile phase B is 0-1.0 min, 1%; 1.0-8.0 min, 1%-99%; 8.0-10.0 min, 99%; 10.0-10.1 min, 99%-1%, 10.1-12.0 min, 1%.

[0029] Furthermore, the conditions for collecting non-targeted lipidomics analysis data using ultra-high performance liquid chromatography include:

[0030] In the positive and negative ion modes, mobile phase A is a 60%-40% solution of acetonitrile-water containing 10 mM ammonium formate; mobile phase B is a 10%-90% solution of acetonitrile-isopropanol containing 10 mM ammonium formate;

[0031] The flow rate was 0.6 mL / min.

[0032] The gradient of mobile phase B is 40% to 100% from 0 to 9.0 min; 100% from 10.0 min; 100% to 40% from 10.0 to 10.2 min; and 40% from 10.2 to 13 min.

[0033] Furthermore, the data acquisition conditions of ultra-high performance liquid chromatography for targeted metabolomics analysis include:

[0034] In the positive ion mode, mobile phase A was 0.1% formic acid in water, and mobile phase B was acetonitrile solution containing 0.1% formic acid; in the negative ion mode, mobile phase A was 0.5 mM ammonium fluoride solution, and mobile phase B was acetonitrile;

[0035] The flow rate was 0.4 mL / min.

[0036] The gradient of mobile phase B is 0-1.0 min, 1%; 1.0-8.0 min, 1%-99%; 8.0-10.0 min, 99%; 10.0-10.1 min, 99%-1%, 10.1-12.0 min, 1%.

[0037] Furthermore, the data acquisition conditions of ultra-high performance liquid chromatography for targeted lipidomics analysis include:

[0038] In the positive and negative ion modes, mobile phase A is a 60%-40% solution of acetonitrile-water containing 10 mM ammonium formate; mobile phase B is a 10%-90% solution of acetonitrile-isopropanol containing 10 mM ammonium formate;

[0039] The flow rate was 0.4 mL / min.

[0040] The gradient of mobile phase B is 40% to 60% from 0 to 9.0 min; 100% from 10.0 min; 60% to 0% from 10.0 to 10.2 min; and 40% from 10.2 to 13 min.

[0041] Furthermore, the electrostatic field orbital trap mass spectrometry parameter conditions in step S2 include: spray voltage, 3600 V; sheath gas flow rate, 60 arb; auxiliary gas flow rate, 25 arb; purge gas flow rate, 2 arb; capillary temperature, 300°C; auxiliary gas heater temperature, 370°C; wherein the floating range of each numerical parameter involved in the above conditions is ±10% of the numerical value.

[0042] Furthermore, in step S3, multivariate statistical analysis is used to convert the raw data into ABF format for deconvolution, peak extraction, registration and compound identification.

[0043] Furthermore, the multivariate analysis in step S3 includes orthogonal partial least squares discriminant analysis.

[0044] Furthermore, the orthogonal partial least squares discriminant analysis screening criteria for differential metabolites in step S3 are: VIP>1, P<0.05, FC≤0.67 or FC≥1.5, and differential metabolites with a relative standard deviation of >30% in the quality control sample are eliminated, and differential metabolites with M / Z>1000Da are not considered.

[0045] Furthermore, in step S5, in the multiple reaction monitoring-information-dependent-enhanced mass spectrometry scanning mode, the mass spectrometry conditions include: data acquisition range, 50Da~1000Da; nebulizer gas, 50Psi; auxiliary heating gas metabolomics analysis is 60Psi, lipidomics analysis is 50Psi; curtain gas, 40Psi; ionization voltage, 4500V; wherein the floating range of each numerical parameter involved in the above conditions is ±10% of the numerical value.

[0046] In summary, compared with the prior art, the present invention achieves the following technical effects:

[0047] 1. The present invention statistically analyzes the ion pairs obtained from non-targeted omics analysis, screens out differential ion pairs, and then uses ultra-high performance liquid chromatography-quadrupole / linear ion trap mass spectrometry and multiple reaction monitoring-information-dependent-enhanced mass spectrometry scanning mode for verification, which greatly shortens the detection time and reduces the "false positive" situation of traditional non-targeted metabolomics and lipidomics, and improves the accuracy of identification, thereby achieving the purpose of accurately screening differential metabolites of brown adipose tissue after arsenic exposure.

