Chemical characterization and metabolic profiling of Xiao-Er-An-Shen Decoction by UPLC-QTOF/MS

Background: Xiao-Er-An-Shen decoction (XEASD), a TCM formula composed of sixteen Chinese medicinal herbs, has been used to alleviate tic disorders (TD) in clinical practice for many years. However, the chemical basis underlying the therapeutic effects of XEASD in the treatment of TD remains unknown. Purpose: The present study aimed to determine the major chemical components of XEASD and its prototype compounds and metabolites in mice biological samples. Methods: The chemical constituents in XEASD were identified using ultra-high Performance liquid chromatography coupled with quadrupole time-of-flight tandem mass spectrometry (UPLC-Q-TOF-MS/MS). Following this, XEASD was orally administered to mice, and samples of plasma, urine, feces, bile, and tissue were collected in order to identify effective compounds for the prevention or treatment of TD. Result: Of the total 184 compounds identified to be discriminated in the XEASD, comprising 44 flavonoids, 26 phenylpropanoids, 16 coumarins, 16 triterpenoids, 14 amino acids, 13 organic acids, 13 alkaloids, 13 ketones, 10 cyclic enol ether terpenes, 7 citrullines, 3 steroids, and 5 anthraquinones, and others. Furthermore, we summarized 54 prototype components and 78 metabolic products of XEASD, measured with biological samples, by estimating metabolic principal components, with four prototype compounds detected in plasma, 58 prototypes discriminated in urine, and 40 prototypes identified in feces. These results indicate that the Oroxylin A glucuronide from Citri reticulatae pericarpium (CRP) is a major compound with potential therapeutic effects identified in brain, while operating positive effect in inhibiting oxidative stress in vitro. Conclusion: In summary, our work delineates the chemical basis underlying the complexity of XEASD, providing insights into the therapeutic and metabolic pathways for TD. Various types of chemicals were explored in XEASD, including flavonoids, phenylpropanoids, coumarins, organic acids, triterpenoid saponins, and so on. This study can promote the further pharmacokinetic and pharmacological evaluation of XEASD.


