Metabolomics Study of Shaoyao Plants Decoction on the Proximal and Distal Colon in Mice with Dextran Sulfate Sodium-Induced Colitis by UPLC-Q-TOF-MS

Introduction

Ulcerative colitis (UC), a major inflammatory bowel disease (IBD),¹ is characterized by mucosal epithelial damage and disruption of intestinal homeostasis and manifests as abdominal pain, diarrhea, and bloody mucus in stools.^2–4 The etiology of UC is complex and includes genetic susceptibility, defective immune responses, and environmental factors’.^5,6 The morbidity rate of IBD is 5–15 per 100,000 persons per year in the Western world, and the incidence of UC is increasing in both developed and developing countries’.^7–11 Currently, the different drugs used to treat UC include aminosalicylates, steroids, immunosuppressants, and biological agents.^12,13 Although these drugs are generally effective, not all patients respond to or achieve sustained remission, and their use has certain limitations, such as an increased risk of infection and malignancy.^14–16 The disadvantages of the current therapies make the prospect of using traditional Chinese medicine to treat UC even more appealing.

Traditional Chinese medicine is a multicomponent, multichannel, and multitarget therapeutic drug widely used in China and has a long history. It has shown great advantages in treating and preventing diseases and has accumulated rich experience.^17,18 Shaoyao decoction (SYD) is a traditional herbal medicine composed of Baishao, Danggui, Huanglian, Huangqin, Gancao, Binglang, Muxiang, Dahuang, and Rougui. SYD has multiple pharmacological activities and is widely used for the treatment of UC. Baishao has anti-inflammatory and antioxidant properties.¹⁹ Danggui has anti-inflammatory and anti-cancer effects.²⁰ Berberine, a major component of Huanglian, acts on many organs and systems of the body and has many biological functions, such as anti-inflammatory, anti-tumor, anti-arrhythmic, and blood lipid regulation.^21–23 Huangqin has a variety of biological activities, including antioxidant, antiviral, antibacterial, anti-inflammatory, and anti-tumor functions.²⁴ Many clinical studies have shown that SYD has a therapeutic effect on UC.^25–28 In addition to the clinical efficacy of SYD, some studies have reported mechanisms of SYD in the treatment of UC. In a dextran sulfate sodium (DSS)-induced colitis mouse model, Chi et al found that SYD could regulate the signal pathway of STAT3 and NF-κB and regulate epithelial cell apoptosis and permeability.²⁹ Wei et al demonstrated that SYD alleviates pyroptosis by regulating the MKP1/NF-κB/NLRP3 pathway.³⁰ Moreover, SYD may regulate the abundance and changes in the gut microbiota and play a role in the treatment of UC.³¹ Although some studies have reported these mechanisms, the pathogenesis of UC is complex, and the mechanism of SYD in the treatment of UC needs to be further clarified.

In recent years, metabolites produced by the intestinal flora have been shown to play a key role in regulating intestinal mucosal homeostasis. These metabolites are produced and secreted into the lumen and regulate the activity of many host cell types, including immune and intestinal epithelial cells.^32–35 These metabolites, including short-chain fatty acids (SCFAs) and lipids, can influence the health and disease state of hosts.³⁶ Studies have found that the content of SCFAs in patients with UC decreased.^37,38 SCFAs can repair intestinal epithelial cells, promote the production of mucin by goblet cells, modify tight junction proteins, and improve the integrity of the intestinal mucosal barrier.^39–41 In addition, studies have found that UC is associated with alterations in lipid metabolism.^42,43 Lipids control cellular processes, such as proliferation, differentiation, migration, and apoptosis.^44,45 Currently, most studies have focused on changes in metabolites in feces, plasma, and urine, and few studies have examined metabolic changes in the colon tissue. However, research showed that the intestinal mucosa absorbs 90–95% of SCFAs produced by the gut microbiota, and SCFAs have the highest concentration in colon tissue.^46–49 Therefore, the study of changes in intestinal metabolites in UC should not be limited to feces but should focus on the colon tissue.

