Atractylodin (purity > 98%) was purchased from Chengdu Push Biotechnology Co., Ltd (Chengdu, China). Sulfasalazine was procured from Shanghai Sine Tianping Pharmaceutical Co., Ltd (Shanghai, China). DSS (36,000–50,000 Da) was obtained from MP Biomedicals (Solon, OH, United States). Dimethyl sulfoxide and LPS O55:B5 were purchased from Merck (Darmstadt, Germany). Mouse TNF-α ELISA kit was purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). Lactate assay kit was obtained from Nanjing Jiancheng Bioengineering Institute. GAPDH assay kit was purchased from GENMED Biotechnology Co., Ltd. (Shanghai, China). Pierce co-immunoprecipitation kit was purchased from Thermo Scientific (Waltham, MA, United States). Antibodies against P38 (#9212S), P-P38 (#4511S), P44/42 MAPK (ERK 1/2) (#4695S), P-P44/42 MAPK (T202/Y204) (#9101S), SAPK/JNK (#9252S), and P-SAPK/JNK (T183/Y185) (#4668S) were purchased from CST (Boston, MA, United States). Antibodies against GAPDH (#60004-1-Ig), IL-1β (#16806-1-AP), and IL-6 (#66146-1-Ig) were purchased from Proteintech (Wuhan, China). Antibodies against β-actin (#K200058^M) was purchased from Solarbio (Beijing, China). MUC2 antibodies (#ab97386) were purchased from Abcam Inc. (Cambridge, MA, United States). Antibodies against TNF-α (#AF7014) were purchased from Affinity (Liyang, China). Anti-Kmal antibody (#PTM-901) was purchased from PTM Biotechnology Co., Ltd. (Hangzhou, China).

Specific-pathogen-free male BALB/c mice (6–8-weeks old, 18–22 g) were purchased from the Experimental Animal Center of China Three Gorges University (Yichang, China) (animal license No. SYXK (E) 2017–0,067). BALB/c mice were housed in SPF animal room with temperature of 24 ± 1°C and humidity of 50–70%. They were subjected to a 12 h:12 h light:dark cycle and allowed to acclimatize to their surroundings for one week. Forty mice were randomized into the following five groups: normal, model, atractylodin-10 mg/kg, atractylodin-20 mg/kg, and Sulfasalazine (SASP), with eight mice per group. From the first to the eighth day, the mice in the normal group were given free access to pure drinking water, whereas those in the other groups were provided access to a 3.5% DSS solution. The pure water and DSS solution were changed every two days. From the first to the seventh day, the mice in the atractylodin group were intraperitoneally injected atractylodin at a dose of 10 or 20 mg/kg, whereas those in the SASP group were intragastrically administered 250 mg/kg of sulfasalazine. On the eighth day, all mice were anesthetized using pentobarbital sodium and blood were withdrawn. The colons of mice were photographed and their lengths were measured. A 1 cm long piece of the colon was cut from 1 cm below the cecum for staining, immunohistochemistry, and immunofluorescence. The rest of the colon was reserved for western blotting.

Mice were sacrificed and the colon sections were rinsed with ice-cold phosphate-buffered saline (PBS). The excess PBS was blotted and the tissue samples were immediately fixed in 10% buffered formalin, dehydrated, and embedded in paraffin. Sections (5 μm thick) were stained using hematoxylin and eosin.

Paraffin sections of mice colon were treated with xylene I for 20 min, xylene II for 20 min, absolute ethanol I for 5 min, absolute ethanol II for 5 min, and 75% alcohol for 5 min, followed by washing with tap water. The sections were stained with AB-PAS C for 15 min and rinsed with tap water until colorless. Next, the sections were stained with AB-PAS B for 15 min, rinsed once with tap water, and twice with distilled water. AB-PAS A was allowed to equilibrate to room temperature, after which the sections were stained for 30 min in the dark and then rinsed for 5 min. The sections were further treated with anhydrous ethanol I for 5 min, anhydrous ethanol II for 5 min, anhydrous ethanol III for 5 min, xylene I for 5 min, and xylene II for 5 min. The transparent slices were sealed with neutral gum and observed using microscopy for image acquisition and analysis.

Individual scores were combined to generate the DAI, which was calculated daily for each mouse. The maximum score was 12 based on a 0–4 scoring system for the following parameters: Score = 0, Weight loss: none, Stool consistency: normal, Blood in stool: none; Score = 1, Weight loss: 1–5%, Stool consistency: loose stools, Blood in stools: presence of blood; Score = 2, Weight loss: 5–10%, Stool consistency: watery diarrhea, Blood in stools: presence of blood; Score = 3, Weight loss: 10–20%, Stool consistency: slimy diarrhea, little blood, Blood in stools: presence of blood; Score = 4, Weight loss: > 20%, Stool consistency: severe watery diarrhea with blood, Blood in stools: gross bleeding (Liu et al., 2020).

