Standard diet rats, male, 6–8 weeks old, weighing 180∼220 g [License No. SCXK (Jing) 2019-0010], were purchased from SiPeiFu (Beijing) Biotechnology Co. The rats were placed in a controlled environment with a temperature of 23 ± 1°C and a humidity of 60 ± 5%, with a 12 h light/dark cycle. They were provided with a standard diet and had access to water at will. A 7 days acclimatization period was conducted under these control conditions before the start of the experiment. At the end of the acclimatization period, they were classified as blank, 0.5, 1, 2, 3, 6, 8, 12, and 24 h using the random number table method. Six rats in each group were given the same dose of MMQ-8 drug by gavage every day. A normal dose of MMQ-8 was given on day 6 with water fasting, and blood samples were collected on day 7. An Agilent ultra-high performance liquid chromatography 1,290 UPLC system with a Waters UPLC BEH C18 column (1.7 μm 2.1 × 100 mm) was used. The column temperature was 55°C, and the injection volume was 5 μL. The flow rate was set at 0.5 mL/min. The mobile phases were 0.1% formic acid aqueous solution (A) and 0.1% formic acid acetonitrile aqueous solution (B). A multi-step linear gradient elution program of 0∼11 min, 85∼25% A; 11–12 min, 25–2% A; 12–14 min, 2–2% A; 14–14.1 min, 2–85% A; 14.1–15 min, 85–85% A; 15–16 min, 85–85% a was performed using a Q Exactive Focus mass spectrometer Combined with Xcalibur software, MS and MS/MS data were acquired using IDA acquisition. In each acquisition cycle, the mass range was 100∼1,500, and the first three of each cycle were filtered to acquire the corresponding MS/MS data further. Sheath gas flow rate: 45 Arb, auxiliary gas flow rate: 15 Arb, capillary temperature: 350°C, full ms resolution: 70,000, ms/ms resolution: 17,500, collision energy: 15/30/45 in NCE mode, spray voltage: 4.0 kV (positive) or 4.0 kV (negative). The raw mass spectrometry readings were imported using XCMS software. Various tasks, including contention time correction, peak identification, extraction, integration and alignment, were performed to improve the accuracy of the data. Substance identification of peaks containing MS/MS data was performed using a self-constructed secondary mass spectrometry database and a matching method based on the corresponding cleavage rules.

Mongolian Medicine Qiqirigan-8 plant ratios were 207 g for K. galanga L; 66 g for I. helenium L.; 66 g for D. costus (Falc.) Kasana and A.K. Pandey; 41 g for R. palmatum L.; 248 g for H. rhamnoides L.; P. longum L. 165 g; B. sappan (L.) Tod for 124 g; S. scrofa L. for 83 g. These plants were mixed proportionally and pulverized using a pulverizer. Moreover, filtered through a 70 mesh sieve. If it could not be filtered, we continued to pulverize and sift the mixture until all the pulverized mixture passed through the filter. Finally, it was carefully dried and stored in a cool, dry place. Clinical administration requires a dose of 3 g of MMQ-8 per dose for an adult weighing 60 kg, which implies a dose of 50 mg/kg body weight. Based on body surface area, the equivalent dose ratio for humans to mice is 12.3. Therefore, the dose of MMQ-8 given to mice is 50 mg/kg of body weight (formula: MMQ-8 50 mg/kg × 12.3 g = 0.615 g/kg). Finally, after calculating the dose for each mouse, the drug was dissolved in 2 ml of H₂O. In this study, MMQ-8-low dose group (0.615 g/kg) and MMQ-8-high dose group (1.23 g/kg) were administered by gavage for 12 weeks, and all treatments were administered by gavage once a day at a fixed time.

C57BL/6J mice, male, 4–6 weeks old, 18–20 g [License No. SCXK (Jing) 2019-0010] were purchased from SiPeiFu Co (Beijing, China). Mice were kept in a normal environment during a 12 h light/dark cycle with ad libitum access to food and water. All mice were randomly divided into four groups (n = 8): control, model, MMQ-8 low dose, and MMQ-8 high dose. Except in the Control group, which was fed a standard diet, the other groups were fed a choline-deficient diet (60% fat-supplied methionine 0.1%, choline-deficient feed, batch no. 20230612 purchased from Jiangsu Xietong Pharmaceutical Bio-engineering Co (Jiangsu, China). Where the Control group did not receive any treatment, the Model group was gavaged with H₂O, MMQ-8 low dose (0.615 g/kg), and MMQ-8 high dose (1.23 g/kg) daily for 12 weeks. All animal experiments were approved by the Animal Protection and Use Committee of Inner Mongolia Medical University (approval number: YKD202301165).

