A method for recognizing a sugar chain of a steviol glycoside

Abstract

The present application relates to a kind of steviol glycoside sugar chain identification method.First, construct virtual monosaccharide chain library;Based on the non-targeted metabolomics method of ultra-high performance liquid chromatography-high resolution mass spectrometry, the metabolomics data of the sample to be measured is obtained under different energies in negative ion mode;The lowest energy when the number of virtual monosaccharide in the secondary mass spectrum of the same metabolite under different energies is not 0 and reaches the maximum from low to high is selected;The fragment with the highest intensity in the secondary mass spectrum under the energy and the difference between mass-to-charge ratio are the minimum sub-ion fragment of virtual monosaccharide, which are identified as aglycone-hydroxyl end sugar chain fragment and aglycone fragment respectively;The hydroxyl and carboxyl end sugar chain composition is identified by searching virtual monosaccharide chain library through the difference between mass-to-charge ratio of aglycone-hydroxyl end sugar chain fragment and aglycone fragment and parent ion and aglycone-hydroxyl end sugar chain fragment respectively, the known aglycone accurate mass-to-charge ratio of steviolbioside is searched by aglycone fragment to identify aglycone, and false positive results are removed by matching degree of virtual glycosyl and hydroxyl end sugar chain.

Description

Technical Field

[0001] This invention relates to the field of analytical chemistry and is a method for recognizing steviol glycoside sugar chains. Background Technology

[0002] Steviol glycosides are the main sweet-tasting components found in stevia. These components are natural sweeteners that emerged after many countries around the world restricted the use of artificial sweeteners. Steviol glycosides have attracted worldwide attention due to the abundance of stevia as a raw material, their good taste, and high sweetness. Countries around the world are actively researching and developing them. Currently, steviol glycosides are widely used in the food industry, pharmaceutical manufacturing, and other fields.

[0003] Steviosides are a class of tetracyclic diterpenoid glycosides. Steviol is currently the most common aglycone of steviol, and it is also the most prevalent aglycone form of steviol in stevia extract. Its structure primarily consists of two glycosylation sites: one at the carboxyl group and the other at the hydroxyl group. Five aglycone forms have been reported in the literature, but almost all glycoside types have two glycosylation sites at one end (carboxyl and hydroxyl). The known glycosyl forms on the chain of steviol glycosides mainly include hexoses, deoxyhexoses, and pentoses, with the longest reported carbon chain being 12 glycosyl units in length. Structurally, the diversity and complexity of steviol glycosides mainly stem from the diversity of the sugar chains connected to the two glycosylation sites, including the diversity of chain lengths and sugar type.

[0004] The length of the sugar chain and the type of sugar moiety in steviol glycosides directly affect their sweetness quality. For example, Rebaudioside A has one more glucose moiety than Stevioside. Studies have shown that Rebaudioside A is considered one of the steviol glycosides with the best sweetness and mouthfeel quality. While Stevioside is sweet, it has a bitter aftertaste. Rebaudioside D (with a sugar chain consisting of two glucose molecules at the carboxyl end and three glucose molecules at the hydroxyl end) has been shown to induce a sweet sensation in humans, while Dulcoside A (with a sugar chain consisting of one glucose molecule at the carboxyl end and one glucose molecule and one rhamnose molecule at the hydroxyl end) is directly related to bitterness in taste sensory evaluations. Therefore, correctly identifying the double-chain structure in steviol glycosides helps in the further characterization of novel steviol glycosides with different chain lengths, and also provides a component basis for removing bitterness and enhancing sweetness in samples.

[0005] In addition, stevia leaf extracts may contain complex and diverse steviol glycoside isomers. The isomerism of these isomers is largely due to glycan isomerism. Glycan isomers affect the chemical properties and flavor of steviol glycosides in stevia, and further affect their absorption, distribution, metabolism, elimination and toxicity in the human body. Therefore, glycan identification is particularly important for the structural elucidation and in-depth exploration of steviol glycoside components in stevia extracts.

