An integrated analysis method for high-coverage detection and multiple reaction monitoring accurate quantification of small molecule compounds in complex samples

Abstract

The application discloses an integrated analysis method for high-coverage detection and multi-reaction monitoring accurate quantification of small molecule compounds in complex samples, which utilizes high performance liquid chromatography-mass spectrometry technology, adopts Zeno SWATH technology, effectively collects mass spectrometry information of low-abundance small molecules, and utilizes MSDIAL to search a database with collected mass spectrometry data to annotate small molecule compounds in complex samples; meanwhile, high resolution-multi-reaction monitoring (HR-MRM) mode is combined to realize synchronous analysis of high-coverage detection and accurate quantification of small molecule compounds in vivo. The method performs excellently in terms of intra-day / inter-day precision and repeatability, more than 90% of characteristics have relative standard deviation (RSD) less than 20%, and the RSD of intra-day / inter-day precision, repeatability and recovery rate of all targeted analysis target objects is less than 20%; the high-throughput method not only significantly saves experimental and data processing time, but also avoids systematic errors of multi-platform data integration, and provides a new paradigm for non-targeting-targeting integrated analysis.

Description

Technical Field

[0001] This invention relates to the fields of analytical chemistry and life sciences, and is a highly efficient integrated method for non-targeted screening and targeted quantification. By analyzing small molecule compounds in human plasma, it can achieve high coverage non-targeted analysis and targeted quantification in a single injection, which greatly improves analytical throughput while saving a lot of experimental and data processing time, and avoids systematic errors in multi-platform data integration. It is of great significance for integrated non-targeted and targeted detection and analysis. Background Technology

[0002] Small molecules in complex systems, including metabolites such as essential nutrients, amino acids, sugars, and fatty acids, as well as gut microbiota metabolites and exogenous exposures, mostly have molecular weights less than 1500 Da. These small molecule compounds can reflect the end state of complex biological systems and are a research hotspot in clinical diagnostics, drug development, cancer treatment, and environmental health risk assessment. However, the concentrations of these small molecule compounds vary greatly, especially endogenous metabolites and exogenous exposures, which often exhibit order-of-magnitude concentration differences. Therefore, a single method is insufficient to meet analytical requirements. In practice, non-targeted and targeted analyses are usually combined, leading to a series of bottlenecks such as limited analytical throughput, increased experimental costs, and high sample consumption. In traditional targeted analyses, multiple response monitoring (MRM) is widely used due to its high sensitivity and specificity, but the number of detectable features is limited. Non-targeted analyses often employ data-dependent acquisition (DDA) and data-independent acquisition (DIA) to improve coverage. However, DDA is limited by insufficient acquisition intensity, while DIA requires deconvolution, limiting its application scope. While targeted metabolomics combines the advantages of high-resolution mass spectrometry and triple quadrupole tandem mass spectrometry, it still suffers from limitations such as slow scan speed and lack of MS2 data for low-abundance or co-eluted metabolites. SWATH technology, by performing secondary full-coverage acquisition of primary ions in different mass windows, can improve MS2 data quality. However, there is still a lack of integrated and efficient analytical methods for simultaneously detecting small molecule compounds with different concentration ranges, especially in metabolism-exposure correlation analysis studies. High-coverage metabolomics detection and high-precision quantitative analysis of the exposure group remain significant challenges. Therefore, an integrated analytical method for high-coverage detection and precise quantification of small molecules is urgently needed. This invention proposes a parallel non-targeted / targeted strategy based on MRM-HR acquisition mode and ZenoSWATH technology, enabling high-coverage non-targeted detection of various small molecule compounds in human plasma and high-precision quantitative analysis of low-abundance target compounds. This method has been applied to a type 2 diabetes mellitus (T2DM) cohort for metabolism-exposure correlation studies, revealing changes in the metabolic profile and potential pathogenic risk factors in T2DM patients compared to healthy controls. Summary of the Invention

[0003] The purpose of this invention is to propose an integrated analytical method for high-coverage detection and precise quantification of small molecule compounds in complex samples using multiple reaction monitoring. This method simultaneously achieves non-targeted high-coverage analysis and targeted precise quantification of small molecule compounds in complex systems, enabling efficient utilization of biological samples and significantly improving analytical throughput.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] An integrated analytical method for high-coverage detection and precise quantification of small molecule compounds in complex samples using high-performance liquid chromatography-mass spectrometry (HPLC-MS) with optimized HPLC conditions and Zeno SWATH and DDA acquisition modes is proposed. This method acquires primary retention time (RT), primary (MS1), and secondary (MS2) data of complex samples. MSDIAL is used to annotate the acquired mass spectrometry data in a database. For low-abundance small molecule compounds, secondary fragments are obtained by purchasing standards. Pseudo-ion pairs (parent ion > daughter ion) are constructed using the parent ion and fragment ions to establish an MRM list of low-abundance small molecule compounds, thus obtaining comprehensive information on small molecule compounds in complex samples. These small molecule compounds include exogenous exposures and endogenous metabolites.

[0006] The optimized liquid phase conditions are as follows:

[0007] In positive ion mode, an ACQUITY UPLC BEH C18 column (Waters, Milford, MA) with a diameter of 2.1 × 100 mm and a column diameter of 1.7 μm was used. The column temperature was set to 50 °C. The aqueous phase (phase A) was water with a volume content of 0.1% formic acid, and the organic phase (phase B) was acetonitrile with a volume content of 0.1% formic acid. The elution gradient was: 0–1 min, 5% B; 1–24 min, 100% B; 24–28 min, 100% B; 28.1–32 min, 5% B.

