Chemical Characterization and Metabolic Profiling of the Compounds in the Chinese Herbal Formula Li Chang Decoction by UPLC-QTOF/MS

Background Li Chang decoction (LCD), a Chinese medicine formula, is commonly used to treat ulcerative colitis (UC) in clinics. Purpose This study aimed to identify the major components in LCD and its prototype and metabolic components in rat biological samples. Methods The chemical constituents in LCD were identified by establishing a reliable ultra-performance liquid chromatography coupled with quadrupole time-of-flight tandem mass spectrometry (UPLC-QTOF/MS) method. Afterwards, the rats were orally administered with LCD, and the biological samples (plasma, urine, and feces) were collected for further analyzing the effective compounds in the treatment of UC. Result A total of 104 compounds were discriminated in LCD, including 26 flavonoids, 20 organic acids, 20 saponins, 8 amino acids, 5 oligosaccharides, 5 tannins, 3 lignans, 2 alkaloids, and 15 others (nucleosides, glycosides, esters, etc.). About 50 prototype and 94 metabolic components of LCD were identified in biological samples. In total, 29 prototype components and 22 metabolic types were detected in plasma. About 27 prototypes and 96 metabolites were discriminated in urine, and 34 prototypes and 18 metabolites were identified in feces. Conclusion The flavonoids, organic acids, and saponins were the major compounds of LCD, and this study promotes the further pharmacokinetic and pharmacological evaluation of LCD.


Introduction
Traditional Chinese medicine (TCM) attracts more attention in the world since it possesses reliable therapeutic efficacy in some complex diseases, especially chronic illness [1]. e chemical composition of Chinese herbal compound is complex, and the composition of the multi-Chinese medicine is crossed, summarized as "multitarget and multicomponent," which is the feature of TCM [2,3]. is characteristic promotes the curative effect and reduces toxicity; however, it brings enormous challenge to figure out the effective components and mechanism for the therapeutic effect [4].
Li Chang decoction (LCD), a Chinese compound prepared from twelve Chinese medicine including Codonopsis Radix (CR), Notoginseng Radix et Rhizoma (NRR), Bletillae research on the effective component and metabolite profiling of LCD is an urgent need.
Ultra-performance liquid chromatography coupled with quadrupole time-of-flight tandem mass spectrometry (UPLC-QTOF/MS) provides a rapid and reliable method to identify the component of natural medicine, which promotes the development of natural medicine component analysis and new drug discovery [9,10]. Herein, we recruited an UPLC-QTOF/ MS method to profile the effective components of LCD, and the unknown components were classified and assigned based on the fragmentation patterns and diagnostic ions of different structural types of components. According to the component characterization result of LCD in vitro, the prototypes in plasma, urine, and feces were further analyzed based on the similarity of mass spectrometry behavior (accurate molecular weight and secondary fragments) and chromatographic behavior (retention time). Metabolites were matched e mass defect filtering (MDF) caused by biotransformation and were further confirmed by MS/MS spectrum analysis.  magnetic beads, the samples were homogenized using tissue grinders (Shanghai Jingxin, Shanghai, China) and centrifuged at 13000 rpm, 4°C, 10 min. About 200 μl supernatant was removed, dried under nitrogen gas, and redissolved in 200 μl acetonitrile (50%). Finally, the samples were centrifuged at 13000 rpm, 4°C, 10 min, and a 2 μl aliquot was injected into UPLC-QTOF-MS. For the urine sample, the mixed urine was centrifuged at 4000 rpm for 10 min, and 1 ml supernatant was loaded on pre-activated Sep-Pak Vac C18 columns (3 cc, 500 mg, Waters, Ireland). After washing with 1 ml ultrapure water and eluting with 1 ml methanol, the elution was collected and centrifuged at 13000 rpm, 4°C, 10 min. About 400 μl supernatant was transferred and dried under nitrogen gas. e residues were redissolved in 400 μl acetonitrile (50%). Finally, the samples were centrifuged at 13000 rpm, 4°C, 10 min, and a 2 μl aliquot was injected into UPLC-QTOF-MS.

