Ursolic acid improves the bacterial community mapping of the intestinal tract in liver fibrosis mice

Experimental design and animal models

All wild-type (WT) C57BL/6 mice used in this study were from the Department of Laboratory Animal Science of Nanchang University. Mice were kept in a specific pathogen-free (SPF) environment with a 12  h light/dark cycle, a room temperature of 22 ± 2 °C, and 55 ± 5% humidity. Mice weighing 25–35 g were randomly divided into a control group, a CCl₄ group and a UA group (n = 8). Mice in the control group were gavaged with olive oil (2 ml/kg) twice a week for 8 weeks. Mice in the CCl₄ group were gavaged with carbon tetrachloride (CCl₄) (1:4 diluted in olive oil, 2 ml/kg) twice a week for 8 weeks, and the mice in the UA group were gavaged with CCl₄ for 4 weeks and then simultaneously with CCl₄ and UA (40 mg/kg/day) for the last 4 weeks. This experimental protocol meets the National Institutes of Health Guide for the Care and Use of Laboratory Animals and has been approved by the Medical Research Ethics Committee of the First Affiliated Hospital of Nanchang University (No. 66265256).

Histological analysis

For the observation of liver tissue damage and fibrosis, the livers were fixed in 4% paraformaldehyde, embedded in paraffin and sectioned for haematoxylin and eosin (H&E) and Masson’s trichrome staining. We randomly selected five fields of view for observation, and a professional pathologist scored the degree of liver fibrosis based on the METAVIR scoring criteria (Poynard, Bedossa & Opolon, 1997) (Table 1).

Serum index test

Immediately after the mice were sacrificed, blood was collected from their hearts. The serum was then analysed using an automated blood biochemistry analyser for the detection of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and total bilirubin (TBIL) (Department of Clinical Laboratory, First Affiliated Hospital of Nanchang University, China).

Genomic DNA extraction and sequencing

Genomic DNA was extracted from the ileal contents and faeces of mice using E.Z.N.A. stool DNA kit (Omega Biotek, Norcross, GA, United States). Primers 338F 5′- ACTCCTACGGGAGGCAGCAG-3′and 806R 5′- GGACTACHVGGGTWTCTAAT-3′were used for polymerase chain reaction (PCR) amplification of the V3–V4 region of the 16S rRNA gene of the extracted DNA. The PCR amplification of 16S rRNA gene was performed as follows: 95 °C for 3 min; 25 cycles at 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s; and a final extension at 72 °C for 10 min. Purified amplicons were pooled in equimolar and paired-end sequenced (2 × 300) on an Illumina MiSeq platform (Illumina, San Diego,USA). The raw 16S rRNA gene sequencing reads were demultiplexed, quality-filtered by Trimmomatic and merged by FLASH. All raw reads were deposited into the open resource database.

Analysis of the 16S function prediction of microbiota

The 16S function prediction removes the influence of the number of copies of the 16S marker gene in the genome of the species by using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) software, which stores the Cluster of Orthologous Groups (COG) information and Kyoto Encyclopedia of Genes and Genomes (KEGG) information corresponding to the greengene id, that is, to normalize the Operational Taxonomic Units (OUT) abundance table, and then to correspond to each OTU. The greengene id obtains the COG family information and KEGG Orthologue (KO) information corresponding to the OTU and calculates the abundance and KO abundance of each COG. According to the information of the COG database, the descriptive information of each COG and its functional information can be parsed from the evolutionary genealogy of genes: non-supervised Orthologous Groups database to obtain a functional abundance spectrum; according to the information in the KEGG database, KO and Pathway information can be obtained, and according to the OTU abundance, the abundance of each functional category can be calculated. Finally, the predicted functional composition profiles were collapsed into levels 1∼3 of COG and KEGG database pathways. The output file was further analyzed using the Statistical Analysis of Metagenomic Profiles (STAMP) software package (Parks et al., 2014).

