Magnetic ligand fishing protocol combined with HPLC-FT-ICR-MS for screening potential α-Glucosidase inhibitors from UCG and in silico analysis

2.1 Chemicals and reagents

UCG was obtained from Consun Pharmaceutical (Horgos) Co., Ltd (Neimenggu, China). α-Glucosidase (EC3.2.1.20, from yeast, lyophilized powder) and tetraethoxysilane (TEOS) were purchased from Adamas Reagent Co., Ltd (Shanghai, China). 4-Nitrophenyl-α-D-gluco pyranoside (pNPG), p-nitrophenol (pNP), glutaraldehyde (50 %), (3-Aminopropyl) triethoxysilane (APTES) was obtained from Sigma-Aldrich (St. Louis, MO, USA). Acarbose (purity > 98 %) and Brilliant Blue G-250 were supplied by Macklin Biochemical Co., Ltd (Shanghai, China). Ferric chloride hexahydrate (FeCl₃·6H₂O), ferrous sulphateheptahydrate (FeSO₄·7H₂O), sodium hydroxide (NaOH), phosphate-buffered saline (PBS, pH 6.8, 0.1 M) and sodium carbonate (Na₂CO₃) were acquired from Ailan Chemical Technology Co., Ltd (Shanghai, China). Reference compounds (purity > 98 %) of Protocatechuic acid, Danshensu, Caffeic acid, Paeoniflorin, Liquiritin, Liquiritigenin, Luteolin, Aloe-emodin, 10,15-Octadecadienoic acid, Rhein, Limonexic acid and Matrine used in this study were from Macklin Biochemical Co., Ltd (Shanghai, China). Reference compound (purity = 99.8 %) of Sanggenol A used in this study was from Chengdu Push Bio-technology Co., Ltd. Reference compound (purity = 98 %) of Glabrone used in this study was from Shanghai yuanye Bio-Technology Co.,Ltd. Purified water was purchased from Wahaha (Hangzhou, China). Formic acid (LC/MS grade) was obtained from Fisher Scientific (Fair Lawn, NJ, USA). Ethanol absolute (HPLC grade), acetonitrile (HPLC grade) and ammonium hydroxide were from Concord Technology Co., Ltd (Tianjin, China).

2.2 Preparation of UCG solution and mixed standard solution

UCG were extracted by ultrasonic extraction method. Details can be seen in S 1.1.

2.3 Instrument and analytical conditions

HPLC-FT-ICR-MS was used for qualitative analysis of UCG extract, HPLC was used for quantitative analysis, and the conditions of the instrument can be found in S 1.2.

2.4 Synthesis of Fe₃O₄@SiO₂@NH₂@α-Glucosidase

Synthesis of Fe₃O₄@SiO₂@NH₂@α-Glucosidase was based on relevent literature with some modifications (Qing et al., 2012). As shown in Fig. 1, The FeCl₃·6H₂O (4.32 g) and FeSO₄·7H₂O (1.6 g) were dissolved in 20 mL of water, followed by the addition of 18 mL of 0.1 g/mL NaOH solution under nitrogen atmosphere. The reaction mixture was under stirring at 70 °C for 1 h. After the reaction, the product was washed with water and phosphate buffer solution (PBS) (0.1 M, pH 6.8) to neutral pH and dried at 60 °C in a vacuum oven to obtain Fe₃O₄.

Close

Fig. 1

Synthesis of Fe₃O₄@SiO₂@NH₂@α-Glucosidase.

Export to PPT

1.5 g of Fe₃O₄ was weighed and dispersed in an 120 mL ethanol solution consisting of water (40 mL) and ethanol (80 mL). Subsequently, 4 mL of tetraethoxysilane (TEOS) and 4.5 mL of ammonia were fully mixed and vigorously stirred in an ice bath for 12 h. At the end of the reaction, the product was washed with water and PBS (0.1 M, pH 6.8) to neutral pH and dried at 60 °C in a vacuum oven to obtain Fe₃O₄@SiO₂.

Accurately weigh 1.0 g of Fe₃O₄@SiO₂ into a three-necked flask, followed by adding 100 mL of ethanol. Then 6 mL of APTES was added and mechanically stirred at 70 °C for 8 h under nitrogen atmosphere. At the end of the reaction, the product was washed three times with water and PBS (0.1 M, pH 6.8) and dried at 60 °C in a vacuum drying oven to obtain Fe₃O₄@SiO₂@NH₂.