Introduction
Tic disorder (TD) is a chronic neuropsychiatric condition characterized by motor and vocal tics lasting at least 1 year, including Tourette syndrome (TS) (McGuire et al., 2020).It typically develops in early childhood, with a global prevalence estimated to be between 0.3% and 1%.In China, the prevalence of TD is approximately 6.1% (Scharf et al., 2015).Children with TD often experience co-occurring psychiatric disorders, such as obsessive-compulsive disorder (OCD), attention-deficit/ hyperactivity disorder (ADHD), mood and anxiety disorders, and compulsive-like conditions such as hair-pulling and pathological skin-picking (Willsey et al., 2017;McGuire et al., 2021).These conditions, along with pain or injury, social isolation or bullying, and emotional problems, can seriously affect the quality of daily life for TD patients (Brander et al., 2021).Unfortunately, current psychotropic drug therapies have limited effectiveness in treating tics and TD, and may cause significant long-term side effects (Osland et al., 2018;Pringsheim et al., 2019).While firstgeneration typical antipsychotics like haloperidol have been used in treating TD by antagonizing dopamine receptor type 2 (D2R) in the brain, they are no longer considered first-line drugs due to potential toxicity (Pringsheim et al., 2019).Therefore, it is necessary and urgent to develop new therapeutics for TD.
Traditional Chinese medicine (TCM) has been recognized as an integral part of modern medicine, serving as a vital resource of natural medicines and playing a significant role in the treatment of various human diseases (Li and Kan, 2017).The distinctive characteristic of TCM preparations is the use of multiple herbs, which contain numerous active ingredients that work synergistically on various targets to treat diseases, thereby enhancing the therapeutic effects and minimizing toxicity (Han et al., 2020).Therefore, it is crucial to elucidate the chemical compositions and metabolite profiles of TCM to facilitate their standardization research.
Ourprevious invivostudieshave demonstratedthatthe mechanism ofactionofXEASDinalleviatingtwitchsymptomsandrelateddisorders is mainly related to reversing abnormal changes in neurotransmitter levels and enhancing the antioxidant status of the mouse brain, while in vitro experiments have illustrated its ability to modulate neuronal growth and antioxidant activity, thereby providing neuroprotective effects (Li et al., 2018;Chen et al., 2019).Both in vitro and in vivo results demonstrated the XEASD-induced increase in plasma cyclic adenosine monophosphate (cAMP) levels and the subsequent phosphorylation of cAMP response element-binding (CREB) protein.Although a quality standard has been established to detect the contents of glycyrrhizin, mauroisoflavone glucoside, ammonium glycyrrhizinate, naringin, 3,6′diglucosyl sucrose, hesperidin and neohesperidin in XEASD, there is still controversy and a lack of understanding regarding the pharmacodynamic material basis of XEASD (Hou-ming. et al., 2018).Hence, a systematic study of the active ingredients and metabolite profiles of XEASD is urgently needed.
Over the last decade, the use of UPLC-QTOF/MS has enabled the rapid and accurate identification of chemical compounds in complex Chinese medicines, natural products and formulas.This has driven the development of natural product analysis and drug design.In the present study, an analytical method of major ingredients based on the UPLC-QTOF/MS system was established.Unknown ingredients were categorized according to the fragmentation patterns and diagnostic ions of different structural types of ingredients.To further characterize XEASD components in vivo, the prototypes were analyzed in plasma, urine, feces, and bile, utilizing feature-based similarity in mass spectrometry response and chromatographic retention time.The relationship between bio-transformation and the role of biotransformed metabolites was identified through mass defect filtration (MDF) and further corroborated by MS/MS spectroscopy.S1).Neochlorogenic acid, Chlorogenic acid, Cryptochlorogenic acid, Esculetin, Caffeic acid, Loganin, Liquiritin, Nicotiflorin, Ferulic acid, Naringin, Cornuside, Hesperidin, Neohesperidin, Baicilin, Isoliquiritin, Calycosin, Glycyrrhizic acid, Limonin, Nobiletin, Obacunone were obtained from Chengdu Alfa Biotechnology Co., Ltd (Sichuang, China).MTT, palmitic acid and L-glucose (Sigma, USA).Oroxylin A glucuronide [Oroxyloside (purity >98%)], purchased from ACMEC Biochemical Co., Ltd.(Shanghai, China).For the stock solution, Oroxylin A glucuronide was dissolved in 100% dimethyl sulfoxide (DMSO).Bovine serum albumin (BSA) was purchased from Sigma-Aldrich (St. Louis, MO).Control groups received the same volume of solvent DMSO.The purity of each standard compound reported by HPLC analysis was more than 98%.All solutions were prepared from Milli-Q water (Milli-Q Ultrapure water systems, Millipore).Acetonitrile (LC-MS Grade, Optima) and formic acid (LC-MS Grade, Thermo Scientific Pierce) were purchased from Fisher Chemicals.

Mice treatment and sample collection
Male ICR mice (18-20 g, n = 8 in each group) were assigned randomly to one of three groups: including a control group for collecting blank bio-samples, a treatment group for collecting plasma, urine, feces, and tissues, and a treatment group for collecting bile.Mice in the control group were administered normal saline (NS) intragastrically.In the treatment group, mice were deprived of food (fasted) for 16 h before the administration of XEASD at the dose of 8 g/kg with free access to drinking water (Chen et al., 2019).Blood samples were collected from the retroorbital plexus of the mice at 0.25, 0.5, 1, 2, 4, 6, 8, and 10 h after treatment and placed into heparin anticoagulant tubes.The tubes were then centrifuged at 3,000 rpm for 10 min to obtain plasma samples, which were combined for each time point and stored at −80 °C until further analysis.Feces, urine, and bile acid samples were collected from each mouse at the indicated time points.This animal study was approved by the Ethics Committees of the Chinese University of Hong Kong (Shenzhen), and was conducted in accordance with the Chinese University of Hong Kong (Shenzhen) animal care regulations (CUHKSZ-AE202206).