[0008] In negative ion mode, an ACQUITYUPLC HSS T3 column (Waters, Milford, MA) 2.1×100mm 1.7μm was used, with the column temperature set at 55℃. Phase A was water containing 6.5mM ammonium bicarbonate, and phase B was water containing 95% (v / v) methanol and 5% water containing 6.5mM ammonium bicarbonate. The elution gradient was: 0–1 min, 2% B; 1–18 min, 100% B; 18–22 min, 100% B; 22.1–26 min, 2% B.

[0009] The process of acquiring mass spectrometry data of complex samples using high-resolution mass spectrometry Zeno SWATH and DDA acquisition modes, and annotating small molecule compounds in complex samples, is as follows:

[0010] First, MS1 information is obtained by performing a full scan based on DDA mode;

[0011] For MS1 information, the scan range of precursor ion m / z was set to 100-1200 Da, the accumulation time was 0.15 s, and the DP and CE in positive and negative ion modes were set to 80 V / -80 V and 10 V / -10 V, respectively.

[0012] For MS2 information, the scan range of product ion m / z was set to 50–1000 Da, DP to 80 V / -80 V, CE to ±35 V with a variation range of 15 V, and the accumulation time was set to 0.03 s.

[0013] The precursor ions were then distributed across 11 SWATHMS acquisition windows to obtain MS2 information;

[0014] ZenoSWATH was used to acquire mass spectrometry data of mixed QC (quality control) plasma samples and to collect data in the database.

[0015] The mass spectrometry data acquired in ZenoSWATH mode were deconvolved, peak extracted, and aligned using MSDIAL. Then, exogenous exposures and endogenous metabolites in complex samples were annotated by searching the database.

[0016] For data acquired in DDA mode, peaks were extracted and aligned using MSDIAL, and exogenous exposures and endogenous metabolites were annotated by searching the database.

[0017] When searching the database, the errors of MS1 and MS2 were set to 0.01Da and 0.02Da, respectively. The similarity of MS2 was evaluated by the average of the dot product, the inverse dot product, and the fragment matching rate.

[0018] Only signals with a signal-to-noise ratio greater than 5 and a total score exceeding 60 are retained for subsequent analysis. The annotation levels are defined based on the overall score and MS2 similarity. Signals with a score of 80 or above and a good match between MS1 ​​and MS2 that is verified in the OSI-SMMS database are classified as Level 1; signals with a score of 80 or above and a good match between MS1 ​​and MS2 are classified as Level 2; and signals with a score below 60 and a partial match between MS2 are classified as Level 3.

[0019] The parameters for the high-resolution mass spectrometry section were set as follows: MS1 scan range of the time-of-flight detector, 100-1000 Da; signal threshold for secondary Zeno enrichment, 20000 cps; curtain gas, 35 psi; ion source gas 1, 50 psi; ion source gas 2, 50 psi; induced collision pyrolysis gas, 7 psi; declustering voltage, 100 V (+) / 80 V (-); collision energy, 10 V (+) / 10 V (-); ion source spray voltage, 5500 V (+) / 4500 V (-); ion source transport temperature, 500 °C (+) / 350 °C (-).

[0020] The mass spectrometry parameters of ZenoSWATH were set as follows: curtain gas, 35 psi; ion source gas 1, 50 psi; ion source gas 2, 50 psi; collision-induced pyrolysis gas, 7 psi; declustering voltage, 100 V (+) / 80 V (-); collision energy, 10 V (+) / 10 V (-); ion source spray voltage, 5500 V (+) / 4500 V (-); ion source transport temperature, 500 °C (+) / 350 °C (-).

[0021] Calculations using the SWATH variable window method showed that the m / z scan ranges for the precursor ion and product ion in positive ion mode were 100-175.9 Da, 174.9-253.4 Da, 252.4-331 Da, 330-460.2 Da, 459.2-511.4 Da, 510.4-556.5 Da, 555.5-696.1 Da, 695.1-764.9 Da, 763.9-798.4 Da, and 797.4-842 Da, respectively. 4 Da, 841.4-1200 Da; the scanning ranges in negative ion mode are 100-249.6 Da, 248.6-288.1 Da, 287.1-315 Da, 314-343.6 Da, 342.6-376.6 Da, 375.6-450.3 Da, 449.3-529 Da, 528-575.7 Da, 574.7-675.2 Da, 674.2-768.7 Da, and 767.7-1200 Da.

[0022] The MRM-MS method for targeted acquisition of information on undetected and low-abundance compounds is as follows: For undetected and low-abundance small molecule compounds (signal-to-noise ratio less than 5), secondary fragments of the small molecule compounds are obtained by purchasing their standards. Pseudo-ion pairs are constructed using the parent ion and fragment ions, and their collision energy (CE) and de-cluster voltage (DP) are optimized on the mass spectrometer: the CE voltage is 5–120 V and the DP voltage is 0–180 V, so that the ion pairs obtain the highest signal response value under the optimal DP and CE values, in order to establish the final MRM list of compounds.

[0023] The MRM list contains basic chemical information of target exogenous substances with a signal-to-noise ratio of less than 5, including CAS number, CID number, chemical formula, retention time, CE, and DP values.

