Lipid detection method, detection kit and detection system for biological sample

Abstract

The invention provides a lipid detection method, a detection kit and a detection system for a biological sample, in the detection method, at least one first internal standard and one control lipid compound are determined for each to-be-detected lipid subclass, and for each control lipid compound, at least one second internal standard and one control lipid compound are determined for each to-be-detected lipid subclass. And selecting a second interior label from the first interior labels and confirming a corresponding relative correction factor. And detecting the sample to be detected by using the liquid chromatography tandem mass spectrometry, searching quantitative control from the control lipid compounds of the lipid subclass to which the lipid compounds belong for each lipid compound in the sample, and calculating the content by using the second internal standard and the relative correction factor. The method is simple to operate and low in detection cost, can accurately determine the content of the lipid compound with a reference substance in a sample, can also provide a more accurate relative content result for the compound without the reference substance, can be used for detecting a large batch of samples, and provides a better test basis for lipidomics research.

Description

Technical Field

[0001] The present invention belongs to the technical field of metabolomics, specifically lipidomics, and relates to a lipid detection method, a detection kit and a detection system for biological samples. Background Art

[0002] Lipids refer to a large class of compounds that are easily soluble in organic solvents and are heterogeneous in chemical composition and structure. They mainly include fatty acids and their naturally occurring derivatives (such as esters or amines), as well as compounds related to their biosynthesis and function. Studies have shown that mammalian cells contain 1,000 to 2,000 types of lipids, and with the continuous development of new technologies and methods, various new lipid molecules are constantly being discovered. The diversity of lipid structures gives lipids a variety of important biological functions. Lipids are not only involved in regulating a variety of life activities, including energy conversion, material transport, information recognition and transmission, cell development and differentiation, and cell apoptosis, but abnormal lipid metabolism is also closely related to certain diseases, such as arteriosclerosis, diabetes, obesity, Alzheimer's disease, and the occurrence and development of tumors.

[0003] Lipidomics, a key branch of metabolomics, studies the structure, function, and metabolic pathways of lipids in organisms, tissues, cells, or body fluids. Lipidomics can comprehensively investigate the relationship between abnormal lipid metabolism and disease, conduct high-throughput, holistic lipid analysis, and identify early, characteristic biomarkers associated with disease.

[0004] Mass spectrometry is currently the most commonly used detection method for lipidomics analysis, characterized by high throughput and high sensitivity. In practical applications, chromatography is generally combined with mass spectrometry, which not only helps to separate lipid compounds but also reduces matrix interference. However, due to the wide variety of lipids and their complex and diverse structures, it is difficult to obtain quantitative data for different types of lipid compounds in a single step using mass spectrometry. In addition, the difficulty in obtaining a large number of lipid reference substances is also a major challenge facing lipid quantitative detection. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, in order to achieve simultaneous quantitative (including semi-quantitative) detection of more lipid compounds, the present invention provides a lipid detection method, detection kit and detection system for biological samples, which can obtain quantitative data of a large number of lipid compounds in biological samples without the need for reference substances.

[0006] In one aspect, the present application provides a method for detecting lipids in a biological sample, the method comprising:

[0007] For each lipid subclass to be tested, at least one first internal standard and one reference lipid compound are determined; for each reference lipid compound, a second internal standard is selected from the first internal standard of the lipid subclass to which it belongs, and a relative correction factor of the reference lipid compound relative to the second internal standard is determined under liquid chromatography tandem mass spectrometry detection conditions;

[0008] adding each second internal standard to the sample to be tested to prepare a sample solution, and detecting the sample to be tested under the liquid chromatography tandem mass spectrometry detection conditions to determine each lipid compound to be tested contained in the sample to be tested;

[0009] For each lipid compound to be tested, one control lipid compound of the lipid subclass to which it belongs is selected as a quantitative control, and the second internal standard and relative correction factor of the quantitative control are used as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested, and the content of the lipid compound to be tested is calculated by the internal standard method.

[0010] In a second aspect, the present application provides a lipid detection kit for a biological sample, the kit comprising: a first internal standard for each lipid subclass to be detected;

[0011] The lipid subclass to be tested includes at least one of the following categories: phosphatidylethanolamine (PE), lysophosphatidylethanolamine (LPE), ceramide (Cer), phosphatidylcholine (PC), lysophosphatidylcholine (LPC), diglyceride (DG), cholesterol ester (CE), triglyceride (TG), sphingomyelin (SM), monoglyceride (MG), phosphatidylinositol (PI), lysophosphatidylinositol (LPI), phosphatidylglycerol (PG), lysophosphatidylglycerol (LPG), fatty acids (FA), phosphatidylserine (PS), lysophosphatidylserine (LPS), phosphatidic acid (PA);

[0012] For each lipid subclass to be tested, the first internal standard included is as follows:

[0013] Serial number Lipid subclasses First internal standard 1 PE 15:0-18:1-d7-PE 2 LPE 18:1-d7LPE 3 Cer Cer(d18:1(d7) / 18:0) 4 PC 16:0PC-d9 5 LPC 18:1-d7LPC 10 MG 18:1-d7MG 6 DG 1,3-17:0d5DG 8 TG 15:0-18:1-d7-15:0TG 7 CE 18:1Chol(d7)ester 9 SM SM(d18:1 / 15:0)-d9 11 PI 15:0-18:1-d7-PI 12 LPI 17:0LPI-d5 13 PG 15:0-18:1-d7-PG 14 LPG 17:0LPG-d5 15 FA C17:0

[0014] Preferably, the kit further comprises: an extraction solvent, wherein the extraction solvent consists of methyl tert-butyl ether, methanol and water, more preferably, the volume ratio of methyl tert-butyl ether, methanol and water is 2-6:0.5-1.5:0.4-1.6, and the optimal volume ratio is 4:1:0.8;

[0015] Preferably, lipid detection is performed using the kit according to the above-mentioned detection method.

[0016] In a third aspect, the present application provides a lipid detection system for a biological sample, the detection system comprising: a liquid chromatography tandem mass spectrometry module, a mass spectrometry data processing module, a storage module, a comparison module, and a calculation module;

[0017] The liquid chromatography tandem mass spectrometry module is used to detect the sample solution added with each second internal standard and obtain mass spectrometry data; the mass spectrometry condition setting of the liquid chromatography tandem mass spectrometry module includes:

[0018] Scanning mode: positive and negative ion switching scanning;

[0019] Scan type: Full scan / data-dependent secondary scan (Full MS / ddMS2);

[0020] A mass spectrometry data processing module processes the mass spectrometry data and annotates the first internal standard, each lipid compound to be tested, and the lipid subclass to which it belongs in the sample to be tested;

[0021] A storage module, for storing each lipid subclass to be measured, a reference lipid compound, a second internal standard and a relative correction factor;

[0022] A comparison module compares each lipid compound to be tested with each control lipid compound of the lipid subclass to which it belongs, and selects the second internal standard and relative correction factor of the control lipid compound with the highest similarity as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested;

[0023] The calculation module calculates the content of each lipid compound to be tested using the internal standard method.

[0024] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0025] (1) For each lipid subclass, a respective internal standard and reference lipid compound are selected to obtain relative correction factors for the reference lipid compound and the corresponding internal standard. A reference lipid compound is selected for each lipid compound to be tested, and then the quantitative internal standard and quantitative correction factor for each lipid compound to be tested are determined. Quantification is performed using the internal standard method. In this way, quantitative data can be obtained even without a reference substance for each lipid compound to be tested. This quantitative result is more accurate than that of the existing general internal standard method.

[0026] (2) It is used for the simultaneous qualitative and quantitative detection of a large number of lipid compounds of different lipid categories, and can obtain qualitative and quantitative data of a large number of lipid compounds in biological samples.