Colon tissue was fixed in 4% (w/v) paraformaldehyde solution, embedded in paraffin, sectioned, dewaxed, and then treated with the antigen. Next, the samples were incubated with primary and secondary antibodies at 4°C. Subsequently, the nuclei were stained using DAPI (Beyotime, Nanjing, China), and the sections were imaged using an Olympus FV 1000 laser confocal microscope (Olympus, Tokyo, Japan) (Shi et al., 2019).

Immunostaining was performed according to previously published protocols (Liu et al., 2020). The primary antibodies F4/80 (Affinity, 28463-1-AP), iNOS (Affinity, 18985-1-AP), ZO-1 (Abcam, ab216880), and occludin (Abcam, ab167161) were used. After incubation with the corresponding secondary antibodies, the cell nucleus was stained using DAPI (Roche, Shanghai, China) for 15°min, following which, the samples were observed and the images were captured using fluorescence microscopy (Olympus, Tokyo, Japan).

RAW264.7 cells were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Cells were cultured in DMEM with 10% fetal bovine serum, 1% nonessential amino acids, and 1% penicillin-streptomycin and incubated in an environment of 5% CO₂ at 37°C. RAW264.7 cells were seeded in 6-well plates or Petri dishes and divided into normal, model, low-dose atractylodin, medium-dose atractylodin, and high-dose atractylodin groups. The low, medium, and high-dose groups were stimulated with LPS for 24 h and treated with 10, 20, and 40 μM of atractylodin, respectively. The normal group was not stimulated with LPS, whereas the model group was subjected to LPS stimulation; both the normal and model groups were not subjected to atractylodin intervention.

Trizol reagent was used to collect cells from the 6-well plates (Guangzhou Jet Bio-Filtration Co., Ltd: TCP-010–006), after which RNA was extracted and purified. The cDNA of each sample was reverse transcribed using a commercial kit (Vazyme Biotech Co., Ltd., Nanjing, China). Then, the reverse transcription product was used as a template to perform real-time polymerase chain reaction (PCR) using a Step One Plus thermal cycler and PowerUp™ SYBR™ Green Master Mix (Vazyme Biotech) following the manufacturer’s instructions. All the primers were referenced from the previous study and synthesized by Invitrogen. The primer sequences are shown in Supplementary Table S1. The final data were analyzed using the 2^-∆∆CT method.

Western blotting was performed according to standard protocols. Briefly, mice colon tissues were homogenized and lyzed with a cleavage buffer containing the cocktail. The total protein was separated using SDS-PAGE and transferred to a polyvinylidene fluoride membrane. The membrane was sealed with skim milk for 2°h, incubated with the primary antibody overnight, and then incubated with the secondary antibody at room temperature for 2 h. An ECL chemiluminescence detection kit was used to detect the protein bands.

The cell-culture medium was collected and centrifuged at 1,000 xg for 20 min at 4°C. The supernatant was used to detect the concentration of TNF-α according to the manufacturer’s instructions provided in the ELISA kit (Linghang et al., 2020).

The cell protein lysates were collected, 3 μM trichostatin A (GLPBIO, Montclair, United States) and 50 mM nicotinamide (Merck, Darmstadt, Germany) were added for protein extraction. First, the GAPDH antibody was immobilized on the coupling resin and then the control agarose resin was used to pretreat the cell lysate. Next, the GAPDH-immobilized antibody was used for immunoprecipitation. After elution, the malonylated antibody was used for SDS-PAGE detection.

The supernatants of RAW264.7 cells treated for 24 h and untreated cells were collected and centrifuged at 1,000 xg for 20 min at 4°C to determine the lactate levels according to the manufacturer’s instructions provided in the lactate assay kit (Nanjing Bioengineering Institute, Nanjing, China). First, 20 μL of sample, 1 ml of working enzyme solution, and 200 μL of chromogenic agent were added to a test tube. Then, the mixture was vortex mixed and incubated at 37°C for 10 min. Then, 2 ml of termination reagent was added to arrest the reaction. Lastly, the OD value was determined at 530 nm using a spectrophotometer.