Statistical analysis

GraphPad Prism 7.0 software was used for image production and output. Each experiment was repeated 3 times to ensure confidence in the results. A one-way analysis of variance (one-way ANOVA), Student’s t test, the Mann–Whitney rank sum test, or the Kruskal-Wallis H test was used to analyse the significant differences between groups using SPSS 23.0 software. P < 0.05 was considered significant.

UA reversed liver damage and fibrosis in CCl₄-treated mice

We validated the CCl₄-induced liver fibrosis model and the effect of UA on it (Gan et al., 2018). H&E and Masson’s trichrome staining on liver sections were used to determine the effect of UA on liver damage and fibrosis in CCl₄-treated mice (Figs. 1A–1F). In CCl₄-treated mice, the normal structure of the liver and the hepatic lobule was destroyed, and collagen deposition occurred, accompanied by inflammatory cell infiltration and hepatocyte necrosis. However, after UA treatment, the structure of the liver nodules and collagen deposition showed a certain improvement, accompanied by less inflammatory cell infiltration and hepatocyte swelling. Histological quantitative analysis also confirmed the improvement of liver injury and fibrosis by UA (Figs. 1G–1H). Moreover, we also measured serum from mice to assess liver function (Figs. 1I–1K). Compared with the control group, the serum ALT, AST, and TBIL levels in mice in the CCl₄ group increased. The ALT, AST, and TBIL levels in mice in the UA group showed a significant decrease compared with the CCl₄ group, indicating the protective effect of UA in liver fibrosis mice. These results suggest that UA can improve CCl₄-induced liver damage and liver fibrosis in mice.

High-throughput sequencing across the different anatomic sites of the mouse intestinal tract

Previous studies on the changes in microbiota in liver fibrosis often used conventional bacterial culture methods (Fouts et al., 2012; Gomez-Hurtado et al., 2011). However, this culture technique has certain limitations, and more than 80% of the bacteria cannot be cultured (Mylotte & Tayara, 2000). As the importance of the microbiota in liver fibrotic diseases increases and more microbiota are discovered, more advanced and accurate microbial identification analysis techniques are needed. Here, we used high-throughput 16S rRNA gene sequencing technology to detect the bacteria in the ilea and faeces of the mouse model. In this study, we used a total of 2,153,777 high-quality sequences with a read length ≥200 for analysis. These high-quality sequences were assigned to 1 domain, 1 kingdom, 17 phyla, 35 classes, 66 orders, 106 families, 226 genera, and 382 species from the ileum mucosa and anal faeces of all the mice (Table 2). Moreover, we counted the number of OTUs common and unique in the sample to visually represent the similarities and differences of the microbiota composition of the samples under different treatments (Figs. 2A–2B). The results showed that in the ileal samples, there were 503 common OTUs in the control group, CCl ₄group, and UA group and 550 common OTUs in stool samples. The total number of OTUs in the faeces is greater than in the ileum, which may suggest a higher degree of microbiota in the faeces.

Effect of UA on the diversity of bacteria in liver fibrosis mice

To better explore the effect of UA on the bacterial microbiota of liver fibrosis, we analysed the diversity and composition of the ileal and faecal intestinal microbiota. We first tested the correlation index for the alpha diversity of the intestinal microbiota. The value of the Chao1 index, the estimated microbial abundance, showed a decrease in CCl₄-treated liver fibrosis mice, whereas the value of this decline increased in liver fibrosis mice after UA treatment (Figs. 3A–3B). The Shannon index, used to assess the diversity of the microbiota, also showed similar changes (Figs. 3C–3D). The value of the Shannon index of bacteria in the CCl₄ group mice was lower than that in the control group. In the UA group, the value of the Shannon index increased. In addition, the Chao1 value in the ileum was lower than that in faeces, and the Shannon index was higher in the faeces, indicating that the diversity and abundance of microbiota damage caused by liver fibrosis was partially improved after UA treatment and that the diversity of the microbiota in the ileum may be slightly higher than that in the faeces, while the abundance is the opposite.