[0048] 2. The present invention uses multiple reaction monitoring-information-dependent-enhanced mass spectrometry scanning mode to perform quasi-targeted metabolomics and lipidomics. Compared with the traditional multiple reaction monitoring mode quasi-targeted technology, it can overcome background interference, improve the signal-to-noise ratio, and still achieve high sensitivity for complex samples, with good quantitative effect. The obtained fragment information peaks are clear, and the structures of differential metabolites can be more accurately identified. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 This is an operation flow chart of Example 1 of the present invention;

[0051] Figure 2It is the non-targeted metabolomics and non-targeted lipidomics orthogonal partial least squares discriminant analysis (OPLS-DA) score graph, permutation test graph, and volcano graph of Example 1 of the present invention; wherein, Figure 2 A is the non-targeted metabolomics OPLS-DA score graph; Figure 2 B is the permutation test result of OPLS-DA analysis, R 2 Y=0.92、Q 2 =0.74 indicates that the model has good separation performance and predictability; Figure 2 C is the metabolomics volcano plot; Figure 2 D is the non-targeted lipidomics OPLS-DA score graph; Figure 2 E is the permutation test result of non-targeted lipidomics OPLS-DA analysis, R 2 Y=0.871、Q 2 =0.516 indicates that the model has good separation performance and predictability; Figure 2 F is the volcano plot of non-targeted lipidomics results;

[0052] Figure 3 A Sankey diagram of the results of multivariate statistical analysis of non-targeted metabolomics and non-targeted lipidomics in Example 1 of the present invention; Figure 3 A is the Sankey diagram of the multivariate statistical analysis results of non-targeted metabolomics; Figure 3 B Sankey diagram of multivariate statistical analysis results of non-targeted lipidomics;

[0053] Figure 4 This is a metabolomics differential metabolite heat map after verification by quasi-targeted metabolomics and lipidomics in Example 1 of the present invention; wherein, Figure 4 A is the result of pseudo-targeted metabolomics; Figure 4 B is the result of pseudo-targeted lipidomics;

[0054] Figure 5 This is the secondary spectrum analysis of the differential substance PE 40:4 obtained by the targeted MRM-IDA-EPI mode in Example 1 of the present invention;

[0055] Figure 6 The differential lipid metabolite pathway analysis and potential biomarker analysis after targeted lipidomics verification in Example 1 of the present invention; Figure 6 A is the pathway analysis result of accurate differential metabolites obtained by targeted lipidomics; Figure 6 B is the result of potential biomarker analysis of accurate differential metabolites obtained by targeted lipidomics. DETAILED DESCRIPTION

[0056] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only embodiments of a part of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0057] The present invention divides the arsenic-exposed mouse interscapular brown adipose tissue samples into two batches, one for non-targeted metabolomics and lipidomics analysis, and the other for quasi-targeted metabolomics and lipidomics analysis. One batch of samples is subjected to non-targeted metabolomics and lipidomics analysis by electrostatic field orbital trap mass spectrometry, metabolites are screened by multivariate statistics, ion pairs of the screened metabolites are extracted, another batch of samples is subjected to quasi-targeted metabolomics and lipidomics analysis based on the extracted ion pairs, further screening is performed by quasi-targeted metabolomics and lipidomics, and finally differential metabolites are generated and pathway and potential biomarker analysis is performed. The flow chart is as follows: Figure 1 shown.

[0058] The experimental methods used in the following examples are conventional methods unless otherwise specified. The materials and reagents used are all commercially available unless otherwise specified.

