Structural annotation method for analyzing positions sn-and C+C of lipid based on liquid chromatography-organic matter ion electron collision excitation mass spectrometry

Abstract

The invention discloses a structure annotation method for analyzing positions sn-and C = C of lipid based on liquid chromatography-organic matter ion electron collision excitation mass spectrometry. According to the method, a proper amount of 0.1 mM sodium acetate is added during lipid redissolution, so that polar lipid and neutral lipid are analyzed at the same time through one-time sample introduction. Meanwhile, a method for automatically annotating the complex organic matter ion electron collision excitation mass spectrum is developed, and the method comprises the steps of (1) extracting mass spectrum information in original data, (2) identifying lipid subclasses, (3) matching lipid total composition, (4) pairing sn-position, (5) positioning C = C position, and (6) determining annotation level. For 34 lipid standard substances in serum, the annotation accuracy rates of the annotation method on sn-and C = C isomer levels are 100% and 82.3% respectively. A total of 1312 sn-isomers and 1033 C = C isomers are annotated from mixed plasma samples of healthy people and patients with mild cognitive impairment of Alzheimer's disease. In a word, the lipid fine structure analysis method developed by the invention has the characteristics of simple pretreatment, high coverage and efficient annotation of the lipid fine structure.

Description

Technical Field

[0001] This invention relates to the fields of analytical chemistry, software development, and clinical medicine. Specifically, this invention relates to establishing a high-coverage liquid chromatography-organic ion electron collision excitation tandem mass spectrometry analytical method, and developing a method for automatically annotating the organic ion electron collision excitation spectra of lipid isomers. This method identifies a group of differentially expressed lipid isomers that can be used to distinguish between healthy individuals and those with early-stage Alzheimer's disease. Background Technology

[0002] Lipids play a vital role in various cellular functions, such as forming cell membranes, transmitting biological neural signals, and providing energy for cellular activities. The lipid database LipidMaps contains 40,000 lipid molecules, categorized into eight major classes and numerous subclasses. Taking phosphatidylcholine (PC) lipid molecules as an example, complete identification requires elucidating the following six structural levels: 1) lipid subclass, 2) acyl chain composition, 3) sn-position of the acyl or ether chain on the glycerol backbone, 4) position of the carbon-carbon double bond (C=C), 5) type and position of substituted functional groups (e.g., methylation and hydroxylation), and 6) stereoisomerism of the carbon-carbon double bond (cis / trans isomerism). Differences in the molecular structure of lipid isomers determine their different properties, leading to differences in the biological functions they perform. For example, increasing the omega-3 / omega-6 ratio has been shown to have beneficial effects on various diseases, including cardiovascular disease, cancer, inflammation, and autoimmune diseases. Given the importance of lipid structure, obtaining detailed structural information across lipid classes is crucial for advancing lipidomics. Liquid chromatography-mass spectrometry (LC-MS) has become a powerful tool for lipidomics analysis, enabling precise characterization and quantification of lipids in complex biological samples. However, traditional dissociation techniques, such as collision-induced dissociation (CID) or high-energy collision dissociation (HCD), do not distinguish between sn-isomers and C=C isomers in lipids.

[0003] In recent years, numerous mass spectrometry-based analytical techniques have significantly improved our ability to distinguish lipid isomers. These reactions include the Pastenò-Büchi (PB) reaction and epoxidation, which accurately determine the C=C position by C=C derivatization and CID or HCD integration. Furthermore, novel ion activation / dissociation techniques have also been enlightening in mapping the sn-positions and C=C positions of lipids, including ozone-induced dissociation (OzID), ultraviolet photodissociation (UVPD), and electron shock excitation of organic ions (EIEIO). These methodological advances have not only broadened the scope of lipidomics analysis but also revealed the link between lipid isomer composition and disease pathology. However, to date, only a few lipid structure characterization strategies have been able to simultaneously analyze neutral and polar lipids in a single experiment, hindering the progress of larger-scale lipidomics studies.