Biological sample preparation
Plasma samples were collected and prepared by mixing approximately 200 μL plasma with 600 μL of acetonitrile containing 0.2% methanoic acid.After vortexing for 2 min, the samples were centrifuged at 13,000 rpm, 4 °C for 10 min.Then, 400 μL of the supernatant was removed, dried under nitrogen gas, and redissolved in 200 μL of 50% acetonitrile/50% water.
Finally, the samples were centrifuged at 13,000 rpm, 4 °C for 10 min.An aliquot of 2 μL sample was injected into UPLC-QTOF-MS.
The urine and feces samples were collected every 2 hours by placing mice in individual metabolic cages.The urine was centrifuged at 1,503 g (4,000 rpm) for 10 min, after which 1.5 mL of the supernatant was loaded onto a C 18 solid-phase extraction (SPE) column (Sep-Pak Vac 3 cc 500 mg, Waters, Ireland).The eluant was dried under nitrogen at room temperature and resuspended in 400 μL acetonitrile/water (1:1, v/v) before analysis.Fecal samples that were prepared by weighing approximately 300 mg of feces was placed in 2-mL polypropylene tube, and two volumes of methanol were then added to mix.Fecal extracts were homogenized with 2 mL tungsten carbide beads using a tissue grinder (Wuhan Servicebio, Wuhan, China) and centrifuged for 10 min (13,000 rpm, 4 °C), 400 μL of the supernatant were dried.The residue was reconstituted with 200 μL 50% acetonitrile in water (v: v), and the aliquot of 10 μL was injected into the LC-MS/MS system.
Mice were anesthetized with urethane (2.0 g/kg) and bile was collected continually via a cannula inserted into the bile duct at 2, 4, 6, 8, and 10 h, which drained into a collection tube.The bile sample was centrifuged at 4,000 rpm for 10 min, twice.Samples were then loaded on a 1.5 µL C18 pre-column (Optimize Technologies) and the procedure was the same format as urine.For the mice in the three groups, tissue, including liver, heart, spleen, kidney, lung, and brain, were collected after mice execution, respectively.A tissue lyser (Wuhan Servicebio, Wuhan, China) was used to homogenize 100 mg of tissues in 800 μL of methanol.After centrifuging the mixture for 15 min at 4 °C at 13,000 rpm, 400 μL of the supernatant were dried under nitrogen gas.The residue was reconstituted with 100 μL 50% acetonitrile in water (v: v), and the aliquot of 10 μL was injected into the LC-MS/MS system.

UPLC-QTOF-MS analysis condition
LC&MS/MS experiments were performed on an exion LC system (AB Sciex, Foster City, CA, USA).An Acquity HSS T3 column (1.8 μm, 2.1 × 150 mm) equipped with a VanGuard precolumn (1.8 μm; Waters Corporation) served for chromatographic separation.The mobile phases used for elution were (A) 0.1% (v/v) formic acid/water and (B) acetonitrile.The UPLC eluting conditions were optimized as follows: 3%-7% acetonitrile for 0-5 min, 7%-30% acetonitrile for 5-12 min, 30%-80% acetonitrile for 12-20 min, 80%-95% acetonitrile for 20-21 min, and 95% acetonitrile for 21-27 min, then back to the initial ratio of 3% B and maintained with additional 10 min for reequilibration.The sample injection volume was 2 μL.MS data were recorded using an AB Sciex X500B QTOF mass spectrometer with an ESI source and operated in both the positive and the negative modes.MS conditions were set as follows: ions spray voltage -4500 V in negative mode and 5,500 V in positive, ion source heater temperature 500 °C, source gas 145 psi, source gas 245 psi, and curtain gas 35 psi.The declustering potential, collision energy and the collision energy spread (CES) were set at 50V, ±35V and 15V, respectively.The initial data was processed on the Sciex OS 1.6.1 platform, followed by metabolite fishing using MetabolitePilot ™ 2.0.4 software (Peak finding strategy combined mass defect filter (MDF), characteristic product ion filter (PIF), and neutral loss filter (NLF).Set MS m/z tolerance at 10 ppm and minimum peak intensity at 1,000 cps.Sample-control ratio was set at 3).