[0024] Advantages of this invention: This invention utilizes an ultra-high performance liquid chromatography-quadrupole time-of-flight tandem mass spectrometry (UHPLC-QMS) platform. A parallel non-targeted / targeted integrated analysis strategy is constructed to achieve efficient, rapid, and integrated acquisition of small molecule compounds in human plasma samples. This method comprehensively acquires small molecule information in plasma based on Zeno SWATH technology and optimizes the collision energy (CE) and declustering voltage (DP) of target molecules. Subsequently, the established LC HR-MRM method is systematically evaluated, including repeatability, intra-day and inter-day precision, and recovery. Finally, the optimized parallel non-targeted / targeted omics method is applied to real samples. This method exhibits excellent performance in intra-day / inter-day precision and repeatability, with over 90% of the features having a relative standard deviation (RSD) of less than 20%, and the RSDs for intra-day / inter-day precision, repeatability, and recovery for all targeted analytes being less than 20%. Furthermore, the lower limit of quantification (LLOQ) of the MRM portion of this method is 0.1–25 ng / mL, and the higher limit of quantification (HLOQ) is 2.5–1000 ng / mL, which can meet the requirements of high-coverage detection and accurate quantitative analysis of small molecules. Attached Figure Description

[0025] Figure 1 The TIC and EIC of the representative standard matrix spiked based on the LC-HR-MRM method are shown. A to B are TIC diagrams of exogenous standard matrix spiked based on the LC-HR-MRM method (A: positive ion mode, B: negative ion mode); C to D are EIC diagrams of exogenous standard matrix spiked based on the LC-HR-MRM method (C: positive ion mode, D: negative ion mode). The final concentration of the exogenous standard is 125 ng / mL.

[0026] Figure 2 Comparison of sensitivity between MRM and HR under low concentration conditions (10 ng / mL). A-B are total ion chromatograms (TIC) of the targeted exposure group under multiple reaction monitoring (MRM) mode; C-D are extracted ion chromatograms (EIC) of the targeted exposure group under high resolution quality (HR) mode. The concentration of mixed standard added to serum is 10 ng / mL.

[0027] Figure 3 The correlation between serum risk factors and differentially expressed metabolites was investigated. The correlation coefficients between significant metabolites and cyclophenyl tranexamic acid (A), hydrocortisone (B), PFOA (C), and PFOS (D) were calculated using PASW Statistics 18 software (SPSS, Chicago, IL). Spearman's r represents the color and thickness of the edges in the network, and -logP describes the size of the nodes in the network. Detailed Implementation

[0028] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings: the embodiments are implemented based on the technical solution of the present invention, and detailed implementation methods and specific operation processes are given, but the protection scope of the present invention is not limited to the following embodiments.

[0029] Literature review revealed that the concentration of exogenous exposures in the human body is far lower than that of endogenous metabolites, and these low-abundance exposures often pose a potential risk to disease development. Exogenous exposures typically include pesticides, herbicides, fungicides, veterinary drugs, and persistent organic pollutants. In recent years, studies on exogenous exposures in plasma have been widely reported, indicating their association with various metabolic diseases, including diabetes. Based on this, we utilize the integrated analytical method established in this invention to attempt an integrated analysis of the metabolic profile and exposure genome of diabetes.

[0030] Example 1

[0031] (1) Specific implementation process of liquid phase condition optimization:

[0032] 100 μL of human mixed quality control plasma (a mixture of QC from 100 different populations) collected in the laboratory was added to 400 μL of LAC / MeOH (1:1, v / v) containing exogenous standards (a mixed analytical solution containing endogenous and exogenous small molecules was constructed, see Table 1). The mixture was vortexed for 60 seconds, and then centrifuged at 14000 g / min and 4 °C for 15 minutes. Two 200 μL aliquots of the supernatant were taken and lyophilized for analysis in positive and negative ion modes, respectively. The lyophilized sample was reconstituted with 40 μL of 20% ACN, centrifuged at 14000 g / min and 4 °C for 15 minutes, and 5 μL of the supernatant was injected for analysis. Ultra-high performance liquid chromatography-quadrupole time-of-flight tandem mass spectrometry (UHPLC-quadrupole time-of-flight tandem mass spectrometry) was used for analysis. TMAnalysis was performed using a 7600 mass spectrometer (AB SCIEX, Framingham, USA). Positive ion mode was achieved using an ACQUITY UPLC BEH C18 column (Waters, Milford, MA) 2.1 × 100 mm 1.7 μm, and negative ion mode using an ACQUITY UPLC HSS T3 column (Waters, Milford, MA) 2.1 × 100 mm 1.7 μm. The liquid chromatography conditions for both positive and negative ion modes were optimized to ensure good separation of small molecule compounds in the plasma. In the final positive ion mode, the column temperature was set to 50 °C, the aqueous phase (phase A) was water containing 0.1% formic acid, and the organic phase (phase B) was acetonitrile containing 0.1% formic acid. The elution gradient was: 0–1 min, 5% B; 1–24 min, 100% B; 24–28 min, 100% B; 28.1–32 min, 5% B. In negative ion mode, the column temperature was set to 55℃. Phase A consisted of water containing 6.5 mM ammonium bicarbonate, and Phase B consisted of water containing 95% methanol and 5% water containing 6.5 mM ammonium bicarbonate. The elution gradient was: 0–1 min, 2% B; 1–18 min, 100% B; 18–22 min, 100% B; 22.1–26 min, 2% B. Mass spectrometry parameters were set as follows: TOF scan range, 100–1000 Da; Curtain gas, 35 psi; Ion source gas, 1.50 psi; Ion source gas, 2.50 psi; CAD gas, 7 psi; Declustering potential, 100 V (+) / 80 V (-); Collision energy, 10 V (+) / 10 V (-); Ion spray voltage, 5500 V (+) / 4500 V (-); Interface heater temperature, 500℃ (+) / 350℃ (-). Figure 1 As shown, Figures A and B are the total ion chromatograms (TIC) of spiked plasma in positive and negative ion modes, respectively. Figures C and D are the extraction ion chromatograms (EIC) of exogenous substances with different structural categories, respectively. It can be seen that under this elution gradient, both endogenous and exogenous small molecules are well separated, indicating that the optimized chromatographic gradient is suitable for the analysis of complex matrices containing multiple chemical structures.