[0004] It is well known that accurate structural characterization of complex lipids requires sophisticated separation techniques and automated data analysis tools. To date, only a limited number of tools have been developed to analyze complex lipid structural data. For example, LipidOA has proven to be an effective tool for locating glycerophospholipid double bonds using PB-MS / MS data. Furthermore, MS-DIAL 5.0 software offers powerful capabilities for processing MS data in both CID and EIEIO modes. Additionally, a previously unpublished, self-programmed offline analysis software has been developed using differential ion mobility spectrometry to decode lipid information collected from EIEIO. However, existing lipid analysis tools are often limited to the identification of single lipid classes or C=C positions, resulting in poor utilization of lipid structural information. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a structural annotation method for resolving the sn- and C=C positions of lipids based on liquid chromatography-organic ion electron collision excitation mass spectrometry (LC-EIEIO-MS / MS). By optimizing the parameters of the LC-EIEIO-MS / MS method, polar and neutral lipids can be analyzed simultaneously in a single injection. Subsequently, the efficient annotation method provided by this invention is used to annotate the obtained comprehensive lipid mass spectrometry data to locate the sn- and C=C positions of complex lipids.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A structural annotation method for lipid sn- and C=C positions based on liquid chromatography-organic ion electron collision excitation mass spectrometry, the method comprising the following steps:

[0008] (1) Preprocess the sample to obtain lipid extracts from the sample;

[0009] (2) The lipid extract was resuspended in a sodium acetate solution containing 0.1 mM to obtain the lipid extract solution to be injected;

[0010] (3) The raw data of the lipid extract in step (2) were obtained by liquid chromatography-organic ion electron collision excitation mass spectrometry.

[0011] (4) Based on the raw mass spectrometry data of the obtained sample lipid extract, the structural positions of lipid sn- and C=C were analyzed and annotated.

[0012] The samples in step (1) include, but are not limited to, plasma, serum, tissue, cells, yeast, microalgae, and other sample types.

[0013] When the sample in step (1) is plasma, serum, tissue, or cells, the pretreatment process is as follows: add 10-100 μL of sample, 100-1000 μL of methanol, and 0.5-5 mL of methyl tert-butyl ether to a centrifuge tube, shake and mix for 5-15 min, then add 100-1000 μL of water, vortex for 1-3 min to form a two-phase system, centrifuge at 8000-15000 g for 10-30 min, and freeze-dry the supernatant to obtain the lipid extract from the sample;

[0014] When the sample in step (1) is yeast or microalgae, the pretreatment process is as follows: add 10-100 μL of sample and 100-1000 μL of methanol to a centrifuge tube, grind it for 2 min at 25 Hz using a tissue homogenizer, add 0.5-5 mL of methyl tert-butyl ether, shake and mix for 5-15 min, then add 100-1000 μL of water, vortex for 1-3 min to form a two-phase system, centrifuge at 8000-15000 g for 10-30 min, take the supernatant and freeze dry to obtain the lipid extract in the sample;

[0015] The 0.1 mM sodium acetate solution in step (2) is a 30%-95% (v / v) aqueous solution of an organic solvent, wherein the organic solvent includes, but is not limited to, one or two of acetonitrile, isopropanol, and methanol.

[0016] In step (3), the liquid chromatography conditions are as follows: mobile phase A is a 30%-90% (v / v) aqueous solution of ammonium acetate in acetonitrile, and mobile phase B is a 30%-90% (v / v) isopropanol in acetonitrile solution in ammonium acetate; the separation column type includes, but is not limited to, reversed-phase columns such as C8, C18 or C30.

[0017] The organic ion electron collision excitation mass spectrometry described in step (3) is performed in EAD fragmentation mode under an electron beam pressure of 10 eV, and data acquisition is carried out in positive ion mode.

[0018] Step (4) Annotate the structural positions of lipid sn- and C=C based on the obtained sample lipid mass spectrometry information, including the following steps:

[0019] 1) Based on the raw data of the lipid extract, extract lipid spectrum information, which includes the mass-to-charge ratio of the parent ion in the primary lipid spectrum information and the mass-to-charge ratio and relative intensity of the daughter ion in the secondary lipid spectrum information;

[0020] 2) Identification of lipid subclasses: Phosphatidylcholine, sphingomyelin, and phosphatidylethanolamine lipids were categorized as known categories; ceramide, diglycerides, and triglycerides lipids were not categorized and were treated as unknown categories.

[0021] 3) Matching total lipid composition: Construct an index library including known and unknown categories for total composition matching; match the mass-to-charge ratio of lipid precursor ions of known and unknown categories according to the theoretical molecular mass in the index library, and annotate the total lipid composition; for lipids labeled with subclasses, the mass-to-charge ratio of precursor ions is annotated by finding the closest theoretical molecular mass in the same subclass; for lipids of unknown subclasses, the minimum difference between the mass-to-charge ratio of the actual precursor ion and the mass-to-charge ratio of the theoretical precursor ion is found in the index library of unknown categories for annotation;

[0022] 4) Determine sn-position: Match characteristic fragment ions based on the established sn-position index library and annotate lipid sn-positions; for phosphatidylcholine, phosphatidylethanolamine and diglyceride lipids, the relative intensity of fragment ions generated by the breakage between C1-C2 of their glycerol backbone is lower than that of fragment ions generated by the breakage between C1-O and C2-O. Therefore, characteristic fragment ions between C1-C2 can be used to determine sn-positions;

[0023] For sphingomyelin and ceramide lipids, odd and even acyl chain fragments in the organic ion electron collision excitation spectrum can be used to solve the problem of the arrangement of the two chains on the backbone and sn-2.