Cell survival and proliferation assays
Cells were evaluated with the 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) assay (Dong et al., 2014;Xu et al., 2018).In a nutshell, 20 μL of MTT solution was added to the culture medium at a final concentration of 0.5 mg mL −1 and incubated at 37 °C for 4 h.Then, the supernatants were aspirated carefully, 150 μL of DMSO was added to each well to dissolve the reaction product insoluble formazan of MTT, and the OD was spectrophotometrically measured using a microplate reader (BioTek, uQuant, Santa Barbara, CA, USA) at a wavelength of 570 nm, with DMSO as a blank.

Statistical analysis
Statistical analysis was conducted using GraphPad Prism 9 statistical software.Mean values were compared between control and treatment groups using One-way ANOVA analysis.A p-value less than 0.05 was considered statistically significant (*), while a p-value less than 0.01 was considered highly significant (**).

Characterization of chemical compounds in XEASD
Representative base peak chromatogram (BPC) of XEASD in the positive ion mode and negative ion mode are shown in Figure 1.Initial analysis identified or tentatively characterized a total of 198 chemical components by UPLC-QTOF-MS (Supplementary Table S2), including 44 flavonoids, 26 phenylpropanoids, 16 coumarins, 16 triterpenoids, 14 amino acids, 13 organic acids, 13 alkaloids, 13 ketones, 10 cyclic enol ether terpenes, 7 citrullines, 3 steroids, and 5 anthraquinones.Pooled outcome findings are summarized in Table 1.ATR calamus was found to be rich in chemical components, including flavonoids, alkaloids, phenylpropanoids, and ketones, with specific characteristic components such as Acoramone and Tatarine C. ACB cholonan was characterized by characteristic bile acid compounds, while triterpenoids were the characteristic components of P. The major characteristic ingredients of CRP were flavonoid.Coumarins were identified in NRR.C contained cyclic enol ether terpenoids such as loganin and morroniside, while AOF had characteristic naphthones.POR contained   The number in the brackets was the repeat compounds.
Frontiers in Pharmacology frontiersin.organthraquinones and phenylpropanoid glycosides as its characteristic components.AR mainly contained flavonoids in this experiment, while CRP had a high degree of similarity in composition with AF, with a large number of flavonoids, as well as a few coumarins and citrulline.PIR and TLF were mainly composed of amino acid components, while HFG contained Hordenine as its characteristic component, in addition to some organic acids and amino acid compounds.ML was difficult to attribute as a curative component, and only 7 compounds were identified in this experiment, but it contained alkaloids, flavonoids, organic acids, phenylpropanoids, and other components.Finally, GF was mainly characterized by organic acids and phenylpropanoids.Figure 2 illustrates the representative structures of each herb.3.2 Fragmentation mechanisms of medicine representative structures

Polygalae Radix -derived compounds
The typical fragmentation pathways of P68 Sibiricaxanthone B are shown in Figure 3C.