[0027] (3) The method is simple to operate, with high accuracy, repeatability and sensitivity, low detection cost, and is suitable for high-throughput detection. It can be used for the detection and analysis of a large number of biological samples, and is particularly suitable for lipidomics experiments on large batches of samples. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 A schematic diagram of a process for lipid detection in a biological sample provided in an embodiment of the present application;

[0029] Figure 2 A schematic diagram of the structure of a lipid detection system for biological samples provided in an embodiment of the present application;

[0030] Figure 3 Scheme 5: Total ion current in positive ion mode under mobile phase;

[0031] Figure 4 Scheme 5: Total ion current in negative ion mode under mobile phase;

[0032] Figure 5 Scheme 6 Total ion current in positive ion mode under mobile phase;

[0033] Figure 6 Scheme 6: Total ion current in negative ion mode under mobile phase;

[0034] Figure 7 AUC graph of compound similarity evaluation model;

[0035] Figure 8 Total ion current of serum sample solution in positive ion mode;

[0036] Figure 9 Total ion current of serum sample solution in positive ion mode;

[0037] Figure 10 Total ion current of tissue sample solution in positive ion mode;

[0038] Figure 11 Total ion chromatogram of tissue sample solution in negative ion mode. DETAILED DESCRIPTION

[0039] The technical solution of the present application is further described below by way of specific embodiments. Those skilled in the art should understand that the embodiments are merely helpful for understanding the present invention and should not be considered as specific limitations of the present application. Unless otherwise specified, the technical terms in this application have the general meanings in the art.

[0040] The terms "first," "second," "third," "fourth," "1," "2," and the like (if any) in the specification, claims, and drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence, nor do they indicate a difference between the two. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the application described herein can be implemented in an order other than that shown or described in the drawings. "Each" means "one," "each," or "a plurality" and can be in the singular or plural.

[0041] A quantitative internal standard refers to an internal standard used for quantitative detection of a compound or for content calculation. A "quantitative correction factor" refers to a relative correction factor used for quantitative detection of a compound. It corresponds to the quantitative internal standard and the quantitative control. Generally, the relative correction factor is calculated by detecting one or more known concentrations of the quantitative control and the quantitative internal standard using a single-point method, a multi-point method, or a standard curve method. When detecting the content of a compound in a sample to be tested, a quantitative internal standard is added to the sample to be tested, the response of the compound and the quantitative internal standard is detected, and the content of the compound is calculated using the internal standard method. The formula for the internal standard method is as follows:

[0042] Ci=Pi / Pr*Cr*RFi Formula (1)

[0043] Where Ci represents the content or concentration of compound i, Cr represents the content or concentration of the quantitative internal standard in the sample to be tested, Pi represents the response of compound i, Pr represents the response of the quantitative internal standard, and RFi represents the quantitative correction factor of compound i. "Response" refers to the detector's response to the sample or compound contained therein after the sample enters the instrument for detection. In this application, it generally refers to the mass spectrometry response. The compound response can be represented by the peak height or peak area of ​​an ion or ion pair generated by the compound ionization in the mass spectrometry data.

[0044] Mass spectrometry data refers to data obtained through detection by a mass spectrometer or deduction by mass spectrometry fragmentation rules, which includes primary mass spectrometry data and / or secondary mass spectrometry data, and can be obtained through detection or existing public data. When a mass spectrometer detects a sample to be tested, sample mass spectrometry data is obtained, including mass spectrometry data of each component in the sample, and the mass spectrometry data of each compound can be extracted from the sample mass spectrometry data. The primary mass spectrometry data of a compound includes the parent ion (also known as the precursor ion), adduct ion and intensity and other related information of the compound, and the secondary mass spectrometry data includes fragment ions (also known as daughter ions) and intensity and other related information. Based on this information, the compound can be identified or annotated. By extracting each ion in the mass spectrometry data, a chromatographic peak with time as the horizontal axis and intensity as the vertical axis can be obtained, and the peak height or peak area can be obtained.

[0045] Lipids are a general term for esters and their derivatives formed by the reaction of fatty acids and alcohols. According to the lipid classification system proposed by the Lipid Metabolites and Pathways Strategy (LIPID MAPS) project funded by the National Institutes of Health (NIH) in 2003, lipids are generally divided into eight major categories: fatty acids, glycerolipids, glycerophospholipids, sphingolipids, sterol lipids, prenol lipids, saccharolipids, and polyketides. However, if lipids are divided into eight major categories, the structures of the individual lipid compounds in each major lipid class still vary greatly. If the individual lipid compounds in each major lipid class share the same internal standard, some lipid compounds will differ too much from the internal standard, resulting in inaccurate calibration results. In order to better ensure quantitative accuracy, lipids are divided into lipid subclasses with more similar structures, such as the lipid subclasses in the table below:

[0046] Table 1 Lipid subclass information

[0047] Serial number Chinese name English name Abbreviation 1 Phosphatidylethanolamine phosphatidylethanolamine PE 2 Lysophosphatidylethanolamine lysophosphatidylethanolamine LPE 3 Ceramide ceramide Cer 4 Phosphatidylcholine phosphatidylcholine PC 5 Lysophosphatidylcholine lysophosphatidylcholine LPC 6 Monoglycerides monoacylglycerol MG 7 diglycerides diacylglycerol DG 8 triglycerides triacylglycerol TG 9 Cholesterol ester cholesterylester CE 10 Sphingomyelin sphingomyelin SM 11 Phosphatidylinositol phosphatidylinositol PI 12 Lysophosphatidylinositol lysophosphatidylinositol LPI 13 Phosphatidylglycerol phosphatidylglycerol PG 14 Lysophosphatidylglycerol lysophosphatidylglycerol LPG 15 fatty acids Fattyacid FA 16 Phosphatidylserine phosphatidylserine PS 17 Lysophosphatidylserine Lysophosphatidylserine LPS 18 Phosphatidic acid Phosphatidicacid PA

[0048] First, the lipid detection method, detection kit, and detection system for biological samples provided in this application are intended to simultaneously quantify various lipid compounds in biological samples and obtain more accurate quantitative results for a wider range of lipid compounds.

[0049] The following describes the technical solution of the present application and the technical effects produced by the technical solution of the present application through several exemplary embodiments. It should be noted that the following embodiments can refer to, draw on, or combine with each other, and the same terms, similar features, and similar implementation steps in different embodiments will not be repeated.

[0050] The instruments and reagents used in this application can be obtained through commercial channels and are not limited or specifically introduced here. The main internal standard and control lipid compounds in the examples of this application are summarized in the following table.

[0051] Table 2 Summary of internal standard and reference lipid compounds

[0052]

[0053]

[0054]

[0055] *: Internal standard.

[0056] The present invention provides a method for detecting lipids in biological samples. Figure 1 , the method includes S1 to S4.

[0057] S1. For each lipid subclass to be tested, determine at least one first internal standard and one reference lipid compound.

[0058] For each lipid subclass to be tested, a respective first internal standard and control lipid compound are selected. For a lipid subclass, the first internal standard generally belongs to that lipid subclass and, therefore, has a similar molecular structure and can serve as an internal standard for the quantification of individual lipid compounds within that lipid subclass. For example, the internal standard can be an isotopic internal standard or a lipid compound of that lipid subclass that is not present in the biological sample. For a lipid subclass, the control lipid compound belongs to that lipid subclass and is selected from available control lipid compounds. This can be a lipid compound present in the biological sample to be tested or the target of detection, or a lipid compound not present in the sample to be tested or the target of detection.

[0059] The lipid subclass to be tested can be determined according to the test requirements and can be one lipid subclass or multiple lipid subclasses. In some embodiments, the lipid subclass to be tested includes one or more of the following: PE, LPE, Cer, PC, LPC, DG, CE, TG, SM, MG, PI, LPI, PG, LPG, FA, PS, LPS, and PA.

[0060] In some embodiments, the lipid subclasses to be tested include: PE, LPE, Cer, PC, LPC, DG, CE, TG, SM, MG, PI, LPI, PG, LPG, and FA.