The Fe₃O₄@SiO₂@NH₂ nanoparticles (10 mg) were combined with 2 mL of 10 % glutaraldehyde solution in a centrifuge tube. After centrifugation at 5000 rpm for 1 h, the samples were subjected to disintegration via ultrasonication for 10 min. The supernatant was discarded, and the magnetic material was washed three times with PBS (0.1 M, pH 6.8). Subsequently, the Fe₃O₄@SiO₂@NH₂ was crosslinked with the α-Glucosidase solution and incubated for 4 h. After magnetic cleaning, the magnetically immobilized enzyme was stored in PBS (0.1 M, pH 6.8) at 4 °C.

2.5 Characterization of Fe₃O₄@SiO₂@NH₂@α-Glucosidase

2.6 Optimization of Fe₃O₄@SiO₂@NH₂@α-Glucosidase

The optimization of four crucial immobilization parameters, including concentration of Glutaraldehyde (GA) (6 %, 8 %, 10 %, 12 %, 14 %, 16 %), ratio of Fe₃O₄@SiO₂@NH₂ to enzyme (10: 1, 20: 1, 30: 1, 40: 1, 50: 1), crosslinking time (0.5 h, 1.0 h, 1.5 h, 2.0 h, 2.5 h, 3.0 h) and the immobilization time (2 h, 4 h, 6 h, 8 h, 10 h, 12 h) were conducted to achieve maximum enzyme activity.

2.7 Performance study

The performance of immobilized enzyme was studied, including pH, temperature and reusability. Details can be seen in S 1.3.

2.8 Enzyme loading of Fe₃O₄@SiO₂@NH₂@α-Glucosidase

The study of α-Glucosidase loading was based on relevent literature with some modifications (Yi et al., 2022). Detailed steps can be seen in S 1.4.

2.9 Free and immobilized α-Glucosidase activity

In the study of enzyme activity, a few modifications was made according to the methodology reported in the literature (Miao et al., 2018). 4-Nitrophenyl-α-D-gluco pyranoside (pNPG) was used as a substrate, and its hydrolysis by α-Glucosidase resulted in the production of p-nitrophenol (pNP). The detailed procedure for examining enzyme activity can be seen in S 1.5.

2.10 Kinetic parameters of free and immobilized α-Glucosidase

The concentrated pNPG solution was diluted using PBS (pH 6.8, 0.1 M) in order to generate a series of solutions with concentrations ranging from 0.2 mM to 2 mM. Subsequently, kinetic studies were conducted on free α-Glucosidase and immobilized α-Glucosidase at 37 °C. The values for the enzymes' kinetic parameters were determined using Eq. (1):

The initial and maximum reaction rates are represented by v and v_max, respectively. [s] represents the concentration of the substrate, while km represents the Michaelis-Menten constant which signifies the enzyme's affinity for the substrate.

2.11 Screening of inhibitors in UCG

The UCG solution was denoted as L0. Firstly, 400 μL of L0 was incubated at 37 °C for 5 min, then 10 mg of immobilized enzyme and Fe₃O₄@SiO₂@NH₂ was added respectively, and mixed at 1000 rpm for 15 min. After separating the solid using a magnet, the magnetic material was eluted with acetonitrile–water (90:10, v/v). The eluent of immobilized enzyme was recorded as L1, and the eluent of Fe₃O₄@SiO₂@NH₂ was recorded as blank. The chemical composition in both L0, L1 and blank were compared by HPLC-FT-ICR-MS to identify the active ingredient.

2.12 Molecular docking

The interaction between α-Glucosidase and the active ingredients was elucidated, docking simulations performed using AutoDock 4.2. The crystal structure of α-Glucosidase (PDB ID: 3A4A, Resolution: 1.60 Å) in pdb format was downloaded from the PDB database (https://www.rcsb.org/). The water and the original ligands (GLC and Ca²⁺) were then removed from the target protein through Pymol 2.5.5 and stored in PDBQT format. The mol2 format of the active ingredient was downloaded from the PubChem database (https://www.pubchem.com/) and stored in PDBQT format. A total number of fifty independent docking runs were performed for each active ingredient. Receptor-ligand interaction analysis and figure rending were performed using Discovery Studio.