Characterization of XEASD-Related xenobiotics in mice biological samples
Based on the chemical characterization of XEASD, the MS/MS fragmentation patterns and retention time were used to analyze the components in plasma, urine, feces, and bile.P53 Loganin was selected as an example, and its XIC in XEASD (Figure 4A) and multiple XICs of 8 bio-samples (Figure 4B) were examined.A peak at 9.35 min was clearly observed, which only appeared in the administered urine of bio-samples and responded at the same retention time as the extracts.Of great significance, the MS/MS spectra (m/z of 435, 227, and 127) of Loganin in XEASD (Figure 4C) and bio-samples (Figure 4D) were found to be comparable.
Based on the previous results, a total of 54 prototypes were detected from plasma, urine, feces, or bile samples.Among them, 2 compounds were detected in plasma, 27 in urine, 44 in feces, and 6 in bile.These compounds in feces may not be absorbed into the bloodstream, which could be still helpful in modulating the fecal microbiota.Elaborate process distribution of the prototypes is shown in Table 2.
The metabolic patterns of phase I and phase II were used as the basis for the similarity of secondary mass spectrometry profiles to achieve rapid screening of metabolite libraries from the matrix that are distinct from the prototype components.This automatic matching the prototype components assist in the identification and annotation of metabolites, as illustrated in Figure 5.For example, the mass deviation between P20 and M1 were Δm = 176.0318,which is consistent with the biotransformation pathway "glucuronidation".Comparing the secondary mass spectra of P20 and M1, it was possible to spot the distinctive fragments produced when the GlcA group was taken out of M1.There is a high similarity between the two secondary profiles, including m/z 121, 103, 93, 91, 77, etc.Therefore, M1 can be matched as one of the metabolites of P20.Following this principle, a total of six metabolites were matched to P20, and their structural association diagrams are presented in Figure 6.
In this regard, further metabolite analysis was performed on the 54 prototype compounds mentioned earlier, of which 41 prototypes could be matched to metabolites, resulting in a total of 78 matched metabolites.Among them, four were detected in plasma, fifty-eight in urine, forty in feces and twenty-three in bile.The detailed distribution of metabolites has been shown in Table 3.   shows the association network between the prototypes and the related metabolites.Metabolites detected in feces are assumed to be metabolized by intestinal flora, while those metabolized by the liver may be detected in bile, plasma, and urine.Detailed biotransformation and annotation of prototype and metabolite components are presented in Supplementary Table S3.
The analysis of tissue distribution of compounds mentioned above showed that a limited number of compounds were distributed in tissues, and none were detected in lung and spleen tissues.Specifically, P135 Oroxylin A glucuronide was the only prototype detected in the brain tissue, and P73 Polygalaxanthone III was metabolized into M8, which was identified in the heart tissue.Moreover, P46 Polygalatenoside A was metabolized into M54, which was detected in the kidney tissue, while P174 Glycyrrhizic acid was metabolized into M76 and identified in the liver tissue (Figure 8).These ingredients are believed to be the underlying chemical basis for the therapeutic effects of XEASD.

Oroxylin A glucuronide operates positive effect in inhibiting oxidative stress in vitro
The results of cell viability assay (Figure 9A) and Annexin-V FITC/PI staining (Figure 9B) showed that 12-hr treatment of Oroxylin A glucuronide (Figure 8) at indicated concentrations (1, 10, and 100 µm) led to an enhanced effect on cell viability, and the apoptotic proportion was decreasing by Oroxylin A glucuronide in a dose-dependent manner, indicating that Oroxylin A glucuronide inhibited oxidative stress in cardiomyocytes with high glucose stimulation.

Discussion
These results of our study indicate that XEASD contains flavonoids, phenylpropanoids, triterpenoids and amino acids as its major chemical constituents.
Our findings build on prior work (Hou-ming. et al., 2018) by highlighting the importance of chemical characterization and metabolic profiling of XEASD.While GAO et al. used a single selective condition (HPLC-MS) to detect eight components (liquiritin, calycosin-7-O-β-Dglucoside, ammonium glycyrrhizinate, naringin, 3,6′-disinapoyl sucrose, hesperidin, neohesperidin, and astragaloside Ⅳ) in XEASD for quality control, our study analysed 54 prototypes of XEASD in various bodily fluids, including plasma, urine, feces, and bile.A total of 78 metabolites were discovered after 41 of these prototypes could be matched to existing metabolites.Overall, our study provides a more comprehensive understanding of the chemical composition and metabolic fate of XEASD.We were able to learn how XEASD is metabolized in the body by examining a variety of prototypes in vivo, which has implications for future studies on theeffectiveness and safety of XEASD.

