Sprague-Dawley male rats (n = 36, weight: 200 ± 20 g) of 6–8-week-old specific-pathogen-free were obtained from the Experimental Animal Center of Anhui Medical University, Hefei, China. The animals were housed in a controlled environment (25 ± 2°C, 50–60% humidity, 12 h/12 h light/dark cycle) and fed with standard rat chow and tap water ad libitum. One week of adaption was offered to all the animals before starting the study. The experimental protocol was reviewed and approved by the Animal Ethics Committees of Anhui Medical University (Approval number: 20170354). All efforts were taken to minimize the animal suffering during experiments.

Rats fasted for 12 h while water was free to access before operation. The IAH model was established according to the method previously described (Gong et al., 2011; Leng et al., 2016). In brief, rats were anesthetized with sodium pentobarbital 40 mg/kg by intraperitoneal injection. Their body was kept warm by a thermostatic blanket. After the abdomen was shaved and sterilized, a disposable venous infusion needle with a micro-infusion pump was punctured on the abdominal cavity for nitrogen gas insufflation. The IAP was raised exceeding 20 mmHg and maintained for 4 h. The sham group was subjected to the same operation without nitrogen gas insufflation. After 4 h of IAH, rats were sacrificed by an overdose of sodium pentobarbital (160 mg/kg). Samples of whole blood, hepatic tissue, jejunum tissue, and feces in the cecum were collected.

Rats were randomly divided into six groups (n = 6 in each group): (1) Control group (sham operation, Con); (2) IAH 4-h group, a sustained elevation in IAP exceeding 20 mmHg for 4 h; (3) PBS group, in which the rats were administered PBS intragastrically once daily for 5 consecutive days; (4) ABX group, in which the rats were administered antibiotics intragastrically to deplete the gut microbiota; (5) FMT-C group, in which the rats received feces oral gavage from the control group rats; (6) FMT-I group, in which the rats received feces oral gavage from the IAH 4-h group rats.

Antibiotics (vancomycin, 100 mg/kg; ampicillin, 200 mg/kg; metronidazole, 200 mg/kg; and neomycin sulfate, 200 mg/kg) were administered to rats intragastrically one time daily for 5 consecutive days to deplete the gut microbiota (Gong et al., 2021). Another group of rats was administered as described above, except the antibiotics were replaced by PBS. FMT in recipient rats was performed according to the several modified methods described previously (Gong et al., 2018; Xia et al., 2021). In brief, the stool contents of the donor rats (control group rats or IAH 4-h group rats) were harvested and resuspended in sterile PBS at 0.05 g/ml and centrifuged at 1,000 rpm for 15 min. The supernatants were aliquoted and frozen at −80°C until used. Recipient rats were orally gavaged with antibiotics as mentioned above for 5 consecutive days to deplete the gut microbiota, and then, an amount of 1 ml donor fecal contents were administered to the recipient rats by gavage once daily for 3 consecutive days.

Blood was obtained via cardiac puncture before the rats were sacrificed. The blood samples were centrifuged at 3,000 × g for 10 min at 4°C to collect the serum. An automated analyzer (HITACHI 3100, Tokyo, Japan) was used to measure the levels of serum aspartate aminotransferase (AST) and alanine aminotransferase (ALT).

