Therapeutic effect of total flavonoids of Sargentodoxa cuneata on ulcerative colitis in mice by correcting gut dysbiosis

2.1 Chemicals and reagents

The 36–50 kDa dextran sulfate sodium (DSS) was obtained from MP Biomedical (Illkirch, France). Reagents (methanol and acetonitrile) used for chromatography were provided by Thermo-Fisher Scientific (FairLawn, NJ, USA). Formic acid and standards of SCFAs and organic acids (≥99.9 % purity) were purchased from Sigma-Aldrich (St. Louis, MO, USA).

2.2 Preparation and composition of TFSc

The herbs used in this study were identified by Prof. Wang Xiangpei of the Guizhou Minzu University. An aqueous extract of Sargentodoxa cuneata was obtained by decocting with water, followed by filtration and concentration (Yu et al., 2023b). Briefly, the Sargentodoxa cuneata herb was weighted, and water was added to submerge the herbs 2–3 cm, soaked for 15 min, boiled over high heat, and then decocted over low heat for 15 min. The mixture was filtered and water was added water, followed by decoction. This procedure was repeated thrice. Finally, the filtrate was combined and concentrated using rotary evaporator. The extract was placed on a D101 macroporous resin column and eluted using different ethanol concentrations (5, 10, 20, 30, 40, 50, 70, and 95 %). After detection by sodium nitrite-aluminum nitrate colorimetric assay, 20 % and 40 % ethanol elution sites enriched in flavonoid components were combined, concentrated to no alcohol and dried to obtain TFSc.

An Agilent SB-C₁₈ column (100 × 2.1 mm, 1.8 μm) was used to analyze the ingredients of TFSc at 40 °C. The mobile phase comprised 0.1 % formic acid aqueous solution (solvent A) and 0.1 % formic acid acetonitrile (solvent B) at 0.35 mL/min. The time program for gradient elution was as follows: 0–9 min, 5–95 % B, 9–10 min; 95 % B, 10–11.10 min; 95–5 % B, 11.10–14 min; 5 % B. Injection volume was 4 μL. A triple quadrupole-linear ion trap mass spectrometer was used to identify the composition of TFSc in positive and negative ion modes. Source temperature of electrospray ionization detector (ESI) was set at 500 °C, and ion spray voltage was set at 5500 V using positive ion mode and at −4500 V using negative ion mode. Instrument tuning and mass calibration were performed in triple quadrupole (QQQ) and LIT modes, respectively. A specific set of multiple-reaction monitoring (MRM) transitions was monitored for each period according to the metabolites eluted within that period. Based on the self-constructed local metabolic database of Biomarker Technologies (Beijing, China), the TFSc compounds were analyzed qualitatively and quantitatively using mass spectrometry.

2.3 Animals

Male C57BL/6 mice (18–22 g) were provided by SPF (Beijing) Biotechnology Co., Ltd (Beijing, China). All animals were kept in rooms maintained at 22–26 °C with a 12 h light, 12 h dark cycle. All animals were treated kindly and provided food and water ad libitum. Experiments were conducted in accordance with the requirements of the Animal Ethics Committee of Guizhou University of Traditional Chinese Medicine (N0. 20220046).

2.4 Mice with DSS-induced UC and drug administration

UC mouse models were constructed in accordance with the well-established methods (Fig. 2a). A total of 28 mice were randomly divided into control, UC, high-dose TFSc (39.02 mg/kg), and low-dose TFSc (19.51 mg/kg) groups. A 19.51 mg/kg dose of TFSc was equivalent to a human clinical dose of 0.25 g/kg of Sargentodoxa cuneata. Except for the animals in control group, which drank pure water, animals in other groups were giving water containing 2 % DSS for seven consecutive days, followed by pure water for another seven days. From days 1–14, the TFSc groups were administered TFSc at two different concentrations by intragastric gavage, whereas the control and UC groups were administered normal water.