[0024] For triglyceride lipids, the combined loss of the sn-1 or sn-3 and sn-2 fatty acid chains produces a pair of doublets with a mass difference of 2.01 Da. The chain length and number of double bonds of sn-1 or sn-3 can be calculated from the doublets. At the same time, the double loss of the sn-1 and sn-3 fatty acyl groups results in a singlet state for the sn-2 fragment ion, thus determining the position of sn-2. Except for lysophosphatidylcholine lipids, all characteristic fragment ions generated by the fatty acyl chains of other types of lipid isomers are summarized to construct an sn-position diagnostic index library for determining sn-position.

[0025] 5) Locating the C=C position: When a double bond exists on the fatty acyl chain, a mass difference of 26.02 Da can be observed instead of the 28.02 Da produced by the carbon-carbon single bond, which can be used to locate the lipid double bond position;

[0026] 6) Determine the annotation level: Classify the annotation results. Lipids annotated at both C=C and sn-positions are classified as Level 1; lipids annotated only at sn-positions or C=C positions are classified as Level 2; lipids not annotated at sn-positions and C=C positions but inferred only from the overall composition are classified as Level 3.

[0027] The lipid types are classified into seven subclasses: lysophosphatidylcholine, phosphatidylcholine, phosphatidylethanolamine, sphingomyelin, ceramide, diglycerides, and triglycerides.

[0028] This method can be used to resolve the fine structure of lipid molecules when analyzing samples using liquid chromatography coupled with EIEIO mass spectrometry.

[0029] The principle of this method is as follows:

[0030] (1) By adding 0.1 mM sodium acetate during reconstitution and using a mobile phase system containing 10 mM ammonium acetate, the lipid extract met the EIEIO mass spectrometry analysis conditions for various subclasses of lipids.

[0031] (2) The diagnostic system includes the following devices: the chromatographic column is an Accucore C30 column (150mm×2.1mm, 2.6μm), the separation system is a Shimadzu LC, and the detection system is an AB SCIEX Zeno TOF 7600 mass spectrometer, using positive ion mode detection;

[0032] (3) By summarizing the fragmentation patterns of various types of lipids under EIEIO positive ion modeling, the fine lipid structure of lipids in the sample is annotated.

[0033] The present invention has the following effects:

[0034] (1) By adding an appropriate amount of 0.1 mM sodium acetate during lipid reconstitution, polar and neutral lipids can be simultaneously analyzed in a single injection. (2) The automated annotation method achieved annotation accuracy of 100% and 82.3% at the sn- and C=C isomer levels, respectively. (3) A total of 1312 sn- isomers and 1033 C=C isomers were annotated from the mixed plasma of healthy individuals and Alzheimer's patients. (4) Nineteen lipid isomers were identified that can be used to distinguish between the healthy group and the early stage of Alzheimer's disease. Attached Figure Description

[0035] Figure 1 To compare different LC conditions and EIEIO-MS for high-efficiency lipid analysis based on 30 lipid standards; (A) comparison of lipid separation efficiency of different column types, showing the optimal column type for high-efficiency lipid analysis; (B) the highest ratio of sodium salt-bound TG to ammonium salt-bound TG in serum at a sodium acetate concentration of 0.1 mM; (C) optimization of accumulation time for MS1 and MS2; (D) optimization of electron beam current and Zeno trap fold to improve lipid ionization efficiency and obtain abundant secondary fragments.

[0036] Figure 2The lipid annotation method based on LC-EIEIO-MS / MS includes: (A) a flowchart of the de novo annotation method; (B) classification of unique fragment patterns based on the polar head groups of three lipids, PC, PE and SM; (C) using PC, SM and TG as examples, constructing a sn-position labeling index library using lipid sn-position diagnostic ions; and (D) double bond localization in monounsaturated and polyunsaturated lipids based on diagnostic ions.