Metabolomics is a technology for the qualitative and quantitative analysis of metabolites in a given organism or biological sample, which can detect small molecular substances and precisely display the changes that occur in the body.^50,51 Presently, the commonly used techniques in metabolomics include nuclear magnetic resonance (NMR), gas chromatography-mass spectrometry (GC-MS), and ultra-performance liquid chromatography combined with quadrupole-time-of-flight mass spectrometry (UPLC-Q-TOF-MS). UPLC-Q-TOF-MS is a powerful metabolomics technology owing to its high resolution and sensitivity.⁵² In addition, non-targeted analysis has a wide application in drug research by probing biomarkers to clarify the underlying mechanisms of drugs.⁵³ In recent years, the widespread use of this technology has provided a new perspective for traditional Chinese medicine research.^54–56

This study aimed to explore the effects of SYD on colon metabolites in a murine model of DSS-induced UC. UPLC-Q-TOF-MS-based untargeted metabolomics was used to analyze the changes in colon metabolites. At the same time, we also compared the differences in the proximal and distal colonic metabolic profiles to provide more precise evidence for SYD in the treatment of UC. Our study provides a new perspective for further elucidation of the mechanism of SYD in the treatment of UC.

Materials and Methods

Chemicals and Reagents

Dextran sulfate sodium (DSS) (M. W.:36,000–50,000) was purchased from MP Biomedicals Corporation (Santa Ana, CA, USA), methanol and acetonitrile were purchased from EMD Millipore Corporation (Darmstadt, Germany), and formic acid was purchased from Shanghai Aladdin Biochemical Technology Limited Liability Company (Shanghai, China).

Preparation of SYD

SYD comprises nine different herbs, including Baishao, Danggui, Huanglian, Huangqin, Gancao, Binglang, Muxiang, Dahuang, and Rougui. Details of their composition are shown in Table 1. All herbs were purchased from the Second Affiliated Hospital of Zhejiang Chinese Medical University (Hangzhou, Zhejiang, China) and accredited by a pharmacologist. The dose of SYD for experimental mice was determined according to the clinical equivalent dosage and study by Chi et al.²⁹ The daily clinical dose of SYD in humans is 107 g. According to the formula of equivalent body surface area between humans and mice,⁵⁷ the dose for a mouse is 17.8 g/kg/day. All raw herbs were soaked in distilled water (1:10, w/v) for 30 min, boiled for 1 h, filtered with gauze, filtrate collected, decocted the residue with distilled water (1:6, w/v) as described above, combined the two filtrates, and concentrated them to 1 g/mL using a rotary evaporator at 60°C. Part of the SYD liquid was stored at −80°C for UPLC-QTOF-MS analysis (Waters Corporation, Ireland).

Table 1 Composition Information of Each Chinese Herbal Medicine in SYD

UPLC-Q-TOF-MS Analysis for SYD

The liquid SYD was thawed at 4°C, diluted with methanol to 100 mg/mL, the mixture was stirred and ultrasonicated for 15 min, and centrifuged at 12,000 x g for 10 min. The supernatant was obtained for UHPLC-Q-TOF-MS analysis (Waters Corporation, Ireland). The SYD liquids were analyzed in negative and positive modes on the ACQUITY UPLC HSS T3 column (100 mm × 2.1 mm, 1.8 µm; Waters Corporation, Ireland) at a flow rate of 0.3 mL/min, sample injection volume of 2.0 μL, autosampler temperature of 4°C and column temperature of 40°C. Mobile phases A and B were 0.1% (v/v) formic acid/water and 0.1% (v/v) formic acid/acetonitrile, respectively. The gradient elution conditions were set as follows: 20% B for 0–2 min, 52% B for 12 min, 65% B for 18 min, 90% B for 22 min, and 20% B for 26–30 min. The mass spectrum parameters of the positive mode were capillary voltages: 3.0 kV, sampling cone: 40.0 V, source temperature: 100°C, desolvation temperature: 500°C, cone gas flow: 100°C, desolvation gas flow: 1000.0 L/h, scan time: 0.10 s, interscan time: 0.014 s, mass range: 50–2000 m/z. The mass spectrum parameters of the negative mode were capillary voltages: 2.5 kV, sampling cone: 40.0 V, source temperature: 100°C, desolvation temperature: 500°C, cone gas flow: 100°C, desolvation gas flow: 1000.0 L/h, scan time: 0.10 s, interscan time: 0.014 s, mass range: 50–2000 m/z.

Animals

Forty male C57BL/6 mice (6–8 weeks old, 20–22 g) were purchased from the Shanghai Slack Laboratory Animal Limited Liability Company (No. SCXK 2017–0005). All mice were housed under standard laboratory conditions (55 ± 10% humidity, 22 ± 2°C, and 12 h light/dark cycle). Animal welfare was performed according to the guidelines for the management and use of laboratory animals (Ministry of Science and Technology, China, 2016). The experimental protocol was approved by the Experimental Animal Ethics Committee of the Zhejiang Chinese Medical University (IACUC-20210531-02).