[0061] In some embodiments, the lipid subclasses to be detected further include: PS, LPS, and PA.

[0062] In some embodiments, for each lipid subclass, the corresponding first internal standard includes the following table:

[0063] Table 3 The first internal standard included in each lipid subclass

[0064] Serial number Lipid subclasses First internal standard 1 PE 15:0-18:1-d7-PE 2 LPE 18:1-d7LPE 3 Cer Cer(d18:1(d7) / 18:0) 4 PC 16:0PC-d9 5 LPC 18:1-d7LPC 10 MG 18:1-d7MG 6 DG 1,3-17:0d5DG 8 TG 15:0-18:1-d7-15:0TG 7 CE 18:1Chol(d7)ester 9 SM SM(d18:1 / 15:0)-d9 11 PI 15:0-18:1-d7-PI 12 LPI 17:0LPI-d5 13 PG 15:0-18:1-d7-PG 14 LPG 17:0LPG-d5 15 FA C17:0 16 PS 15:0-18:1-d7-PS 17 LPS 17:0LPS-d5 18 PA 15:0-18:1-d7-PA

[0065] When it is necessary to detect a corresponding lipid subclass or lipid compounds within a lipid subclass, the corresponding internal standard can be selected according to the table above. The first internal standards for these lipid subclasses in the table above are not randomly selected. The inventors conducted extensive experiments to study the relative calibration factors of the internal standards and the lipid compounds within the lipid subclass to which they belong, as well as the stability of the internal standards. Ultimately, an internal standard with a relative calibration factor within the range of 0.1-10 for most target lipid compounds within the lipid subclass to which it belongs and good stability was obtained as the first internal standard.

[0066] S2. For each reference lipid compound, select a second internal standard from the first internal standard of the lipid subclass to which it belongs, and determine the relative correction factor of the reference lipid compound relative to the second internal standard under liquid chromatography tandem mass spectrometry detection conditions.

[0067] When there are multiple first internal standards for each lipid subclass, one that is more similar in structure to the reference lipid compound and whose relative correction factor is closer to 1 is generally selected from the first internal standards as the second internal standard.

[0068] It should be noted that in some embodiments, isotopic internal standards are relatively expensive. To account for this cost, a single primary internal standard is selected for each lipid subclass, and the secondary internal standard is the same as the primary internal standard. For each lipid subclass, the relative calibration factor between each reference lipid compound and its secondary internal standard must be within the range of 0.1-10, with the closest to 1 being optimal. After extensive experimentation, we unexpectedly obtained the secondary internal standard and reference lipid compounds shown in the table below. Under liquid chromatography-tandem mass spectrometry detection conditions, the relative calibration factors between each reference lipid compound and the secondary internal standard are within the range of 0.1-10.

[0069] Table 4 First internal standard and reference lipid compounds for each lipid subclass

[0070]

[0071]

[0072] The relative correction factor can be obtained by preparing a reference lipid compound and a second internal standard into mixed reference solutions of various concentrations, detecting them using a liquid chromatography tandem mass spectrometer, and calculating the relative correction factor using a single-point method, a multi-point method, or a standard curve method.

[0073] In some embodiments, under liquid chromatography tandem mass spectrometry detection conditions, the reference lipid compounds in each lipid subclass, their second internal standards, and relative correction factors are obtained by the standard curve method as shown in the following table.

[0074] Table 5 Reference lipid compounds in each lipid subclass and their second internal standards and relative correction factors

[0075]

[0076]

[0077] Liquid chromatography tandem mass spectrometry was used here to simultaneously acquire mass spectrometric data for a large number of lipid compounds.

[0078] In some embodiments, the liquid chromatography conditions of the liquid chromatography tandem mass spectrometer include: using an XBridgeBEH C18 2.1×150 mm, 2.5 μm chromatographic column or other equivalent chromatographic column, sampling 10 mM ammonium formate in acetonitrile water (mobile phase A) and 10 mM ammonium formate in isopropanol acetonitrile (mobile phase B) for gradient elution.

[0079] More preferably, the gradient elution procedure is as follows:

[0080] Table 6 Gradient elution program

[0081]

[0082]

[0083] The liquid chromatography conditions are suitable for the simultaneous detection of the 18 lipid subclasses listed in the table above, so that the lipid compounds in each lipid subclass have better retention time and chromatographic peak shape, good separation effect, and are easier to ionize under the ESI ion source, resulting in higher mass spectrometric response.

[0084] In some embodiments, the parameter settings of the mass spectrometry conditions of the liquid chromatography tandem mass spectrometry include: using positive and negative ion switching scanning, and the scan type is full scan / data dependent secondary scan (Full MS / ddMS2). Different lipids have different ionization characteristics. Some lipids easily form charged ions in ESI positive ion mode, such as PE, LPE, PC, LPC, SM, MG, DG, TG, CE, etc., while some lipids are more likely to form charged ions in ESI negative ion mode, such as PG, LPG, PI, LPI, FA, PS, LPS, PA, etc. In order to detect more lipids simultaneously, it is necessary to select positive and negative ion switching scanning or inject two needles and scan with positive and negative ions respectively. Positive and negative ion switching scanning is preferred. In this way, a sample solution can be injected once to obtain mass spectrometry data in both positive and negative modes at the same time. Full scan / data-dependent secondary scan can obtain secondary mass spectrometry data while performing a full scan to obtain primary mass spectrometry data. In this way, primary and secondary mass spectrometry data can be used to better identify substances and ensure the accuracy of the identification results. At the same time, primary mass spectrometry data can be used for quantitative calculations, and qualitative and quantitative analysis can be achieved in one test.

[0085] In some embodiments, the parameter setting of the mass spectrometry conditions of the liquid chromatography tandem mass spectrometry also includes: setting 3 or 4 HCD collision energies between 15% and 50%, preferably, the HCD collision energy is set to 15%, 25%, 30%, and 40%. The HCD collision energy will affect the number or abundance of fragmented ions of each lipid compound. In order to obtain more fragment ions, different lipids have different requirements for HCD collision energy. For example, experiments have found that FA is difficult to break up and obtain characteristic fragment ions at 15% HCD collision energy, while PE is completely broken down at 90% HCD collision energy, making it difficult to find characteristic fragment ions. It is difficult to obtain characteristic fragment ions of the substance if it cannot be broken or is broken too much, resulting in inaccurate identification results. Therefore, in order to simultaneously detect multiple lipids, it has been found through a large number of experiments that selecting 3 or 4 HCD collision energies with a certain gap between 15% and 50% can better obtain characteristic secondary mass spectrometry data of various lipid compounds. The optimal HCD collision energy is 15%, 25%, 30%, and 40%. At this time, characteristic secondary mass spectrometry data (i.e., containing molecular ions and fragment ions, and the molecular ion peak intensity is less than 1 / 2 of the highest fragment ion) can be obtained for each lipid compound of the 18 lipid subclasses in the above table, thereby making the qualitative results more accurate.

[0086] S3. Add each second internal standard to the biological sample to be tested to prepare a sample solution, and detect the biological sample to be tested under liquid chromatography tandem mass spectrometry detection conditions to determine each lipid compound to be tested contained in the biological sample to be tested.

[0087] The present application detects and calculates the content of each lipid compound in the biological sample to be tested by the internal standard method. A second internal standard is added to the biological sample to be tested, and a sample solution is prepared. The biological sample to be tested is detected under liquid chromatography tandem mass spectrometry detection conditions, and then the sample mass spectrometry data can be obtained. The sample mass spectrometry data is analyzed to identify the lipid compound to be tested contained in the biological sample to be tested. Here, the primary mass spectrometry data and / or secondary mass spectrometry data of the compound detected in the sample can be matched with the mass spectrometry data of the compound in a commercial database or a self-built database, and the one with the greatest similarity is selected to annotate and mark the lipid subclass to which it belongs.