[0037] Figure 3 The results of the lipid isomer annotation method based on EIEIO-MS / MS spectroscopy in Example 1 are shown. (A) shows the EIEIO-MS / MS spectrum of the precursor ion at m / z 760.5903, annotated as PC 16:0 / 18:1(n-9); (B) shows PC 18:1(n-9) / 16:0 at m / z 760.5896; and (C) shows PC 18:1(n-9) / 16:0 at m / z 851.7164.

[0038] TG16:1(n-7) / 18:1(n-9) / 16:1(n-7), (D) is TG at m / z 851.7214.

[0039] 18:1(n-9) / 16:1(n-7) / 16:1(n-7), (E)-(H) are the corresponding fragment diagrams used to verify the annotation results.

[0040] Figure 4 To evaluate the annotation method of the present invention using plasma samples containing 34 lipid standards, the following were included: (A) detailed information on the identification of the 34 lipid standards in EIEIO and CID modes; (B) a comparison of the accuracy of lipid annotation for chain composition, sn-, and C=C positions; and (C) a comparison of the number of lipids identified in plasma samples by the LC-EIEIO-MS / MS method and the LC-CID-MS / MS method.

[0041] Figure 5 To investigate the changes in lipid isomers associated with MCI-AD using the LC-EIEIO-MS / MS method. (A) Comparison of annotation results between LC-EIEIO-MS / MS and LC-CID-MS / MS methods; (B) Annotation results of different subclass lipids at sn-position (outer circle) and C=C position (inner circle); (C) Lipid characteristic volcano plots of the total composition of NC and MCI-AD samples; (D) Heatmaps of 19 pairs of lipid isomers used to distinguish between NC and MCI-AD. Detailed Implementation

[0042] Example 1

[0043] 1. Analytical Methods

[0044] 1.1 Processing of plasma samples

[0045] Add 300 μL of methanol to a 2 mL centrifuge tube, then add 40 μL of the intra-group serum mixture and 1 mL of methyl tert-butyl ether, and vortex for 10 min. Then add 300 μL of water, vortex for 1 min to form a two-phase system, centrifuge at 10000 g for 10 min, collect 800 μL of the supernatant, freeze-dry, and store at -80℃. Before injection, use 50 μL of acetonitrile / isopropanol / water (65:30:5, v / v / v / ) containing 0.1 mM sodium acetate. Figure 1 B) Vortex the solution for 2 minutes and perform positive ion mode detection.

[0046] 1.2 Instrument Conditions

[0047] The liquid chromatography system used was a Shimadzu LC system. The chromatographic column was an Accucore C30 column (150 mm × 2.1 mm, 2.6 μm). Figure 1 A) Column temperature: 50℃, flow rate: 0.3 mL / min. Mobile phase: 60% (v / v) aqueous solution of 10 mM ammonium acetate in acetonitrile (phase A) and 90% (v / v) isopropanol-acetonitrile solution of 10 mM ammonium acetate (phase B). Gradient: Initial gradient of 40% B, maintained for 1.5 min, then linearly increased to 45% B at 3 min, to 60% B at 11.5 min, to 80% B at 11.6 min, to 85% B at 20 min, then mobile phase B was increased to 97% within 0.5 min and maintained for 2 min. It was then decreased to 40% B within 0.1 min, and the system was equilibrated for 26 min.

[0048] Mass spectrometry conditions: AB SCIEX Zeno TOF 7600 mass spectrometer, positive ion mode. Ion source spray voltage: 5.5 kV for ESI(+); interface heater temperature: 500 °C for ESI(+); Gas 1, Gas 2, and curtain gas: ESI(+): 50, 50, 35 psi, respectively; ESI(-): 50, 50, 35 psi, respectively; mass spectrometry scan range: ESI(+): 100-1500 Da; TOF MS accumulation time: 0.1 s; in IDA mode, the top 10 most abundant precursor ions were selected for secondary acquisition; TOF MS / MS accumulation time: 0.065 s; reaction time: 60 ms (e.g., ...). Figure 1 As shown in C), the electron beam voltage is 10 eV, the electron beam current is 8000 nA, and the Zeno threshold is 400000 cps (as shown in C). Figure 1 (as shown in D).