Animal Model Establishment and Treatment Procedures

After the seven days of acclimatization, 40 C57BL/6 mice were randomly divided into three groups: the normal control (NC) group (n = 10), UC model (UC) group (n = 15), and SYD treatment (SYD) group (n = 15). To establish the UC model, the mice in the UC and SYD groups received drinking water containing 3% (w/v) DSS for seven days,⁵⁸ whereas those in the NC group were given distilled water. After the successful establishment of the murine model of UC, the mice in the SYD group were given 17.8 g/kg SYD once a day by gastric gavage for seven days,²⁹ and the mice in the NC and UC groups were given equal amounts of distilled water at the same time and in the same manner.

Disease Activity Index (DAI)

Body weight, stool consistency, and stool blood levels of mice were recorded daily, and the calculation of the DAI score was based on an earlier study, which was the average score of weight loss, stool consistency, and occult/gross bleeding (Table 2).⁵⁹

Table 2 Criteria for Disease Activity Index

Sample Collection and Preparation

After seven days of treatment, the mice were euthanized using CO2, and the lengths of the colons were photographed and measured. For colon, tissue from the ileocecal junction to the rectum was divided into two equal halves of approximately 2.5–4 cm in length representing proximal and distal colon.^60,61 The proximal colon was morphologically distinct from the remaining parts of the tissue by transverse folds that project into the lumen.⁶² Then, the proximal colon and distal colon of each mouse were cut into pieces; one piece of the proximal colon and distal colon was fixed in 4% paraformaldehyde, and the remaining pieces from a colon were stored at −80 °C for subsequent use.

Histopathological Analysis

The colons fixed in 4% paraformaldehyde were embedded in paraffin, stained with hematoxylin and eosin (H&E) staining, and examined under light microscopy. Histopathological scores were calculated according to damage to the epithelial mucosa and inflammatory infiltration, as shown in Table 3.⁵⁹

Table 3 Evaluation of Histopathological Score

Untargeted Analysis Using UPLC-Q-TOF-MS

Preparation of Metabolomic Samples

Before analysis, the colon samples were freeze dried separately in liquid nitrogen and ground into a powder,100 mg of each sample was collected, thawed at 4°C, and aliquoted with 1 mL methanol to precipitate proteins. The mixture was stirred for 15 min, then centrifuged at 12,000 × g for 10 min, and the supernatant was obtained for UHPLC-Q-TOF-MS analysis. In addition, quality control (QC) samples were prepared, which were hybrids of 10 μL of each extraction mixture.⁵⁴

UHPLC-Q-TOF-MS Conditions

All samples were analyzed in negative and positive modes on an ACQUITY UPLC HSS T3 column (100 mm × 2.1 mm, 1.8 µm; Waters Corporation, Ireland) at a flow rate of 0.3 mL/min, sample injection volume of 2.0 μL, autosampler temperature of 4°C, and column temperature of 40°C. Mobile phases A and B were 0.1% (v/v) formic acid/water and 0.1% (v/v) formic acid/acetonitrile, respectively. The gradient elution conditions were set as follows: 20% B for 0–2 min, 52% B for 12 min, 65% B for 18 min, 90% B for 22 min, and 20% B for 26–30 min. The mass spectrum parameters of the positive mode were capillary voltages: 3.0 kV, sampling cone: 40.0 V, source temperature: 100°C, desolvation temperature: 500°C, cone gas flow: 86°C, desolvation gas flow: 815.0 L/h, scan time: 0.10 s, interscan time: 0.014 s, mass range: 50–2000 m/z. The mass spectrum parameters of the negative mode were capillary voltages: 2.5 kV, sampling cone: 40.0 V, source temperature: 100°C, desolvation temperature: 500°C, cone gas Flow: 100°C, desolvation gas flow: 1000.0 L/h, scan time: 0.10 s, interscan time: 0.014 s, mass range: 50–2000 m/z. Leucine enkephalin was used as the lock mass control compound (m/z 556.2771 in the positive mode, m/z 554.2615 in the negative mode). Before beginning the analysis, a blank sample was injected three times, after which a blank sample was injected eight times, and the QC sample was injected ten times. All the samples were randomly sequenced.