Serum alanine aminotransferase (ALT, C009-2, Nanjing Jiancheng Biological Co., Ltd., China), aspartate amino-transferase (AST, C010-2, Nanjing Jiancheng Biological Co., Ltd., China), and hepatic triglyceride (TG, A110-1, Nanjing Jiancheng Biological Co., Ltd., China) levels were assayed by using commercially available diagnostic kits according to the manufacturer’s instructions.

Histopathologic evaluation

Tissue samples from the liver and colon were paraffin-embedded, sliced to a thickness of 5 μm, and put on separate slides. The sections were stained with HE (Hematoxylin, G1004, Solarbio, China; Eosin, BA-4024, Basso, China) and ABPAS (G1285, Solarbio, China), respectively, and observed by light microscopy. HE staining of the liver was used for NAFLD activity score (NAS) (Qu et al., 2023). Steatosis was scored from 0 to 3 (0: < 5% steatosis; 1: 5–33%; 2: 34–66%; 3: > 67%). Hepatocyte ballooning was scored from 0 to 2 (0: normal hepatocytes, 1: normal-sized with pale cytoplasm, 2: pale and enlarged hepatocytes, at least 2-fold). Lobular inflammation was scored from 0 to 3 based on foci of inflammation counted at 20X (0: none, 1: < 2 foci; 2: 2–4 foci; 3: R4 foci). NAS was calculated as the sum of steatosis, hepatocyte ballooning and lobular inflammation scores.

Colon sections were prepared according to the procedure described above, and the sections were incubated with anti-ZO-1 (21773-1-AP, proteintech, China) primary antibody overnight at 4°C, followed by incubation with fluorophore-coupled secondary antibody (5220-0336, SeraCare, China) for 1 h. Cell nuclei were stained with DAPI (C0065, Solarbio, China). Cell nuclei were stained with DAPI (C0065, Solarbio, China). Images were acquired using a fluorescence microscope (BX51, Olympus, Japan).

RNA was extracted from mice colon tissues using TRIzol reagent (Lot#B04, ProbeGene, China) according to the manufacturer’s instructions, and RNA was reverse-transcribed into cDNA using a reverse transcription kit (FSQ-201, TOYOBO, Japan). Next, qRT-PCR was performed on a QuantStudio3 Real-Time PCR Instrument (A28132, ThermoFisher, United States) using SYBR Green real-time PCR Master Mix (QPK-201, TOYOBO, Japan). Details of the messenger RNA-specific primers for Occludin, ZO-1 are shown in Table 1. Relative mRNA levels were determined using the comparative Ct method, with GAPDH mRNA as the reference gene, and Equation 2^–ΔΔCT.

Five mice feces were taken from each of the three groups, the Control group, the Model group and the MMQ-8 group, and DNA was isolated in a sterile environment. The DNA samples that passed the test were randomly interrupted by a Covaris ultrasonic crusher to form fragments with a length of about 350 bp. and inserted into metagenome shotgun sequencing libraries. Paired-end sequencing was performed using the Illumina HiSeq high-throughput sequencing platform. The raw data were subjected to quality control and host filtering to obtain Clean Data. Then, functional annotation and abundance analysis of metabolic pathways (KEGG) were performed; the α-diversity of fecal bacteria was calculated based on the normalized OTU table of R package Vegan; based on the species abundance table and functional abundance table, abundance clustering analyses such as PCA, PcoA, and NMDS, and multivariate statistical analysis of LEfSe, as well as comparative metabolic pathway analyses, were carried out to excavate the differences in the species compositions and functional compositions among samples; and the differences in species and functional compositions among samples were explored. Functional composition differences.

A 100 μL serum sample from five mice in each of the Control, Model, and MMQ-8 groups was placed in an EP tube, and 400 μL of 80% methanol in water was added to the tube. The sample was vortexed and shaken, allowed to stand for 5 min on an ice bath, and centrifuged at 15,000 g for 20 min at 4°C. A certain amount of supernatant was diluted with mass spectrometry-grade water to 53% methanol; the supernatant was collected by centrifugation at 15,000 g for 20 min at 4°C and injected into the LC-MS for analysis. The data (.raw) files were imported into CD 3.3 library software for processing. The data were pre-processed, identified with metabolites, and finally analyzed statistically. Data were transformed and subjected to principal component analysis (PCA) and partial least squares discriminant analysis (OPLS-DA) using metabolomics data processing software meta X. The criteria for differential metabolite screening were VIP > 1, P-value < 0.05 and FC ≥ 1.2 or FC ≤ 0.5. The KEGG database was used to study the functions and metabolic pathways of metabolites, which were considered to be enriched when x/n > y/n and significantly enriched when the P-value of the metabolite pathway < 0.05. Screening of KEGG pathways related to NASH was performed using the microbiology platform¹ for histogram plotting.