[0049] 2. De novo annotation method based on LC-EIEIO-MS / MS

[0050] In the established high-coverage LC-EIEIO-MS / MS system, seven lipid subclasses generate abundant characteristic fragments in positive ion mode, including lysophosphatidylcholine (LPC), phosphatidylcholine (PC), phosphatidylethanolamine (PE), sphingomyelin (SM), ceramide (Cer), diacylglycerol (DG), and triacylglycerol (TG). Based on the abundant characteristic ions generated by these lipid classes, this invention develops an annotation method for efficient resolution of EIEIO spectra. This annotation method comprises six consecutive steps: 1) extracting MS information; 2) identifying lipid subclasses; 3) matching the overall composition; 4) determining sn-positions; 5) locating C=C positions; and 6) determining the annotation level. The flowchart of the annotation method in this invention is shown below. Figure 2 A.

[0051] 1) Extraction of mass spectrometry information: To obtain comprehensive structural information, MS2 data and corresponding MS1 information were directly extracted from the raw data, including the mass-to-charge ratio (m / z) and retention time (t) of the parent ion. R The data includes the values ​​of daughter ions and their relative intensities. By prioritizing the extraction of MS2 information, it can be ensured that all EIEIO spectral data are available for subsequent annotation.

[0052] 2) Identification of lipid subclasses: such as Figure 2 As shown in Figure A, lipid classes can be distinguished using the characteristic ions of the head groups generated by PC, PE, and SM. Although the head groups of protonated PC and SM produce the same fragment ion at m / z 184.07, PC also produces specific ions at m / z 224.10 and 226.08, while SM produces specific ions at m / z 225.10 and m / z 253.10. To avoid reducing annotation efficiency, lipids without obvious head group characteristic peaks, such as DG, Cer, and TG, are not annotated by class.

[0053] 3) Matching Overall Composition: Following the category identification in the previous step, we obtained precursor ions for both known and unknown categories. In the third step, we first performed overall composition matching by creating an index library including known categories (LPC, PC, PE, and SM) and unknown categories (Cer, DG, and TG). The index library was obtained by calculating the lipid m / z corresponding to different addition ion types based on the theoretical molecular masses in the Lipid Maps database. This library contains 1190 lipid molecules across 13 lipid subclasses. For lipids labeled with subclasses, the precursor ion was annotated by finding the best-matching theoretical m / z within the same subclass. For lipids in unknown subclasses, the lipid corresponding to the smallest difference between the actual m / z and the theoretical m / z in the unknown index library was annotated. Notably, in both cases, the m / z difference between the theoretical and actual precursor ions was limited to 0.01 Da to minimize false matches.

[0054] 4) Determining the sn-position: Previous studies have shown that in EIEIO, the relative intensity of the fragment ion generated by the C1-C2 breakage of the glycerol backbone of PC, PE, and DG lipids is lower than that of the fragment ions generated by the C1-O and C2-O breaks of the two acyl chains. Therefore, the fragment ion peak generated by the C1-C2 breakage can be used to determine the sn-position. For SM and Cer lipids, odd and even acyl chain fragments in EIEIO are used to solve the problem of the arrangement of the two chains on the backbone and sn-2. For TG lipids, the co-loss of the sn-1 or sn-3 and sn-2 fatty acid chains will produce a pair of doublets with a mass difference of 2.01 Da. The chain length and number of double bonds of sn-1 or sn-3 can be calculated from the doublets. At the same time, the double loss of the sn-1 and sn-3 fatty acyl groups will result in a singlet state of the sn-2 fragment ion, thereby determining the position of sn-2. Figure 2 (B) All diagnostic ions generated by the fatty acyl chains of lipid isomers other than lysophosphatidylcholine lipids are included in the index library to determine the sn-position. The distribution of acyl chains in the index library varies depending on the lipid class, with the composition of sn-2 acyl chains typically being more diverse than that of sn-1 acyl chains.

[0055] 5) Locating the C=C position: EIEIO can precisely locate the C=C isomers of various lipids without C=C derivatization. This is because in EIEIO mode, an energy of 10 eV can break carbon-carbon single bonds, while carbon-carbon double bonds are difficult to break. Therefore, a distinct "V-shape" or a mass difference of 26.02 Da can be observed at the C=C position. Figure 2 C). Typically, carbon-carbon cleavage on saturated fatty acyl chains forms a series of fragment ions consisting of CH2 (14.014 Da), resulting in a 28.03 Da mass difference between adjacent carbon-carbon single bonds, which is significantly different from the 26.02 Da mass difference observed when double bonds are present. This rule applies to both monounsaturated and polyunsaturated lipids. When analyzing polyunsaturated fatty chains, adjacent double bonds are separated by three carbons to reduce the number of double bond candidates. Based on information about C=C positions in lipid mapping databases and these rules, we established a diagnostic ion library for double bond positions at C=C positions with an unsaturation level ≤6.

