Chaihuang Qingyi Huoxue granule ameliorates severe acute pancreatitis by modulating gut microbiota and repairing the intestinal mucosal barrier

CH (Lot: 20220721) was purchased from the Affiliated Traditional Chinese Medicine Hospital of Southwest Medical University. The extraction methods and quality control were described in detail in a previous study (Zhao et al., 2021). Sodium taurocholate (T9034) was purchased from Sigma (New Mexico, USA). Sodium pentobarbital (170108) was purchased from Red Rock Reagent Factory (Tianjin, China). HE detection kit (C0105S) was purchased from Beyotime (Shanghai, China). The quantitative end-point chromogenic tachypieus amebocyte lysate assay kit (EC80545S) was purchased from Xiamen Houshiji (Xiamen, China). ZO-1 antibody (21773-1-AP) and occludin antibody (66378-1-lg) were purchased from Proteintech (Wuhan, China). High-efficiency RIPA lysis buffer (PC101), Bicinchoninic Acid Assay (BCA) kit (ZJ102), and high-sensitivity Electrochemiluminescence (ECL) detection kit (SQ201) were purchased from Epizyme (Shanghai, China). GAPDH antibody (AF7021) and horseradish peroxidase (HRP)-conjugated secondary antibody (S0001) were purchased from Affinity (Jangsu, China).

48 SPF male Sprague-Dawley rats (250 ± 20 g) were obtained from the Animal Experimental Research Centre of Southwest Medical University and housed under pathogen-free conditions (a regular 12/12 h light/dark cycle, 22 ± 2°C temperature, and 40-70% relative humidity) throughout the experiment. All experiments were approved by the Animal Ethics Committee of Southwest Medical University (No. 20231016-014). All rats had free access to standard rodent chow and water and were acclimated to the environment for 1 week before the formal experiment, and then were randomly divided into a control group (SHAM group), a model group (SAP group), and an intervention group (CH group), with 16 rats in each group, and the groups were divided into two subgroups of 8 rats each, a 12h group and a 24h group, according to the time after administration by gavage (12 hours and 24 hours). Rats in the SAP and CH groups were induced as SAP models by retrograde injection of 3.5% sodium taurocholate into the biliopancreatic duct (Hu et al., 2022), and the SHAM group received the same surgical intervention except that they were given 0.9% saline instead. The CH group received CH (4.4 g/kg) by gavage 6 hours after modeling. The SHAM and SAP groups received equal amounts of saline by gavage. Rats were sacrificed after 12 and 24 hours. Blood, pancreas, colon, and fresh feces samples were collected for further analysis.

Fresh pancreas and colon tissues were promptly fixed in a 4% paraformaldehyde solution, then dehydrated and embedded in paraffin. The embedded samples were stained with hematoxylin-eosin (HE) staining solutions. Histopathological evaluation was performed under a light microscope in a double-blind approach with the Schmidt pathology scoring criteria (Schmidt et al., 1992) for pancreatic tissue and the Chiu pathology scoring criteria (Chiu et al., 1970) for colonic tissue.

Serum amylase and lipase levels were measured using a fully automated biochemical analyzer (Olympus, Japan). Serum endotoxin was determined by a quantitative end-point chromogenic tachypieus amebocyte lysate assay kit.

The paraffin-embedded sections of colon tissue were deparaffinized, rehydrated, and sealed with bovine serum albumin. The sections were successively incubated with primary antibodies of ZO-1 (1:200), occludin (1:200) and the corresponding secondary antibodies, stained with 3,3'-diaminobenzidine (DAB) and hematoxylin, and visualized under a light microscope (Leica, Germany). Finally, the results were analyzed using the Image J software (NIH, USA).

Rats colon tissues were lysed in high-efficiency RIPA lysis buffer on ice. Isolated total intestine protein was quantified by the BCA detection method, 50 µg of protein per well was loaded for 7.5% SDS-PAGE gel electrophoresis and then transferred to a PVDF membrane. The membrane was blocked in 5% skim milk for 2 hours at room temperature and incubated with the specified primary antibodies (ZO-1, 1:5000 and occludin, 1:5000) at 4°C overnight, using GAPDH (1:20000) serving as an internal reference protein. The membrane was then incubated with HRP-conjugated secondary antibody (1:5000) at room temperature for 1 h. Signals were developed using the high-sensitivity ECL detection kit and images were captured using the TOUCH IMAGER™ (e-Blot, China). Image J software (NIH, USA) was used to analyze the grey intensity of the protein bands.