Untargeted Metabolomics Analyses

All data were obtained using MassLynx V4.1 software (Waters Corporation, Milford, MA, USA) and then imported into Progenesis QI V2.0 software (Waters Corporation, Milford, MA, USA) for preprocessing. The Human Metabolome Database (HMDB, http://www.HMDB.ca), Lipid Maps (http://www.lipidmaps.org), NIST (https://www.nist.gov/), and ChemSpider (http://www.chemspider.com/) were used to speculate and identify the compounds. The data stabilized as normal distributions were imported into the SIMCA-p14.1 software, and Pareto scaling was selected for multivariate statistical analysis, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA).⁵⁴ The quality of the OPLS-DA model was assessed using 10-fold cross-validation, CV-ANOVA, 200 permutation tests, and Q2 values. Differential metabolites were screened using a combination of univariate and multivariate analyses. Metabolites with variable importance in the projection (VIP) ≥ 1 in the OPLS-DA analysis and a P value less than 0.05 in two-sample t-tests were considered significantly differential metabolites between the two groups. The significantly altered metabolites were analyzed using MetaboAnalyst 5.0.⁶³ Finally, the Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used to confirm the metabolic pathways caused by these significantly altered metabolites.⁶⁴

Statistical Analysis

Statistical software SPSS 21.5 (IBM Corporation, USA) was used for the data analysis and processing. The data are expressed as mean ± standard deviation. The Shapiro–Wilk test was performed to assess the data distribution. The ANOVA statistical test was used to compare the data among multiple groups when the data were normally distributed, and the least significant difference post-hoc test was performed following the ANOVA statistical test. The Kruskal–Wallis test, which is a non-parametric test, was used when the data were non-normally distributed. Statistical significance was set at P<0.05.

Results

Analysis of the Components of SYD

UPLC-Q-TOF-MS was used to identify the major chemical components of SYD. Twelve compounds were distinguished: gallic acid, liquiritin, glycyrrhizin, arecoline, baicalin, trans-ferulic acid, chlorogenic acid, berberine, licoisoflavone A, quercetin, formononetin, and kaempferol. The ion chromatograms are shown in Figure S1. The Identification results of the major chemical components of SYD are shown in Table S1. Three batches of SYD were tested, and the active gradients of SYD were essentially the same, indicating that the results were reliable and robust. The identification results for the other two batches of SYD are shown in Figures S2 and S3.

SYD Alleviates DSS-Induced UC in Mice

To confirm the therapeutic effect of SYD on UC, we assessed colitis in mice based on weight, DAI score, and colon length. Compared with the NC group, the weight of the mice that received DSS significantly decreased (Figure 1A). On day 7 of DSS induction, the DAI score of mice that received DSS was significantly higher than that of the mice in the NC group (Figure 1B). These results suggested that the UC mouse model was successfully established. After administration of SYD, the weight loss and DAI score improved (Figure 1A and C). Furthermore, we measured the colon length (Figure 1D) and observed a reduction of approximately 14.52% and 0.50% in the UC and SYD groups, respectively, compared with that of the NC group. The colon length of mice in the UC group was significantly shorter than those in the NC (P<0.001) and SYD groups (P<0.001). There were no differences between the NC and SYD groups (Figure 1E).

Figure 1 The weight, disease activity index (DAI) score, and colon length of mice in each group. (A) The changes in weight of mice throughout the experiment; (B) The DAI score of the 7th day in each group; (C) The DAI score of the 14th day in each group; (D) Typical colon anatomy of mice in each group; (E) The colon length of mice in each group. The red arrow means the first day of inducing UC by dextran sulfate sodium (DSS) and the blue arrow indicates the first day of medication. Compared with the NC group, ***P<0.001, **P<0.01, *P<0.05; compared with the UC group, ###P<0.001, ##P<0.01, #P<0.05.

SYD Improved Mucosal Inflammation of Colonic Tissues

Typical histopathological manifestations in the proximal colon and distal colon of mice in each group are shown in Figure 2A and B. The colons of mice in the NC group showed complete colonic epithelium and no inflammatory cell infiltration. In the UC group, the colons of mice showed high levels of inflammatory cell infiltration in the intestinal mucosa. In addition, there was a loss of intact epithelium, and a large number of epithelial cells were destroyed. In the SYD group, destruction of the intestinal wall and inflammatory infiltration were milder than in the UC group.

Figure 2 Histopathological changes and scores of colonic tissues of mice in each group. (A) Typical histopathological performance of distal colon (× 5 and × 20); (B) Typical histopathological performance of proximal colon (× 5 and × 20); (C) The histopathological score of distal colon; (D) The histopathological score of proximal colon; (E) The histopathological score between proximal colon and distal colon. *P<0.05; ***P<0.001.