[0006] Liquid chromatography-mass spectrometry (LC-MS) has been widely used for the targeted analysis of steviol glycosides in stevia. However, current research mainly focuses on targeted quantitative studies. For the qualitative analysis of non-targeted steviol glycosides, current analysis is limited to matching with limited standards or secondary mass spectrometry libraries and studying based on characteristic fragments. There is still no mature and convenient method for identifying glycan structures. Summary of the Invention

[0007] This invention relates to a method for identifying steviol glycoside glycan chains. First, a virtual monosaccharide library is constructed. Then, based on ultra-high performance liquid chromatography-high resolution mass spectrometry (UHPLC-HDMS) non-targeted metabolomics, metabolomics data of the sample to be tested are acquired at different energies in negative ion mode. The lowest energy at which the number of virtual monosaccharides in the secondary mass spectra of the same metabolite is not zero and reaches its maximum value from low to high is selected. At this energy, the fragment with the highest intensity in the secondary mass spectrum and the smallest daughter ion fragment whose mass-to-charge ratio difference is equal to that of the virtual monosaccharide are identified as aglycone-hydroxyl-terminal glycan chain fragments and aglycone fragments, respectively. The hydroxyl and carboxyl-terminal glycan chain compositions are identified by searching the virtual monosaccharide library for the mass-to-charge ratio difference between the aglycone-hydroxyl-terminal glycan chain fragments and the aglycone fragments, as well as between the parent ion and the aglycone-hydroxyl-terminal glycan chain fragments. Aglycone identification is performed by searching for the precise mass-to-charge ratio of known steviol glycoside aglycones using the aglycone fragments. False positive results are removed by the matching degree between the virtual glycans and the hydroxyl-terminal glycan chains.

[0008] The technical solution adopted by the present invention to achieve the above objective is: a method for recognizing steviol glycosides, comprising the following steps:

[0009] S1: Construct a virtual monosaccharide library of steviol glycosides;

[0010] S2: Based on high performance liquid chromatography-high resolution mass spectrometry, a non-targeted metabolomics method is used to acquire non-targeted metabolomics data of the biological sample to be tested under different energy conditions in negative ion ionization mode; the non-targeted metabolomics data includes the precise mass, retention time, intensity and corresponding tandem mass spectrometry information of metabolites.

[0011] S3: Calculate the difference in mass-to-charge ratio between any two fragments in the secondary mass spectrum to obtain the mass of at least one sugar form among the dehydrated hexose and dehydrated pentose, which is the number of virtual monosaccharide groups; metabolites with the same mass-to-charge ratio and retention time in metabolomics data obtained at different energies are the same metabolites;

[0012] S4: Compare the number of virtual monosaccharides in the secondary mass spectra of the same metabolite at different energies, and select the lowest energy corresponding to the maximum value of the number of virtual monosaccharides from low to high. This energy is the optimal fragmentation energy for identifying the glycan information of the glycoside.

[0013] S5: In the secondary mass spectrum under the optimal fragmentation energy condition, the fragment with the highest intensity is the aglycone-hydroxyl-terminal sugar chain fragment of steviol glycoside.

[0014] S6: Compare the mass-to-charge ratios of all daughter ion fragments with a mass-to-charge ratio difference equal to that of a virtual monosaccharide in the secondary mass spectrum. The fragment with the smallest mass-to-charge ratio is the steviol glycoside aglycone fragment.

[0015] S7: Search the aglycone library for the mass-to-charge ratio of known aglycones in the negative ion mode of stevioside components by the mass-to-charge ratio of stevioside aglycone fragments. One or more aglycones with the same mass-to-charge ratio are identified as the corresponding aglycone candidates for the glycoside.

[0016] S8: By searching the virtual monosaccharide chain library based on the difference in mass-to-charge ratio between the fragment with the highest intensity in the secondary mass spectrum and the steviol glycoside aglycone fragment in S6, the resulting sugar chain composition is the sugar chain composition of the hydroxyl terminus of the steviol glycoside.

[0017] S9: The difference in mass-to-charge ratio between the primary precursor ion and the aglycone-hydroxyl-terminal glycan fragment in S5 is used to search a virtual monosaccharide library, and the resulting glycan composition is the carboxyl-terminal glycan composition of the steviol glycoside.

[0018] S10: By matching the virtual monosaccharides found in S4 with the hydroxyl-terminal monosaccharide chain composition and number identified in S8, false positive results are removed, and the annotation of steviol glycosides is completed.

[0019] In S1, non-targeted metabolomics data and energy-resolved secondary mass spectrometry of biological samples under different energy conditions are acquired with an initial energy of 10 eV and an interval of 5 eV.

[0020] In S2, the virtual monosaccharide chain library includes three types of sugars: hexose, deoxyhexose, and pentose, and the mass number formed by dehydration condensation when the total chain length of the single chain is 1 to 6.

[0021] S3 is as follows:

[0022] First, the intensity of all fragments is normalized by using the strongest fragment in the secondary mass spectrum to obtain the relative intensity of all fragments;

[0023] Fragments with relative intensities less than a threshold are filtered out. The difference in mass-to-charge ratio between any two daughter ion fragments in the secondary mass spectrum is calculated and screened. The sum of all theoretical glycosyl counts for at least one of the following glycosyl types—a dehydrated glucose, a dehydrated deoxy hexose, or a dehydrated pentose—is defined as the virtual monosyl count, which is ≥1.