2.13 Inhibition assay

The α-Glucosidase inhibition activity of the inhibitor extracted from UCG extract was verified by pNPG method, but with a slight modification (Miao et al., 2018). IC₅₀ is determined by varying the concentration of inhibitors while maintaining a constant amount of pNPG (2 mM). Briefly, three groups were set for the assay: blank group, sample group and control group (n = 3). The blank group contained no sample or enzyme, while the control group had no sample. In the sample group, α-Glucosidase (50 μL) was mixed with the sample solution (50 μL) and substrate (2 mM pNPG, 50 μL). The mixture was then incubated for 30 min at 37 °C. Finally, 0.2 M Na₂CO₃ (50 μL) was added into the system and immediately removed 100 μL solution from 96-well plate. The absorbance at 405 nm was determined using a microplate reader. The optical density (OD) value was calculated as the percentage of inhibition. The calculation of the inhibition rate was performed using Eq. (2):

The absorbance values of the sample group, blank group and control group are represented by As, Ab and Ac, respectively.

2.14 Data analysis strategy

According to the exact molecular weights measured, the possible molecular formulas were calculated by Data Analysis 4.4 software (Bruker Corp., Karlsruhe, Germany) within the specified mass error range (within ± 3 ppm) and purity score range (more than 85 %). The chemical composition of UCG was finally determined by matching each possible molecular formula and its fragment information of secondary mass spectrometry with the in-house database or the reference compounds. An in-house database of chemical compounds which includes names, molecular formulas, theoretical mass, chemical structures and the MS/MS fragmentation profiles was established according to Chemspider (https://www.chemspider. com/), SciFinder (https://scifinder.cas.org/), PubMed (https://www.ncbi.nlm.nih.gov/pubmed), Massbank (https://massbank.eu/), CNKI (https://www.cnki.net/) and Web of Science (https://www.webofknowledge.com) (Liu et al., 2022; Zhang et al., 2021a; Ma et al., 2014; Wang et al., 2019; Li et al., 2022; Wang et al., 2015; Gong et al., 2020; Yan et al., 2016; Li et al., 2014b; Jiang et al., 2019; Zhang et al., 2021b; Sheng et al., 2021; Zhong et al., 2015).

3.1 Qualitative analysis of components in UCG

Fig. 2 reveals the base peak ion chromatograms (BPCs) of UCG solution in positive and negative ion modes, together with the BPC of the mixed standard solution in negative ion mode. On the basis of its specific molecular weight, retention duration and in-house database, the chemical structure and cleavage patterns were inferred. As a consequence, 142 components in all were found in UCG, including 57 flavonoids, 21 terpenoids, 24 phenylpropanoids, 5 phenols, 8 alkaloids, 8 anthraquinones, 10 organic acids and 9 other ingredients. 9 components, including Danshensu, Caffeic acid, Paeoniflorin, Liquiritin, Liquiritigenin, Luteolin, Aloe-emodin, 10, 15-Octadecadienoic acid and Rhein, were identified by matching molecular weights, retention time and MS/MS fragment data to the references. Molecular weight-specific extracted ion chromatograms (EICs) were shown in the supplementary material Fig. S1. The results of each chemical structure are shown in Fig. S2. The fragmentation pathway of typical compounds is shown in Fig. S3. The formula, retention time, measured mass, theoretical mass, ion mode, error value and MS/MS are shown in Table S1. UCG component analysis of studies can be found in S 2.1.

Close

Fig. 2

UCG's positive (A) and negative (B) ion base peak chromatograms (BPC), and the reference compounds in positive ion mode (C). 1-Danshensu, 2-Caffeic acid, 3-Paeoniforin, 4-Liquiritin, 5-Liquiritigenin, 6-Luteolin, 7-Aloe-emodin, 8–10,15-Oct adecadienoic, 9-Rhein.