[0033] (2) Acquisition of panoramic information on potential small molecule compounds in plasma.

[0034] Specific implementation: Take 100 μL of the above mixed plasma, add 400 μL of LACN / MeOH (1:1, v / v) extraction reagent, vortex for 60 seconds, centrifuge the mixture at 14000 g / min and 4℃ for 15 minutes, take two equal 200 μL supernatants, freeze-dry them, and use them for positive and negative ion mode analysis respectively. Add 40 μL of 20% ACN to the freeze-dried sample for reconstitution, centrifuge at 14000 g / min and 4℃ for 15 minutes, and inject 5 μL of the supernatant for analysis.

[0035] Ultra-high performance liquid chromatography-quadrupole time-of-flight tandem mass spectrometry analysis: using ultra-high performance liquid chromatography-HPLC (AB SCIEX, Framingham, USA)-ZenTOF TM Analysis was performed using a 7600 mass spectrometer (AB SCIEX, Framingham, USA) in positive ion mode, employing an ACQUITY UPLC BEH C column. 18 The column (Waters, Milford, MA) was 2.1 × 100 mm 1.7 μm, with a column temperature set at 50 °C. The aqueous phase (phase A) was water containing 0.1% formic acid, and the organic phase (phase B) was acetonitrile containing 0.1% formic acid. The elution gradient was: 0–1 min, 5% B; 1–24 min, 100% B; 24–28 min, 100% B; 28.1–32 min, 5% B. In negative ion mode, an ACQUITY UPLC HSS T3 column (Waters, Milford, MA) was used, 2.1 × 100 mm 1.7 μm, with a column temperature set at 55 °C. Phase A was water containing 6.5 mM ammonium bicarbonate, and phase B was water containing 95% methanol and 5% water containing 6.5 mM ammonium bicarbonate. The elution gradient was: 0–1 min, 2% B; 1–18 min, 100% B; 18–22 min, 100% B; 22.1–26 min, 2% B.

[0036] The mass spectrometry parameters were set as follows: TOF scan range, 100-1000 Da; Curtain gas, 35 psi; Ion source gas 1, 50 psi; Ion source gas 2, 50 psi; CAD gas, 7 psi; Declustering potential, 100 V (+) / 80 V (-); Collision energy, 10 V (+) / 10 V (-); Ion spray voltage, 5500 V (+) / 4500 V (-); Interface heater temperature, 500℃ (+) / 350℃ (-).

[0037] First, MS1 information was acquired using a full scan in DDA mode. For MS1, the scan range of precursor ion m / z was set to 100-1200 Da, the accumulation time was 0.15 s, and the DP and CE in positive and negative ion modes were set to 80 V / -80 V and 10 V / -10 V, respectively. For MS2, the scan range of product ion m / z was set to 50-1000 Da, DP was set to 80 V / -80 V, CE was set to ±35 V with a variation range of 15 V, and the accumulation time was set to 0.03 s. Then, the precursor ions were distributed across 11 SWATH MS acquisition windows calculated by SWATH variable window Calculator_V1.1 to acquire high-quality MS2 information. Ultimately, the mass spectrometry parameters for ZenoSWATH were set as follows: Curtain gas, 35 psi; Ion source gas 1, 50 psi; Ion source gas 2, 50 psi; CAD gas, 7 psi; Declustering potential, 100 V (+) / 80 V (-); Collision energy, 10 V (+) / 10 V (-); Ion spray voltage, 5500 V (+) / 4500 V (-); Interface heater temperature, 500 °C (+) / 350 °C (-). Finally, using SWATH variable window Calculator_V1.1, the scan ranges of the precursor ion m / z and product ion m / z in positive ion mode were 100-175.9 Da, 174.9-253.4 Da, 252.4-331 Da, 330-460.2 Da, 459.2-511.4 Da, 510.4-556.5 Da, 555.5-696.1 Da, 695.1-764.9 Da, 763.9-798.4 Da, and 79... The scanning ranges in negative ion mode are 7.4-842.4 Da and 841.4-1200 Da, respectively: 100-249.6 Da, 248.6-288.1 Da, 287.1-315 Da, 314-343.6 Da, 342.6-376.6 Da, 375.6-450.3 Da, 449.3-529 Da, 528-575.7 Da, 574.7-675.2 Da, 674.2-768.7 Da, and 767.7-1200 Da.