[0024] In S7, when searching the aglycone library based on the mass-to-charge ratio of aglycone fragments, the [MH] of the aglycone is calculated. - Ions, deviation less than 10 ppm.

[0025] In S8 or S9, the number of sugar saccharides in the virtual monosaccharide chain is limited to 1 to 6, the quality deviation is 10 ppm, and when there are multiple sugar chain candidates, all candidate results are retained and false positive results are removed.

[0026] S10 is as follows: The sum of the number of virtual sugars that are consistent with the sugar type and number in the hydroxyl-terminal virtual monosaccharide chain is divided by the number of sugars in the hydroxyl-terminal sugar chain to obtain the matching degree between the virtual monosaccharide and the hydroxyl-terminal sugar chain. Candidates with a matching degree less than 0.75 are deleted.

[0027] A steviol glycoside glycan recognition device includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement the steviol glycoside glycan recognition method when the computer program is executed.

[0028] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for recognizing steviol glycosides.

[0029] The present invention has the following beneficial effects and advantages:

[0030] This invention's method, when performing important differential pathway and pathway analyses, deeply mines the massive amount of secondary mass spectrometry information of metabolites obtained from non-targeted metabolomics. First, it fully utilizes existing qualitative metabolomics methods to comprehensively annotate the structural information of known metabolites. Second, for undetermined differential metabolites that are beyond the capabilities of the aforementioned methods, this method infers and assigns the structural type and pathway information of unknown metabolites based on existing mature category inference methods. It is no longer limited by the qualitative capabilities of existing methods for differential metabolite identification, nor does it require one-to-one matching and mapping of exact metabolite names and structures in the KEGG database. It analyzes important differential pathways based on compound categories and their pathway assignments. Simultaneously, this method fully utilizes the structural and secondary mass spectrometry information of metabolites that were successfully identified earlier but showed no significant differences in inter-group comparisons, as well as the predicted categories and pathway information of differential metabolites, to assist in obtaining structural information of unknown metabolites and confirming and revising metabolites with known categories and pathways. Finally, the structural and pathway information of all differential metabolites and their changes between groups are used for the identification and analysis of differential pathways. Attached Figure Description

[0031] Figure 1 Flowchart for steviol glycoside glycan recognition;

[0032] Figure 2 Comparison of the energy required to break down the two sugar chains of steviol glycosides;

[0033] In this context, 2A represents the energy-intensity curve of the characteristic ion of the aglycone-carboxyl terminus glycan chain, 2B represents the energy-intensity curve of the characteristic ion of the aglycone-hydroxy terminus glycan chain, and 2C represents the structure of the three asymmetric glycan chains of steviol and the description of the glycosylation sites of steviol.

[0034] Figure 3Secondary mass spectra of Rebaudioside A at different energies and the corresponding number of virtual monosaccharides; where 3A represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 10%, 3B represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 15%, 3C represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 20%, 3D represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 25%, 3E represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 30%, 3F represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 35%, and 3G represents the secondary mass spectrum of Rebaudioside A at a relative collision energy of 40%. The secondary mass spectrum of A and the corresponding number of virtual monosaccharides, 3H represents the secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides when the relative collision energy is 45%, and 3I represents the secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides when the relative collision energy is 50%.

[0035] Figure 4 Energy-virtual monosaccharide number curves of Rebaudioside A at different energies and corresponding fragmentation processes; where 4A represents the energy-virtual monosaccharide number curves of Rebaudioside A at different energies, and 4B represents the fragmentation processes of Rebaudioside A at different energies.

[0036] Figure 5 Identification of Rebaudioside A aglycone fragments and sugar chains;

[0037] Figure 6 Fragmentation curves of Rebaudioside A at different energies for aglycone and aglycone-sugar residue characteristic ions;

[0038] Figure 7 Steviosides are commonly found in aglycone forms containing disaccharide sylation sites;

[0039] Figure 8 Secondary mass spectra of Rebaudioside A at different absolute fragmentation energies and corresponding virtual monosaccharide number and energy-virtual monosaccharide number curves;