[0059] Example 1

[0060] (1) Sample extraction

[0061] Fourteen specific pathogen-free C57BL / 6 male mice were randomly divided into a normal group and an arsenic-exposed group. The feed of the normal group was prepared according to the formula of GB 14924.3-2010, and the feed of the arsenic-exposed group was prepared by simulating arsenic in rice. The specific ratios of iAsV (arsenic acid), iAsIII (arsenic acid), MMA (monomethylarsenic acid) and DMA (dimethylarsenic acid) were 7.3%, 72.7%, 1.0% and 19.0%, respectively. The interscapular brown adipose tissue of mice was collected after continuous exposure for 14 months. The metabolites of brown adipose tissue samples of the normal group and the arsenic-exposed group were extracted using the Matyash method. Methanol and methyl tert-butyl ether (MTBE) were precooled at -20℃. 50 mg of brown adipose tissue was placed in 225 μL of precooled methanol and homogenized in a high-throughput grinder for 2 minutes at a speed of 1350 rpm. Then 750 μL of precooled methyl tert-butyl ether was added, and the mixture was vortexed for 10 seconds and shaken at 4℃ for 5 minutes. Then add 188 μL of room temperature water, vortex for 20 seconds to induce phase separation, centrifuge at 14000g for 2 minutes, take two 350 μL portions of the upper low-polarity phase for lipidomics, two 125 μL portions of the lower polar phase for metabolomics, and combine the rest to form quality control samples, which are blown dry with nitrogen. Low-polar substances are reconstituted with 110 μL of a 9:1 mixture of methanol: toluene, and high-polar substances are reconstituted with 110 μL of acetonitrile. Quality control samples are injected after several biological samples in each group. (2) Non-targeted metabolomics and lipidomics detection

[0062] 3 μL of the resuspended sample was injected into a Vanquish UHPLC system (Thermo Scientific, Waltham, MA, USA) equipped with a Waters Acquity UPLC HSS T3 column (1.8 μm, 2.1 × 100 mm) and a Waters Acquity VanGuard CSH C18 precolumn (5 mm × 2.1 mm; 1.7 μm). The Thermo Scientific QExactive Focus benchtop LC-MS / MS combines quadrupole precursor ion selection and high-resolution accurate mass (HR / AM) Orbitrap mass analyzer, and the spectra were collected using the Data-Dependent Acquisition (DDA) method.

[0063] UPLC parameters: column temperature and flow rate were set at 65 °C and 0.4 mL / min, respectively.

[0064] In the positive ion mode of metabolomics, the mobile phase A was water containing 0.1% formic acid, and B was acetonitrile containing 0.1% formic acid. In the negative ion mode, the mobile phase A was water containing 0.5 mM ammonium fluoride, and B was acetonitrile. The flow rate was 0.4 mL / min, and the B phase was 1% from 0 to 1 min, 1%-99% from 1 to 8 min, 99% from 8 to 10 min, 99% from 10.0 to 10.1 min, and 1% from 10.1 to 12 min.

[0065] The lipidomics flow rate was set at 0.4 mL / min, and the mobile phases in both positive and negative modes were: mobile phase A contained 10 mM ammonium formate in acetonitrile-water 60%-40% solution; mobile phase B was 10 mM ammonium formate in acetonitrile-isopropanol 10%-90% solution. Phase B was 40%-100% from 0 to 9.0 min, 100% at 10.0 min, 100% from 10.0 to 10.2 min, 100%-40% from 10.2 to 13 min, 40%.

[0066] MS parameters: spray voltage (IS), 3.6 kV; sheath gas flow rate, 60 arb; auxiliary gas flow rate, 25 arb; sweep gas flow rate, 2 arb; capillary temperature, 300 °C; auxiliary gas heater temperature, 370 °C. The following acquisition parameters were used for MS1 analysis: resolution, 60000, automatic gain control target, 1e6; scan range 150-1700 m / z; spectral data type, center type. Data-dependent MS / MS parameters: resolution, 15000; automatic gain control target, 1e5; cycle count, 3; TopN, 3; isolation window, 1.0 m / z; fixed first mass, 70.0 m / z; NCE / stepped nce, 20, 30, 40; spectral data type, center point; minimum automatic gain control target, 8e3; response value threshold, 1.6e5; isotope exclusion, on; dynamic exclusion, 3.0 s.

[0067] (3) Multivariate statistical analysis

[0068] All mass spectrometry raw data files were converted to ABF format using ABF converter (https: / / www.reifycs.com / AbfConverter / ). Deconvolution, peak extraction, alignment, and compound identification were performed using MS-DIALver.4.00 software.

