Metabolomics study of Angelica sinensis (Oliv.) Diels on the abnormal uterine bleeding rats by ultra‐performance liquid chromatography–quadrupole–time‐of‐flight mass spectrometry analysis

Abstract The objective of this study was to explore the effects and underlying intervention mechanisms of Angelica water extract (AWE) on abnormal uterine bleeding (AUB) based on serum metabolomics. Firstly, the concentration of main active substances in AWE was determined and the chemical components were identified by UPLC‐Q‐Exactive Orbitrap‐MS/MS. A drug‐induced abortion model was established by mifepristone and misoprostol. After administration AWE (2.16 g/kg) for 7 days, the coagulation function, serum hormone levels, H&E staining, and immunohistochemistry observation of uterus were detected. In addition, serum metabolites profiles were performed on ultra‐performance liquid chromatography–quadrupole–time‐of‐flight mass spectrometry (UPLC‐Q‐TOF‐MS). The contents of ferulic acid, senkyunolide A, and ligustilide in AWE were 0.7276, 0.0868, and 1.9908 mg/g, respectively. Twenty‐six compounds were identified in AWE. It was found that AWE was effective in regulation of coagulation function and promoting endometrial recovery. Meanwhile, the levels of E2, Pg, and HCG and the expression of ERα, Erβ, and PR were down‐regulated in AUB model and up‐regulated by the treatment of AWE. Twenty‐one potential biomarkers were eventually identified by multivariate statistical analysis. Study indicated that glycerophospholipid, sphingolipid, amino acids, retinol metabolism and primary bile acid biosynthesis were the main related metabolic pathways involved for the treatment of AUB by AWE. The results showed that AWE has potential therapeutic effect on AUB by altering the metabolic aberrations.


| INTRODUC TI ON
In the early 1970s and 1980s, medical abortion became an alternative way for early termination of pregnancy (Regina et al., 2011).
Mifepristone combined with misoprostol is the preferred clinical approach for the induction of abortion (Klaira & Paul, 2020), but the severe side effects of incomplete medical abortion still reached 15% (Ma et al., 2016). Abnormal uterine bleeding (AUB) is a common characteristic in incomplete medical abortion, with an incidence of approximately 3%-30% among reproductive-aged women (Munro et al., 2018). The use of estrogen, tranexamic acid, multi-dose compound contraceptive, and multi-dose progesterone regimen are common clinically available non-surgical options for AUB (Bradley & Gueye, 2015), but these treatments may cause multiple side effects (Yujie et al., 2018).
Angelica sinensis (Oliv.) Diels (Danggui) was widely used as a functional food or a dietary supplement in Asia, Europe, and America. It is also a famous traditional Chinese medicine for the treatment of anemia, dysmenorrhea, premenstrual, menopausal syndrome, and other gynecological diseases (Ma et al., 2015). Previous research indicates that polysaccharides, volatile oils, and organic acids are the main bioactive ingredients of A. sinensis. (Jin et al., 2016;Wei et al. 2016). The results of pharmacological studies indicated that A. sinensis can replenish and invigorate blood, prevent pain, and moisten the intestines (Ma et al., 2016). In addition, A. sinensis also have antiarrhythmic effects, enhanced immune function, cardioprotective effects, anti-atherosclerotic effects, and inhibiting platelet aggregation (Gu et al., 2016). However, few reports focus on the underlying therapeutic effects and mechanisms of A. sinensis for AUB.
Metabolomics is a technique to study the metabolites and their dynamic changes before and after being stimulated or disturbed for the biological system (e.g., after a specific gene variation or environmental change). Metabolomics has been widely used in plant molecular phenotype (Showkat et al., 2019), drug safety Yu et al., 2018), molecular pathology (Yang & Lao, 2019), mechanism of drug action (Su et al., 2020), and disease diagnosis (Karakioulaki & Stolz, 2019). In addition, it was also used to evaluate the impact of storage environment on food quality (Guo et al., 2019). Metabolomics is considered to be a useful strategy to explain the underlying mechanisms of TCM for the treatment of diseases. It emphasizes the study objects (humans or animals) as a unified whole, which is in accordance with the principle of integrity and dynamics of Traditional Chinese Medicine (TCM). The analysis methods of metabolomics mainly include nuclear magnetic resonance spectroscopy (NMR), gas chromatography-mass spectrometry (GC-MS), and liquid chromatography-mass spectrometry (LC-MS). Among all of the analytical technologies, UPLC-Q-TOF-MS is becoming a key technology in biomarker discovery (Gika et al., 2014).
In the present study, the concentration of main active substances in AWE was determined and the chemical components were identified by UPLC-Q-Exactive Orbitrap-MS/MS. The coagulation function, serum hormone levels, H&E staining, and immunohistochemistry observation of uterus were detected after the intervention of AWE. An UPLC-Q-TOF-MS method and multivariate analysis were applied to identify the potentially authentic biomarkers. The purpose of this study was to reveal the underlying mechanisms of AWE for the treatment of AUB and to provide a theoretical basis for clinical application.  (batch No. 20,200,528) was brought from Bioss Co., Ltd. (Peking, China). PBS (batch No. 20,190,307) was brought from Zsbio Co., Ltd. (Peking, China). DAB (batch No. 07,062,019) was brought from Baso Co., Ltd. (Zhuhai, China).