[0040] In this spectrum, 8A represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 10 eV and the corresponding number of virtual monosaccharides; 8B represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 15 eV and the corresponding number of virtual monosaccharides; 8C represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 20 eV and the corresponding number of virtual monosaccharides; 8D represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 25 eV and the corresponding number of virtual monosaccharides; 8E represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 30 eV and the corresponding number of virtual monosaccharides; 8F represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 35 eV and the corresponding number of virtual monosaccharides; 8G represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 40 eV and the corresponding number of virtual monosaccharides; and 8H represents the secondary mass spectrum of Rebaudioside A at an absolute collision energy of 45 eV and the corresponding number of virtual monosaccharides. The secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides are shown below. 8I represents the secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides when the absolute collision energy is 50 eV. 8J represents the secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides when the absolute collision energy is 55 eV. 8K represents the secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides when the absolute collision energy is 60 eV. 8L represents the secondary mass spectrum of Rebaudioside A and the corresponding number of virtual monosaccharides when the absolute collision energy is 65 eV.

[0041] Figure 9 Energy-virtual monosaccharide number curves of Rebaudioside A at different absolute fragmentation energies. Detailed Implementation

[0042] The following detailed description of the implementation of the present invention is provided in conjunction with the accompanying drawings and tables: This embodiment is implemented based on the technical solution of the present invention, and provides detailed implementation methods and specific operation processes, but the protection scope of the present invention is not limited to the following embodiments.

[0043] This invention relates to a method for identifying steviol glycoside glycan chains. First, based on ultra-high performance liquid chromatography-high resolution mass spectrometry (UHPLC-HDMS) non-targeted metabolomics, metabolomics data of the sample to be tested are acquired under different energy conditions in negative ion mode. A virtual monosaccharide library is generated. The number of virtual monosaccharides in the spectra is calculated. The number of virtual monosaccharides for the same metabolite at different energies is compared, and the lowest energy corresponding to the maximum value of the number of virtual monosaccharides is selected. The secondary mass spectra at this energy are analyzed, and the fragment with the highest intensity is the aglycone-hydroxyl-terminal glycan fragment. All daughter ion fragments in the spectra whose mass-to-charge ratio difference is equal to that of the virtual monosaccharide are compared, and the fragment with the smallest mass-to-charge ratio is the steviol glycoside glycan. Aglycone fragments; candidate aglycones are identified by searching for the precise mass-to-charge ratio of common aglycone types in stevioside compounds under negative ion conditions using aglycone fragments; hydroxyl-terminal glycan composition is identified by searching a virtual monosaccharide library using the mass-to-charge ratio difference between the fragment with the highest intensity in the spectrum and the aglycone fragment; carboxyl-terminal glycans are identified by searching a virtual monosaccharide library using the mass-to-charge ratio difference between the primary parent ion and the fragment with the highest daughter ion in the spectrum; finally, reliable glycan candidates are selected based on the matching degree of the found virtual monosaccharides and virtual monosaccharide chains. The final identification results of steviol glycosides include mass-to-charge ratio, retention time, aglycone candidates, hydroxyl-terminal glycan candidates, and carboxyl-terminal glycan candidates.

[0044] The technical solution adopted in this invention has the following steps:

[0045] S1: Construct a virtual monosaccharide library of steviol glycosides;

[0046] S2: Based on high performance liquid chromatography-high resolution mass spectrometry, a non-targeted metabolomics method is used to acquire non-targeted metabolomics data of the biological sample to be tested under different energy conditions in negative ion ionization mode; the non-targeted metabolomics data includes the precise mass, retention time, intensity and corresponding tandem mass spectrometry information of metabolites.

[0047] S3: Calculate the difference in mass-to-charge ratio between any two fragments in the secondary mass spectrum to obtain the mass of at least one sugar form among the dehydrated hexose and dehydrated pentose, which is the number of virtual monosaccharide groups; metabolites with the same mass-to-charge ratio and retention time in metabolomics data obtained at different energies are the same metabolites;

[0048] S4: Compare the number of virtual monosaccharide groups in the spectra of the same metabolite at different energies, and select the lowest energy corresponding to the maximum value of the number of virtual monosaccharide groups from low to high. This energy is the optimal fragmentation energy of the carboxyl terminus of the steviol glycoside.

[0049] S5: In the secondary mass spectrum under the optimal fragmentation energy condition, the fragment with the highest intensity is the aglycone-hydroxyl-terminal sugar chain fragment of steviol glycoside.

[0050] S6: In the second-order spectrum under the optimal fragmentation energy condition, the fragment with the highest intensity is the steviol glycoside aglycone-terminal sugar chain fragment.

[0051] S7: Search the aglycone library for the precise mass-to-charge ratio of known aglycones in the negative ion mode of stevioside components by the mass-to-charge ratio of stevioside aglycone fragments. One or more aglycones with the same mass-to-charge ratio are identified as the corresponding aglycone candidates for the glycoside.