FIGURE 8
Herbal ingredients form XEASD and major compounds with potential therapeutic effects identified in various organs.
Frontiers in Pharmacology frontiersin.org15 Yang et al. 10.3389/fphar.2023.1219866through activation of the ERK/CREB signaling pathway.P107 Hesperidin (Hajialyani et al., 2019)  Naringenin (Goyal et al., 2022), P87 Liquiritin (Qin et al., 2022), P101 Rhoifolin (Brinza et al., 2020;Mai et al., 2022), P146 Calycosin (Deng et al., 2021), P97 Naringin (Ahmed et al., 2019;Zhou et al., 2019), P56 Sweroside (Yang et al., 2020;Brinza et al., 2022) and P174 Glycyrrhizic acid (Cao et al., 2020) have been demonstrated to possess therapeutic effects in a variety of neurological disorders, such as antidepressants, as well as protecting the liver.Furthermore, P85 Eriocitrin (Meng et al., 2022) and P168 Monohydroxy-tetramethoxyflavone (Zheng et al., 2020), have shown beneficial effects on microbiota and bacterial diversity, thereby improving wasting muscle atrophy or ameliorating splenomegaly-related diseases.In contrast, to recent work that focused solely on in vitro chemical composition analysis and network pharmacology analysis of a TCM-derived product (Jingxin Zhidong Formula) for TD treatment (Tian et al., 2022), our study analyzed in vivo blood components and tissue distribution, as well as correlated in vitro chemical composition prototypes and metabolized metabolites.These compounds may have potential therapeutic applications for various conditions and represent important quality control markers for XEASD.These findings shed new light on the mechanisms underlying neurodegeneration in TD diseases.For instance, P135 Oroxylin A glucuronide was detected in brain tissue, and in vitro experiments have demonstrated that it can effectively inhibit high glucose-stimulated cardiomyocytes, which was tightly associated with oxidative stress (Chen et al., 2023), while in vitro and in vivo studies have shown that the main mechanism of XEASD for TD is closely related to the inhibition of oxidative stress.
However, identification of active compounds is a critical first step in the mechanistic analysis of how herbal macrobiotics affect various aspects of body functions, once these key compounds have been identified, further extensive work is required.For example, it is important to determine whether prototypical compounds involved in metabolic reactions are closely linked to the progression of specific diseases.Another crucial objective is the identification of target molecules affected by the active components and the key signaling pathways involved in their biological functions.The present study exemplifies the identification and characterization of cleavage patterns of potentially key compounds or metabolites in the context of specific diseases and herbal formulas, providing valuable insights for the development of targeted therapies.

Conclusion
The potentially major compounds identified in XEASD were flavonoids, phenylpropanoids, triterpenoids, and amino acids.The chemical basis underlying the beneficial effects of XEASD against TD may be attributed to active compounds such as P46 Polygalatenoside A, P73 Polygalaxanthone III, P135 Oroxylin A Glucuronide, P174 Glycyrrhizic acid as well as the metabolites of liver metabolisms and fecal metabolites.In conclusion, this study highlights the need for further pharmacokinetic and pharmacological evaluation of XEASD.

FIGURE 2
FIGURE 2Representative structures of each medicine of XEASD.
with m/z of 171, and [M- C 8 H 16 O 2 -3H 2 O-H] + with m/z of 139.Typical MS/MS fragmentation patterns of P166 are illustrated in Figure 3M. Figure 7

FIGURE 4
FIGURE 4 Identification of prototypes in bio-samples, and P53 Loganin is taken as an example.(A) multiple XICs of Loganin in XEASD and bio-samples, A-Administration,B-Blank.; (B) MS/MS spectrum of Loganin in XEASD; (C) MS/MS spectrum of Loganin in urine.

FIGURE 9
FIGURE 9 Effect of Oroxylin A glucuronide on the proliferation and apoptosis in H9C2 cells.(A) H9C2 cells were treated with Oroxylin A glucuronide (1, 10, and 100 µm) for 12 h and cell viability was measured by MTT assay.(B) The apoptosis-inducing effect of Oroxylin A glucuronide on H9C2 cells was tested by Annexin-V FITC/PI double-staining assay.Each experiment was done independently at least three times and all the data is quantified as the mean ± SEM. **p-value < 0.01 and ***p-value < 0.001 compared with the model group and ###p-value < 0.001 compared with the control group.

TABLE 1
Chemical component of XEASD.
Triterpenoid acids generally respond in the positive mode.Pinicolic acid E (P172, [M + H] + at m/z of 471.3467) exhibited fragment ions at [M-H 2 O+ H] + at m/z of 453, [M-CH 2 O 2 +H] + at m/ z of 407, [M-C 9 H 14 O 2 + H] + at m/z of 317, and [M-H 2 O-C 6 H 12 -C 9 H 16 O+H] + at m/z of 235.The typical fragmentation pathways of P172 Pinicolic acid E are shown in Figure 3K.

TABLE 2 (
Continued) Distribution of prototype compounds in vivo.