[0038] To annotate more small plasma molecules, we used Zeno SWATH technology to acquire mixed plasma QC mass spectrometry data and collected data from multiple databases, including the MoNA database (https: / / mona.fiehnlab.ucdavis.edu) and the NIST Mass Spectrometry Library (2023 version). https: / / www.nist.gov / srd / nist-standard-reference-database-1a), OSI-SMMS Databases included MS-DIAL4, the internal exposure group database, and relevant published literature. The raw data acquired in the Zeno SWATH mode were then deconvolved, peaks extracted and aligned using MSDIAL, and then the database annotations for exogenous exposures and endogenous metabolites were searched. For data acquired in the DDA mode, peak extraction and alignment were also performed using MSDIAL before searching the database annotations. During the database search, the errors for MS1 and MS2 were set to 0.01 Da and 0.02 Da, respectively. MS2 similarity was evaluated using the average values ​​of dot-product, reverse dot-product, and matched fragment ratio. To reduce false positives, only signals with a signal-to-noise ratio greater than 5 and a total score greater than 60 were retained for subsequent analysis. Annotation levels were defined based on the overall score and MS2 similarity: scores above 80 with good MS1-MS2 matching and validation in the OSI-SMMS database were classified as Level 1; scores above 80 with good MS1-MS2 matching were classified as Level 2; and scores below 60 with partial MS2 matching were classified as Level 3. To ensure the accuracy of the annotations, all annotated compounds were manually verified for structure-preservation relationships and characteristic fragments before statistical analysis.

[0039] (3) MRM-MS for targeted acquisition of information on low-abundance compounds: For low-abundance small molecule compounds (especially exogenous exposures), their secondary fragments are obtained by purchasing their standards, pseudo-ion pairs are constructed by the parent ion and fragment ions (parent ion > daughter ion), and the data are analyzed in Triple Quad. TMCE and DP were optimized on an AB SCIEX 6500 instrument (USA). First, DP optimization (0–180V) was performed, followed by CE optimization (5–120V). This ensured that the ion pairs (parent ion > daughter ion) achieved the highest signal response values ​​under optimal DP and CE conditions, thus establishing the final MRM list (Table 2). The MRM list includes basic chemical information (CAS, CID, molecular formula, etc.) of the target low-abundance exogenous substances, as well as retention times and optimized CE and DP values. Finally, based on the LC HR-MRM method, a parallel non-targeted / targeted integrated detection and analysis strategy was established. HR-MS was used for high-coverage acquisition of endogenous metabolite information, while MRM-MS was used for targeted acquisition of low-abundance compound information, thereby achieving integrated non-targeted and targeted detection in a single injection. The final parameters for the HR-MS section were set as follows: TOF scan range, 100-1000 Da; Zeno plusing threshold of TOF MS / MS scan, 20000 cps; Curtain gas, 35 psi; Ion source gas 1, 50 psi; Ion source gas 2, 50 psi; CAD gas, 7 psi; Declustering potential, 100V(+) / 80V(-); Collision energy, 10V(+) / 10V(-); Ionspray voltage, 5500V(+) / 4500V(-); Interface heater temperature, 500℃(+) / 350℃(-). Specific MRM-MS parameters are shown in Table 2.

[0040] Compared with high-resolution mass spectrometry, the MRM mode significantly improved the response intensity of ion pairs. At a low concentration of 10 ng / mL, over 70% of the 210 low-abundance exogenous substances detected showed a more than 3-fold increase in response intensity in MRM mode compared to HR mode, highlighting the advantage of MRM-HR for screening low-concentration small molecules. Figure 2 (A~2D).

[0041] Example 2

[0042] This study included 72 Chinese patients with type 2 diabetes mellitus (T2DM) and 72 healthy controls. Plasma samples were collected from them after overnight fasting. Their biological and clinical information, including sex, age, BMI, alcohol consumption history, smoking history, education level, and income level, were collected and summarized in Table 3.

[0043] (1) The ultra-high performance liquid chromatography-quadrupole time-of-flight tandem mass spectrometry (UHPLC-QQS-MS / MS) platform was used to efficiently and rapidly collect exogenous exposures and endogenous metabolites in human plasma samples to obtain raw data.

[0044] Specific implementation process:

[0045] Sample pretreatment: 100 μL of plasma from each diabetic patient was collected and 400 μL of an acetonitrile / methanol (ACN / MeOH) extractant with an internal standard (as shown in Table 4) at a volume ratio of 1:1 was added. After vortexing for 60 seconds, the plasma mixture was centrifuged at 14000 g / min and 4 °C for 15 minutes. Two 200 μL aliquots of the supernatant from each plasma sample were lyophilized. 40 μL of 20% (v / v) acetonitrile (ACN) was added to each lyophilized sample for reconstitution. The samples were then centrifuged at 14000 g / min and 4 °C for 15 minutes, and the supernatant was collected. 5 μL of the supernatant from each sample was injected for analysis. Then, an equal volume (50 μL in this case) of the supernatant from each sample was mixed and used as the quality control (QC) sample. To eliminate matrix interference and determine stable quantitative results, the 100 μL of plasma in the sample pretreatment was replaced with acetonitrile standards containing the following concentration gradients (0.1, 0.25, 0.5, 1, 2.5, 5, 10, 25, 50, 100, 125, 250, 500 and 1000 ng / mL) to construct a quantitative linear curve.