2.5 Disease activity index (DAI) scoring and sampling

The animals were weighed every third days and on the last day. DAI was assessed on days 7, 10, and 13; and DAI scores included weight loss, stool consistency, and rectal bleeding scores. Weight loss was scored as 0–0 %, 1–1 %–5%, 2–5 %–10 %, 3–10 %–20 %, 4–>20 %, stool consistency as (0–negative; 1–malleable stool; 2–semi-sloppy stool; 3–loose stool; 4–severe loose stool); and rectal bleeding as (0–negative; 1–bloody stool visible to the naked eye (+); 2–++; 3–+++; 4–>+++). On day 14, animals of all the groups were sacrificed by cervical dislocation and the length of the colon tissue was measured. Moreover, the organ indices of heart, liver, spleen, lung, kidney, and thymus were also calculated. The colon and liver tissues were immobilized in 4 % paraformaldehyde solution, and the intestinal contents were collected for further experiments.

2.6 Histological evaluation

Colon and liver tissues were fixed with 4 % paraformaldehyde, processed, and stained using hematoxylin and eosin (HE) staining. Colon tissue damage index (TDI) scores comprised ulcer depth, degree of inflammatory cell infiltration, and depth of inflammatory cell infiltration scores. The following scores were allocated to the different damage parameters: depth of the ulcer (0, negative; 1, epithelium; 2, mucosal lamina propria; 3, mucosal muscularis propria), degree of inflammatory cell infiltration (0, negative; 1, mild; 2, moderate; 3, severe), and depth of inflammatory cell infiltration (0, negative; 1, mucosal layer; 2, mucosa and submucosa; 3, whole colon). Liver TDI scores consisted of cellular enlargement, cytoplasmic vacuolation, and narrowing of the hepatic sinusoids scores. The following scores were allocated to the different TDI parameters of liver: cellular enlargement (0, negative; 1, mild; 2, moderate; 3, severe), cytoplasmic vacuolation (0, negative; 1, mild; 2, moderate; 3, severe), narrowing of hepatic sinusoids (0, negative; 1, mild; 2, moderate; 3, severe).

2.7 Immunohistochemistry (IHC)

A wax block of colon tissue was taken and after deparaffinized using conventional methods and antigen was repaired with a citrate buffer solution. The sections were then incubated in 3 % hydrogen peroxide solution (hydrogen peroxide: purified water = 1:9) for 15 min at room temperature in dark to block endogenous peroxidase activity. The endogenous peroxidase activity was blocked by using 5 % goat serum for containment, and antibodies against tumor necrosis factor (TNF)-α (1:200, Affinity, Jiangsu, China) were added dropwise to the sections and incubated at 4 °C overnight. Next, secondary antibodies were added and incubated at 37℃ for 30 min, color developed using diaminobenzidine (DAB), restrained with Harris hematoxylin, and the sections were prepared for image acquisition. Hematoxylin-stained nuclei were blue, and positive staining with DAB appeared brownish-yellow.

2.8 16S rRNA gene sequencing of the intestinal flora

The genomic DNA of each group was extracted from the fecal samples and prepared for PCR amplification of selected genes in specific regions of the primer using specific primers containing barcodes. Fluorescence was quantified using a Quant-iT PicoGreen dsDNA assay kit and a microplate reader as the fluorescence quantification instrument. (BioTek, FLx800, USA). Amplicon library size and quantity were assessed using Agilent Bioanalyzer (Agilent, United States), and sequencing was performed on Illumina platform. NovaSeq 6000 SP kit reagent was used for sequencing. The raw sequences were de-primed, quality-filtered, de-noised, spliced, and de-chimerised using the Divisive Amplicon Denoising Algorithm 2 (DADA2) in FASTQ format. Operational taxonomic unit (OTU) analysis was performed with >97 % similarity. The abundance information, α- and β-diversity, along with community structure and species composition of the samples were analyzed using multivariate analysis. Furthermore, differences in biologically relevant taxonomic biomarkers among the groups were analyzed using linear discriminant analysis effect size (LEfSe).