The histopathological score of mice in the UC group was significantly higher than that of mice in the NC group (P<0.001). The histopathological score of mice in the SYD group was significantly lower than that of mice in the UC group (P<0.001) (Figure 2C and D). In addition, in the UC group, the histopathological scores of distal colon were higher than proximal colon (P<0.05), while there was no difference in pathological scores between the proximal colon and distal colon in the NC group and SYD group (Figure 2E).

Analysis of Proximal Colon and Distal Colon Metabolite Profiles

UPLC-Q-TOF-MS was performed to identify the metabolic profiles of the proximal colon and distal colon samples collected from mice. As shown in Figure 3, the total ion current chromatograms (TICs) of the QC samples highly overlapped, indicating that the instrument was reliable and stable. Figure 4 indicates the representative TICs plots of each group, respectively.

Figure 3 The total ion current chromatograms (TICs) of the quality control (QC) samples. (A) TICs of QC samples in the distal colon under positive ion mode; (B) TICs of QC samples in the distal colon under negative ion mode; (C) TICs of QC samples in the proximal colon under positive ion mode; (D) TICs of QC sample in the proximal colon under negative ion mode.

Figure 4 Indicates the representative total ion current chromatograms (TICs) plots of each group. (A) TICs of the distal colon under positive ion mode; (B) TICs of the distal colon under negative ion mode; (C) TICs of the proximal colon under positive ion mode; (D) TICs of the proximal colon under negative ion mode.

PCA was used to distinguish the inherent trends within the colon metabolic data of each group. The score maps of PCA were shown in Figure 5, the tendency for segregation of the NC, UC, and SYD groups was observed, indicating significant variations in metabolites within the three groups. The loading maps of PCA were shown in Figure S4. Moreover, OPLS-DA was performed to force the classification of each component and facilitate the identification of differences and similarities between groups. The OPLS-DA models was validated by cross validation and CV-ANOVA. Validation parameters of OPLS-DA models (Q2 and R2) were arranged in Table 4, so were the p-values tested by CV-ANOVA. The p-values were <0.05 in all models, suggesting the models were valid. The OPLS-DA results showed that in the positive and negative ion modes, samples in each group were significantly clustered, and each group was significantly separated from another (Figure 6). The validation plots were shown in Figure 7.

Table 4 Validation Parameters of OPLS-DA Models for Colon of Mice

Figure 5 Multivariate statistical analysis of plasma metabolic profiles in mice of each group. (A and B) Three-dimensional plots of principal component analysis (PCA) of distal colon metabolic profile in mice of each group under negative and positive ion mode; (C and D) Three-dimensional plots of PCA of proximal colon metabolic profile in mice of each group under negative and positive ion mode.

Figure 6 The orthogonal partial least squares discriminant analysis (OPLS-DA) diagrams between groups. (A and B) The OPLS-DA diagrams of distal colon between NC and UC group under negative and positive ion mode; (C and D) The OPLS-DA diagrams of distal colon between UC and SYD group under negative and positive ion mode; (E and F) The OPLS-DA diagrams of proximal colon between NC and UC group under negative and positive ion mode; (G and H) The OPLS-DA diagrams of proximal colon between UC and SYD group under negative and positive ion mode.

Figure 7 The validation of the orthogonal partial least squares discriminant analysis (OPLS-DA) model between groups. (A and B) The validation of the OPLS-DA model of distal colon between NC and UC group under negative and positive ion mode; (C and D) The validation of the OPLS-DA model of distal colon between UC and SYD group under negative and positive ion mode; (E and F) The validation of the OPLS-DA model of proximal colon between NC and UC group under negative and positive ion mode; (G and H) The validation of the OPLS-DA model of proximal colon between UC and SYD group under negative and positive ion mode.

Metabolites with VIP scores >1 in the OPLS-DA model and P<0.05 were considered significantly altered metabolic biomarkers. The volcano plots showed the differential metabolites between the comparison groups (Figure 8). In the distal colon samples, 102 significantly differential metabolites were identified between the NC and UC groups, and 54 significantly differential metabolites were identified between the UC and SYD groups (Tables S2 and S3). We further found that phosphatidic acid (PA) 20:0/12:0 and lysophosphatidylethanolamine (lysoPE) 0:0/22:5(4Z,7Z,10Z,13Z,16Z) significantly decreased after the induction of UC. After SYD intervention, the levels of these metabolites significantly increased. Moreover, three metabolites significantly increased in mice in the UC group compared with that in mice in the NC group; these metabolites were lactosylceramide (d18:1/12:0), erythrodiol 3-palmitate, and lysophosphatidylcholine (lysoPC) 22:2(13Z,16Z). After SYD treatment, these metabolites significantly decreased, as shown in Table 5 and Figure 9.