Effect of UA on the bacterial composition of liver fibrosis mice

Liver fibrosis often affects the composition of the intestinal microbiota through the intestinal axis, further aggravating primary liver disease (Fouts et al., 2012; Zhu et al., 2012). Therefore, we next analysed the changes in intestinal microbiota composition during liver fibrosis and the effect of UA on these changes. Principal coordinate analysis (PCoA) plots show that the bacterial composition of the control group, CCl₄ group and UA group belongs to different communities (Figs. 4A–4B). We also compared the composition of the microbiota between different groups. At the phylum level (Figs. 4C–4F), Firmicutes, which includes probiotic bacteria such as Lactobacillus, and Bacteroidetes decrease in CCl₄-treated mice. However, decreased Firmicutes and Bacteroidetes in CCl₄-induced liver fibrosis mice increased after UA treatment. Elevated Verrucomicrobia in the CCl₄ group showed a certain decrease in the UA group. At the genus level (Figs. 5A–5D), compared with the control group, Lachnospiraceae decreased in the CCl ₄group, while Akkermansia increased. This change was reversed in the UA group. Moreover, linear discriminant analysis effect size (LefSe) is a software for discovering high-dimensional biomarkers and revealing genomic features. We used LefSe to reveal the composition of the microbiota of the ilea and faeces of mice in different groups (Figs. 6–7). The results show Firmicutes enrichment in liver fibrosis mice undergoing UA treatment. However, there are separately enriched species in the ileal or faecal microbiota, indicating that UA can improve the disorder of intestinal microbiota caused by liver fibrosis, but the changes in the sites of different intestines are not completely consistent.

Prediction and analysis of microbiota function

Changes in the microbiota under different body conditions indicate some changes in metabolic function. We predicted the function of microbial changes in different animal models by querying the COG and KEGG databases. The COG database included pathways for bacterial and archaeal genomes. Compared with control mice, the enriched pathways were “Cell wall/membrane/envelope biogenesis”, “Intracellular trafficking, secretion, and vesicular transport”, and “Secondary metabolites biosynthesis, transport and catabolism” in the bacteria of the ilea of mice in the CCl₄ group. After UA treatment, enriched pathways were transformed into “Carbohydrate transport and metabolism”, “Transcription”, “Defense mechanisms”, and “Translation, ribosomal structure and biogenesis” (Figs. 8A–8B). In faeces bacteria, liver fibrosis mice enriched pathways were “Intracellular trafficking, secretion, and vesicular transport”, “Signal transduction mechanisms”, and “Secondary metabolites biosynthesis, transport and catabolism”. The enriched pathways of faeces in UA-treated liver fibrosis mice were “Translation, ribosomal structure and biogenesis”, “Transcription”, “Defense mechanisms”, “Replication, recombination and repair” and “Energy production and conversion” (Figs. 8C–8D).

The KEGG database mainly contains information about metabolic signalling pathways. We performed functional predictions on the microbiota of different parts by searching the KEGG database. In the bacteria of the ilea of mice in the CCl ₄group, the main metabolic KEGG terms were “Genetic Information Processing”, “Amino Acid Metabolism”, and “Cellular Processes and Signaling”. Compared with the CCl₄ group, the main KEGG terms were “Membrane Transport”, “Cell Motility”, “Replication and Repair”, “Nucleotide Metabolism”, and “Translation” in the bacteria of the ilea of mice in the UA group (Figs. 9A–9B). Then, in the functional prediction of faecal microbiota, CCl₄-induced liver fibrosis mice mainly included the metabolic pathways “Cellular Processes and Signaling”, “Poorly Characterized”, and “Signaling Molecules and Interaction”. The main metabolic pathways in UA-treated liver fibrosis mice were “Replication and Repair”, “Translation”, “Nucleotide Metabolism”, and “Enzyme Families”, similar to the control mice (Figs. 9C–9D).