Export to PPT

3.2 Characterization of Fe₃O₄@SiO₂@NH₂@α-Glucosidase

3.2.1

3.2.1 SEM-EDX analysis

The SEM images of the Fe₃O₄, Fe₃O₄@SiO₂, Fe₃O₄@SiO₂@NH₂ and Fe₃O₄@SiO₂@NH₂@α-Glucosidase were shown in Fig. 3. Their particle diameters were 20 nm, 66 nm, 77 nm and 83 nm, respectively. As shown in Fig. 3b, an increase in particle size accompanied by a more pronounced manifestation of the spherical and smooth structure was observed after the modification with TEOS. These findings prove the successful coating of SiO₂ layer onto the surface of Fe₃O₄ (Cheng et al., 2019). As shown in Fig. 3c, after the modification with APTES, Fe₃O₄ magnetic nanoparticles become more agglomerate. In addition, As shown in Fig. 3d, in Fe₃O₄@SiO₂@NH₂@α-Glucosidase, a large number of APTES and α-Glucosidase act as a “bridge”, and a membrane was found on the surface of nanomaterials. These results indicated that the surface modification of every step was successful. The result of EDX is shown in Fig. 3e. The coexistence of Fe (pink) and O (green) in Fe₃O₄@SiO₂@NH₂@α-Glucosidase indicated the successful synthesis of Fe₃O₄. The existence of Si (cyan) was due to the successful modification of TEOS. The existence of N (yellow) indicated the successful modification of APTES and the active NH₂ was successfully introduced. The SEM-EDX analysis reveals the successful synthesis of Fe₃O₄@SiO₂@NH₂.

Close

Fig. 3

SEM images of Fe₃O₄ (a), Fe₃O₄@SiO₂ (b), Fe₃O₄@SiO₂@NH₂ (c), Fe₃O₄@SiO₂@NH₂@ α-Glucosidase (d), and EDX mapping of different elements in the Fe₃O₄@SiO₂@NH₂@α-Glucosidase (e).

Export to PPT

3.2.2

3.2.2 FT-IR analysis

FT-IR spectra of Fe₃O₄, Fe₃O₄@SiO₂ and Fe₃O₄@SiO₂@NH₂@α-Glucosidase are shown in Fig. 4. The peak at about 560 cm⁻¹ was ascribed to the vibrational absorption of Fe-O (Yang et al., 2019). The peaks observed at about 460 cm⁻¹, about 789 cm⁻¹ and about 1086 cm⁻¹ were assigned to the bending vibration of O-Si-O, symmetric stretching vibration of Si-O-Si and stretching vibration of Si-O-Fe; and the decrease in the intensity of the Fe-O adsorption vibration both indicated the generation of Fe₃O₄@SiO₂ (Shi et al., 2018). The presence of a new peak at 2155 cm⁻¹ was observed in the spectrum of Fe₃O₄@SiO₂@NH₂@α-Glucosidase represents C⚌N stretching vibration of Schiff base, which confirmed the linking of the –NH₂ group with the aldehyde group (–HCO) of GA (Agrawal et al., 2016). The overlapping of –NH and –OH stretching vibrations was assigned to a broad and strong band observed at about 3424 cm⁻¹ (Shi et al., 2018, Amirjani et al., 2022, Fan et al., 2022). The absorption peak observed at about about 1628 cm⁻¹ was assigned to the bending mode of absorbed molecule water on Fe₃O₄ (Wang et al., 2021b). In summary, the FT-IR analysis comfirmed the effective synthesis of Fe₃O₄@SiO₂@NH₂.

Close

Fig. 4

FT-IR spectra of sample: Fe₃O₄ (a); Fe₃O₄@SiO₂ (b); Fe₃O₄@SiO₂@NH₂@α-Glucosidase (c).

Export to PPT

3.3 Enzyme loading of Fe₃O₄@SiO₂@NH₂

We optimized the four key parameters and studied the performance of immobilized enzyme, and the result was presented in S 2.1 and S 2.2. Then, the efficiency of immobilization was initially examined. The quantification of enzyme loading on Fe₃O₄@SiO₂@NH₂ was determined to be 240.1 μg/mg, indicating successful immobilization of α-Glucosidase on Fe₃O₄@SiO₂@NH₂. The relation curve between the concentration and absorbance of the pNP solution was shown in Fig. 5a, and the activities of both immobilized and free enzyme were quantified. It was observed that the activity of the immobilized enzyme accounted for 33 % of the free enzyme, which was consistent with the literature (Guo et al., 2023). The decrease in enzyme activity can be attributed to the absence of the unbound form of the N-terminal region of α-Glucosidase.

Close

Fig. 5

The standard curve of p-nitrophenol (pNP) (a), and Lineweaver-Burk plots of the free and immobilized α-Glucosidase (b).