2.9 Measurement of SCFAs and organic acids in intestinal contents

The intestinal content sample was accurately weighed, water was added, and the sample was homogenized for 3 min using zirconia beads (B B24, Next Advance, Inc., NY, USA). The sample was then treated with methanol, and standard internal substances were added. After centrifugation, the supernatant was analyzed using Biomek 4000 workstation (California, USA). The samples were centrifuged and subjected to UPLC-MS/MS analysis. UPLC-Xevo TQ-S system (MA, USA) was used for analysis. A Waters ACQUITY UPLC BEH C₁₈ column (2.1 × 100 mm, 1.7 µM) was used for separation at 40 °C. Gradient elution program [0–1 min, 12 % B; 1–2.8 min, 12–16.5 % B; 2.8–4.2 min, 16.5–24 % B; 4.2–5.6 min, 24–30 % B; 5.6–8.5 min, 30–100 % B; 8.5–10 min, 100 % B; 10–11 min, 100–1 % B; and 11–12 min, 1–12 % B (solvent A, 0.1 % formic acid; solvent B, acetonitrile)] was performed at 0.4 mL/min for the whole analytic process. Injection volume was 5 µL. The ESI ion source temperature was set at 150 ℃ using negative ion mode and the target ions were monitored.

2.10 Correlation analysis of intestinal microbiota with SCFAs and organic acids

SCFAs and organic acids have been found to be closely related to the intestinal flora and are important metabolites of intestinal flora (Lee et al., 2020). Spearman’s correlation coefficients were calculated to identify bacteria-metabolite correlations.

2.11 Statistical analysis

For analysis of pharmacodynamic indicators, intestinal microbiota, SCFA, and organic acids, data were presented as mean ± standard deviation (SD). One-way ANOVA and least significant difference (LSD) test was used to analyze the results among the groups using SPSS 20.0 software (Chicago, Illinois, USA). The α-diversity differences between samples of different groups were analyzed using Wilcoxon's rank-order analysis. P < 0.05 represented a statistical difference, while P < 0.01 indicated more significant statistical difference.

3.1 Analysis of the chemical constituents of TFSc using UPLC-MS

Broad-targeted metabolomic analysis identified 304 flavonoid components from TFSc. The major flavonoid components included flavanols, flavonols, flavones, dihydroflavonoids, dihydroflavonols, proanthocyanidins, chalcones, and isoflavones (Fig. 1, Table S1). The top 100 compounds in terms of metabolic abundance of the identified components are shown in Table 1. All the components were validated based on their accuracy and fragments. Components namely rutin, proanthocyanidin B2, epicatechin, catechin and quercetin have been reported previously (Lai, 2019, Zhang et al., 2021, Zhan et al., 2022).

Close

Fig. 1

Total ion chromatogram (TIC) and multiple-reaction monitoring (MRM) detection of multimodal maps of total flavonoids of Sargentodoxa cuneata (TFSc). (a) TIC in positive ion mode for TFSc sample. (b) TIC in negative ion mode for TFSc sample. (c) Positive ion mode for TFSc sample. (d) Negative ion mode for TFSc sample.

Export to PPT

Close

Fig. 2

Total flavonoids of Sargentodoxa cuneata (TFSc) improved the manifestations of mice with UC. (a) The process of this animal experiment. (b) Body weight changes. (c) Effects of TFSc on heart, liver, spleen, lung, kidney and thymus index scores. (d) The colon tissue and length. (e) DAI scores. Data are presented as mean ± SD (n = 7 per group). ^#P < 0.05 vs. the control group, ^##P < 0.01 vs. the control group, *P < 0.05 vs. the UC group, ^**P < 0.01 vs. the UC group.

Export to PPT

3.2 TFSc ameliorated colon and liver injury in mice with DSS-induced UC

Mice in the UC group experienced dramatic weight loss during days 10–13 (Fig. 2b). TFSc (39.02 and 19.51 mg/kg) inhibited weight loss in mice (Fig. 2b). Unexpectedly, TFSc significantly decreased the DAI scores on day 3 (Fig. 2e). However, on day 13, TFSc only demonstrated a statistically insignificant reduction in DAI scores, suggesting that the animals were already in the recovery phase of UC. DSS significantly shortened the colon length compared to that in the control animals (P < 0.01) (Fig. 2d). TFSc (39.02 and 19.51 mg/kg) improved the colon length. To preliminarily examine the effects of TFSc on the major organs of mice with UC, we measured the organ indices in each group of mice. TFSc (39.02 and 19.51 mg/kg) decreased the heart index of mice with UC, and 19.51 mg/kg TFSc also significantly decreased the liver index. TFSc showed some improvement in the immune organs (spleen and thymus), but the difference was not statistically significant. Moreover, improvements were also seen in the lung and kidney (Fig. 2c).