UPLC-QTOF-MS Analysis Condition.
e separation equipment for this assay was Sciex Exion LC, and the chromatographic column was Waters Acquity HSS T3 (2.1 × 150 mm, 1.7 μm). e temperature was set at 35°C, and the flow rate was 0.3 ml/min. e mobile phases were 0.1% formic acid in water (A) and acetonitrile (B), with the optimized gradient as follows: 0-5 min from 3% B to 8% B, 5-11 min from 8% B to 30% B, 11-20 min from 30% B to 80% B, 20-21 min from 80% B to 95% B, 21-25 min was maintained at 95% B, and then back to the initial ratio and re-equilibration for 7 min. e 5600 QTOF mass spectrometer (AB Sciex, Foster City, CA, USA) equipped with an ESI ion source was operated in positive and negative modes, and the mass range was m/z of 100-1250. e details of mass spectrometry conditions were summarized as follows: gas 1 and gas 2, 45 psi; curtain gas, 35 psi; heat block temperature, 500°C; ion spray voltage, −4.5 kV in negative mode and 5.5 kV in positive; declustering potential, 50V; collision energy, ±35 V; and the collision energy spread (CES), ±15 V. Sciex OS 1.6.1 was the basal data processing platform, and MetabolitePilot 2.0.4 software was applied for further metabolite fishing.

Characterization of Chemical Compounds in LCD.
e base peak chromatograms of LCD in negative and positive ion modes are shown in Figure 2. A total of 104 chemical components, including 20 saponins, 26 flavonoids, 5 tannins, 20 organic acids, 8 amino acids, 2 alkaloids, 5 oligosaccharides, and 3 lignans, were identified or tentatively characterized by UPLC-QTOF-MS. As the result of chemical composition classification is summarized in Table 1, CR mainly contained alkaloid compounds and oligosaccharides, while NRR was characterized by saponins. Besides, the major constituents of SF were flavonoids. GRR contains saponins and flavonoids, and CPRR was as characterized by the C21 type steroidal saponins. e characteristic ingredients of TP were flavonoids and organic acids. CF was characterized by the component of tannins; AMR contains organic acids and esters.

Characterization of LCD-Related Xenobiotics in Rat Biological Samples.
According to the compound characterization of LCD, the fragmentation patterns of mass spectrometry (accurate molecular weight and secondary debris) and retention time of chromatography were adopted to analyze the components in plasma, urine, and feces. P59 ginsenoside Rg1 is taken as example, as shown in the XIC of LCD ( Figure 5(a)) and multiple XICs of 6 bio-samples ( Figure 5(b)), and a peak at 13.4 min was clearly observed in administration of bio-samples but not in the blanks.
Importantly, the MS/MS spectra (m/z of 621, 441, 423, 405, and 203) of ginsenoside Rg1 in LCD ( Figure 5(c)) and biosamples ( Figure 5(d)) were similar. Based on the above principles, a total of 50 components were matched in biological samples, and these components would play a key role in explaining the mechanism of LCD in the future. In particular, flavonoids (P43, P46, and P50) and saponins (P55 and P72) deserved higher attention as the five components were observed in all three bio-samples besides that were common to organisms (P1, P11, P15, P24, P31,  and P68). In addition, 12 compounds were just observed in the fecal sample, mainly including some alkaloids (P25 and P65), flavonoids (P37, P42, P45, P52, and P70), saponin (P74), and other small molecules (P6, P9, P35, and P40).      ese compounds may not be absorbed into the blood, but are still effective in regulating gut microbiota. e detailed information about the distribution of components in plasma, urine, and feces is summarized in Table 2.

Evidence-Based Complementary and Alternative Medicine
Furtherly, the phase I and phase II metabolic regularity, as well as the similarity of secondary mass spectrum profile, was used to identify the metabolite. ose metabolites were annotated through automatic matching with prototype components by MetabolitePilot Software. Briefly, Metabo-litePilot operated prototype-metabolite matching through mass defect filter (MDF), characteristic product ion filter (PIF), and neutral loss filter (NLF). As shown in Figure 6, the mass defect from P50 to M70/71 was -148 Da with the biotransformation named "loss of C 6 H 10 O 4 and O (hydrolysis, phase I) + ketone formation (phase I)." Furthermore, neutral loss of glycosides and methylene was both observed in the MS/MS spectra of P50 and M70/71, which implied the similar skeleton. at was to say, these compounds were structurally related, and M70/71 could be the metabolites of P50. As a result, a total of 107 metabolites were matched with 25 prototypes in plasma, urine, or feces. e network of prototype-metabolite matching is drawn as in Figure 7. e details involving the distribution and biotransformations of metabolites are listed in Table 3. It was worth noting that although some prototypes have not been observed in biosamples, they still are effective through metabolites. For example, P28 hamamelitannin produced 14 metabolites that were all detected in urine, and 5 were found in plasma and 2 in feces. It could be metabolized in the gut, and metabolites were furtherly absorbed into the bloodstream. In total, 29 prototype components and 22 metabolites were detected in plasma. About 27 prototypes and 96 metabolites were detected in urine, and 34 prototypes and 18 metabolites were detected in feces.
ese substances were considered to constitute the pharmacodynamic substance basis of LCD.

Data Availability
e data used to support the findings of this study are included within the article.

Conflicts of Interest
None.