Table 5 Metabolites of Distal Colon Regulated by SYD

Figure 8 The volcano plots between groups. (A and B) The volcano plots of distal colon between NC and UC group under negative and positive ion mode; (C and D) The volcano plots of distal colon between UC and SYD group under negative and positive ion mode; (E and F) The volcano plots of proximal colon between NC and UC group under negative and positive ion mode; (G and H) The volcano plots of proximal colon between UC and SYD group under negative and positive ion mode.

Figure 9 The abundance of differential metabolites in distal colon. (A) The abundance of PA(20:0/12:0) in each group; (B) The abundance of LysoPE(0:0/22:5(4Z,7Z,10Z,13Z,16Z)) in each group; (C) The abundance of Lactosylceramide (d18:1/12:0) in each group; (D) The abundance of Erythrodiol 3-palmitate in each group; (E) The abundance of LysoPC(22:2(13Z,16Z)) in each group. *P<0.05, **P<0.01, ***P<0.001.

In the proximal colon samples, 112 significantly differential metabolites were identified between the NC and UC groups, and 75 significantly differential metabolites were identified between the UC and SYD groups (Tables S4 and S5). We further found that palmitic acid, cyclonormammein, monoacylglyceride (MG) 0:0/18:2(9Z,12Z)/0:0, 13S-hydroxyoctadecadienoic acid, and ceanothine C were significantly decreased after the induction of UC. After SYD intervention, the levels of these metabolites significantly increased. Moreover, tetracosahexaenoic acid, phosphatidylserine (PS) 18:0/22:6(4Z,7Z,10Z,13Z,16Z,19Z), and diglyceride (DG) 14:0/20:2(11Z,14Z)/0:0 significantly increased in mice in the UC group when compared with that in mice in the NC group. After SYD treatment, these metabolites significantly decreased, as shown in Table 6 and Figure 10. Heatmaps of proximal colon and distal colon were used to visualize these significantly different metabolites (Figure 11A and B).

Table 6 Metabolites of Proximal Colon Regulated by SYD

Figure 10 The abundance of differential metabolites in proximal colon. (A) The abundance of Palmitic acid in each group; (B) The abundance of Cyclonormammein in each group; (C) The abundance of MG(0:0/18:2(9Z,12Z)/0:0) in each group; (D) The abundance of 13S-hydroxyoctadecadienoic acid in each group; (E) The abundance of Ceanothine C in each group; (F) The abundance of Tetracosahexaenoic acid in each group; (G) The abundance of PS(18:0/22:6(4Z,7Z,10Z,13Z,16Z,19Z)) in each group; (H) The abundance of DG(14:0/20:2(11Z,14Z)/0:0) in each group. *P<0.05, **P<0.01, ***P<0.001.

Figure 11 The heat map among groups. (A) The heat map of distal colon among groups; (B) The heat map of proximal colon among groups.

To further determine the metabolic pathways regulated by SYD, we performed the KEGG pathway enrichment analysis. In the distal colon samples, four pathways differed in the SYD and UC groups. The metabolic pathways included glycosphingolipid biosynthesis (globo and isoglobo series), sphingolipid metabolism (glycosphingolipid biosynthesis-ganglio series), and glycosphingolipid biosynthesis (lacto and neolacto series) (Figure 12A). These results indicate that SYD may regulate these pathways in the treatment of UC, and among them, the glycosphingolipid biosynthesis (lacto and neolacto series) pathway was the main metabolic pathway regulated by SYD. In the proximal colon samples, four pathways differed in the SYD and UC groups. The metabolic pathways include the biosynthesis of unsaturated fatty acids, fatty acid elongation, fatty acid degradation, and fatty acid biosynthesis (Figure 12B). These results indicate that SYD may regulate these pathways in the treatment of UC, and among them, the biosynthesis of unsaturated fatty acids pathway was the main metabolic pathway regulated by SYD.

Figure 12 The metabolic pathways regulated by SYD. (A) The metabolic pathways of the distal colon regulated by SYD; (B) The metabolic pathways of the proximal colon regulated by SYD.