[0056] 6) Determining Annotation Levels: EIEIO maps are typically generated from mixtures of lipid isomers. Therefore, ranking the annotation levels of the identified results helps in screening key lipid isomers. Lipids accurately annotated at both the C=C and sn- positions are classified as level 1. Lipids annotated only at the sn- or C=C positions are classified as level 2. Lipids inferred solely from the overall composition without characteristic fragment ions are classified as level 3.

[0057] Example of lipid isomer annotation. Taking a pair of sn-isomers, PC 16:0 / 18:1(n-9) and PC 18:1(n-9) / 16:0, as examples, the identification process of the isomers using the annotation method of this invention is illustrated. First, high abundance fragment ions of these two isomers were observed at m / z 184.072, 224.10, and 226.084, which can be identified as head groups generated from PC lipids. Second, by comparing the precise mass with the MS1 index library, the complete precursor ion can be attributed to PC 34:1 within a tolerance of 0.01 Da. Next, fragment ions at m / z 479.34 and 505.35 indicate that the two fatty acyl chains of PC have compositions of 16:0 and 18:1. Furthermore, the cleavage of the C1-C2 glycerol backbone also yielded fragment ions at m / z 491.3359 (…). Figure 3 A) and m / z465.321 ( Figure 3 Characteristic fragments in B) indicate that the sn-2 positions are 18:1 and 16:0 aliphatic chains, respectively. Subsequently, in the two EIEIO spectra, a mass difference of 26.02 Da between fragment ions at m / z 621.44 and 647.45 was observed, indicating that the double bond position is located at n-9. In summary, Figure 3 The EIEIO map in A was identified as PC 16:0 / 18:1(n-9). Figure 3 The EIEIO map in B was identified as PC18:1(n-9) / 16:0, with an identification result of Level 1. Figure 3 E and Figure 3 The above results can be confirmed by these two lipid structure diagrams of F.

[0058] Furthermore, the annotation method of this invention can also be used to identify isomers of TG-type lipids. First, after extracting the lipid information, in the second step, lipids that do not contain significant features of head groups such as TG in the EIEIO spectrum are defined as unknown subclasses. Second, based on the MS1 reference library, the total composition of precursor ions with sodium adduct ions is annotated as TG 50:3. Next, as... Figure 3As shown in C, only a pair of double peaks were observed at m / z 317.2434 / 319.2261, which can be attributed to the C16:1 acyl chains at the sn-1 and sn-3 sites. Furthermore, only a single peak was observed at m / z 331.2593, indicating that the sn-2 position in TG is a C18:1 acyl chain. Figure 3 D also showed similar results, with a single peak observed at m / z 303.2290, indicating a C16:1 fatty chain at the sn-2 position. However, two sets of doublets were observed at m / z 319.26 / 321.24 and m / z 345.26 / 347.24, indicating that the sn-1 and sn-3 positions of the TG molecule are replaced by C18:1 or C16:1 acyl chains. Since this method cannot distinguish between the sn-1 and sn-3 positions of TG, interchangeable underscores were used to denote lipids identified at the sn isomer level. Therefore, in step four... Figure 2 The EIEIO atlas annotation for C is TG 16:1 / 18:1 / 16:1. Figure 3 The EIEIO spectrum annotation in D is TG 18:1_ / 16:1 / _16:1. In the fifth step, a repeatability quality difference of 26.02 Da was observed at positions n-9 and n-7 in the EIEIO spectrum. Therefore, the final identification result is TG 16:1(n-7) / 18:1(n-9) / 16:1(n-7)( Figure 3 C) and TG 18:1(n-9)_ / 16:1(n-7) / _16:1(n-7)( Figure 3 D), the assessment level is Level 1. Figure 3 G and Figure 3 The corresponding structural diagram shown in H verifies the annotation results.

[0059] 3. Evaluation of lipid fine structure annotation methods

[0060] Seven lipid internal standards were used to evaluate the analytical characteristics of the LC-EIEIO-MS / MS method in terms of linearity, limit of detection (LOD), limit of quantitation (LOQ), recovery, and precision. The results are shown in Table 1. Within a concentration range of 4–5 orders of magnitude, the linear regression coefficient (R0) of the lipid standards was [missing information]. 2All values ​​exceeded 0.99. LOD and LOQ were determined with signal-to-noise ratios (S / N) of 3 and 10, respectively, within the ideal ranges of 1.00–13.64 and 7.69–45.5 ng / mL. Except for diacylglycerol (DG) 15:0–18:1–d7, the recoveries of the seven lipid ISs in six replicates at low, medium, and high concentrations were between 70% and 120%, which is acceptable for complex biological matrices. Intra-day RSDs ranged from 0.4% to 14.2%, and all inter-day RSDs were below 25%, meeting the requirements for lipidomics analysis. In summary, these data demonstrate that the LC-EIEIO-MS / MS method of this invention has strong analytical performance and is suitable for sample analysis.