[0069] The detailed parameter settings are as follows: MS1 tolerance, 0.005Da; MS2 tolerance, 0.01Da; minimum peak height, 20000 amplitude; mass slice width, 0.1Da; smoothing method, linear weighted moving average; smoothing level, 5 scans; minimum peak width, 10 scans. In positive ion mode metabolomics and lipidomics analysis, [M+H] +, [M+NH4] + , [M+Na] + , [2M+H] + , [2M+NH4] + , [2M+Na] + was included in the adduct ion set, [MH] - , [M+Cl] - , [M+FA-H] - , [2M-H] - For negative ion metabolomics, [MH] - , [M+Cl] - , [M+Hac-H] - For negative ion mode lipidomics. Compounds were annotated using accurate precursor ion masses and MS / MS spectra against the MassBank database.

[0070] Statistical analysis was performed by median, logarithmic transformation, and Pareto normalization. OPLS-DA, T-test, and calculation of fold difference (FC) were performed using R.

[0071] Screening of differential metabolites The differential metabolites were screened according to the OPLS-DA variable projection importance VIP>1, combined with T test P<0.05, and the difference fold FC≤0.67 or FC≥1.5, while the differential metabolites with MZ>1000Da and RSD>30% in QC samples were not considered, as shown in Tables 1 and 2.

[0072] Table 1 Non-targeted lipidomics multivariate statistical screening of metabolite ion pairs

[0073]

[0074]

[0075]

[0076]

[0077] Table 2 Non-targeted metabolomics multivariate statistical screening of metabolite ion pairs

[0078]

[0079] (4) Ion pair extraction

[0080] The R package MRM-Ion Pair Finder and manual verification were used to construct a database of differential substance ion pairs for the screened differential metabolites. The ion pairs were finally determined by the dominant precursor ions and the characteristic product ions with the highest intensity and relative reference. Based on the results of non-targeted metabolism and lipidomics, a pseudo-targeted omics method of UHPLC-QTrap-MS / MS was established, and the MRM-IDA-EPI scanning mode was used in positive and negative ion modes to simultaneously obtain MRM relative quantitative information and secondary spectra.

[0081] (5) Establishment and testing of targeted metabolomics and lipidomics methods

[0082] 3 μL of the resuspended sample was injected into the Shimadzu Nexera UHPLC LC-30A connected to the AB SCIEX QTRAP 5500 system, equipped with an IonDrive Turbo V ion source and an ACQUITY UPLC BE H C18 column (1.7 μm, 2.1×100 mm). The metabolite ion pair data screened from the non-targeted metabolomics and lipidomics data were input into the software in positive and negative ion modes and detected using the MRM-IDA-EPI mode to obtain the pseudo-targeted metabolomics and lipidomics data. The parameters for pseudo-targeted metabolomics and pseudo-targeted lipidomics are as follows.

[0083] UPLC parameters: column temperature and flow rate were set at 65 °C and 0.4 mL / min, respectively.

[0084] In the positive ion mode of metabolomics, the mobile phase A was water containing 0.1% formic acid, and B was acetonitrile containing 0.1% formic acid. In the negative ion mode, the mobile phase A was water containing 0.5 mM ammonium fluoride, and B was acetonitrile. The flow rate was 0.4 mL / min, and the B phase was 1% from 0 to 1 min, 1%-99% from 1 to 8 min, 99% from 8 to 10 min, 99% from 10.0 to 10.1 min, and 1% from 10.1 to 12 min.

[0085] The mobile phases in both positive and negative modes of lipidomics were: mobile phase A was a 60%-40% solution of acetonitrile-water containing 10 mM ammonium formate; mobile phase B was a 10%-90% solution of acetonitrile-isopropanol containing 10 mM ammonium formate, with phase B at 40%-100% from 0 to 9.0 min, 100% at 10.0 min, 100% from 10.0 to 10.2 min, and 40% from 10.2 to 13 min.