[0088] S4. For each lipid compound to be tested in the biological sample to be tested, select one of the control lipid compounds of the lipid subclass as a quantitative control, use the second internal standard of the quantitative control and the relative correction factor as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested, and calculate the content of the lipid compound to be tested by the internal standard method.

[0089] When a control lipid compound identical to the lipid compound to be measured is present, it serves as the quantitative control, and the corresponding second internal standard serves as the quantitative internal standard. The relative correction factor calculated above between the corresponding control lipid compound and the second internal standard serves as the quantitative correction factor for the lipid compound to be measured. The content is calculated using the mass spectrometric response of the lipid compound to be measured, the mass spectrometric response of the quantitative internal standard, and the quantitative correction factor using the internal standard method, using the same formula (1).

[0090] When there is no control lipid compound identical to the lipid compound to be tested, the lipid compound to be tested is compared with each control lipid compound of the lipid subclass to which it belongs, and the control lipid compound with the highest similarity is selected as the quantitative control. The second internal standard of the quantitative control and the relative correction factor are used as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested, and the content of the lipid compound to be tested is calculated by the internal standard method as above.

[0091] For example, suppose the lipid compound to be measured is 14:0PE, which belongs to the PE class. Assume the internal standard for PE is 15:0-18:1-d7-PE, and the reference lipid compounds for PE are 14:0PE, 16:0PE, 18:0PE, and 18:0-20:4PE. Therefore, the internal standard for 14:0PE is 15:0-18:1-d7-PE, and the control for 14:0PE is 14:0PE. Assume the lipid compound to be measured is 18:1PE, which belongs to the PE class. Therefore, 18:0PE, which has the highest structural similarity to 18:1PE, is selected as the control for quantification, and 15:0-18:1-d7-PE is the internal standard for quantification. The relative correction factor between 18:0PE and 15:0-18:1-d7-PE is the quantification correction factor for 18:1PE.

[0092] For a test lipid compound, when determining which reference lipid compound of its lipid subclass is more similar to it, the structures of the two compounds can be compared and the one with the most similar chemical structure can be selected. In some embodiments, a fingerprint spectrum of the compound can be constructed, and the chemical structure similarity can be determined by comparing the similarity of the fingerprint spectrum.

[0093] In some embodiments, the mass spectrometry data of the lipid compound to be tested and each control lipid compound, such as the primary mass spectrometry data and / or the secondary mass spectrometry data, can be compared, and the similarity can be calculated using a similarity algorithm such as cosine similarity and Jaccard similarity, and the one with the highest similarity can be selected.

[0094] In some embodiments, comparing the lipid compound to be tested with each control lipid compound of the lipid subclass to which it belongs, and selecting the control lipid compound with the greatest similarity as a quantitative control, comprises:

[0095] Calculating structural similarity data and mass spectrometry data similarity data between the lipid compound to be tested and each reference lipid compound of the lipid subclass to which it belongs;

[0096] The structural similarity data and mass spectrometry data similarity data are input into a pre-built compound similarity evaluation model, a similarity evaluation result is output, and a control lipid compound with the highest similarity is selected as a quantitative control for the lipid compound to be tested.

[0097] Chemical structure similarity data refers to data that reflects the degree of similarity between the chemical structures of two compounds, obtained by comparing or calculating the chemical structures of the two compounds. For example, based on the chemical structure, the two compounds are decomposed to obtain the atomic groups (or atoms) and the corresponding number of each compound; the quotient of the number of each atomic group of the two compounds is calculated by non-directional division, and the quotient value is 0 to 1. For example, aspartic acid and asparagine are decomposed into each atomic group and the corresponding number according to the molecular structure, as shown in the following table:

[0098] Table 7 Comparison of chemical structure of aspartic acid and asparagine

[0099] name -COOH -NH2 -CH- -CH2- -CO- Aspartic acid 2 1 1 1 0 Asparagine 1 2 1 1 1

[0100] The quotient of the number of atomic groups of the two compounds is calculated by non-directional division, and the structural similarity data is obtained as shown in the following table:

[0101] Table 8 Structural similarity data of aspartic acid and asparagine

[0102] name -COOH -NH2 -CH- -CH2- -CO- Aspartic Acid vs Asparagine 0.5 0.5 1 1 0

[0103] Directionless division refers to dividing the values ​​of the same variable in two sets of data, where the numerical values ​​of the numerator and denominator are not fixed. Here, the two sets of data are the atomic groups (variables) and corresponding numbers (values) of the two compounds, such as the atomic groups and corresponding numbers of aspartic acid and asparagine in Table 1. When performing directionless division on the molecular structure data of aspartic acid and asparagine, for each atomic group, the one with the larger number of corresponding numbers is used as the denominator, and the one with the smaller number of corresponding numbers is used as the numerator, so that the quotient is between 0 and 1. For example, the quotient of -COOH is 1 / 2 = 0.5, and the quotient of -NH2 is 1 / 2 = 0.5. For atomic groups with a corresponding number of 0, the quotient is 0.

[0104] Mass spectral data similarity data refers to data obtained after similarity calculation of the mass spectral data of two compounds that can reflect the degree of similarity between the two mass spectral data, including one or more similarity data such as cosine similarity, Jaccard similarity, spectral entropy, etc.

[0105] In one embodiment, the biological sample to be tested is serum. To extract various lipid compounds in the serum, preferably, the sample solution is prepared by adding internal standards corresponding to the respective lipid subclasses to be tested to the biological sample to be tested, specifically comprising: taking a serum sample, adding the respective second internal standards; extracting at least once with a first extraction solvent, concentrating the extract, and reconstituted with a reconstitution solvent to obtain a sample solution.

[0106] Preferably, the first extraction solvent consists of methyl tert-butyl ether, methanol and water. More preferably, the volume ratio of methyl tert-butyl ether, methanol and water is 2-6:0.5-1.5:0.4-1.6, and the optimal volume ratio is 4:1:0.8.

[0107] In some embodiments, the resolvent consists of isopropanol and methanol. More preferably, the volume ratio of isopropanol to methanol is 1:1.

[0108] The first extraction solvent and the resolvent are obtained through a large number of experiments and optimization, and can be used for target

[0109] The present invention provides a lipid detection kit for biological samples, which includes the first internal standard or the second internal standard for each lipid subclass to be detected. These internal standards can be added to the sample to be tested, and lipid detection can be performed according to the lipid detection method described above.

[0110] In some embodiments, the kit further comprises: an extraction solvent, wherein the extraction solvent consists of methyl tert-butyl ether, methanol and water. More preferably, the volume ratio of methyl tert-butyl ether, methanol and water is 2-6:0.5-1.5:0.4-1.6, and the optimal ratio is 4:1:0.8.

[0111] The kit includes the reagents required for lipid detection, and the reagents are stable and can be stably stored at -20°C for more than one year. The use of this kit can greatly simplify the sample preparation process, reduce experimental costs, and facilitate the detection of large quantities of samples and commercial production.

[0112] like Figure 2 , an embodiment of the present application provides a lipid detection system for a biological sample, the detection system comprising: a liquid chromatography tandem mass spectrometry module, a mass spectrometry data processing module, a storage module, a comparison module, and a calculation module;

[0113] The liquid chromatography tandem mass spectrometry module is used to detect the sample solution added with each second internal standard and obtain mass spectrometry data; the mass spectrometry condition setting of the liquid chromatography tandem mass spectrometry module includes:

[0114] Scanning mode: positive and negative ion switching scanning;

[0115] Scan type: Full scan / data-dependent secondary scan (Full MS / ddMS2);

[0116] A mass spectrometry data processing module processes the mass spectrometry data and annotates the first internal standard, each lipid compound to be tested, and the lipid subclass to which it belongs in the sample to be tested;

[0117] A storage module, for storing each lipid subclass, a reference lipid compound, a second internal standard, and a relative correction factor;

[0118] A comparison module compares each lipid compound to be tested with each control lipid compound of the lipid subclass to which it belongs, and selects the second internal standard and relative correction factor of the control lipid compound with the highest similarity as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested;

[0119] The calculation module calculates the content of each lipid compound to be tested using the internal standard method.