Export to PPT

3.4 Kinetic study

The Km is often used to evaluate the affinity of the enzyme for a specific substrate. The lower Km suggested higher level of affinity. The Lineweaver-Burk plots for Fe₃O₄@SiO₂@NH₂@α-Glucosidase and free α-Glucosidase (Fig. 5b) were as follows: 1/v = (0.13007 ± 0. 00262) (1/[S]) + (0. 21727 ± 0.00776) with R² = 0.9980 and 1/v = (0.07633 ± 0.00171) (1/[S]) + (0.09559 ± 0.00507) with R² = 0.9975, individually. And the values of Km were determined as 0.59 mM and 0.80 mM, respectively. The results showed that the affinity of α-Glucosidase was enhanced after immobilization. The reason is that the immobilized enzyme active site exhibited enhanced substrate accessibility and a reduced diffusion barrier effect compared with the free enzyme (Arana-Peña et al., 2021). The affinity changes obtained were consistent with some literatures (Li et al., 2020b; Wan et al., 2021b).

3.5 Identification of α-Glucosidase inhibitors

The base peak ion chromatograms (BPCs) of L0, L1 and blank are shown in Fig. 6. On the basis of their specific molecular weight, retention duration and in-house database, the chemical structures of inhibitors were inferred, which was shown in Table 1. As a consequence, 20 compounds were identified, including 9 flavonoids, 3 phenylpropanoids, 3 anthraquinones, 2 organic acids, 1 alkaloid, 1 terpenoid and 1 other class.

Close

Fig. 6

The base peak ion chromatograms (BPCs) of L0 (A), L1 (B) and Blank (C).

Export to PPT

3.6 Molecular docking

To further verify whether the 20 compounds had α-Glucosidase inhibitory, Molecular docking technique was employed to examine the affinity of the candidate compounds with α-Glucosidase. The docking results are shown in Table 2, To improve the screening accuracy, the affinity < -8.0 kcal/mol was used as the screening condition for potential inhibitors in this study.

Limonexic acid, Sanggenol A, Glabrone, Matrine exhibited the high affinity with α-Glucosidase and, thus, displayed potential inhibitory activity against α-Glucosidase. The protein–ligand interactions of the 4 compounds were presented in Fig. 7 and illustrated in Table 3. As shown in Fig. 7a, strong hydrophobic interactions between Limonexic acid and active-site amino acid residues play an important role in their binding. The following residues, namely Lys 156, Tyr 158, Phe 303, Leu 313, Arg 315 and Arg 442, were found to be involved in both bonding and non-bonding interactions with Limonexic acid. The main characteristic of Limonexic acid is that it is composed of oxygen-containing functional groups, including ester group, ether group, carbonyl group, oxhydryl group. The ester group was bound with Lys 156 and Arg 442 via conventional hydrogen bonds, while the oxhydryl group was also bound with Leu 313 via conventional hydrogen bonds. These hydrogen bonds play a important role in enhancing the stability of the protein–ligand complex. Furthermore, the oxhydryl group played a crucial role in stabilizing the complex between the protein and ligand by generating an electrostatic attraction effect. The ether group was bound with Arg 442 via carbon hydrogen bond and Arg 442 constitutes a crucial part of the protein's residues responsible for regulating its active conformation. And six-membered ring and its substituted methyl groups were bound with Tyr 158, Phe 303 and Arg 315 via pi-alkyl bonds and alkyl bonds. The pi-alkyl interaction plays a crucial role in enhancing the stability of the protein–ligand complexes. As shown in Fig. 7b, the docked conformation of the Sannggenol A depicting the possible interactions with various amino acids of 3A4A: Lys 156, Tyr 158, Gly 161, Leu 177, Leu 313, Ala 418, Iso 419 and Glu 422. The methyl group showed a number of hydrophobic interactions with Lys 156, Tyr 158 and Leu 177. Furthermore, the benzene ring was bound with Lys 156, Ala 418 and Iso 419 via hydrophobic interactions, pi-alkyl interactions and alkyl interactions. In addition, the electropositive hydrogen atom of the oxhydryl group and the carbonyl group segment were bound with Lys 156, Tyr 158, Gly 161, Leu 313 and Glu 422 via strong hydrogen bonding. The presence of hydrogen bonding plays a crucial role in preserving the stability of the conformation. These interactions allow Sannggenol A and 3A4A to bind firmly together. As shown in Fig. 7c, Glabrone exhibited interactions with the subsequent amino acids via both bonding and non-bonding interactions: Lys 156, Tyr 158, Val 232, Pro 312, Leu 313, Phe 314 and Asn 415. The methyl group was involved in hydrophobic interactions with Val 232. The benzene ring was involved in important hydrophobic interactions:pi-pi T-shaped interaction with Phe 314, pi-alkyl and alkyl interactions with Val 232 and Leu 313, respectively. Moreover, carbon hydrogen bond was formed between carbon–carbon double bond and Pro 312. The electropositive hydrogen atom and electronegative oxygen atom of oxhydryl group and electronegative oxygen atom of carbonyl group were bound with with Pro 312, Tyr 158, Asn 415 and Lys 156. These high-affinity interactions allow Glabrone to be firmly bound to Glabrone. As shown in Fig. 7d, the conformation of Matrine, when docked with 3A4A, displayed strong interactions with several amino acids: Lys 156, Tyr 158, Arg 315 and Asp 242. The pyrimidine ring was involved in important hydrophobic interactions: pi-alkyl and alkyl interactions with Tyr 158 and alkyl interactions with Lys 156. Moreover, ortho-position of pyridine N part was bound to Asp 242 via Carbon hydrogen bond. The electronegative oxygen atom of carbonyl group was bound with Arg 315 via hydrogen bond. These interactions allow Matrine and 3A4A to bind firmly together. It can be seen that the hydrogen bond and hydrophobic interaction between ligand and enzyme may be beneficial to enhance the stability of inhibitor-enzyme complex, and thus play an important role in the inhibitory activity of α-Glucosidase.