[0052] S8: By searching the virtual monosaccharide chain library based on the difference in mass-to-charge ratio between the fragment with the highest intensity in the secondary mass spectrum and the steviol glycoside aglycone fragment in S6, the resulting sugar chain composition is the sugar chain composition of the hydroxyl terminus of the steviol glycoside.

[0053] S9: The difference in mass-to-charge ratio between the primary precursor ion and the aglycone-hydroxyl-terminal glycan fragment in S5 is used to search a virtual monosaccharide library, and the resulting glycan composition is the carboxyl-terminal glycan composition of the steviol glycoside.

[0054] S10: By matching the virtual monosaccharides found in S4 with the hydroxyl-terminal monosaccharide chain composition and number identified in S8, false positive results are removed, and the annotation of steviol glycosides is completed.

[0055] In the first step, non-targeted metabolomics data and energy-resolved secondary mass spectrometry of biological samples under different energy conditions were obtained with 10 eV as the starting energy and 5 eV as the interval.

[0056] In the second step, the virtual monosaccharide chain library consists of three types of sugars: hexose, deoxyhexose, and pentose. The mass number formed by dehydration condensation when the total chain length of the single chain is 1 to 6 is ;

[0057] In the third step, the intensity of all fragments is first normalized by the strongest fragment in the secondary mass spectrum to obtain the relative intensity of all fragments. Then, fragments with relative intensities less than a threshold are filtered out. The difference in mass-to-charge ratio between any two daughter ion fragments in the secondary mass spectrum is calculated and screened. The sum of all theoretical glycosyl counts for at least one of the following glycosyl types is defined as the virtual monosyl count, which must be ≥1.

[0058] In step seven, when searching the aglycone library based on the mass-to-charge ratio of aglycone fragments, the [MH]- ion of the aglycone is calculated with a deviation of less than 10 ppm.

[0059] In steps eight and nine, the number of sugar saccharides in the virtual monosaccharide chain is limited to 1 to 6, the quality deviation is 10 ppm, and when there are multiple sugar chain candidates, all candidate results are retained and false positive results are removed.

[0060] In step 10, the sum of the number of virtual sugars that match the sugar type and number in the hydroxyl-terminal virtual monosaccharide chain is divided by the number of sugars in the hydroxyl-terminal glycan chain to determine the matching degree between the virtual monosaccharide and the hydroxyl-terminal glycan chain. Candidates with a matching degree less than 0.75 are deleted.

[0061] Example 1

[0062] This invention relates to a method for identifying glycan chains of steviol glycosides based on liquid chromatography-high resolution energy-resolved mass spectrometry, such as... Figure 1 As shown, firstly, metabolomics data of the biological samples to be tested were acquired under different energy conditions in negative ion mode using the ultra-high performance liquid chromatography-high resolution mass spectrometry non-targeted metabolomics method; a virtual monosaccharide library was generated; the difference in mass-to-charge ratio between the two fragments in the spectrum was calculated, and the difference was counted as the number of masses of dehydrated glucose, dehydrated deoxyhexose, and dehydrated pentose, which is the number of virtual monosaccharides. If the number of virtual monosaccharides is 0 from beginning to end, it is considered not to be a steviol glycoside compound and no further analysis is performed; the number of virtual monosaccharides in the spectra of the same metabolite at different energy levels was compared, and the lowest energy corresponding to the maximum value of the number of virtual monosaccharides was selected from low to high. This energy is the optimal fragmentation energy for identifying the glycan information of the glycoside. For secondary mass spectra with two maximum values ​​of the number of virtual monosaccharides, the latter energy was selected. Further analysis was conducted on the secondary mass spectra at the highest energy level for the virtual monosaccharide group with the largest number of virtual monosaccharide groups. In the secondary mass spectra under the optimal fragmentation energy condition, the fragment with the highest intensity was the aglycone-hydroxyl-terminal sugar chain fragment. The difference in mass-to-charge ratio between the spectra represented all the daughter ion fragments of the virtual monosaccharide group, and the smallest fragment was the aglycone fragment. Candidate aglycones could be identified by searching for the precise mass number of common aglycone types in stevia compounds using aglycone fragments. The composition of the hydroxyl-terminal sugar chain was identified by searching the virtual monosaccharide chain library using the difference in mass-to-charge ratio between the strongest fragment in the spectrum and the aglycone fragment. The carboxyl-terminal sugar chain could be identified by searching the virtual monosaccharide chain library using the difference in mass-to-charge ratio between the primary parent ion and the highest daughter ion fragment in the spectrum. Finally, reliable glycoside candidates were selected based on the matching degree between the largest virtual monosaccharide group and the hydroxyl-terminal virtual monosaccharide chain found.