[0046] Ultra-high performance liquid chromatography-quadrupole time-of-flight tandem mass spectrometry analysis: using ultra-high performance liquid chromatography-HPLC (AB SCIEX, Framingham, USA)-ZenTOF TMTargeted data acquisition of exposed pollutants and non-targeted detection of metabolites were performed using a 7600 mass spectrometer (AB SCIEX, Framingham, USA) (Table 2). Based on the optimized chromatographic conditions of this invention, in positive ion mode, the column temperature was 50°C, the aqueous phase (phase A) was water containing 0.1% (v / v) formic acid, and the organic phase (phase B) was acetonitrile containing 0.1% (v / v) formic acid. The elution gradient was: 0–1 min, 5% B; 1–24 min, 100% B; 24–28 min, 100% B; 28.1–32 min, 5% B. In negative ion mode, the column temperature was 55°C, phase A was water containing 6.5 mM ammonium bicarbonate, and phase B was water containing 95% (v / v) methanol and 5% (v / v) water containing 6.5 mM ammonium bicarbonate. The elution gradient was: 0–1 min, 2% B; 1–18 min, 100% B; 18–22 min, 100% B; 22.1–26 min, 2% B. Mass spectrometry parameters were set as follows: TOF scan range, 100–1000 Da; Curtain gas, 35 psi; Ion source gas 1, 50 psi; Ion source gas 2, 50 psi; CAD gas, 7 psi; Declustering potential, 100 V (+) / 80 V (-); Collision energy, 10 V (+) / 10 V (-); Ion spray voltage, 5500 V (+) / 4500 V (-); Interface heater temperature, 500 °C (+) / 350 °C (-). MRM-MS was used to target and collect information on exogenous exposures in plasma. Specific MRM-MS parameters are shown in Table 2.

[0047] (2) Comprehensive characterization of metabolic profiles

[0048] Specific implementation process: The collected sample QC data is annotated with metabolites using database searches (here, the MoNA database (https: / / mona.fiehnlab.ucdavis.edu) and the NIST mass spectrometry library (2023 version) are used). https: / / www.nist.gov / srd / nist-standard-reference-database-1a), OSI-SMMSDatabases (MS-DIAL4 database, self-built exposure database based on standards, and published literature, etc.) were used. The raw peak area data of internal standards and metabolites in all samples were exported, and the raw peak areas of 698 metabolites were corrected using 21 internal standards to obtain relative peak area data (the ratio of the peak area of ​​a metabolite to that of an internal standard). In this embodiment, 698 metabolites with an RSD less than 30% in the QC samples were included in subsequent analysis. Nonparametric analysis (Wilcoxon signed-rank test) showed that 174 metabolites had significant differences between the diabetic and healthy control groups (p-value less than 0.05). Compared with the healthy control group, the levels of fatty acids, carnitine, and lysophosphatidylcholine were significantly increased in the T2DM group. Subsequently, MetaboAnalyst 6.0 (https: / / www.metaboanalyst.ca) was used to analyze the metabolic pathways of diabetes. The results showed that the synthesis of unsaturated fatty acids and glycerophospholipid metabolism were the main metabolic pathways, indicating that lipid metabolism changes significantly in diabetes. Then, binary logistic regression analysis was further performed using the R language program (R 4.2.1) and RStudio software (Reference 4. Racine, JS, RStudio: a platform-independent IDE for R and Sweave. Journal of Applied Econometrics 2012, 27(1), p.167-172.). The results showed that a total of 67 metabolites were screened as potential risk metabolites for T2DM, most of which were significantly elevated in the disease group and had an OR value greater than 1.

[0049] 3) Targeted quantitative analysis of the exposed group

[0050] Specific implementation process: Mass spectrometry data of plasma samples from diabetic and healthy controls were collected using a parallel non-targeted metabolomics / targeted exposure omics strategy. To ensure experimental accuracy, a matrix-spiked linear analysis (using fetal bovine plasma as the matrix) was interpolated during the actual sample analysis. Fourteen linear samples with different final concentrations of the exposed substance were used to construct linear curves and determine the limit of quantitation. The concentrations of exogenous exposed substances in the linear samples were distributed in a gradient pattern: 0.1, 0.25, 0.5, 1, 2.5, 5, 10, 25, 50, 100, 125, 250, 500, and 1000 ng / mL. Each linear sample contained 210 exposed substance concentrations of equal value. After data collection, the relative peak area of ​​the exposed substance was obtained by dividing its peak area by the peak area of ​​its corresponding internal standard. This led to the construction of a regression equation y = ax + b between the relative peak area (y) of the exposed substance and the actual added concentration (x) in the linear data, where a is the slope of the linear equation and b is the intercept (constant). In this example, the R0 values ​​for all exposed substances... 2All peak area data were greater than 99%. The raw peak area data of the tested samples were exported and corrected using an internal standard to obtain relative peak area data for the risk exposures. Then, a standard linear curve was used to quantitatively analyze the internally standard-corrected relative peak area data to obtain the quantitative results of the risk exposures. Results below the minimum quantitation limit or missing values ​​in the quantitative data of risk exposures were then invalidated by assigning null values, with the assignment set to half of the minimum quantitation limit, for subsequent statistical analysis. Subsequently, the detection rate of the exposure in all samples was calculated (detection rate = number of detected samples / total number of samples) to determine whether it was a high-frequency exposure (an exposure with a detection rate greater than 30% was defined as a high-frequency exposure). Finally, the quantitative data of high-frequency exposures that met the QC sample RSD less than 30% were selected. Nonparametric tests (a method of inferring the overall distribution pattern using sample data when the population variance is unknown or poorly known) were performed using R language program (R 4.2.1) and RStudio software to identify exposures with significant differences between healthy and diseased populations (satisfying p < 0.05). Using quantitative data of exposures with significant differences, the hazard ratio (OR) of exposures with significant differences was further calculated by binary logistic regression analysis using the R language program (R 4.2.1) and RStudio software (Reference 4. Racine, JS, RStudio: a platform-independent IDE for R and Sweave. Journal of Applied Econometrics 2012, 27(1), p.167-172.) (Reference 5. Blanco, JF; da Casa, C.; Pablos-Hernández, C.; González-Ramírez, A.; Julián-Enríquez, JM; A., 30-day mortality after hip fracture surgery: Influence of postoperative factors. PLoS one 2021, 16(2), e0246963.), initially identified potential exposure risk factors. Then, correlation analysis was used to test whether there was a significant correlation between potential exposure risk factors and the disease.