[0120] This lipid detection system is used to perform the aforementioned detection method, including liquid chromatography-tandem mass spectrometry detection and mass spectrometry data processing, quantitative internal standardization, determination of quantitative correction factors, and obtaining the final content results. The prepared sample solution is placed in the system for testing, and the final result is directly obtained. The high degree of automation further simplifies the operation process, facilitates cost reduction, and facilitates large-scale testing.

[0121] In some embodiments, the comparison module includes the compound similarity evaluation model. The structural similarity data and mass spectrometry data similarity data of each lipid compound to be tested and each control lipid compound of the lipid subclass to which it belongs are input into the model to obtain a similarity score, and the control lipid compound with the highest score is obtained as a quantitative control for the lipid compound to be tested.

[0122] The lipid detection method for biological samples provided in the embodiments of the present application is described below with reference to preferred examples.

[0123] Example 1

[0124] Determine reference lipid compounds, internal standards, and relative correction factors for each lipid subclass

[0125] Based on available literature, public databases, and laboratory test data, we selected 18 lipid subclasses and multiple lipid compounds commonly found in animal samples. Specifically, the lipid subclasses included: PE, LPE, Cer, PC, LPC, DG, CE, TG, SM, MG, PI, LPI, PG, LPG, FA, PS, LPS, and PA. Based on the availability of lipid compounds, we purchased control substances to preliminarily identify control lipid compounds. For each lipid subclass, we selected at least one internal standard.

[0126] Each control lipid compound is mixed with an internal standard to prepare a mixed control solution, which is then subjected to LC-MS detection. The LC-MS is optimized to obtain LC-MS conditions that enable simultaneous detection of each compound.

[0127] 1. Study on detection conditions of liquid chromatography tandem mass spectrometry

[0128] To simultaneously detect various lipid compounds, we optimized the liquid chromatography-tandem mass spectrometry detection conditions, including liquid chromatography separation parameters such as mobile phase, elution gradient, flow rate, and column temperature, as well as mass spectrometer parameters such as ion voltage, gas flow rates, column temperature, and HCD collision energy. The mobile phase and HCD collision energy have a greater impact on the detection results, as shown below.

[0129] (1) Mobile phase investigation

[0130] Table 9 List of mobile phases under each scheme

[0131]

[0132] Take the mixed control solution and test it according to the mobile phase of schemes 1-6 respectively, and optimize the chromatographic parameters such as flow rate.

[0133] The experiment found that: Scheme 1 could not detect the lipid compounds in FA. Scheme 2 could detect FA, but the chromatographic peak shape was poor and the mass spectrometry responses of CE, TG, etc. were low. Scheme 3 could detect FA, but the chromatographic peak shape was severely tailing. Scheme 4 could not detect CE-type compounds. Under Schemes 5 and 6, both the internal standard and the reference lipid compounds could be detected, and the peak shape was better and the mass spectrometry response was higher. The effect under Scheme 6 was better, as shown in the following figure. Figure 3-Figure 4 These are the total ion currents of positive and negative ions under Scheme 5, Figure 5-Figure 6 The total ion currents of positive and negative ions are shown respectively under the conditions of Scheme 6. Therefore, the mobile phases of Schemes 5 and 6 can be selected, with Scheme 5 being preferred, for the simultaneous detection of 18 lipid subclasses.

[0134] (2) HCD collision energy

[0135] Different HCD collision energies were set, and mass spectrometry detection was performed on 18 lipid subclasses respectively to check the fragmentation of each lipid compound under different HCD collision energies.

[0136] First, individual lipid compounds were tested at different HCD collision energies using a syringe pump to determine their optimal HCD collision energy range. Further optimization was then performed to determine the optimal HCD collision energy. The experiments revealed that lipid compounds in FA, CE, and MG were difficult to fragment, while those in PE and Cer were easily fragmented, being completely fragmented at lower HCD collision energies.

[0137] Different HCD collision energies were further set to test the mixed control solution. The HCD collision energies were: ① 25%, 30%; ② 15%, 25%; ③ 30%, 40%; ④ 15%, 30%, 90%; and ⑤ 15%, 35%, 65%; ⑥ 15%, 25%, 30%, 40%. The results showed that: ① and ② conditions failed to yield rich fragment ion information for FA compounds; ③ conditions made it difficult to identify characteristic fragment ions for PE compounds; ④, ⑤, and ⑥ conditions all yielded molecular ions and characteristic fragment ions for various lipid compounds, with ⑥ providing the best results. In addition to obtaining molecular ions and characteristic fragment ions for each compound, the peak intensity of the molecular ion was less than 1 / 2 of the highest fragment ion, providing richer fragment ion information and more accurate identification results.

[0138] In addition, the present application adopts full scan / data-dependent secondary scan (Full MS / ddMS2). After the mass spectrometer performs a full scan, the parent ion (intensity-dependent) list is selected from the full scan mass spectrum (primary mass spectrometry data) for secondary scan to obtain secondary mass spectrometry data. That is, when performing the secondary scan, only compounds with high parent ion response can be scanned to obtain secondary mass spectrometry information. The mobile phase of the above scheme 6 better ensures the primary mass spectrometry response of the 18 lipid compounds to be tested, and can obtain a more intense parent ion, which is then scanned by the secondary scan. Combined with the HCD collision energy under the conditions of ⑥, rich secondary mass spectrometry information is obtained, making it suitable for simultaneous qualitative and quantitative detection of 18 lipid compounds.

[0139] The optimal LC-MS detection conditions finally obtained are as follows:

[0140] An ultra-high performance liquid chromatography (manufacturer: Thermo, model: Vanquish) was coupled with a mass spectrometer (manufacturer: Thermo, model: OE 120).

[0141] Liquid phase parameters

[0142] Chromatographic column: Waters BEH C18, 1.7 μm, 2.1*100 mm, Part No.: 186002352

[0143] Injection volume: 2 μL

[0144] Column temperature: 50°C

[0145] Gradient elution was performed using mobile phase A and mobile phase B. Mobile phase A: ACN: H2O = 6:4 (10 mM NH4COOH); mobile phase B: ISO: ACN = 9:1 (10 mM NH4COOH). The gradient elution program was as follows.

[0146] Table 10 Gradient elution program

[0147] Time (min) Mobile phase A (%) Mobile phase B (%) 0.0 70 30 5.0 57 43 5.1 50 50 14.0 30 70 14.1 30 70 21.0 1 99 24.0 1 99 25.0 70 30 30.0 70 30

[0148] Mass spectrometry parameters

[0149] Ion source type: ESI; positive and negative ion switching scanning was adopted, the positive ion voltage was 3500 V, and the negative ion voltage was 2400 V; sheath gas: 30 Arb; auxiliary gas (Aux): 10 Arb; ion transfer tube temperature (Ion Transfer Tube Temp): 325°C; nebulizer temperature (Vaporizer Temp): 300°C.

[0150] Full scan performance parameters: Orbitrap Resolution: 60,000; Scan Range: 150-1200.

[0151] ddMS2 scanning performance parameters: HCD collision energy: 15%, 25%, 30%, 40%; ion trap resolution: 30,000.

[0152] 2. Determine the relative correction factor

[0153] According to the table below, take each internal standard and reference lipid compound to prepare a series of mixed reference solutions (Cal1 to Cal11).