Total genome DNA from collected feces was extracted and monitored by Equalbit dsDNA HS Assay Kit. The V3–V4 hypervariable regions of the 16S rDNA sequence were selected for generating amplicons and following taxonomy. A linker with an index was added to the end of the PCR product of 16S rDNA by PCR for NGS sequencing, and the library was purified with magnetic beads. The library was quantified to 10 nM, and PE250/FE300 paired-end sequencing was performed according to the Illumina MiSeq/NovaSeq (Illumina, San Diego, CA, United States) instrument manual. A quality filtered, purified, chimeric sequenced, VSEARCH clustering sequence was used for operational taxonomic unit (OTU) clustering (the standard of sequence similarity is set to 97%). Then, the ribosomal database program classifier Bayesian algorithm was used to analyze the representative sequences of OTU species taxonomy. The community composition of each sample was statistically analyzed under different species classification levels. Based on the OTU analysis results, the method of random sampling sample sequences was used; Chao1 index and Shannon alpha diversity index were calculated; community species abundance, diversity of rarefaction curves, and rank-abundance graph were drawn. Principal components analysis (PCA) was performed based on the sample OTU abundances table. Principal coordinates analysis (PCoA) and non-metric multidimensional scaling (NMDS) were calculated based on the distance between the Bray–Curtis matrix. Linear discriminant analysis effect size (LEfSe) was used to compare the hierarchy of evolution between-group differences of microbial community structure and species, which were shown by the species and the branching tree diagram. Metastats gap analysis was used to present the species abundance differences of microbial communities.

The liver tissues of rats were fixed in 10% formaldehyde, embedded in paraffin, and sectioned into 4 μm thick layers. After staining with hematoxylin and eosin (H&E), the pathological changes of images were detected using the TissueFAXSi-plus imaging system (TissueGnostics, Vienna, Austria). The histological changes of the hepatic sections were evaluated by pathologists via a blind test. The histological score of the liver was graded from 0 to 4 based on the severity of the inflammatory and necrosis process. The score was calculated by accumulating all the scores of each parameter and a maximum score was 12 (Ko et al., 2020).

The expression of claudin-1, occludin, and ZO-1 in the paraffin-embedded sections was detected by immunohistochemical staining. Briefly, the sections were incubated with claudin-1 (1:200, 28674-1-AP, Proteintech, Rosemont, IL, United States), occludin (1:200, 27260-1-AP, Proteintech, Rosemont, IL, United States), and ZO-1 (1:500, ab221547, Abcam, CA, United States) antibody overnight at 4°C after receiving antigen recovery. Then, incubation with goat anti-rabbit IgM (1:200, ab2891, Abcam, CA, United States) for 1 h was carried out. Images were scanned by the TissueFAXSi-plus imaging system (TissueGnostics, Vienna, Austria).

The rat hepatic tissues were extracted with RIPA lysis buffer (Beyotime, Jiangsu, China) to obtain protein samples. The concentration of protein was detected by the bicinchoninic acid assay kit (Beyotime, Jiangsu, China). Equal volumes of protein samples were fractionated by SDS-PAGE and then transferred to the PVDF membranes. Later, membranes were blocked in 5% skimmed milk for 1 h at room temperature. After incubation with primary antibody ERK1/2 (1:5,000, ab184699, Abcam, CA, United States), p-ERK 1/2 (1:1,000, ab201015, Abcam, CA, United States), JNK (1:2,000, ab208035, Abcam, CA, United States), p-JNK (1:5,000, ab76572, Abcam, CA, United States), p38 (1:2,000, 66234-1-Ig, Proteintech, Rosemont, IL, United States), p-p38 (1:2,000, 28796-1-AP, Proteintech, Rosemont, IL, United States) overnight at 4°C, HRP-conjugated secondary antibodies (1:1,000, SA00001-1/SA00001-2, Proteintech, Rosemont, IL, United States) were incubated for 1 h at room temperature. Protein bands were visualized by the BeyoECL Plus assay kit (Beyotime, Jiangsu, China). The quantification of target protein was analyzed using ImageJ software (National Institutes of Health, Bethesda, MD, United States).

All data were statistically analyzed by GraphPad Prism 8.4. The results were expressed as mean ± SD and evaluated using a two-tailed Student’s t-test. p < 0.05 was considered statistically significant between groups.

Figure 1 shows that the levels of ALT and AST in plasma were significantly increased in the IAH 4-h group (p < 0.05, Figures 1A,B). The appearance of the liver after H&E staining implied that rats in the IAH 4-h group also presented with more severe hepatic damage (Figures 1C–E).