[0051] In this embodiment, 20 high-frequency exposures with a detection rate greater than 30% were identified. Their detection rates, RSD distributions, concentrations, fold changes, and significance p-values ​​are shown in Table 5. The quantitative results of these 20 high-frequency exposures were then analyzed using nonparametric tests. The results showed that 5 exposures were significantly higher in the disease group than in the healthy group. Further binary logistic regression analysis was used to calculate the risk values ​​of these differentially exposed exposures. Four of these exposures—PFOA, PFOS, cyclamic acid, and hydrocortisone—were identified as significant and thus potential risk factors.

[0052] 4) Exposure-metabolism association analysis based on non-targeted metabolomics / parallel targeted exposureomics strategies

[0053] Based on simultaneously collected metabolomics and exposure information, an exposure-metabolic correlation study was conducted. In this case, metabolic reprogramming in the diabetic group mainly manifested as perturbations in lipid metabolism, with cyclohexanoic acid, hydrocortisone, PFOA, and PFOS as potential risk factors. Spearman correlation analysis of differential metabolites and exposure risk factors showed that cyclohexanoic acid, hydrocortisone, PFOA, and PFOS were significantly positively correlated with lipids such as LPC, PC, FA, and CAR. Notably, most fatty acids (especially long-chain fatty acids) and most LPCs, as well as long-chain acylcarnitines, were positively correlated with cyclohexanoic acid, hydrocortisone, PFOA, and PFOS, while the correlation with hydrocortisone was limited. Furthermore, some SMs and Cers were positively correlated with cyclohexanoic acid, PFOA, and PFOS, and DGs were positively correlated with hydrocortisone. Figure 3 (A~3D). Considering the positive correlation between PFOA and PFOS accumulation in plasma and elevated blood lipids, it can be concluded that PFOA and PFOS are potential risk factors for T2DM by inducing perturbations of LPCs, SMs, Cers, DGs, and CARs. Furthermore, this study also found that cyclohexanoic acid, as an artificial sweetener, is positively correlated with lipid metabolism, revealing a potential pathogenic mechanism by which cyclohexanoic acid is positively correlated with T2DM.

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061]

[0062]

[0063]

[0064]

[0065]

[0066]

[0067]

[0068] Table 3. Epidemiological Information of the Cohort

[0069]

[0070]

[0071]

[0072] Table 5. Quantitative range, detection rate, average concentration, fold change, significance, and distribution characteristics of high-frequency exposures in T2DM and healthy control groups.

[0073]

[0074] Note:Significant exposures with fold change over 1 were bolded andasterisked

Claims

1. An integrated analytical method for high-coverage detection and precise quantification of small molecule compounds in complex samples using multiple reaction monitoring, characterized in that, High-performance liquid chromatography-mass spectrometry (HPLC-MS / MS) was used with optimized HPLC conditions and ZenoSWATH and DDA acquisition modes to acquire primary retention time (RT), primary (MS1), and secondary (MS2) data of complex samples. MSDIAL was then used to search databases to annotate small molecule compounds in the complex samples. For low-abundance small molecule compounds, secondary fragments were obtained by purchasing standards. Pseudo-ion pairs (parent ion > daughter ion) were constructed using the parent ion and fragment ions to establish a MRM list of low-abundance small molecule compounds, thus obtaining a comprehensive overview of small molecule compounds in the complex samples, including both exogenous exposures and endogenous metabolites.

2. The integrated analysis method according to claim 1, characterized in that: The optimized liquid phase conditions were as follows: in positive ion mode, an ACQUITYUPLC BEH C18 column (Waters, Milford, MA) 2.1×100mm 1.7μm was used, with the column temperature set to 50℃. The aqueous phase (phase A) consisted of water with a volume content of 0.1% formic acid, and the organic phase (phase B) consisted of acetonitrile with a volume content of 0.1% formic acid. The elution gradient was: 0–1 min, 5% B; 1–24 min, 100% B; 24–28 min, 100% B; 28.1–32 min, 5% B. In negative ion mode, an ACQUITYUPLC HSS T3 column (Waters, Milford, MA) 2.1×100mm 1.7μm was used, with the column temperature set at 55℃. Phase A was water containing 6.5mM ammonium bicarbonate, and phase B was water containing 95% methanol and 5% water containing 6.5mM ammonium bicarbonate. The elution gradient was: 0–1 min, 2% B; 1–18 min, 100% B; 18–22 min, 100% B; 22.1–26 min, 2% B.