Close

Fig. 7

2D, 3D molecular docking graphs of Limonexic acid (a), Sanggenol A (b), Glabrone (c), Matrine (d) with 3A4A.

Export to PPT

3.7 Inhibitory activity of the potential inhibitors

The dose–response curves in Fig. 8 illustrate the inhibitory effects of the ligands and acarbose on α-Glucosidase activity. It was determined that the IC₅₀ values of Acarbose, Limonexic acid, Sanggenol A, Glabrone and Matrine were measured at 16.57 μM, 5.91 μM, 8.52 μM, 10.91 μM, 11.51 μM, respectively, which showed that the inhibitory of Limonexic acid, Sanggenol A, Glabrone and Matrine surpasses that of Acarbose. Results of molecular docking and inhibition assays show that these 4 compounds exhibit promising prospects as α-Glucosidase inhibitors in field drug.

Close

Fig. 8

The inhibition curves of Limonexic acid, Sanggenol A, Glabrone, Matrine, and Acarbose.

Export to PPT

3.8 Quantitative analysis

Quantitative analysis of the 4 compounds screened were carried out by HPLC. We determined the presence of 4 four compounds by comparison to the references, as shown in Fig. 9.

Close

Fig. 9

Chromatograms of sample (A), reference substances (B). 1-Limonexic acid; 2-Matrine; 3-Sanggenol A; 4-Glabrone.

Export to PPT

4.1 Selection of sample solution preparation

During the preparation of the sample solution, different extraction solvents (methanol, 50 % methanol, ethanol and 50 % ethanol), different extraction methods (reflux, ultrasound) and different extraction times (20, 30 and 40 min) were used to investigate the preparation method of UCG sample solution. The results showed that various spectral peaks was significant, and the overall peak shape was well separated with 50 % ethanol. After comprehensive consideration, it was determined to use 50 % ethanol solution as the extraction solvent, and the ultrasonic time was 30 min.

4.2 Selection of chromatographic conditions

This study investigated the mobile phase by gradient elution using methanol–water, acetonitrile-0.1 % formic acid and acetonitrile-0.1 % phosphoric acid, respectively. The results showed that when acetonitrile-0.1 % formic acid was used as the mobile phase, the separation effect of each chromatographic peak was better, and the baseline was stable. Finally, acetonitrile-0.1 % formic acid was determined as the mobile phase system for this experiment. Then, the column temperatures (20, 25, 30 and 35 °C) were investigated. The results showed that the column temperatures were 35 °C, and the chromatographic peaks were well separated. Therefore, the above chromatographic conditions were used in the experiment to ensure the method's accuracy.