| Preparation of the water extract of Angelica sinensis
The water extract of A. sinensis (AWE) was extracted by heating reflux method. Thirty grams of crude herbal drugs was added to purified water ten times and extracted for twice (v/w), each extraction time was 30 min. The concentration of AWE was 0.6 g/ml after filtration (expressed by the weight per mL of crude drugs).
The MS operating parameters were as follows: the ion mode was positive; ion spray voltages, 3.5 kV; turbo spray temperature, 320℃; and m/z range, 100-1000. The main chemical constituents of A. sinensis were identified according to the exact molecular mass, the cleavage fragments of MS2, the mz cloud, mzVault 2.0 MS database, and literature review.

| Determination of ferulic acid, senkyunolide A, and ligustilide in AWE by UPLC-MS
The AWE was mixed with 50% methanol (1:1) and filtered through a 0.22 µm membrane filter. The concentration of AWE was 0.022 g/ ml after filtration (expressed by the weight per mL of crude drugs).

| Histopathological examination
The uterine tissues were immediately dissected after the experiment, removed fat and connective tissue, and fixed in 4% paraformaldehyde solution. Then, the uterine tissues were dehydrated at 4℃ for 24-48 hr, conventionally paraffin embedded, sectioned at 4 µm, and stained with hematoxylin-eosin (HE). Pathological changes of the endometrium were observed and photographed under a microscope.

| Measurement of serum hormone levels
The serum levels of progesterone (Pg), estradiol (E 2 ), and human chorionic gonadotrophin (HCG) were measured according to the instructions of manufacturer of the ELISA kits, respectively.

| Detection of plasma coagulation function
Collected blood (3 ml) from the abdominal aorta with sodium citrate at a mass concentration of 3.8 g/L (anticoagulant: blood = 1:9) was centrifuged to obtain plasma. Prothrombin time (PT), thrombin time (TT), activated partial thrombin activity time (APTT), and fibrinogen (FIB) were determined by a hemagglutination analyzer.

| Protein distribution analyses by immunohistochemistry
The uterine tissues were dehydrated, defatted, and conventionally paraffin embedded. Then, the uterine tissues were sectioned at 4 µm, deparaffinized, and rehydrated. After that, the tissue sections were incubated with 3% H 2 O 2 , repaired in antigen recovery solution, and then sealed at room temperature for 20 min after drip-adding normal goat serum blocking solution. Then, the samples were incubated at 4 ℃ overnight with primary inhibitors (dilution 1:200): estrogen receptor α (ERα), estrogen receptor β (ERβ), and progesterone receptor (PR). One day after incubation, the cells were washed with PBS three times for 5 min each time, and then, drip-added and incubated the secondary antibody at room temperature for 1 hr. SABC was added and incubated at 37℃ for 1 hr. Stained the proteins to dark-brown by immersion in diaminobenzidine. The slices were rinsed with deionized water for 10 min. Hematoxylin was used for counterstaining, and hydrochloric acid alcohol was differentiated. Routine dehydration, transparentizing, sealing, and microscopy were performed.

| Serum sample preparation
Added 400 µl anhydrous acetonitrile containing internal standard into 100 µl serum sample and vortex mixed for 3 min. Then, centrifuged the mixture at 12,000 rpm for 10 min to obtain the supernatant and putted it into a sampling vial.

| Data processing
The partial least-squares discriminant analysis (PLS-DA) and orthogonal partial least-squares discriminant analysis (OPLS-DA) were used for data analysis. The database used to identify the potential biomarkers was as follows: https://hmdb.ca/, http://www.lipid maps. org/, http://www.genome.jp/kegg/, http://metlin.scrip ps.edu/. All results were described as the mean ± standard deviation (SD). Oneway analysis of variance (ANOVA) was used to analyze study data for significance comparison.

| Identification of major compounds in AWE
The representative chromatography was shown in Figure 1.