HE staining revealed that the mucosal, submucosal, muscular, and outer membranes of the colon were disrupted and infiltrated with inflammatory cells in mice with UC (Fig. 3). TFSc (39.02 and 19.51 mg/kg) reduced colonic tissue damage and inflammatory cell infiltration, as well as the colon TDI score (P < 0.05) (Fig. 3). In addition, cellular enlargement, cytoplasmic vacuolation, narrowing of the hepatic sinusoids, and slight inflammatory infiltrates were observed in mice with UC. This is consistent with the literature reporting that UC causes liver damage (Guan et al., 2022). TFSc improved the aforementioned pathological changes and decreased the liver TDI scores. These results indicate that TFSc ameliorated both colon and liver injuries in mice with UC.

Close

Fig. 3

Representative HE and IHC histological sections, TDI scores of colons and livers (200 × and 400 × magnification), and the integrated optical density (IOD)/area values of TNF-α in the colons (400 × magnification). Data are presented as mean ± SD (n = 6 per group). ^#P < 0.05 vs. the control group, ^##P < 0.01 vs. the control group, *P < 0.05 vs. the UC group, ^**P < 0.01 vs. the UC group.

Export to PPT

3.3 TFSc decreased the levels of TNF-α in the colon tissues in mice with UC

After induction by DSS, the levels of TNF-α in the mouse colon increased significantly (P ≤ 0.05), indicating the presence of immunoinflammatory injury in the mouse colon tissues. After treatment with TFSc, the levels of TNF-α in the colon of the animals reduced significantly, indicating that TFSc has a favorable anti-inflammatory protective effect on mice with UC (Fig. 3).

3.4 TFSc regulated gut microbiota dysbiosis in mice with UC

16S rRNA gene sequencing demonstrated that DSS treatment reduced the Chao1 (Fig. 4a), Simpson (Fig. 4b), and Shannon (Fig. 4c) indices. The TFSc dose of 39.02 mg/kg increased the Simpson and Shannon indices, suggesting recovery of intestinal microbial richness and diversity. However, the 19.51 mg/kg TFSc dose had less effect on intestinal microbial richness and diversity. Furthermore, PCoA plot of the groups revealed that the control, UC, and TFSc-treated groups were distributed across different spatial regions (Fig. 4d). The Venn graph and number of taxa in the gut microbiota are shown in Fig. 4e. These results suggest that TFSc modulates the diversity and relative abundance of intestinal flora in mice with UC.

Close

Fig. 4

Total flavonoids of Sargentodoxa cuneata (TFSc) affected the intestinal flora in mice with UC. (a) chao1 index. (b) Simpson index. (b) Shannon index. (d) PCoA analysis results of control group (C), UC group (M), and high- (H1) and low-dose (H2) TFSc groups. (e) The Venn graph. (n = 7 per group). *P < 0.05 vs. the control group, ^**P < 0.01 vs. the control group, and ^***P < 0.001 vs. the control group.

Export to PPT

The heatmap shows that the flora significantly altered by TFSc intervention primarily included Anaeroplasma, Turicibacter, Parabacteroides, Anaerotruncus, Mucispirillum, Corynebacterium, Odoribacter, Staphylococcus, Akkermansia, Streptococcus, Shigella, Proteus, Prevotella, and Lactobacillus (Fig. 5a). Fig. 5b and c show the bar plots of the gut microbiota at family and genus levels, respectively. TFSc significantly affected abundance of Bacteroides, Shigella, Allobaculum, Lactobacillus, Ruminococcus, Prevotella, and [Prevotella] at the genus level.

Close

Fig. 5

Total flavonoids of Sargentodoxa cuneata (TFSc) affected the relative abundance of intestinal flora in mice with UC. (a) The heatmap analysis. (b, c) The relative abundance of intestinal flora at the family and genus levels. The control group: C, UC group: M, high-dose TFSc group (H1), and low-dose TFSc group (H2) (n = 7 per group).