[0086] Mass spectrometry parameters: electrospray ionization source (ESI) was used, and the data acquisition range was m / z 50Da~1000Da; nebulizer gas (GS1), 50Psi; auxiliary gas (GS2) 60Psi (lipidomics negative ion mode 50Psi); curtain gas (CUR), 40Psi; ionization voltage (IS), 4500V; nebulizer temperature (TEM), 500℃; collision gas (CAD), High; declustering voltage (DP), 80V; collision cell exit potential (CXP), 13V; EPI scan 50~950m / z; information dependent acquisition (IDA) trigger threshold, 500; collision energy spread (CES), 20V.

[0087] SCIEX OS software was used for peak extraction to exclude metabolites with signal-to-noise ratio <3, peak value <10,000, and RSD >30% in QC samples. The remaining metabolites were the accurate differential metabolites after targeted validation, as shown in Tables 3 and 4.

[0088] (6) Results Analysis

[0089] Non-targeted metabolomics test results Figure 2 As shown, Figure 2 A is the result of non-targeted metabolomics test. The control group in group A was separated from the experimental group in group B. The permutation test result of OPLS-DA analysis was R 2 Y is 0.837, Q 2 is 0.609( Figure 2 B) shows that the model is reliable; Figure 2 D is the non-targeted lipidomics OPLS-DA score map, the control group and the experimental group are obviously separated, R 2 Y is 0.684, Q 2 is 0.487( Figure 2 E) indicates that the model is reliable. Figure 2 C is the volcano plot of non-targeted metabolomics results, Figure 2 F is the volcano plot of the non-targeted lipidomics results, showing metabolites with P < 0.05, fold difference FC ≤ 0.67 or FC ≥ 1.5. Figure 3 A is a non-targeted metabolomics, in which 23 metabolites were obtained through multivariate statistical analysis, of which 14 were lipid small molecules. Figure 3 B Non-targeted lipidomics obtained 116 lipid metabolites through multivariate statistical analysis, including 66 glycerophospholipids.

[0090] Figure 4 A is the result of the quasi-targeted metabolomics. The heat map of accurate high-polarity differential metabolites shows the changes of 9 accurate differential metabolites. Figure 4 B is the result of targeted lipidomics. The accurate low-polarity differential metabolite heat map shows the changes in the 64 accurate differential metabolites screened out. Figure 5PE 40:4 was screened out by non-targeted lipidomics. Due to the non-targeted characteristics, the secondary spectrum was very complex and difficult to identify. A clear secondary spectrum was obtained by pseudo-targeted MRM-IDA-EPI mode, and the branch chain information was further analyzed to obtain PE (22:2 (13Z, 16Z) / 18:2 (9Z, 12Z)). Figure 6 A is the pathway analysis of accurate differential metabolites obtained by targeted lipidomics. The five most important pathways are dopaminergic synapses, actin cytoskeleton regulation, sphingolipid signaling pathways, retrograde endocannabinoid signaling, and circadian rhythm regulation, indicating that arsenic exposure may not only cause specific metabolic changes, but also have a wide range of effects on multiple biological systems and functions, causing or exacerbating a variety of health problems, providing an in-depth perspective for understanding the toxicological mechanisms of arsenic. Figure 6 B and is the potential biomarker analysis of the accurate differential metabolites obtained by targeted lipidomics. The metabolites with ROC analysis AUC>0.9 include NAE 21:1, FA 20:5, DG 36:5 and LPE 20:5, indicating that these four metabolites have high accuracy, good sensitivity and specificity as potential biomarkers of arsenic-exposed brown adipose tissue.

[0091] Table 3 Metabolite ion pairs screened by targeted lipidomics

[0092]

[0093]

[0094]

[0095] Table 4 Metabolite ion pairs for targeted metabolomics screening

[0096]

[0097] In summary, the present invention first uses non-targeted metabolomics and lipidomics technology to collect non-targeted data, analyzes the data to screen differential metabolites, then extracts metabolite ion pairs, and further screens the extracted metabolite ion pairs through targeted MRM-IDA-EPI mode. There are 23 high-polarity differential metabolites detected by non-targeted metabolomics, and 9 more after quasi-targeted detection. There are 116 low-polarity differential metabolites screened by non-targeted lipidomics, and 64 more after quasi-targeted detection, which reduces the false positive phenomenon of traditional non-targeted metabolomics and lipidomics, and improves the accuracy of detecting differential metabolism of brown fat.