[0154] Table 11 Concentration of mixed reference solution in a series of concentrations (μg / mL)

[0155]

[0156]

[0157]

[0158] The detection was performed under the above-mentioned LC-MS detection conditions. The peak area of ​​the primary mass spectrometry response (parent ion) and the retention time was used as the ordinate, and the concentration was used as the abscissa to construct a standard curve. The correlation coefficient of the standard curve of each compound was greater than 0.99. The ratio of the slope of each control lipid compound to the slope of the internal standard was calculated as the relative correction factor. The results were repeated three times. The results are shown in the following table:

[0159] Table 12 Summary of relative correction factors

[0160]

[0161]

[0162]

[0163] It should be noted that the above table shows the final preferred results. Compounds that do not meet the requirements have been eliminated during the research process. For example, arachidonic acid-d11 in FA has been eliminated. If arachidonic acid-d11 is used as the internal standard of FA, under the above LC-MS detection conditions, the relative correction factors of arachidonic acid-d11 and C12:0, C14:0, C16:1, C18:1, C18:2, C16:0, and C18:0 are 92.87, 42.91, 42.54, 10.58, 8.89, 22.59, and 15.87, respectively. Most of them exceed the range of 10, with large errors.

[0164] Example 2

[0165] This embodiment provides a compound similarity evaluation model and a construction method thereof, the construction method comprising steps S100 and S200.

[0166] S100, obtain training data set

[0167] The molecular structures and reference mass spectrometry data (secondary mass spectrometry data) of small molecule compounds are collected through the MoNA public database and self-built database.

[0168] Obtain positive data, negative data, and extended data. The following uses the known compound leucine (leu), its related compound glycyleucine (gly-leu), and the unrelated compound adenosine triphosphate (ATP) as examples to illustrate the process of obtaining positive data, negative data, and extended data.

[0169] 1. Structural similarity data

[0170] The molecular structures of leu, gly-leu, and ATP were obtained in SMILES format and can be queried through the Biodeep database (https: / / query.biodeep.cn / ) built by the applicant.

[0171] Table 13

[0172]

[0173] For each compound, we traverse each atom according to the SMILES structural formula and classify the atoms according to their type and connections to other atoms. Specifically, different types of atoms are assigned different categories; the same type of atoms, but different types or numbers of other atoms to which they are connected, are assigned different categories. For example, if two C atoms are both single-bonded to another C and three H atoms, the two C atoms are considered the same. If one C is single-bonded to another C and three H atoms, and the other C is double-bonded to another C and single-bonded to two H atoms, the two C atoms are considered different. This method can be used to determine the atomic class and number of each compound. The data for each atomic class of the two compounds are then calculated using non-reverse division to obtain the structural similarity data for the two compounds.

[0174] Here, atoms are used as the smallest unit, so there is no repeated statistics when splitting the structure. It is more suitable for splitting and comparing the structural features of a large number of complex compounds.

[0175] By splitting and calculating the directionless division of the SMILES structures of Leu, Gly-Leu, and ATP, the atomic categories, atomic numbers, and pairwise structural similarity data of these three compounds are shown in the following table.

[0176] Table 14

[0177]

[0178]

[0179] *: According to SMILES representation, the atom in bold and underline is the target atom, for example in is the target atom, C is Single-bonded atoms, and Connected with 3 Hs.

[0180] 2. Mass spectrometry data similarity data

[0181] When collecting mass spectrometry data, multiple mass spectrometry data sets exist for Leu, Gly-Leu, and ATP, including mass spectrometry data in positive ion mode and reference mass spectrometry data in negative ion mode. The specific numbers are shown in the table below. When comparing two mass spectrometry data sets, the mass spectrometry data in the same ion mode should be used as a reference, and the similarity between the two should be calculated.

[0182] Table 15

[0183] name Number of mass spectrometry data in positive ion mode Number of mass spectrometry data in negative ion mode leu 6 9 gly-leu 7 9 ATP 2 6

[0184] For example, in positive ion mode, for any one of the six mass spectrometry data of Leu, the similarity with the other five is calculated separately, thus obtaining 30 sets of mass spectrometry data similarity data. When calculating the similarity between the six mass spectrometry data of Leu and the seven mass spectrometry data of Gly-Leu, the similarity between any one of the six mass spectrometry data of Leu and the seven mass spectrometry data of Gly-Leu is calculated separately, thus obtaining 42 sets of mass spectrometry data similarity data; when calculating the similarity between the seven mass spectrometry data of Gly-Leu and the six mass spectrometry data of Leu, the similarity between any one of the seven mass spectrometry data of Gly-Leu and the six mass spectrometry data of Leu is calculated separately, thus obtaining another 42 sets of mass spectrometry data similarity data. The method for obtaining the similarity of mass spectrometry data between other identical or different compounds is the same.

[0185] 3. Training Dataset

[0186] Positive data: The compound's own mass spectrometry data similarity data and structural similarity data, with the input label being 1. Six mass spectrometry data of Leu in positive ion mode can yield 30 sets of positive data.

[0187] Extended data: Mass spectrometry similarity data and structural similarity data for Leu and the related compound Gly-Leu, with an input label of 0.5. Six mass spectrometry data for Leu and seven mass spectrometry data for Gly-Leu in positive ion mode yield 84 sets of extended data.

[0188] Negative data: mass spectrometry data similarity data and structural similarity data of Leu and ATP, which are non-related compounds, with the input label being 0. In positive ion mode, 6 mass spectrometry data of Leu and 2 mass spectrometry data of ATP can generate 24 sets of negative data.

[0189] The following table shows some of the positive data, extended data, and negative data obtained.

[0190] Table 16

[0191]

[0192]

[0193] Note: In the table, CSF, CSR, J, and SES represent the positive cosine similarity score, negative cosine similarity score, Jaccard similarity, and spectral entropy similarity, respectively. These methods all calculate the similarity between two mass spectra. When calculating cosine similarity, one of the two mass spectra is used as a reference, and the other is aligned to this reference. During alignment, m / z values ​​not found in the reference are discarded. Using one of the two mass spectra as a reference, two cosine similarity scores, CSF and CSR, are obtained. J is calculated by dividing the number of identical m / z values ​​in the two mass spectra by the total number of m / z values. SES Same as Li Y, Kind T, Folz J, Vaniya A, Mehta SS, Fiehn O. Spectral entropyoutperforms MS / MS dot product similarity for small-molecule compound identification. Nat Methods. 2021Dec; 18(12):1524-1531.doi:10.1038 / s41592-021-01331-z.Epub 2021Dec 2. PMID: 34857935; PMCID: PMC11492813. Calculated.

[0194] Finally, we obtained about 100,000 sets of data, of which about 20,000 were positive data, about 40,000 were extended data, and about 40,000 were negative data. 80% of this data was used as the training data set and 20% as the test data set.

[0195] S200, establishing a fully connected neural network model, inputting the training data set into the fully connected neural network model for training, and obtaining a compound similarity evaluation model.

[0196] In the preset fully connected network model, the input dimension information N is the above-mentioned reference mass spectrum data similarity data and structural similarity data features, specifically including: mass spectrum data similarity data represented by CSF, CSR, J, SES, etc. and The compound similarity evaluation model was constructed based on the atomic structure of the compound. There were six hidden layers, with the number of neurons in each layer being 2*N, 8*N, 8*N, 16*N, N, and 2 (N being the input dimension). Based on the aforementioned training dataset and the pre-set fully connected network model, 10,000 iterations of training were performed to obtain the compound similarity evaluation model.

[0197] The ROC curve of the compound similarity evaluation model obtained after training is as follows: Figure 7The AUC value was 0.846, demonstrating excellent performance for evaluating the similarity between two compounds. The mass spectrometry similarity data and structural similarity data for Leu and itself, Leu and gly-Leu, and Leu and ATP were input into the compound similarity evaluation model, and the average similarity results were output, as shown in Table 16. These results are consistent with expectations, confirming the accuracy of the model.

[0198] Example 3

[0199] Detection of lipid compounds in serum

[0200] 1. Sample solution preparation

[0201] Internal standard solution: Take each internal standard substance according to the table below, dissolve and dilute it with methanol to prepare the internal standard solution.