Discussion

SYD is a traditional Chinese medicine prescription for treating many diseases, including UC and colorectal cancer.^29,30,65,66 In this study, we constructed a murine model of UC induced by DSS and proved that SYD was effective in UC mice, as it ameliorated the symptoms of UC, increased weight, improved colonic shortening, and reduced colonic damage. UPLC-Q-TOF-MS-based untargeted metabolomics was used to analyze the metabolite profiles of the proximal and distal colon and monitor and capture potential metabolic responses and biomarkers. Biomarkers have been used to label changes in systems, organs, tissues, cells, and subcellular structures or functions.⁶⁷ Although there are reports of potential blood biomarkers for UC, few studies have focused on UC biomarkers in tissue extracts, especially from the colon. Studies have demonstrated that most SCFAs are absorbed by the intestinal mucosa and have the highest concentration in colon tissue.^46–49 Therefore, studies on changes in intestinal metabolites in UC should focus on the colon tissue. Through metabolomic analysis of colon tissues, we identified five and eight metabolites that were up- or down-regulated by SYD in the distal colon and proximal colon, respectively.

In the distal colon, the levels of PA (20:0/12:0) and lysoPE (0:0/22:5(4Z,7Z,10Z,13Z,16Z)) decreased, and that of lactosylceramide (d18:1/12:0), erythrodiol 3-palmitate, and lysoPC (22:2(13Z,16Z)) increased in UC mice. However, the content of these metabolites tended toward normal levels after SYD treatment. LysoPE is a lysoglycerophospholipid derived from phosphatidylethanolamines. The role of lysoPE in IBD was suggested when the rectal application of lysoPE reduced inflammation and mucosal damage in 2,4,6-trinitrobenzene sulfonic acid (TNBS)-induced colitis in rats.⁶⁸ Recently, Zou et al found that the levels of lysoPE 16:0 and lysoPE 18:1 decreased in fecal samples from patients with UC compared with those from controls, which is consistent with our results. In addition, they demonstrated that oral supplementation with lysoPE 18:1 could regulate the expression of occludin, claudin-1, and ZO-1, and ameliorate intestinal permeability.⁶⁹ In our study, we found that lysoPE levels in the distal colon tissue decreased after UC modeling, whereas SYD increased lysoPE levels, indicating that SYD can regulate the intestinal mucosal barrier by regulating the level of lysoPE, thereby improving intestinal mucosal damage. Phosphatidylcholine (PC) is a phospholipid that accounts for more than 70% of total phospholipids in the intestinal mucus layer, and it has been shown to possess anti-inflammatory effects.^70,71 In contrast, lysoPC derived from PC has been reported to be pro-inflammatory.^72,73 LysoPC exacerbates inflammatory response by promoting the expression of growth factors, chemotaxis factors, and endothelial cell adhesion molecules.⁷² It increases intestinal permeability and bacterial translocation.⁷⁴ Patients with UC were found to have decreased PC levels and increased lysoPC levels, and PC supplementation alleviated inflammation.^75,76 In TNBS-induced colitis mice and DSS-induced colitis mice, lysoPC was elevated in serum samples. PC and lysoPC levels are strongly associated with UC, and the lysoPC/PC ratio is particularly important in improving intestinal permeability.^77,78 Our results confirmed that SYD could regulate lysoPC levels and exert an anti-inflammatory effect in the distal colon. Lactosylceramide (LacCer) is a glycosphingolipid.⁷⁹ Glycosphingolipids are a class of membrane molecules that regulate membrane fluidity, receptor protein function, and cell adhesion.⁸⁰ LacCer plays an important role in the human body by taking part in the conduction of various signals.^81,82 It was found to be elevated primarily in Crohn’s disease (CD) in a study of serum metabolomics in children with IBD, and it was suggested that serum LacCer may be a potential biomarker that could differentiate children with IBD from healthy individuals and distinguish the subtypes of IBD.⁸³ Another study also demonstrated that serum LacCer 18:1/16:0 could significantly discriminate CD from UC in children.⁸⁴ Both studies involved children; therefore, more studies are required to demonstrate the role of LacCer in adult IBD patients. Interestingly, in inflamed colon tissue of adult patients with UC, the level of LacCer was increased,⁸⁵ which was consistent with our results. Therefore, the role of LacCer in UC warrants further study.