[0098] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for accurately detecting differential metabolites in brown adipose tissue exposed to arsenic, characterized in that: The steps include: S1: The Matyash method was used to extract arsenic-exposed brown fat metabolites to obtain metabolite samples; S2: Perform non-targeted metabolomics and non-targeted lipidomics analysis on the metabolite samples described in S1 using ultra-high performance liquid chromatography-electrostatic field orbitrap mass spectrometry to obtain raw data; S3: The original data described in S2 were searched in the database and multivariate analysis was performed to screen out the differential metabolites of arsenic-exposed brown adipose tissue; S4: Use MRM-Ion Pair Finder to construct a differential metabolite ion pair database for the differential metabolites described in S3; S5: Ultra-high performance liquid chromatography-quadrupole-linear ion trap mass spectrometry was used to perform quasi-targeted lipidomics and quasi-targeted metabolomics detection on the differential metabolite ions described in S4, and the MRM-IDA-EPI mode was used in the positive and negative ion modes to obtain the secondary spectra and MRM relative quantitative information; S6: Analyze the secondary spectra obtained in S5, use SCIEX OS software for peak extraction, obtain relative quantitative data, perform multivariate statistical analysis, and screen for accurate differential metabolites; The ultra-high performance liquid chromatography data acquisition conditions for the targeted metabolomics analysis include: In the positive ion mode, mobile phase A was 0.1% formic acid in water, and mobile phase B was acetonitrile solution containing 0.1% formic acid; in the negative ion mode, mobile phase A was 0.5 mM ammonium fluoride solution, and mobile phase B was acetonitrile; The flow rate was 0.4 mL / min. The gradient of mobile phase B is 0-1.0min, 1%; 1.0-8.0min, 1%-99%; 8.0-10.0min, 99%; 10.0-10.1min, 99%-1%, 10.1-12.0min, 1%; The ultra-high performance liquid chromatography data acquisition conditions for the targeted lipidomics analysis include: In the positive and negative ion modes, mobile phase A is a 60%-40% solution of acetonitrile-water containing 10 mM ammonium formate; mobile phase B is a 10%-90% solution of acetonitrile-isopropanol containing 10 mM ammonium formate; The flow rate was 0.4 mL / min. The gradient of mobile phase B is 0-9.0min, 40%-100%; 10.0min, 100%; 10.0-10.2min, 100%-40%; 10.2-13min, 40%; The electrostatic field orbital trap mass spectrometry parameter conditions in step S2 include: spray voltage, 3600V; sheath gas flow rate, 60arb; auxiliary gas flow rate, 25arb; purge gas flow rate, 2arb; capillary temperature, 300°C; auxiliary gas heater temperature, 370°C; wherein the floating range of each numerical parameter is ±10% of the numerical value; The multivariate analysis in step S3 includes orthogonal partial least squares discriminant analysis; The orthogonal partial least squares discriminant analysis screening conditions for differential metabolites are: variable projection importance>1, P<0.05, FC≤0.67 or FC≥1.5, and differential metabolites with relative standard deviation>30% in quality control samples are eliminated; In step S5, the multiple reaction monitoring-information-dependent-enhanced mass spectrometry scanning mode is adopted, and the mass spectrometry conditions include: data acquisition range, 50Da~1000Da; nebulizer gas, 50Psi; auxiliary heating gas metabolomics analysis is 60Psi, lipidomics analysis is 50Psi; curtain gas, 40Psi; ionization voltage, 4500V; wherein the floating range of each numerical parameter is ±10% of the numerical value.

2. The method according to claim 1, characterized in that The accurate differential metabolites described in S6 include low-polarity differential metabolites obtained by quasi-targeted lipidomics analysis and high-polarity differential metabolites obtained by quasi-targeted metabolomics analysis.

3. The method according to claim 2, characterized in that The low polarity differential metabolites and mass spectrum information are shown in Table 1 below: Table 1 The high polarity differential metabolites and mass spectrum information are shown in Table 2: Table 2 4. The method according to claim 1, characterized in that: Before multivariate statistical analysis in step S3, the non-raw data are first converted into ABF format for deconvolution, peak extraction, registration and compound identification.

Images