[0202] Table 17 Internal standard solution concentration

[0203] Serial number Lipid subclasses internal standard Concentration (μg / mL) 1 PE 15:0-18:1-d7-PE 20 2 LPE 18:1-d7LPE 20 3 Cer Cer(d18:1(d7) / 18:0) 20 4 PC 16:0PC-d9 40 5 LPC 18:1-d7LPC 40 6 DG 1,3-17:0d5DG 40 7 CE 18:1Chol(d7)ester 40 8 TG 15:0-18:1-d7-15:0TG 10 9 SM SM(d18:1 / 15:0)-d9 20 10 MG 18:1-d7MG 40 11 PI 15:0-18:1-d7-PI 20 12 LPI 17:0LPI-d5 20 13 PG 15:0-18:1-d7-PG 20 14 LPG 17:0LPG-d5 20 15 FA C17:0 40

[0204] Sample solution: 100 μL of serum sample (pooled sera from multiple healthy individuals) was placed in a 2 mL centrifuge tube and 20 μL of internal standard solution was added. 130 μL of methanol and 600 μL of methyl tert-butyl ether were added, and the mixture was vortexed for 60 s. 120 μL of water was added, and the mixture was vortexed for 60 s. The mixture was placed on ice for 10 min. The mixture was centrifuged at 12,000 rpm at 4°C for 10 min, and 500 μL of the supernatant was transferred to another 2 mL centrifuge tube. 500 μL of methyl tert-butyl ether / methanol (4:1, v / v) solution was added to the lower residue, and the mixture was vortexed for 60 s. The mixture was centrifuged at 12,000 rpm at 4°C for 10 min, and 500 μL of the supernatant was transferred to the combined extracts. The extracts were concentrated to dryness under vacuum. 200 μL of isopropanol / methanol (1:1, v / v) solution was added to the residue, and the mixture was vortexed for 60 s. The mixture was centrifuged at 12,000 rpm at 4°C for 10 min, and 120 μL of the supernatant was transferred to a sample vial.

[0205] The sample solution was tested under the same liquid chromatography tandem mass spectrometer detection conditions as in Example 1 to obtain mass spectrum data, such as Figure 8 and Figure 9 The total ion currents of positive and negative ions in serum sample solution.

[0206] LipidSearch software (V5.1) was used to process the raw mass spectrometer data (*.raw format) individually, performing lipid annotation, peak alignment, and peak filtering to generate a raw peak area list. Parameters were set as Rt Correctiontol.min = 0.25 and S / N Threshold = 3. A list containing lipid name, mass-to-charge ratio (m / z), retention time (RT), and peak intensity was generated.

[0207] According to the mass spectrometry data of the lipid compound to be tested, the relative correction factor, the content of the quantitative internal standard in the sample solution and its mass spectrometry data, the content of the lipid compound to be tested, Ci, is calculated according to the following formula:

[0208] Ci=Pi / Pr*Cr*RFi

[0209] Wherein, Cr is the internal standard concentration; Pi is the peak area of ​​the lipid compound to be detected; and Pr is the peak area of ​​the internal standard.

[0210] For lipid compounds to be tested with reference substances (such as the lipid compounds in Table 2), Ci is calculated using the internal standard corresponding to its own reference lipid compound and the relative correction factor.

[0211] For lipid compounds to be tested without a reference substance, calculate the cosine similarity of the mass spectral data between the test lipid compound and each reference lipid compound of the corresponding lipid subclass. Select the reference lipid compound with the highest similarity as the quantitative control for the test lipid compound. Calculate the content using the internal standard and relative correction factor of this quantitative control.

[0212] Example 4

[0213] Detection of lipid compounds in serum

[0214] The same method as in Example 3 is used, except that for a lipid compound to be tested without a reference substance, the molecular structure similarity and mass spectrometric similarity data between the lipid compound to be tested and each reference lipid compound of the corresponding lipid subclass are calculated. The similarity evaluation model constructed in Example 2 is then input to output the similarity evaluation results, and the reference lipid compound with the highest similarity is selected as the quantitative control for the lipid compound to be tested. The content is calculated based on the internal standard and relative correction factor of this quantitative control.

[0215] Example 5

[0216] Detection of lipid compounds in tissue samples

[0217] Same as Example 3, except that the sample to be tested is rabbit kidney, and cryogenic grinding is required in the preparation of the sample solution. Figure 10 and Figure 11 This is the total ion current diagram of positive and negative ions in the tissue sample solution.

[0218] Analytical method validation

[0219] Specificity verification confirmed that each internal standard could be detected simultaneously under the conditions of Example 1 without interference.

[0220] 1. Exclusivity

[0221] Take blank samples (use water instead of serum and treat in the same way) and serum sample solutions, test them separately, extract the primary chromatogram of each internal standard, check the peak area, and the results are as follows.

[0222] Table 18 Specificity test results

[0223]

[0224]

[0225] There was an interference peak of C17:0 in the blank sample, and the peak area ratio of C17:0 to the blood sample solution was 2.7%, and the interference could be ignored.

[0226] 2. Linear

[0227] Prepare a standard curve or standard working curve according to standard methods and perform linear regression analysis using the least squares method. Use at least five standard points (excluding blanks). The concentration range should cover as many orders of magnitude as possible. For accurate quantitative methods, the linear correlation coefficient (r) of the linear regression equation should be no less than 0.99. The first internal standard for each lipid subclass is used as a representative of each lipid subclass.

[0228] Table 19 Linearity and range

[0229]

[0230] 3. Accuracy and precision

[0231] Three concentration levels of spiked serum samples were prepared to evaluate the precision and accuracy of the analytical method. The low concentration (L) should be between three times the limit of quantification (Cal11 in Table 11) and 10% of the calibration curve range (Cal1 to Cal11 in Table 11), the medium concentration (M) should be 30%-50% of the calibration curve range, and the high concentration (H) should be above 70% of the calibration curve range. The spiked recovery was calculated and the recovery was between 60% and 150%.

[0232] The results of at least five injections of spiked serum samples at three concentration levels within the batch and between batches at different times were examined. The quality control samples at the three concentrations should meet the requirement that more than 2 / 3 of the samples RSD ≤ 15% within the batch and between batches (low concentrations can be relaxed to 20%).

[0233] Table 20 Spike recovery results

[0234]

[0235]

[0236]

[0237] Table 21 Intra-batch precision results (RSD%)

[0238]

[0239]

[0240] Table 22 Inter-batch precision results (RSD%)

[0241]

[0242]

[0243] From the above results, it can be seen that the lipid detection method of the biological sample of the present application can simultaneously detect multiple lipid subclasses, the method has good linearity and repeatability, and has a certain degree of accuracy.

[0244] According to the lipid detection method of the biological sample of the present application, human serum samples and mouse kidney tissue are detected. 3175 lipid compounds can be detected in the serum sample, wherein the number of compounds belonging to 15 lipid subclasses of PE, LPE, Cer, PC, LPC, DG, CE, TG, SM, MG, PI, LPI, PG, LPG, FA is 2217, and 3226 lipid compounds can be detected in the mouse kidney tissue sample, wherein the number of compounds belonging to these 15 lipid subclasses is 1714. This shows that the number of lipid compounds in the biological sample detected by the lipid detection method of the biological sample of the present application is many, and at least 15 lipid subclasses can be detected simultaneously, and relatively accurate quantitative results are given. The method of the present application is simple to operate, and low cost, and can carry out large-scale sample detection, is conducive to promotion and use, and provides a better experimental basis for lipidomics research.

[0245] The preferred embodiments of the present invention are described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various simple modifications can be made to the technical solution of the present invention, and these simple modifications all fall within the scope of protection of the present invention.

Claims

1. A method for detecting lipids in biological samples, characterized in that: include: For each lipid subclass to be tested, at least one first internal standard and one reference lipid compound are determined; For each reference lipid compound, a second internal standard is selected from the first internal standard of the lipid subclass to which it belongs, and a relative correction factor of the reference lipid compound relative to the second internal standard under liquid chromatography tandem mass spectrometry detection conditions is determined; adding each second internal standard to the biological sample to be tested to prepare a sample solution, and detecting the sample to be tested under the liquid chromatography tandem mass spectrometry detection conditions to determine each lipid compound to be tested contained in the biological sample to be tested; For each lipid compound to be tested, one control lipid compound of the lipid subclass to which it belongs is selected as a quantitative control, and the second internal standard and relative correction factor of the quantitative control are used as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested, and the content of the lipid compound to be tested is calculated by the internal standard method.