| Determination of ferulic acid, senkyunolide A, and ligustilide in AWE
The representative chromatography was shown in Figure 2. Ferulic acid was linear over a concentration range of 3.10-154.90 µg/ml, senkyunolide A was linear at 1.33-40.00 µg/ml, and ligustilide was linear at 2.04-254.65 µg/ml. The concentration of the mixed reference solution was taken as the abscisic coordinate, and the peak area was taken as the ordinate for linear regression. Typical equation of calibration curve for ferulic acid was y = 14333x + 63,294 (r = 0.9970); the curve of senkyunolide A was y = 343389x−340450 (r = 0.9978); and that of ligustilide was y = 88,686x + 876,354 (r = 0.9953). The contents of ferulic acid, senkyunolide A, and ligustilide were 0.7276, 0.0868, and 1.9908 mg/g, respectively.

| AWE improved the histopathological damage
Microscopic examination explored that the endometrium of the P group was significantly thickened, the uterine cavity was small, the glands and blood vessels in the lamina propria were rich and dilated, and some of them were hyperemic (Figure 3a). Compared with the P group, the endometrium was thin, with local defects, and the lamina propria was mainly showed densely distributed blood vessels with slight congestion in the M group ( Figure 3b). The endometrium in the AWE group was rich in blood vessels and loose in the stroma, but the symptoms were less severe than those in the model group ( Figure 3c).

| Measurement of serum hormone levels
Comparing with the P group, the E 2 , Pg, and HCG levels were significantly decreased in the M group (p < .01). The E 2 and Pg levels were significantly increased with the treatment of AWE (p < .05) (Figure 4a-c).

| Effects of AWE on blood coagulation function in rats
Compared with the P group, the APTT and TT levels were significantly longer and FIB level was significantly lower in the M group (p < .05). The APTT and TT levels were significantly lower, and FIB level was significantly longer with the treatment of AWE (p < .05).

F I G U R E 1
The total ion chromatograms of samples by UPLC-MS

| Effects of AWE on ERα, Erβ, and PR levels in rats
Expression of ERα, Erβ, and PR was reduced in the M group when compared to the P group (p < .01). The expression of ERα and ERβ was increased with the treatment of AWE (p < .05).

| Data quality assurance of UPLC-Q-TOF-MS
The typical total ion current chromatograms of each group of serum samples were shown in Figure 7.

| Differential metabolites between the AUB and the pregnancy rats
The PLS-DA and OPLS-DA analyses explored that there was an obvious separation between the P, M, and AWE groups (Figure 8a,b). The S-plots indicated the contribution of different metabolites variables between the P and M groups (Figure 8c). Fourteen significantly differential metabolites were shown in Table 2  with the P group (Figure 9). The metabolism pathways of glycine, serine and threonine, glyoxylate and dicarboxylate, glycerophospholipid, primary bile acid biosynthesis, glutathione, and sphingolipid were significantly altered in the M group ( Figure 10).

| Differential metabolites between AUB and AWE treatment rats
There was a significant separation between the M and AWE groups in the OPLS-DA model (Figure 11a), indicating that AWE had an effect on the metabolic profile of AUB rats. The S-plots indicated the contribution of different metabolites variables between the M and AWE groups (Figure 11b). Twenty-one significantly differential metabolites were shown in Table 3, and there showed that the specific changes of the relative content in these specific metabolites ( Figure 12). LysoPC (20:5), Glycine, N-Acetyl-leukotriene E4, PC  Figure 13).

F I G U R E 5
The four coagulation indices of each experimental group. The results were presented as the mean ± SD, n = 6. *p < .    (Hua et al., 2017). Qin Fan et al. found that ferulic acid could scavenge PPH-and ABTS-free radicals, while ligustilide exhibited scavenging capacity for ABTS-free radicals (Fan et al., 2020).

Zi-wen Yuan et al. found that A. sinensis intervention could signifi-
cantly relieve blood stasis syndrome in rats (Yuan et al., 2019). In this study, APTT and TT levels were significantly lower in AWE group (p < .05), the APTT and TT levels were significantly lower, and FIB level was significantly longer with the treatment of AWE, indicating that A. sinensis can regulate blood coagulation function of AUB rats.
Metabolomics is a large-scale research technology that uses modern analytical method to assess the creature physiological status in different conditions (Wei et al., 2020). The metabolic profiles reflect an individual's function state at a certain point, which