GAPDH activity was determined using colorimetry. Briefly, 3 ml of GENMED cleaning solution was added to both treated and untreated RAW264.7 cell cultures to rinse the cell surface after 24 h of treatment. Next, 3 ml of GENMED cleaning solution was added to the cells and centrifuged at 300 xg for 5 min at 4°C. The supernatant was discarded and 500 μL of GENMED lysis solution was added, mixed well, vortexed for 15 s, incubated in ice for 30 min, and centrifuged at 16,000 xg for 5 min at 4°C. Next, 500 μL of the supernatant was taken to 1.5 ml tube. About 10 μL of the sample was used for protein quantification. Briefly, 150 μL of GENMED buffer solution was taken in a 96-well plate and 20 μL of GENMED reaction solution was added to each well. Then, 20 μL of GENMED substrate solution was added, mixed well, and incubated at 25°C for 3 min. Lastly, GENMED negative solution or a sample containing 20 μg of protein was added to the corresponding wells and mixed thoroughly. Absorbance of the samples was determined at 340 nm using a microplate reader.

Fecal genomic DNA was extracted from 0.1 g of frozen fecal samples using cetyltrimethyl ammonium bromide or sodium dodecyl sulfate and the purity and concentration were determined using agarose gel electrophoresis. Amplicon libraries covering the V3–V4 hypervariable regions of the bacterial 16S-rDNA gene were amplified using primers 338F: 5′-ACT​CCT​ACG​GGA​GGC​AGC​A-3′, and 806R: 5′-GGACTACHVGGGTWTCTAAT-3′. PCR was performed using a 25 μL mixture containing 5 μL of 5× buffer, 2 μL of 2.5 mM dNTPs, 1 μL of each primer (5 μM), 0.25 μL of Fast Pfu polymerase, 1 μL of template DNA, 14.75 μL of ddH₂O. PCR was conducted as follows: initial denaturation for 3 min at 95°C followed by 27 cycles of 30 s at 95°C, 30 s for annealing at 55°C, 45 s for elongation at 72°C, and a final extension at 72°C for 10 min. PCR products were detected using 2% agarose gel electrophoresis. The library was constructed using TruSeq Preparation DNA PCR-Free sample preparation kit. The library was quantified using Qubit and Q-PCR, qualified, and sequenced using NovaSeq 6,000.

Statistical analysis was performed using GraphPad v8.0. Data are presented as mean ± standard deviation. Student’s t-test or one-way ANOVA followed by Bonferroni test was used to compare two independent variables; ns, not significant, ^# p < 0.05, ^## p < 0.01, *p < 0.05, **p < 0.01.

The weight of mice that consumed DSS-infused water decreased significantly from the third day compared to those in the normal group. The weight loss of the ATL (10 or 20 mg/kg) intervention groups and SASP group was lower than that in the model group (Figure 1C). Compared to that of the normal group, the disease activity index (DAI) score of the model group was significantly different from the third day (p < 0.05) and significantly different from the fourth day (p < 0.01). Compared to that of the model group, the DAI scores of ATL (10 or 20 mg/kg) and SASP groups increased gradually, and there was a significant difference between the ATL (10 or 20 mg/kg) and model groups from the third day (p < 0.05). A significant difference was found between the SASP and model groups on the fifth day (p < 0.05) (Figure 1D). After DSS therapy, the colon lengths of mice were found to be significantly shortened (p < 0.01). The mice in the ATL (10 or 20 mg/kg) and SASP groups did not exhibit colon shortening. Compared to the mice in the model group, those in the ATL (10 or 20 mg/kg) and SASP groups exhibited significant changes (p < 0.01) (Figures 1E,F). After treatment with DSS, the colonic glands of mice were obviously disordered, the goblet cells were obviously absent, and considerable inflammatory infiltration was observed compared to mice in the normal group. After the administration of ATL (20 mg/kg), the arrangement of glands was more orderly than that in the model group; the morphology of goblet cells recovered and inflammatory infiltration was significantly reduced (Figure 1G).

The number of goblet cells in the colon of mice that consumed DSS decreased significantly compared to that in the normal group. After treatment with ATL (20 mg/kg), the number of colonic goblet cells increased significantly (Figure 2A). The expression of MUC2 in the colons of DSS-induced mice was found to be significantly decreased. After atractylodin (20 mg/kg) treatment, the expression of MUC2 in the colon increased significantly (Figure 2B). The expression of ZO-1 and occludin in the colons of DSS-induced mice decreased significantly. After treatment with ATL (20 mg/kg), the expression of tight junction proteins ZO-1 and occludin increased significantly (Figures 2C,D).