[0063] We will use Rebaudioside A as a typical inference example to illustrate this process:

[0064] First, we used asymmetric steviol glycoside standards with carboxyl and hydroxyl-terminal glycan chains to demonstrate that the more easily fragmented glycoside formation site in steviol glycosides is its carboxyl site. The structures of the three asymmetric steviol glycoside standards, Stevioside, Rebaudioside F, and Rebaudioside D, are shown below. Figure 2 As shown in Figure 2C, in its energy-resolved secondary mass spectrum, the characteristic ions of the hydroxyl-terminal glycan chain and the carboxyl-terminal glycan chain are Aglycone-R2 and Aglycone-R1, respectively. Figure 2As shown in Figure 2B, the strongest fragment of Aglycone-R2 in the spectrum can be obtained at 25% of the NCE energy value, indicating that the fragmentation of the carboxyl-terminal sugar chain can be achieved at low energy. In contrast, as Figure 2 As shown in Figure 2A, strong Aglycone-R1 fragments are almost impossible to obtain in the spectrum, indicating that the sugar chain at this site is not easily preserved.

[0065] We use Rebaudioside A as a typical example to illustrate this method. First, we obtain all secondary mass spectra of Rebaudioside A under negative ion mode at 10%–50% (NCE, collision energy normalized) energies, such as… Figure 3 As shown in Figures 3A-3I, we calculated the difference in mass-to-charge ratio between two fragments in each secondary mass spectrum based on the secondary daughter ion fragments. This difference represents the number of virtual monosaccharides in common glycosyl forms of steviol glycosides, including hexose, deoxyhexose, and pentose. The results showed that the number of virtual monosaccharides in the spectra at 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, and 50% NCE were 1, 0, 0, 3, 3, 3, 2, and 2, respectively. At 15%, 20%, and 25% NCE, although one sugar was neutrally lost in the secondary mass spectrum, the two daughter ion fragments differing by one glycosyl group did not exist. Further analysis of the fragmentation process of steviol glycosides at different energies is shown below. Figure 4As shown in B, this indicates that the strongest fragment is relatively stable at this point; further fragmentation requires continued energy absorption. Before these three energies, there is a small peak that drops from 1 to 0, meaning that at 10% NCE energy, the difference between the two fragments is exactly one glucose unit. This indicates that the glycosyl group can fragment at lower energies. From the structure of Rebaudioside A, it's easy to see that at 10% NCE energy, the fragmented glycosyl group is the carboxyl-terminal sugar chain. When the energy increases to 30%, Rebaudioside... In A, the more difficult-to-break sugar moieties begin to break down, and the number of virtual monosaccharides in the hydroxyl-terminal sugar chain in the spectrum increases. However, due to the slow increase in energy, the strongest fragment is not completely broken down. Therefore, at this point, the highest sugar chain is still the fragment of the more difficult-to-break aglycone-hydroxyl-terminal sugar chain. As the energy continues to increase, although the number of virtual monosaccharides in the hydroxyl-terminal sugar chain becomes more obvious in terms of intensity, the intensity of the characteristic fragment of the aglycone-hydroxyl-terminal sugar chain decreases due to the continuous breakage of sugar moieties. When the terminal sugar moieties are completely broken down, only two sugar moieties remain in the sugar chain. By analogy, the number of virtual monosaccharides in the spectrum will shorten as the chain breaks down, from 3 to 2, from 2 to 1, and then from 1 to 0, until finally only the strongest aglycone fragment remains. If the resolution of energy-resolved mass spectrometry is sufficient, the change between the two energies can be observed. Therefore, we consider the lowest energy at which the virtual monosaccharide in the sugar chain reaches its strongest point to be the critical energy value at which the fragile sugar chain completely breaks down while the less fragile sugar chain is almost completely preserved. The strongest ion in the spectrum best reflects the composition of the hydroxyl-terminated sugar chain. Figure 4 As shown in Figure 4A, we believe that 30% NCE is the optimal fragmentation energy for rebaudioside A to infer the composition of the sugar chain. At this point, the maximum number of virtual monosaccharide units, nmax, is 3, which is exactly the number of sugar units in the hydroxyl-terminal sugar chain.