[0061] The annotation method in this invention was evaluated. Thirty-four lipid standards were added to the intragroup mixed serum matrix, and the automated annotation method of this invention was compared with the identification results based on MS-DIAL software. Figure 4 As shown in A and 4B, this method achieved annotation rates of 100% and 82.35% for lipid sn-positions and C=C positions, respectively, significantly higher than the annotation results (32% and 2.94%) achieved using the MS-DIAL-based CID mode. Detailed information on the annotation of 34 standards using the self-developed annotation method and MS-DIAL can be found in [link to documentation]. Figure 4 C. A total of 1284 lipid molecules were identified in the matrix-spiked samples in EIEIO mode, while only 367 lipid molecules were characterized in CID mode.

[0062] Example 2

[0063] Plasma samples were collected from 32 healthy individuals as the control group (NC) and from 30 patients with mild Alzheimer's disease (MCI-AD) as the disease group. All volunteers signed informed consent forms before plasma sample collection. Collected samples were processed according to the plasma sample preparation method described in Example 1. 50 μL of supernatant was taken from each sample and mixed thoroughly to prepare quality control samples for lipid identification. The quality control samples were analyzed by LC-EIEIO-MS / MS in positive ion mode. Compared to the traditional CID method, which identified only 335 lipid molecules, the annotation method described in this invention identified 1493 lipid isomers in EIEIO mode. Figure 5 A). These isomers contain 1312 sn-position isomers and 1033 C=C position isomers, respectively. TG and PC are the two lipid subclasses with the most annotations. Figure 5 B). Detailed information on some lipid isoforms based on the annotations of this invention is shown in Table 2.

[0064] To differentiate between MCI-AD and NC, preliminary analysis of the identified lipids at the total composition level was performed in EIEIO mode, using the peak area of ​​precursor ions to quantify the lipids. Volcano plot analysis using Origin revealed significant changes in 38 lipids (p<0.05). Figure 5 C). Further, the top 10 lipids with the greatest differences were selected for relative quantification at the sn and C=C positions, yielding the relative contents of 19 lipid isomers in the two sample groups. A heatmap of the relative contents of these 19 lipid isomers was generated using the MetaboAnalyst online analysis platform, showing a significant difference between the NC group and the MCI-AD group (p<0.05). Figure 5 D).

[0065] Table 1. Results of linearity, limit of detection (LOD), limit of quantitation (LOQ), extraction recovery, intra-day and inter-day precision of the methods used in this study for seven lipid internal standards.

[0066]

[0067] Table 2. Partial annotation results of plasma lipids in Example 2 based on the EIEIO spectral annotation method of the present invention.

[0068]

Claims

1. A structural annotation method for lipid sn- and C=C positions based on liquid chromatography-organic ion electron collision excitation mass spectrometry, characterized in that: The method includes the following steps: (1) Preprocess the sample to obtain lipid extracts from the sample; (2) The lipid extract was resuspended in a sodium acetate solution containing 0.1 mM to obtain the lipid extract solution to be injected; (3) The raw data of the lipid extract in step (2) were obtained by liquid chromatography-organic ion electron collision excitation mass spectrometry. (4) Based on the raw mass spectrometry data of the obtained sample lipid extracts, the structural positions of lipid sn- and C=C were analyzed and annotated.

2. The structural annotation method according to claim 1, characterized in that: The samples in step (1) include, but are not limited to, plasma, serum, tissue, cells, yeast, microalgae, and other sample types.

3. The structural annotation method according to claim 1, characterized in that: When the sample in step (1) is plasma, serum, tissue, or cells, the pretreatment process is as follows: Add 10-100 μL of sample, 100-1000 μL of methanol, and 0.5-5 mL of methyl tert-butyl ether to a centrifuge tube, shake and mix for 5-15 min, then add 100-1000 μL of water, vortex for 1-3 min to form a two-phase system, centrifuge at 8000-15000 g for 10-30 min, and freeze-dry the supernatant to obtain the lipid extract from the sample; When the sample in step (1) is yeast or microalgae, the pretreatment process is as follows: add 10-100 μL of sample and 100-1000 μL of methanol to a centrifuge tube, grind it for 2 min at 25 Hz using a tissue homogenizer, add 0.5-5 mL of methyl tert-butyl ether, shake and mix for 5-15 min, then add 100-1000 μL of water, vortex for 1-3 min to form a two-phase system, centrifuge at 8000-15000 g for 10-30 min, and freeze-dry the supernatant to obtain the lipid extract from the sample.