The levels of pro-inflammatory cytokines, TNF-α, IL-1β, and IL-6, were significantly increased in mice that consumed DSS-infused water. After treatment with ATL (20 mg/kg), the levels of the pro-inflammatory cytokines were found to be significantly decreased (Figure 3A). Moreover, in DSS-induced mice, F480 and iNOS were determined to be significantly increased, whereas these levels decreased significantly after atractylodin treatment (Figure 3B). In DSS-induced mice, expression of the MAPK pathway proteins, such as phosphorylated p38, phosphorylated JNK, and phosphorylated ERK, was significantly upregulated in the colon tissue (Figure 3C). After treatment with ATL (20 mg/kg), phosphorylated p38 (Figure 3D), phosphorylated JNK (Figure 3E), and phosphorylated ERK levels (Figure 3F) in the colon tissue were significantly downregulated.

Cell viability assays showed that atractylodin was not cytotoxic to RAW264.7 cells at a concentration of 0–40 μM (Figure 4A). The mRNA levels of TNF-α, IL-1β, IL-6, and iNOS in RAW264.7 cells stimulated by LPS (100 ng/ml) for 24 h were significantly increased. The expression of IL-1β, IL-6, and iNOS decreased by varying degrees after 24 h of atractylodin intervention (Figures 4B,C–E). The MAPK pathway proteins, including phosphorylated p38, phosphorylated JNK, and phosphorylated ERK, in the LPS-induced macrophages were upregulated (Figure 4F). After treatment with different concentrations of atractylodin, phosphorylated p38 (Figure 4G), phosphorylated JNK (Figure 4H), and phosphorylated ERK (Figure 4I) in the macrophages were downregulated.

TNF-α was found to be highly expressed in the supernatant and lysate of LPS-treated macrophages; however, its levels were decreased by varying degrees after treatment with different concentrations of atractylodin. The expression of TNF-α protein significantly decreased after treatment with 20 μM (p < 0.05) atractylodin and significantly decreased at 40 μM (p < 0.01) (Figures 5A,B) of atractylodin. Lactate levels in the supernatant of the LPS-treated macrophages were found to be significantly increased, whereas treatment with different concentrations of atractylodin led to a decrease. Lactate levels were significantly decreased after treatment with 20 μM (p < 0.05) of atractylodin and extremely significantly decreased upon treatment with 40 μM (p < 0.01) of the compound (Figure 5C). Molecular docking experiments with atractylodin and GAPDH revealed binding free energy of 5.42 kcal/mol (Figure 5D) and suggested that GAPDH was likely an anti-inflammatory target of atractylodin. We investigated the effects of atractylodin on GAPDH activity and found that its activity decreased gradually with an increase in atractylodin concentration (Figure 5E). However, different concentrations of atractylodin did not affect the expression of the GAPDH protein (Figure 5G). In LPS-treated macrophages, malonylation of the total protein was significantly enhanced, and we found that atractylodin could inhibit this process (Figure 5F). Results from the immunoprecipitation assay revealed that atractylodin could significantly inhibit malonylation of the GAPDH protein (Figure 5H).

The cumulative species curve revealed that nearly 5,000 species of intestinal microorganisms were present in the three samples, which could essentially cover the common species (Figure 6A). The number of common intestinal bacteria in the normal and atractylodin (20 mg/kg) group was significantly higher than that in the normal and model group (Figure 6B). The grade curve at the species level showed that the abundance of the intestinal microflora was the most in the normal group and the least in the model group. The abundance of intestinal microflora in the atractylodin (20 mg/kg) group was midway between the normal and model groups (Figure 6C). Cluster analysis revealed that the microflora composition of the normal and model groups was significantly separated, whereas that of the atractylodin (20 mg/kg)-treated groups was close to the composition of the normal group (Figure 6D). Principal component analysis and principal coordinate analysis showed that the distribution of intestinal flora in the normal and model groups was well separated, whereas the atractylodin (20 mg/kg) and normal groups were closely clustered (Figure 6E). The distribution maps of the top 20 genera and top 20 phyla with the highest abundance among different groups of mice in this study are shown in Figures 6F,G, respectively. The abundance of Firmicutes decreased significantly after DSS treatment but increased significantly in the atractylodin (20 mg/kg) group. After DSS treatment, the abundance of Proteobacteria and Deferribacteres increased significantly, whereas treatment with atractylodin (20 mg/kg) resulted in a significant decrease in their abundance (Figure 6H). After DSS modeling, the abundance of pathogenic bacteria, including Helicobacter, Desulfovibrio, Bacteroides, Flavonifractor, and Mucispirillum increased significantly, while that of beneficial bacteria, including Alistipes and Muribaculum, decreased significantly. Atractylodin (20 mg/kg) treatment not only decreased the abundance of these pathogenic bacteria but also increased the abundance of beneficial bacteria including Akkermansia (Figure 6I).