The above gut microbiota and non-targeted metabolomics data were subjected to Spearman correlation analysis using the cloud platform².

Data are expressed as mean ± SEM. Non-parametric Kruskal Wallis one-way analysis of variance (ANOVA) was used to compare the ranking data between different groups. When the difference is significant (P < 0.05), then the Mann-Whitney U test is performed, which is determined by pairwise comparison of significant differences in each group.

To further clarify the effective material basis of MMQ-8 to improve NAFLD, we analyzed MMQ-8-treated serum samples and observed the number of peaks and their responses in the chromatograms (Figures 1A, B). Compounds were identified by considering retention times, molecular ions, major fragment ions, and information from published articles and online databases. Subsequently, 17 potentially active components of MMQ-8 were screened. These identified compounds mainly belonged to the flavonoid class. Table 2 gives detailed information about the identified compounds in the serum samples, including retention times, precise molecular weights, etc.

To investigate the pharmacological effects of MMQ-8 on NAFLD and liver injury, we used a choline-deficient diet-induced NAFLD model. C57BL/6J mice were randomly assigned to Control, Model, MMQ-8 low-dose, and MMQ-8 high-dose groups for 12 weeks (Figure 2A). Choline-deficient diet feeding significantly increased liver/body weight (Figure 2B), serum AST (Figure 2C), and ALT (Figure 2D) levels leading to hepatic injury as compared to the Control group. Importantly, MMQ-8 treatment significantly reduced hepatic ALT levels in NAFLD mice, with a protective effect against liver injury. In addition, the gross appearance of the peritoneal cavities (Figure 2E) and liver TG (Figure 2F) results showed that MMQ-8 could attenuate hepatomegaly and reduce hepatic steatosis in NAFLD mice. Pathological analysis of liver HE staining showed hepatic steatosis, hepatocyte balloon-like lesions, and significant lobular inflammatory infiltration in the Model group mice compared with the Control group. In contrast, liver steatosis and lobular inflammation were significantly reduced after MMQ-8 intervention (Figure 2G), and the NAS activity score was decreased (Figure 2H), which was consistent with previous reports (Narenmandula et al., 2022). In conclusion, these results suggest that MMQ-8 can alleviate hepatic steatosis and liver injury in NAFLD mice.

In order to further study the repair effect of MMQ-8 on the intestinal epithelial mucosal barrier of NAFLD, HE staining of the colon was performed (Figures 3A, B). It was found that the colonic surface structure was complete in the model group, but inflammatory cell infiltration could be seen. However, the villi of the colonic mucosa of MMQ-8 mice were more neatly arranged than that of the model group, and the inflammatory cell infiltration was reduced. The trend of the results of AB-PAS staining (Figures 3C, D) was similar to that of HE staining. Compared with the Control group, the epithelial cells and goblet cells in the superficial layer of the colonic mucosa of mice were full and rounded, while the epithelial cells and goblet cells in the superficial layer of the colonic mucosa of the mice in the Model group were shrunken and their number appeared to be reduced to different degrees. However, MMQ-8 treatment restored the reduced number of goblet cells compared to the Model group. Immunofluorescence staining (Figure 3E) showed that the fluorescent expression of ZO-1 was reduced in the colon of mice in the Model group compared with the Control group. In contrast, the expression of ZO-1 was significantly enhanced in the colon of both MMQ-8 groups. In summary, MMQ-8 may significantly promote the repair of intestinal damage.

Therefore, we further analyzed the colon-intestinal barrier tight junction-related genes by qRT-PCR. The results showed that the expression of ZO-1 (Figure 4A) and Occludin mRNA (Figure 4B) was significantly suppressed in the colon of the Model group. In contrast, MMQ-8 treatment significantly increased expression in the colon, suggesting a possible improvement in intestinal barrier tight junctions.

To explore whether MMQ-8 improves the gut microbiota to attenuate NAFLD, we used fecal metagenomics. Firstly, the gut microbiota of the Control, Model, and MMQ-8 groups was subjected to the Veen diagram (Figure 5A), and it was found that the three groups of samples shared 591,301 species. Based on gut microbiota abundance, the results of principal component analysis (PCA) (Figure 5B) showed a clear separation of the three groups, with MMQ-8 being closer to the Control group, suggesting that the microbiota species compositions were closer to each other. In addition, the α-diversity index of the Model group was relatively low, and the α-diversity index of the intestinal microbiota (Figure 5C) was not significantly improved after MMQ-8 treatment, suggesting that MMQ-8 did not have a significant effect on the diversity of the various microbiota in the intestinal tract. PcoA (Figure 5D) and NMDs (Figure 5E) analyses were next performed to assess the β-diversity among the three groups, and there were significant differences among the three groups, suggesting that the three groups exhibited different microbiota clustering. Thus, MMQ-8 treatment-induced changes in gut microbial community structure in NAFLD mice without affecting microbiota α-diversity. Notably, MMQ-8 treatment significantly modulated microbiota abundance at the phylum level (Figure 5F), genus level (Figure 5G) and family (Figure 5H) level.