4. The structural annotation method according to claim 1, characterized in that: The 0.1 mM sodium acetate solution in step (2) is a 30%-95% (v / v) aqueous solution of an organic solvent, wherein the organic solvent includes, but is not limited to, one or two of acetonitrile, isopropanol, and methanol.

5. The structural annotation method according to claim 1, characterized in that: In step (3), the liquid chromatography conditions are as follows: mobile phase A is a 30%-90% (v / v) aqueous solution of ammonium acetate in acetonitrile, and mobile phase B is a 30%-90% (v / v) isopropanol in acetonitrile solution in ammonium acetate; the separation column type includes, but is not limited to, reversed-phase columns such as C8, C18 or C30. The organic ion electron collision excitation mass spectrometry described in step (3) is performed in EAD fragmentation mode under an electron beam pressure of 10 eV, and data acquisition is carried out in positive ion mode.

6. The structural annotation method according to claim 1, characterized in that, Step (4) Annotates the structural positions of lipid sn- and C=C based on the obtained sample lipid mass spectrometry information, including the following steps: 1) Based on the raw data of the lipid extract, extract lipid spectrum information, which includes the mass-to-charge ratio of the parent ion in the primary lipid spectrum information and the mass-to-charge ratio and relative intensity of the daughter ion in the secondary lipid spectrum information; 2) Identification of lipid subclasses: Phosphatidylcholine, sphingomyelin, and phosphatidylethanolamine lipids were categorized as known categories; ceramide, diglycerides, and triglycerides lipids were not categorized and were treated as unknown categories. 3) Matching total lipid composition: Construct an index library including known and unknown categories for total composition matching; match the mass-to-charge ratio of lipid precursor ions of known and unknown categories according to the theoretical molecular mass in the index library, and annotate the total lipid composition; for lipids labeled with subclasses, the mass-to-charge ratio of precursor ions is annotated by finding the closest theoretical molecular mass in the same subclass; for lipids of unknown subclasses, the minimum difference between the mass-to-charge ratio of the actual precursor ion and the mass-to-charge ratio of the theoretical precursor ion is found in the index library of unknown categories for annotation; 4) Determine sn-position: Match characteristic fragment ions based on the established sn-position index library and annotate lipid sn-positions; for phosphatidylcholine, phosphatidylethanolamine and diglyceride lipids, the relative intensity of fragment ions generated by the breakage between C1-C2 of their glycerol backbone is lower than that of fragment ions generated by the breakage between C1-O and C2-O. Therefore, characteristic fragment ions between C1-C2 can be used to determine sn-positions; For sphingomyelin and ceramide lipids, odd and even acyl chain fragments in the organic ion electron collision excitation spectrum can be used to solve the problem of the arrangement of the two chains on the backbone and sn-2. For triglyceride lipids, the combined loss of the sn-1 or sn-3 and sn-2 fatty acid chains produces a pair of doublets with a mass difference of 2.01 Da. The chain length and number of double bonds of sn-1 or sn-3 can be calculated from the doublets. At the same time, the double loss of the sn-1 and sn-3 fatty acyl groups results in a singlet state for the sn-2 fragment ion, thus determining the position of sn-2. Except for lysophosphatidylcholine lipids, all characteristic fragment ions generated by the fatty acyl chains of other types of lipid isomers are summarized to construct an sn-position diagnostic index library for determining sn-position. 5) Locating the C=C position: When a double bond exists on the fatty acyl chain, a mass difference of 26.02 Da can be observed instead of the 28.02 Da produced by the carbon-carbon single bond, which can be used to locate the lipid double bond position; 6) Determine the annotation level: Classify the annotation results. Lipids annotated at both C=C and sn-positions are classified as Level 1; lipids annotated only at sn-positions or C=C positions are classified as Level 2; lipids not annotated at sn-positions and C=C positions but inferred only from the overall composition are classified as Level 3.

7. The structural annotation method according to claim 1, characterized in that, The lipids are classified into seven subclasses: lysophosphatidylcholine, phosphatidylcholine, phosphatidylethanolamine, sphingomyelin, ceramide, diglycerides, and triglycerides.

Images