According to the manufacturer’s instructions, total microbial genomic DNA was extracted from all groups of fecal samples using the CTAB/SDS method and the purity determined on agarose gels. The V3-V4 region of the 16S rDNA gene was amplified with 341F-806R primers. The PCR products were purified using the AxyPrepDNA gel recovery kit (AXYGEN, USA). Sequencing libraries were generated using the NEB Next^®Ultra™DNA Library Prep Kit for Illumina (NEB, USA) and sequenced on the Illumina NovaSeq 6000 platform. Library preparation and sequencing were conducted at Shanghai Zhongke New Life Biotechnology Co., Ltd. (Shanghai, China).

Sequences with 97% identity were clustered into operational taxonomic units (OTUs) using Uparse software, and then diversity analysis, difference analysis, correlation analysis, and functional prediction analysis were performed. Alpha and beta diversity analyses were performed using the diversity plugin in the R language (R vegan package). Wilcoxon rank-sum test and ANOVA test (R stat package) were used to identify bacterial taxa with different abundances between groups at the phylum and genus level. Linear discriminant analysis (LDA) effect size (LEfSe) method was used to evaluate the influence of each differentially abundant taxon. Microbial function was predicted and analyzed by phylogenetic investigation of communities using reconstruction of unobserved states (PICRUSt) analysis based on Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (R ggpicrust2 package). LEfSe and PICRUSt analyses were performed in (https://www.bioincloud.tech/↗) (Gao et al., 2024). Spearman correlation analysis was used to assess the correlation between important bacterial taxa, SCFAs and disease phenotype.

Fecal samples were sequentially mixed with 0.5% phosphoric acid and ethyl acetate, centrifuged to obtain the supernatant, and 4-methylpentanoic acid was added as an internal control. GC-MS analysis was performed using a 7890 B gas chromatograph-5977 B mass spectrometer (Agilent, USA) with a capillary column (DB-FFAP, 30 m × 0.25 mm × 0.25 µm, Agilent, USA). The temperature was initially set at 90°C, then increased to 160°C at 10°C/min, and finally to 240°C at 40°C/min and held at this temperature for 5 min. The injector, ion source, quadrupole, and transfer line temperatures (°C) were 250, 230, 150, and 250, respectively. The helium carrier gas flow rate was maintained at 1 mL/min. MS conditions were electron impact ionization source, SCAN/SIM scan mode, and electron energy 70 eV.

Data were expressed as the mean ± standard error of mean (SEM), examined using SPSS 26.0 (SPSS Inc., USA), and performed with either R (version 4.3.3). P < 0.05 was considered to be statistically significant.

Serum amylase and lipase are important markers in the diagnosis of SAP. In the same subgroups at 12h and 24h, both serum amylase (Figure 1A) and lipase (Figure 1B) levels were significantly higher in the SAP group compared to the SHAM group (P < 0.05). In contrast, serum amylase and lipase levels were substantially lower in the CH group compared to the SAP group (P < 0.05). In addition, according to the HE staining results (Figure 1C), in the same subgroups at 12h and 24h, the pancreatic tissue structure was significantly damaged in the SAP group compared to the SHAM group, with pancreatic acinar cell necrosis, interstitial edema, inflammatory cell infiltration, local vascular congestion, and significantly higher pathological scores (P < 0.05). In contrast, pancreatic pathological damage was significantly attenuated and pathological scores (Figure 1D) were significantly decreased in the CH group compared to the SAP group (P < 0.05). In conclusion, CH could alleviate pancreatic damage and thus alleviate the symptoms of SAP.

According to the results of HE staining of colonic tissues (Figure 2A), in the same subgroups at 12h and 24h, the SAP group showed significant damage to the colonic mucosal tissue structure, absence of glands and crypts, necrosis of mucosal epithelial cells, detachment of intestinal villi, infiltration of inflammatory cells, and significantly higher colonic pathological scores (Figure 2B) compared to the SHAM group (P < 0.05). In contrast, the colonic pathological scores were significantly decreased in the CH group compared to the SAP group (P < 0.05).

The TJ proteins ZO-1 and occludin are key markers of intestinal epithelial integrity. The IHC results showed that in the same subgroups at 12h and 24h, the expression of both ZO-1 (Figures 2C, D) and occludin (Figures 2E, F) was significantly decreased in the SAP group compared to the SHAM group (P < 0.05). In contrast, the expression of both ZO-1 and occludin was significantly improved in the CH group compared to the SAP group (P < 0.05). Western blot analysis further confirmed that rats with SAP had downregulated the ZO-1 (Figures 3A, C) and occludin (Figures 3B, D) expression in the colonic tissue of compared to the SHAM group (P < 0.05). The levels of ZO-1 and occludin were increased in the CH group compared to the SAP group (P < 0.05). In addition, in the same subgroups at 12h and 24h, the endotoxin levels (Figure 3E) were significantly higher in the SAP group compared to the SHAM group (P < 0.05). In contrast, it was significantly lower in the CH group compared to the SAP group (P < 0.05). These results suggested that CH ameliorated the intestinal mucosal barrier damage in SAP by maintaining TJ protein expression and reducing intestinal permeability.