Export to PPT

Furthermore, LEfSe analysis was used to determine the differential intestinal flora in each group. A total of 81 taxa were identified from phylum to genus, including 25 taxa in the control group, 26 in the UC group, 11 in the high-dose TFSc group, and 19 in the low-dose TFSc group (Fig. 6). Differential flora in the 39.02 mg/kg TFSc group included Dehalobacterium, Peptococcaceae, Epsilonproteobacteria, Helicobacteraceae, Campylobacterales, Faecalibacterium, Rikenellaceae, Ruminococcus, Clostridiales, Clostridia, and Dehalobacteriaceae. Differential flora in the 19.51 mg/kg TFSc group included [Paraprevotellaceae], Alphaproteobacteria, RF32, YS2, 4C0d_2, Erysipelotrichaceae, Clostridium, Peptostreptococcaceae, rc4_4, Desulfovibrio, Sutterella, Alcaligenaceae, Desulfovibrionales, Desulfovibrionaceae, Deltaproteobacteria, Betaproteobacteria, Burkholderiales, RF3, ML615J_28, and [Prevotella] (Fig. 6b).

Close

Fig. 6

The results of LEfSe analysis. (a) The cladogram results. (b) The LDA score (log 10) of each group. The control group: C, UC group: M, high-dose TFSc group (H1), and low-dose TFSc group (H2) (n = 7 per group).

Export to PPT

3.5 TFSc affected the levels of SCFAs and organic acids, and results of Spearman’s correlation analysis

Compared to the control group, isobutyric acid, ethylmethylacetic acid, and succinic acid levels were reduced (P < 0.01), while malonic acid levels were increased in mice with UC (P < 0.05). The 39.02 mg/kg TFSc group demonstrated increased levels of isocaproic acid, 3-hydroxyisovaleric acid, and glutaric acid levels (P < 0.05) in the intestines. The 19.51 mg/kg TFSc group exhibited significantly increased levels of isobutyric acid and ethylmethylacetic acid (P < 0.01). Further, TFSc (39.02 and 19.51 mg/kg) significantly decreased the levels of malonic acid and succinic acid (P < 0.05) in the intestinal contents (Fig. 7).

Close

Fig. 7

Total flavonoids of Sargentodoxa cuneata (TFSc) affected the levels of SCFAs and few organic acids in the intestinal contents in mice with UC. The data are represented as mean ± SD (n = 6–7 per group). ^#P < 0.05 vs. the control group, ^##P < 0.01 vs. the control group, *P < 0.05 vs. the UC group, ^**P < 0.01 vs. the UC group.

Export to PPT

Moreover, a significant correlation of intestinal flora with differential SCFAs and organic acids was observed in each group at family (Fig. 8a, c, e) and genus levels (Fig. 8b, d, f). Spearman’s correlation analysis suggested that malonic acid, valeric acid, ethylmethylacetic acid, and isobutyric acid were associated with different gut flora (Fig. 8a, b). Succinic acid, hexanoic acid, glutaric acid, 3-hydroxyisovaleric acid, and isocaproic acid in high-dose TFSc group were associated with Bifidobacteriaceae, Peptococcaceae, Helicobacteraceae, Corynebacteriaceae, [Odoribacteraceae], Deferribacteraceae, Anaeroplasmataceae, Bifidobacterium, Proteus, Candidatus_Arthromitus, Anaeroplasma, Corynebacterium, Odoribacter, Rikenella, and Mucispirillum (Fig. 8c, d). Differential SCFAs and organic acids in the low-dose TFSc group were also found to be strongly associated with the intestinal flora (Fig. 8e, f).

Close

Fig. 8

Correlation analysis results. (a, b) The correlation analysis between the intestinal flora and SCFAs and few organic acids levels in the UC group at the family and genus levels. (c, d) The correlation analysis between the intestinal flora and SCFAs and few organic acids levels in the high-dose TFSc group at the family and genus levels. (e, f) The correlation analysis between the intestinal flora and SCFAs and few organic acids levels in the low-dose TFSc group at the family and genus levels. Significant differences were indicated at *P < 0.05, ^**P < 0.01 and ^***P < 0.001.

Export to PPT

Comparing the results of high- and low-dose TFSc groups (Fig. 8) demonstrated that TFSc exerts ameliorative effects on colon and liver injuries by modulating Bifidobacteriaceae, Helicobacteraceae, Corynebacteriaceae, [Odoribacteraceae], Bifidobacterium, Proteus, Corynebacterium, Odoribacter, and Rikenella, thereby affecting SCFAs and organic acids.