2. The detection method according to claim 1, wherein The lipid subclass to be tested includes at least one of the following categories: phosphatidylethanolamine (PE), lysophosphatidylethanolamine (LPE), ceramide (Cer), phosphatidylcholine (PC), lysophosphatidylcholine (LPC), diglyceride (DG), cholesterol ester (CE), triglyceride (TG), sphingomyelin (SM), monoglyceride (MG), phosphatidylinositol (PI), lysophosphatidylinositol (LPI), phosphatidylglycerol (PG), lysophosphatidylglycerol (LPG), fatty acids (FA), phosphatidylserine (PS), lysophosphatidylserine (LPS), phosphatidic acid (PA); Preferably, the lipid subclasses to be tested include: phosphatidylethanolamine (PE), lysophosphatidylethanolamine (LPE), ceramide (Cer), phosphatidylcholine (PC), lysophosphatidylcholine (LPC), diglyceride (DG), cholesterol ester (CE), triglyceride (TG), sphingomyelin (SM), monoglyceride (MG), phosphatidylinositol (PI), lysophosphatidylinositol (LPI), phosphatidylglycerol (PG), lysophosphatidylglycerol (LPG), fatty acids (FA); Preferably, the lipid subclasses to be detected further include: phosphatidylserine (PS), lysophosphatidylserine (LPS), and phosphatidic acid (PA).

3. The detection method according to claim 2, wherein For each lipid subclass to be tested, the first internal standard includes the following:

4. The detection method according to claim 2, wherein For each lipid subclass to be tested, the first internal standard and reference lipid compounds include those shown in the following table:

5. The detection method according to claim 2, wherein The reference lipid compounds and their second internal standards contained in each lipid subclass to be tested and the relative correction factors under the liquid chromatography tandem mass spectrometry detection conditions are as follows:

6. The detection method according to claim 1, wherein The liquid chromatography conditions of the liquid chromatography tandem mass spectrometry detection conditions include: Column: XBridge BEH C18 2.1×150 mm, 2.5 μm; Gradient elution was performed using mobile phases A and B: mobile phase A: 10 mM ammonium formate in acetonitrile with a volume ratio of 6:4 for acetonitrile to water; mobile phase B: 10 mM ammonium formate in isopropanol and acetonitrile with a volume ratio of 9:1 for isopropanol and acetonitrile; Preferably, the gradient elution comprises: Preferably, the mass spectrometry conditions of the liquid chromatography tandem mass spectrometry detection conditions include: Scan type: positive and negative ion switching scan; Detection method: Full scan / data-dependent secondary scan (Full MS / ddMS2); More preferably, the mass spectrometry parameter setting of the liquid chromatography tandem mass spectrometry further includes: HCD collision energy: 15%, 25%, 30%, 40%.

7. The detection method according to claim 1, wherein The biological sample to be tested is serum; The step of adding each second internal standard to the biological sample to be tested to prepare the sample solution specifically includes: taking a serum sample, adding each second internal standard; extracting at least once with an extraction solvent, concentrating the extract, and then re-dissolving with a re-solvent to obtain a sample solution; Preferably, the extraction solvent consists of methyl tert-butyl ether, methanol and water. More preferably, the volume ratio of methyl tert-butyl ether, methanol and water is 2-6:0.5-1.5:0.4-1.6, and the optimal ratio is 4:1:0.

8.

8. The detection method according to claim 1, wherein For each lipid compound to be tested, one of the control lipid compounds of the lipid subclass is selected as a quantitative control, including: The lipid compound to be tested is compared with each control lipid compound of the lipid subclass to which it belongs, and the control lipid compound with the greatest similarity is selected as the quantitative control.

9. The detection method according to claim 8, wherein The method of comparing the lipid compound to be tested with each control lipid compound of the lipid subclass to which it belongs, and selecting the control lipid compound with the greatest similarity as a quantitative control, comprises: Obtaining structural similarity data and mass spectrometry data similarity data between the lipid compound to be tested and each reference lipid compound of the lipid subclass to which it belongs; The structural similarity data and mass spectrum similarity data are input into a pre-built compound similarity evaluation model, a similarity evaluation result is output, and a control lipid compound with the highest similarity is selected as a quantitative control for the lipid compound to be tested.

10. The detection method according to claim 9, wherein The compound similarity evaluation model is constructed according to the following steps: Obtaining a training data set, wherein the training data set includes positive data, extended data, and negative data; The positive data includes mass spectrometry similarity data and structural similarity data between the known compound and itself, and the label is 1; The extended data includes mass spectrometry similarity data and structural similarity data of known compounds and related compounds, with labels ranging from 0.1 to 0.

9. The negative data includes mass spectrometry data similarity data and structural similarity data of known compounds and non-related compounds, and the label is 0; A fully connected neural network model is established, and the training data set is input into the fully connected neural network model for training to obtain a compound similarity evaluation model; Preferably, the mass spectrum data similarity data includes at least one of cosine similarity, Jaccard similarity, and spectral entropy of the mass spectrum data of two compounds; Preferably, the structural similarity data is characterized by comparative data of individual atoms or atomic groups of the two compounds, or comparative data of molecular fingerprints of the two compounds.

11. A lipid detection kit for biological samples, characterized in that: The kit includes: a first internal standard for each lipid subclass to be measured; The lipid subclass to be tested includes at least one of the following categories: phosphatidylethanolamine (PE), lysophosphatidylethanolamine (LPE), ceramide (Cer), phosphatidylcholine (PC), lysophosphatidylcholine (LPC), diglyceride (DG), cholesterol ester (CE), triglyceride (TG), sphingomyelin (SM), monoglyceride (MG), phosphatidylinositol (PI), lysophosphatidylinositol (LPI), phosphatidylglycerol (PG), lysophosphatidylglycerol (LPG), fatty acids (FA); For each lipid subclass to be tested, the first internal standard included is as follows: Preferably, the kit further comprises: an extraction solvent, wherein the extraction solvent consists of methyl tert-butyl ether, methanol and water, more preferably, the volume ratio of methyl tert-butyl ether, methanol and water is 2-6:0.5-1.5:0.4-1.6, and the optimal volume ratio is 4:1:0.8; Preferably, the kit is used to perform lipid detection according to the detection method of any one of claims 1-10.

12. A lipid detection system for biological samples, characterized in that: The detection system includes: a liquid chromatography tandem mass spectrometry module, a mass spectrometry data processing module, a storage module, a comparison module, and a calculation module; The liquid chromatography tandem mass spectrometry module is used to detect the sample solution added with each second internal standard and obtain mass spectrometry data; the mass spectrometry condition setting of the liquid chromatography tandem mass spectrometry module includes: Scanning mode: positive and negative ion switching scanning; Scan type: Full scan / data-dependent secondary scan (Full MS / ddMS2); A mass spectrometry data processing module processes the mass spectrometry data and annotates the second internal standard, each lipid compound to be tested, and the lipid subclass to which it belongs in the biological sample to be tested; A storage module, for storing each lipid subclass to be measured, a reference lipid compound, a second internal standard and a relative correction factor; A comparison module compares each lipid compound to be tested with each control lipid compound of the lipid subclass to which it belongs, and selects the second internal standard and relative correction factor of the control lipid compound with the highest similarity as the quantitative internal standard and quantitative correction factor of the lipid compound to be tested; A calculation module calculates the content of each lipid compound to be tested using the internal standard method; Preferably, the lipid detection system is used to perform lipid detection according to the detection method of any one of claims 1-10.

Images