[0066] Further analysis was performed on the spectra under 30% NCE conditions, such as... Figure 5 As shown, we found that the strongest fragment in the spectrum is the characteristic fragment of the aglycone-hydroxy-terminal sugar chain. Since the hydroxy-terminal sugar chain is partially fragmented at this energy, the mass difference between the two fragments in the secondary mass spectrum best reflects the sugar group composition of the hydroxy-terminal sugar chain. Therefore, among all the daughter ion fragments with a mass difference equal to that of the sugar group, the fragment with the smallest mass-to-charge ratio can be inferred to be the aglycone fragment, because the aglycone is relatively stable and difficult to further fragment to obtain a virtual monosaccharide.

[0067] from Figure 6As we can see from the graph, Rebaudioside A has five characteristic ions of the aglycone and aglycone-sugar residue at different energies. These five characteristic curves can further support our glycan identification based on energy-resolved mass spectrometry. The [Steviol-4Glc-H]- characteristic ion rapidly decreases from about 60% to 0 at low energies as the carboxyl-terminal glycan chain completely breaks down. Then, the [Steviol-3Glc-H]- characteristic ion remains relatively stable at low energies, showing the strongest intensity in the spectrum. When the energy is increased to 30% NCE, [Steviol-3Glc-H]- is slowly broken down into [Steviol-2Glc-H]- and [Steviol-Glc-H]-. As the sugar moiety on the glycan chain gradually breaks down, only the strongest characteristic fragment of the aglycone ion remains. From the graph, we can calculate that the matching degree between the virtual sugar moiety and the glycan chain of this glycoside is 1 (3 / 3), which is greater than the minimum threshold of 0.75 for removing false positive results when inferring the glycan chain. The correct inference result is ultimately retained.

[0068] Ultimately, the identification of the aglycone was confirmed by searching for common aglycone forms in steviol glycosides. The reported aglycone types and their corresponding structures in steviol glycosides are as follows: Figure 7 As shown. The carboxyl-terminal glycan chain is obtained by calculating the neutral loss of the strongest characteristic fragment, because under this optimal fragmentation condition, the carboxyl-terminal glycan chain is almost completely fragmented. By the difference between the parent ion and the characteristic fragment of the aglycone-hydroxyl-terminal glycan chain, the glycosyl information of the other chain can be obtained, which is a glucose.

[0069] like Figure 8 As shown, by analyzing the fragmentation of Rebaudioside A at different absolute energies, we found that when the fragmentation energy increased from 15 eV to 70 eV, the corresponding number of virtual monosaccharides in the secondary mass spectra were 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 3, 3, respectively. The energy-virtual monosaccharide number curves at different absolute fragmentation energies are shown below. Figure 9 As shown, we found that due to the large molecular weight of Rebaudioside A, although the chain length information of its hydroxyl terminus can be obtained under absolute energy, the required absolute energy is high, and its optimal fragmentation energy is about 65-70 eV. For many glycosides in stevia, some even with a molecular weight greater than 1000, the optimal absolute energy value may exceed the recommended threshold suggested by the instrument. Secondly, steviol glycosides are relatively sluggish in response to the fragmentation effect of absolute energy. When the energy is increased from 35 eV to 60 eV, only one sugar group on the hydroxyl terminus chain is fragmented, making it difficult to obtain information on all sugar groups on the chain. In fact, the result may be misjudged due to a brief virtual monosaccharide intermediate peak. Therefore, it is finally recommended to use relative fragmentation energy to infer the sugar chain of steviol glycosides.

[0070] Table 1. Glycan identification results of 13 steviol glycoside standards

[0071]

[0072] As shown in Table 1, we finally applied this judgment process to the glycan identification of 13 steviol glycoside standards. We found that 12 out of the 13 standards were correctly identified, while only one steviol glycoside, Stevioside, failed to be identified as a glycoside. This was because the virtual monosaccharide (a dehydrated hexose) did not match the inferred result (hydroxyl terminus of 0), resulting in a score of 0 and a false negative inference result. There are two possible reasons for this result. First, the energy resolution does not meet the requirements of Stevioside, as the energy range is greater than the fragmentation rate of the glycan at the hydroxyl terminus. The other reason is the quality of the spectrum. The low concentration of the standard sample affected the quality of the secondary mass spectrum, and some fragment information was masked by the matrix.