An amount of 1,113,391 valid reads were acquired from 12 specimens by MiSeq sequencing (Con group and IAH 4 h group, n = 6 each group). The sequencing for our samples was deep enough to obtain most of the OTUs due to the rarefaction curves almost reaching the saturation plateau (Figure 2A). The Venn diagram (Figure 2B) result showed that 12 unique OTUs in the control group and 2 unique OTUs in the IAH 4-h group and 274 universal OTUs in both groups were detected. The OTUs abundance and alteration of gut microbiota in both groups were revealed by the OTU rank curve and OTU heatmap (Figures 2C,D). The alpha diversity of microbiota presented in this study illustrated that ACE (Figure 2E), Chao1 (Figure 2F), Shannon (Figure 2G), and Simpson (Figure 2H) indices were significantly lower in the IAH 4-h group than that in the control group (p < 0.05), indicating that the richness and diversity of gut flora in the IAH 4-h group were lower than that in the control group. The beta diversity of microbiota in the IAH 4-h group was different from the control group based on the unweighted pair-group method with arithmetic mean (UPGMA) analysis (Figure 3A), NMDS analysis (Figure 3B), PCoA (Figure 3C), and PCA (Figure 3D).

The changes in microbial composition at the class, family, genus, order, phylum, and species levels during IAH were presented in this study (Figures 3E,F and Supplementary Figure 1). At the phylum level, the most dominant phyla of the control group and the IAH 4-h group were Firmicutes, Bacteroidota, and Proteobacteria species. IAH increased the relative abundance of Firmicutes and Proteobacteria species and decreased the relative abundance of Bacteroidota species. Additionally, the relative differences in the microbial composition were further revealed by the levels of other microbial compositions. The LEfSe shows that the relative abundance of Clostridia, Lachnospiraceae, Gammaproteobacteria, Proteobacteria, Escherichia_Shigella, Enterobacteriaceae, Micrococcaceae, Rothia, Oscillospiraceae, and Family_XIII_UCG_001 was significantly higher in the IAH 4-h group (LDA score > |3|, Figures 3G,H).

As shown in Figure 4C, the IAH-induced liver injury was significantly ameliorated in the ABX group. Compared with the PBS group, the levels of plasma ALT and AST in the ABX group were decreased (p < 0.05, Figures 4A,B), and the HE scores of the liver in the ABX group were also declined (p < 0.05, Figures 4D,E).

To facilitate certification of our assumption that IAH-related gut microbiota dysbiosis promotes liver damage, an FMT operation was performed. First, the rats received antibiotics mentioned above once daily for 5 consecutive days to deplete gut microbiota, and then they received donor feces resuspended in sterile PBS from control or IAH-intervened rats for 3 days (Figure 5A). The rats that received feces of the IAH-intervened group exhibited more severe hepatic injury than those that received the feces of the control group after IAH intervention (p < 0.05, Figures 5B–F). Hence, we summarized that gut microbiota plays a key role in IAH-related hepatic damage.

The expression levels of tight junction proteins were investigated in this study, including claudin-1, occludin, and ZO-1. The immunohistochemical staining results indicate that the expression levels of claudin-1, occludin, and ZO-1 proteins were decreased after the rats intervened by IAH for 4 h and recovered after antibiotics were administered. Furthermore, the expression levels of the three intestinal tight junction proteins were lower in the FMT-I group than that in the FMT-C group (Figure 6A).

Figure 6 shows that the phosphorylation levels of JNK and p38 proteins were significantly higher in the IAH 4-h group than that in the control group (p < 0.05, Figures 6C,D), while no statistical difference was observed in the phosphorylation level of ERK1/2 protein between the two groups (Figure 6B). The phosphorylation levels of JNK and p38 proteins were lower in the ABX group compared with the PBS group and higher in the FMT-I group compared with the FMT-C group (p < 0.05, Figures 6C,D). The phosphorylation levels of ERK1/2 protein were also not statistically different (Figure 6B).