Next, linear discriminant analysis effect size (LDA Effect Size, LEfSe) identified several families (LDA scores > 4) that distinguished the three gut microbiota groups. Among them, Erysipelotrichaceae was significantly dominant in the Model group, and Akkermansiaceae was significantly dominant in the MMQ-8 group (Figure 6A). At the phylum level, there was a significant increase in the relative abundance of Bacillota (Figure 6B) and a significant decrease in the relative abundance of Bacteroidota (Figure 6C) in the model group compared to the control group. In contrast, MMQ-8 intervention reversed the above changes and significantly reduced the Bacillota/Bacteroidota ratio in NAFLD mice (Figure 6D). At the genus level, MMQ-8 significantly increased Prevotella (Figure 6E) abundance and significantly down-regulated Acetatifactor (Figure 6F) abundance. Porphyromonadaceae abundance was significantly up-regulated at the family level (Figure 6G).

Furthermore, functional prediction of the gut microbiota showed significant changes in metabolic pathways after MMQ-8 intervention. Energy metabolism, amino acid metabolism, and lipid metabolism were changed in the Model group compared to the Control group. Significantly, MMQ-8 treatment reversed these changes. Compared with the Model group, metabolic pathways such as amino acid metabolism, bile acid metabolism, butyric acid metabolism and sphingolipid metabolism, glycerophospholipid metabolism, fatty acid degradation, and PPAR signaling pathway appeared to be significantly altered after MMQ-8 treatment (Figure 7A). Increased levels of Porphyromonadaceae, Prevotella, and Bacteroidota after MMQ-8 treatment were generally positively correlated with lipid metabolism, butyric acid metabolism, and amino acid metabolism and negatively correlated with the abundance of Bacillota, Acetatifactor, and Erysipelotrichaceae (Figure 7B). Therefore, it is hypothesized that MMQ-8 regulates the gut microbiota, thereby affecting metabolic pathways in NAFLD mice.

To further explore the metabolic mechanisms on which MMQ-8 ameliorates NASH, we performed serum untargeted metabolomics. Principal component analysis (PCA, Figure 8A) and orthogonal partial least squares discriminant analysis (OPLS-DA, Figure 8B) showed that different metabolic changes were found in the serum of MMQ-8 and Model group mice. Scatterplot visualization of the two groups of differentially expressed metabolites using P-value, and Fold Change (Figure 8C) revealed that MMQ-8 significantly up-regulated 103 metabolites and down-regulated 82 metabolites compared with the Model group. We then performed a KEGG enrichment analysis of these differential metabolites and found that unsaturated fatty acid synthesis was the most significant (Figures 8D, E).

Mongolian Medicine Qiqirigan-8 intervention significantly regulated the expression of metabolites related to glutathione metabolism, butyric acid metabolism, sphingolipid metabolism, and glycerophospholipid metabolism in NAFLD mice. Where MMQ-8 treatment resulted in an overall increase in circulating glutathione (Figure 9A). Fumaric acid and 2-Oxoglutaric acid were significantly increased in butyric acid metabolism (Figure 9B). Glycerophospholipid metabolism (Figure 9C) in MMQ-8 significantly down-regulated LPC levels and up-regulated PC levels after intervention. In addition, sphingolipid metabolism (Figure 9D) significantly increased in adenosine metabolites. Thus, the above results showed that MMQ-8 significantly improved metabolic pathways.

To further explore the potential effects of MMQ-8 on the gut microbiota and serum metabolites in NAFLD mice, we performed Spearman correlation analyses to link the altered metabolites to the gut microbiota. Specifically, MMQ-8 up-regulation of Porphyromonadaceae, Prevotella, and Bacteroidota as a whole was negatively correlated with down-regulated LPC levels in the glycerophospholipid metabolism pathway and positively correlated with up-regulated PC levels (Figure 10A). The opposite relationship was observed for Bacillota, Acetatifactor, which was down-regulated after MMQ-8 intervention. In addition, Porphyromonadaceae, Prevotella, and Bacteroidota up-regulated after MMQ-8 intervention were positively correlated with Fumaric acid and 2-oxoglutaric acid, Adenosine, and L-Glutathione. In contrast, Bacillota and Acetatifactor, which were down-regulated after MMQ-8 intervention, were negatively correlated with all of the above metabolites (Figure 10B).