Claims

1. A method for recognizing steviol glycoside glycans, characterized in that, Includes the following steps: S1: Construct a virtual monosaccharide chain library of steviol glycosides, wherein the virtual monosaccharide chain library includes three types of sugars: hexose, deoxyhexose, and pentose, and the mass number formed by dehydration condensation when the total chain length of the single chain is 1 to 6. S2: Based on high performance liquid chromatography-high resolution mass spectrometry, a non-targeted metabolomics method is used to acquire non-targeted metabolomics data of the biological sample to be tested under different energy conditions in negative ion ionization mode; the non-targeted metabolomics data includes the precise mass, retention time, intensity and corresponding tandem mass spectrometry information of metabolites. S3: Calculate the difference in mass-to-charge ratio between any two fragments in the secondary mass spectrum as the mass of at least one sugar form among the dehydrated hexose and dehydrated pentose, which is the number of virtual monosaccharides. If the number of virtual monosaccharides is 0 from beginning to end, it is considered not to be a steviol glycoside compound. Metabolites with the same mass-to-charge ratio and retention time in metabolomics data obtained at different energies are the same metabolites. S4: Compare the number of virtual monosaccharides in the secondary mass spectra of the same metabolite at different energies, and select the lowest energy corresponding to the maximum value of the number of virtual monosaccharides from low to high. This energy is the optimal fragmentation energy for identifying the glycan information of the glycoside. For secondary mass spectra with two maximum values ​​of the number of virtual monosaccharides, select the secondary mass spectrum at the larger energy with the largest number of virtual monosaccharides for further analysis. S5: In the secondary mass spectrum under the optimal fragmentation energy condition, the fragment with the highest intensity is the aglycone-hydroxyl-terminal sugar chain fragment of steviol glycoside. S6: Compare the mass-to-charge ratios of all daughter ion fragments with a mass-to-charge ratio difference equal to that of a virtual monosaccharide in the secondary mass spectrum. The fragment with the smallest mass-to-charge ratio is the steviol glycoside aglycone fragment. S7: Search the aglycone library for the mass-to-charge ratio of known aglycones in the negative ion mode of stevioside components by the mass-to-charge ratio of stevioside aglycone fragments. One or more aglycones with the same mass-to-charge ratio are identified as the corresponding aglycone candidates for the glycoside. S8: By searching the virtual monosaccharide chain library based on the difference in mass-to-charge ratio between the fragment with the highest intensity in the secondary mass spectrum and the steviol glycoside aglycone fragment in S6, the resulting sugar chain composition is the sugar chain composition of the hydroxyl terminus of the steviol glycoside. S9: The mass-to-charge ratio difference between the primary precursor ion and the aglycone-hydroxyl-terminal glycan fragment in S5 is used to search a virtual monosaccharide library, and the resulting glycan composition is the carboxyl-terminal glycan composition of this steviol glycoside. S10: The matching degree between the virtual monosaccharides found in S4 and the hydroxyl-terminal monosaccharide chains identified in S8 is used to remove false positive results and complete the annotation of steviol glycosides. The sum of the number of virtual monosaccharides that match the type and number of sugars in the hydroxyl-terminal virtual monosaccharide chain is divided by the number of sugars in the hydroxyl-terminal sugar chain to obtain the matching degree between the virtual monosaccharides and the hydroxyl-terminal sugar chains. Candidates with a matching degree less than 0.75 are deleted.

2. The method for recognizing steviol glycosides according to claim 1, characterized in that... In S1, non-targeted metabolomics data and energy-resolved secondary mass spectrometry of biological samples under different energy conditions are acquired with an initial energy of 10 eV and an interval of 5 eV.

3. The method for recognizing steviol glycosides according to claim 1, characterized in that, S3 is as follows: First, the intensity of all fragments is normalized by using the strongest fragment in the secondary mass spectrum to obtain the relative intensity of all fragments; Fragments with relative intensities less than a threshold are filtered out. The difference in mass-to-charge ratio between any two daughter ion fragments in the secondary mass spectrum is calculated and screened. The sum of all theoretical glycosyl counts for at least one of the following glycosyl types—a dehydrated glucose, a dehydrated deoxy hexose, or a dehydrated pentose—is defined as the virtual monosyl count, which is ≥1.

4. The method for recognizing steviol glycosides according to claim 1, characterized in that... In S7, when searching the aglycone library based on the mass-to-charge ratio of aglycone fragments, the [MH] of the aglycone is calculated. - Ions, deviation less than 10 ppm.

5. The method for recognizing steviol glycosides according to claim 1, characterized in that... In S8 or S9, the number of sugar saccharides in the virtual monosaccharide chain is limited to 1 to 6, the quality deviation is 10 ppm, and when there are multiple sugar chain candidates, all candidate results are retained and false positive results are removed.

6. A steviol glycoside glycan recognition device, characterized in that, It includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement, when executing the computer program, a method for recognizing steviol glycosides as described in any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements a method for identifying steviol glycoside sugar chains as described in any one of claims 1-5.

Images