F I G U R E 9
The relative intensities of the examined metabolites obtained from P and M groups. The results were presented as the mean ± SD, n = 6. *p < .05 versus pregnant group, ** p < .01 versus pregnant group F I G U R E 1 0 Summary of pathway analysis of P and M groups is consistent with the integrality and systematicness of traditional Chinese medicine (Bao et al., 2017). This technique has been used to study the effects of Chinese medicine syndrome patterns (Fengxia et al., 2010;Li et al., 2016). As an important part of the human body, blood contains abundant information and is often used as the matrix for metabolomics research. Therefore, the UPLC-Q-TOF-MS metabolomics platform and multivariate statistical analysis method were used to assess the rats' serum to reveal the mechanism of AWE in AUB caused by incomplete abortion. In our study, we identified 21 potential biomarkers that were directly or indirectly correlated for the therapeutic effects of AWE in AUB caused by incomplete abortion, mainly including amino acids, retinol, fatty acids, and lysophospholipids. We have found that the serum of the incomplete medical abortion rats showed altered metabolism mainly in amino acid, retinol, lipid metabolism and primary bile acid biosynthesis pathways.
Amino acids are necessary for embryonic growth and development. In our study, we found that the level of some amino acids including glycine, d-leucine, and l-valine was abnormal in the AWE  group. Glycine was closely associated with inflammation among these amino metabolites (Angélica et al., 2014). Glycine promotes myelin phagocytosis and the production NO and TNFα, which may affect immunological processes of inflammatory diseases (Carmans et al., 2010). In the inflammation, the TNFα induces protein decomposition and catabolism to up-regulate the urea synthesis (Louise et al., 2013 (Miao et al., 2018). In our study, we found that the glycine, D-leucine, and L-valine serum concentrations in the AWE group were higher than those in the M group, which may indicate that Angelica sinensis may affect the inflammatory process in AUB rats.
Lipid metabolism mainly consists of linoleic acid, glycerophospholipid, and sphingolipid metabolism. Oleic acid, which is closely related to leukotriene synthesis, is considered to be a key factor in tissue injury, inflammation, and vasoconstriction. Kim et al. found that the water extract of Angelica sinensis has an anti-inflammatory effect via the NO-burst/calcium-mediated JAK-STAT pathway (Young-Jin et al., 2018). Yao et al. found that the anti-inflammatory activity of the volatile oil of Angelica sinensis mainly through regulating glycine and arachidonic acid metabolic disorders (Yao et al., 2015). And in our study, after AWE intervention, the AWE group showed significant upregulation of N-Acetyl-leukotriene E4, 20-Oxo-leukotriene E4, and Leukotriene D5, which indicated that the effect of AWE on rats with AUB may involve the regulation of arachidonic acid metabolic network disorders. Lyso PC, a phospholipid, can regulate vascular tone and induce endothelial dysfunction.
It is reported that saturated fatty acids can induce the expression of cyclooxygenase (Akito et al., 2009) and promote the synthesis of PGF2α (Helliwell et al., ). In this study, LysoPC ( that the fragmentation rate of human preimplantation embryos medium with S1P was significantly lower and the embryos quality was better (Antoine et al., 2010). In our study, we found that the serum concentration of sphingosine in the AWE group was higher than that in the M group, which may be because AWE has a regulatory effect on sphingosine.
All-trans-retinoic acid belongs to vitamin A (retinol), which is a basic nutrient required for mammalian reproduction. This molecule plays a vital role in promoting the normal development of embryos, regulating cell proliferation and differentiation, and maintaining normal cell differentiation and immune system function integrity. Ueda et al. showed that chemical signaling induced by the use of retinoic acid can promote the differentiation of embryonic stem cells into neurons (Ueda et al., 2018). In our study, we found that the level of all-trans-retinoic acid was down-regulated in the AWE group. In summary, AWE could adjust the abnormal metabolism state of amino acid, retinol, and lipid metabolism with AUB induced by incomplete medical abortion.

| CON CLUS IONS
An UPLC-Q-TOF-MS-based serum metabolomic approach was applied to investigate the mechanisms of AWE for the treatment of AUB. Twenty-one potential biomarkers were eventually identified,

F I G U R E 1 3 Summary of pathway analysis of M and AWE groups
which may be involved in the intervention mechanism of AWE for the treatment of AUB. Some metabolic pathways including primary bile acid biosynthesis, glycerophospholipid metabolism, pentose and glucuronate interconversions, glutathione metabolism, glyoxylate and dicarboxylate metabolism, sphingolipid metabolism, glycine, serine and threonine metabolism, retinol metabolism, ascorbate, and aldarate metabolism were altered by AWE treatment. The significantly reversed the metabolic aberrations in AUB group by AWE facilitates to support the therapeutic effect and potential mechanisms of AWE on AUB induced by incomplete medical abortion.

ACK N OWLED G M ENTS
This research was funded by Health and Family Planning Commission of Chengdu-Key disciplines of clinical pharmacy and Sichuan Medical Research Project (grant number: S18056, S19030).

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.