In the proximal colon, the level of palmitic acid, cyclonormammein, MG (0:0/18:2(9Z,12Z)/0:0), 13S-hydroxyoctadecadienoic acid, and ceanothine C decreased, whereas that of tetracosahexaenoic acid, PS (18:0/22:6(4Z,7Z,10Z,13Z,16Z,19Z)), and DG (14:0/20:2(11Z,14Z)/0:0) increased in UC mice. The levels of these metabolites tended to normalize after SYD treatment. The 13S-hydroxyoctadecadienoic acid (13S-HODE) is a derivative of linoleic acid and the predominant regulator of inflammation in different cell systems, such as endothelial cells.^86–88 However, the role of 13S-HODE in UC has not been studied thoroughly. Interestingly, a recent study found that black raspberries increased the levels of 13S-HODE in interleukin (IL)-10 KO mice and played a role in the treatment of UC.⁸⁹ This is consistent with our findings; SYD increased the level of 13S-HODE in the proximal colon, indicating that it plays a role in the treatment of UC mainly by regulating13S-HODE levels in the proximal colon. PS is participates in inflammation and can initiate platelet activation and apoptosis. It is considered an upstream immune checkpoint that suppresses the development of immunity.⁹⁰ Researchers also found that an increased level of PS was related to the activation of platelets, which could increase thrombotic risk in patients with IBD.⁹¹ In addition, Iwatani et al reported a significant increase in PS with respect to IBD activity,⁹² which is consistent with our findings. We found that in UC mice, the levels of PS increased, and SYD decreased PS levels, demonstrating that SYD may reduce thrombotic risk in UC. Regarding palmitic acid and DG (14:0/20:2(11Z,14Z)/0:0), palmitic acid is a long-chain saturated fatty acid found in animals, plants, and microorganisms. At the intestinal level, palmitic acid induces monocyte activation and stimulates pro-inflammatory responses.^93,94 Ghezzal et al found that the intake of palm oil provokes an increase in intestinal permeability and loss of several tight junction proteins. They also found that palmitic acid increased the expression and secretion of pro-inflammatory cytokines and exerted its harmful effects on epithelial cells.⁹⁵ DG (14:0/20:2(11Z,14Z)/0:0) is a diglyceride and diacylglycerol, which can inhibit the secretion of bile acids and is used to prevent and treat diarrhea.⁹⁶ Murase et al have reported that diacylglycerol can inhibit the occurrence of colitis.⁹⁷ Our study found that the palmitic acid level in the proximal colon decreased and that of DG (14:0/20:2 (11Z, 14 Z)/0:0) increased after UC modeling. This indicates that the inflammation in the proximal colon was not severe; however, metabolic disorders still existed. SYD can regulate the levels of palmitic acid and DG (14:0/20:2(11Z,14Z)/0:0) to normal levels, indicating a specific role of SYD in regulating the disturbance of intestinal metabolites.

Relevant reports on the role of PA (20:0/12:0), erythrodiol 3-palmitate, cyclonormammein, MG (0:0/18:2(9Z,12Z)/0:0), and ceanothine C in UC are few; therefore, the role of these metabolites requires further study. Further analysis of the metabolic pathways regulated by SYD revealed that, in the distal colon, SYD regulated the glycosphingolipid biosynthesis (lacto and neolacto series) pathway. In the proximal colon, the pathway for the biosynthesis of unsaturated fatty acids was the main metabolic pathway regulated by SYD. From the differential metabolites and metabolic pathways, we found that the metabolite environments in the proximal colon and distal colon of the DSS-induced colitis model mice were different, and there were different metabolic disorders present. We observed that in the DSS-induced colitis mouse model, the levels of inflammatory metabolites in the distal colon were higher than that in the proximal colon, indicating that inflammation in the distal colon was more serious than that in the proximal colon. The effect of SYD in the treatment of UC is mediated by the regulation of metabolites in the distal colon. Our findings are also helpful in understanding the pathogenesis of UC, as UC is characterized by chronic mucosal inflammation beginning in the rectum and extending proximally to a variable distance. Moreover, previous studies have found that the assessment of the severity of inflammation in the left-sided colon had excellent accuracy for the detection of right-sided histologic disease activity.⁹⁸ Our findings are helpful for more accurate follow-up studies on UC and suggest that the effects of drugs on the proximal colon and distal colon should be studied separately to provide more precise evidence.

Conclusion

Our study found that different metabolic disorders exist in the proximal colon and distal colon and demonstrated that SYD intervention could ameliorate DSS-induced colonic damage, which may be attributed to SYD targeting specific metabolites and pathways. Our findings may help clarify the detailed mechanism of SYD in the treatment of UC and provide an important reference for further studies.