3. The integrated analysis method according to claim 1, characterized in that: The process of acquiring mass spectrometry data of complex samples using high-resolution mass spectrometry Zeno SWATH and DDA acquisition modes, and annotating small molecule compounds in complex samples, is as follows: First, MS1 information is obtained by performing a full scan based on DDA mode; For MS1 information, the scan range of precursor ion m / z was set to 100-1200 Da, the accumulation time was 0.15 s, and the DP and CE in positive and negative ion modes were set to 80 V / -80 V and 10 V / -10 V, respectively. For MS2 information, the scan range of product ion m / z was set to 50–1000 Da, DP to 80 V / -80 V, CE to ±35 V with a variation range of 15 V, and the accumulation time was set to 0.03 s. The precursor ions were then distributed across 11 SWATHMS acquisition windows to obtain MS2 information; Mass spectrometry data of mixed QC plasma samples were acquired using ZenoSWATH and collected in the database; The mass spectrometry data acquired in ZenoSWATH mode were deconvolved, peak extracted, and aligned using MSDIAL. Then, exogenous exposures and endogenous metabolites in complex samples were annotated by searching the database. For data acquired in DDA mode, peaks were extracted and aligned using MSDIAL, and exogenous exposures and endogenous metabolites were annotated by searching the database. When searching the database, the errors of MS1 and MS2 were set to 0.01Da and 0.02Da, respectively. The similarity of MS2 was evaluated by the average of the dot product, the inverse dot product, and the fragment matching rate. Only signals with a signal-to-noise ratio greater than 5 and a total score exceeding 60 are retained for subsequent analysis. The annotation levels are defined based on the overall score and MS2 similarity. Signals with a score of 80 or above and a good match between MS1 ​​and MS2 that is verified in the OSI-SMMS database are classified as Level 1; signals with a score of 80 or above and a good match between MS1 ​​and MS2 are classified as Level 2; and signals with a score below 60 and a partial match between MS2 are classified as Level 3.

4. The integrated analysis method according to claim 1 or 3, characterized in that: The parameters for the high-resolution mass spectrometry section were set as follows: MS1 scan range of the time-of-flight detector, 100-1000 Da; signal threshold for secondary Zeno enrichment, 20000 cps; curtain gas, 35 psi; ion source gas 1, 50 psi; ion source gas 2, 50 psi; induced collision pyrolysis gas, 7 psi; declustering voltage, 100 V (+) / 80 V (-); collision energy, 10 V (+) / 10 V (-); ion source spray voltage, 5500 V (+) / 4500 V (-); ion source transport temperature, 500 °C (+) / 350 °C (-).

5. The integrated analysis method according to claim 1 or 3, characterized in that: The mass spectrometry parameters of ZenoSWATH were set as follows: curtain gas, 35 psi; ion source gas 1, 50 psi; ion source gas 2, 50 psi; collision-induced pyrolysis gas, 7 psi; declustering voltage, 100 V (+) / 80 V (-); collision energy, 10 V (+) / 10 V (-); ion source spray voltage, 5500 V (+) / 4500 V (-); ion source transport temperature, 500 °C (+) / 350 °C (-). Calculations using the SWATH variable window method showed that the m / z scan ranges for the precursor ion and product ion in positive ion mode were 100-175.9 Da, 174.9-253.4 Da, 252.4-331 Da, 330-460.2 Da, 459.2-511.4 Da, 510.4-556.5 Da, 555.5-696.1 Da, 695.1-764.9 Da, 763.9-798.4 Da, and 797.4-842 Da, respectively. 4 Da, 841.4-1200 Da; the scanning ranges in negative ion mode are 100-249.6 Da, 248.6-288.1 Da, 287.1-315 Da, 314-343.6 Da, 342.6-376.6 Da, 375.6-450.3 Da, 449.3-529 Da, 528-575.7 Da, 574.7-675.2 Da, 674.2-768.7 Da, and 767.7-1200 Da.

6. The integrated analysis method according to claim 1, characterized in that: The MRM-MS method for targeted acquisition of information on undetected and low-abundance compounds (signal-to-noise ratio less than 5) is as follows: For undetectable and low-abundance small molecule compounds, secondary fragments of the small molecule compounds were obtained by purchasing their standards. Pseudo-ion pairs were constructed using the parent ion and fragment ions, and their collision energy (CE) and declustering voltage (DP) were optimized on a mass spectrometer: CE voltage 5–120 V, DP voltage 0–180 V. This allows the ion pairs to achieve the highest signal response values ​​under optimal DP and CE values, thus enabling the establishment of the final MRM list. The MRM list contains basic chemical information of target exogenous substances with a signal-to-noise ratio of less than 5, including CAS number, CID number, chemical formula, retention time, CE, and DP values.

7. The integrated analysis method according to claim 1, characterized in that: The complex sample includes, but is not limited to, one or more of the following: plasma, urine, feces, bile acids, etc., from biological complex systems.

8. The integrated analysis method according to claim 1 or 3, characterized in that: The databases mentioned include the MoNA database (https: / / mona.fiehnlab.ucdavis.edu) and the NIST mass spectrometry library (2023 version). https: / / www.nist.gov / srd / nist-standard-reference-database-1a), OSI-SMMS Databases, MS-DIAL4 database, self-built exposure database based on standards, and published literature, etc.

9. The integrated analysis method according to any one of claims 1-8 can be used for metabolomics and exocomics studies based on complex biological samples such as human plasma, urine, and feces. It is particularly suitable for conducting metabolism-exposure association analysis studies and can achieve high-throughput, high-coverage, and high-precision analysis.

Images