In this study, a total of 1838753 original microbiota sequences were obtained by 16S rDNA sequencing. The sequences of the obtained samples were clustered and species annotated, and a total of 1942 OTUs were obtained, with 97% similarity between the samples. The rarefaction curve can reflect the rationality of sequencing samples, and the species accumulation curve represents the relationship between the sample capacity and the number of species. The two curves in this study tended to be flat, indicating that the amount of sequencing data was reasonable and could be further analyzed (Figures 4A, B). The Venn diagram (Figure 4C) reflects the differences between each group at the level of OTUs, with 289, 208, and 260 in the SHAM, SAP, and CH groups.

The observed-spacies, Chao1, ACE, Shannon, and Simpson indices were used to reflect the α diversity (Liu et al., 2019), which reflects species richness and evenness (Figure 4E). The richness and diversity of the gut microbiota were significantly lower in the SAP group compared to the SHAM group (P < 0.05), whereas they were significantly higher in the CH group compared to the SAP group (P < 0.05). In addition, β diversity was also evaluated by principal coordinate analysis (PCoA), non-metric multidimensional scaling (NMDS) analysis based on Bray-Curtis distances between microbial genera, analysis of similarities (ANOSIM), and hierarchical clustering tree, which showed that the gut microbiota was significantly clustered among the three groups at the OTUs level, with the distribution of the CH group closer to the SHAM group (Figures 4D, F-H).

The effect of CH on potential metabolic pathways in the gut microbiota of SAP rats was predicted using PICRUSt analysis based on 16S rDNA gene sequencing data and the KEGG database. The KEGG pathways of the three groups of rats were mainly involved in two major metabolic pathways, namely metabolism and genetic information processing. The functional categories were mainly related to carbohydrate metabolism, energy metabolism, lipid metabolism, etc. (Figure 5A). The differences in KEGG level 2 and level 3 within the KEGG level 1 metabolic categories between the groups (Figure 5B) showed that the differences in the metabolic functions of the gut microbiota of the rats in the SAP and SHAM groups were mainly related to amino acid metabolism (alanine, aspartate and glutamate metabolism, lysine biosynthesis, histidine metabolism), carbohydrate metabolism (butanoate metabolism), and so on. In contrast, CH intervention brought the predicted gut microbiota function of SAP rats closer to that of rats in the SHAM group. In summary, these results suggested that CH alleviated SAP-induced gut microbiota dysbiosis.

At the phylum level, the three groups were mainly composed of Firmicutes, Bacteroidetes, and Proteobacteria, which accounted for 98.40%, 94.67%, and 92.82% of the gut microbiota in the SHAM, SAP, and CH groups, respectively (Figure 6A). Among them, the SHAM group was dominated by Firmicutes (57.12%) and Bacteroidetes (36.47%), which are important components for the production of SCFAs and maintenance of intestinal homeostasis. Compared to the SHAM group (Figure 6B), the percentage of Firmicutes (38.79%) in the SAP group was significantly decreased (P < 0.05), and the percentage of Proteobacteria (41.89%), which produce more pathogenic bacteria, was significantly higher (P < 0.05). However, CH treatment can significantly reverse these changes (p < 0.05; Figure 6C). In addition, the comparison of the three groups by ANOVA test showed differences in 5 phyla (Figure 6D).

The composition of the gut microbiota changed significantly at the genus level (Figure 7A), where the bacterial genera of Ruminococcus 1, Parabacteroides, and Lachnospiraceae NK4A136 group were reduced in the SAP group compared to the SHAM group, and Escherichia-Shigella, Escherichia, Enterococcus, and Enterobacter were significantly increased (Figure 7B). Compared to the SAP group, and the CH group had an increased percentage of potentially beneficial bacteria such as Ruminococcus 1, Lachnospiraceae NK4A136 group, Parabacteroides, and Lactobacillus, which maintain intestinal homeostasis, promote the production of SCFAs, protect the intestinal mucosal barrier, and inhibit inflammation, and a decreased percentage of pathogenic bacteria Escherichia-Shigella, Enterococcus, and Enterobacter (Figure 7C). Furthermore, the comparison of the three groups by ANOVA test showed differences in 18 genera (Figure 7D).

In addition, LDA and LEfSe analyses were performed to identify taxonomic biomarkers in the gut microbiota. The LDA bar chart (Figure 8A) presented the key biomarkers for each group. At the same time, the cladogram (Figure 8B) illustrated the taxonomic hierarchy of these biomarkers, summarized from the phylum to the species level. 11, 6 and 5 biomarkers were identified in the SHAM, SAP and CH groups, respectively, including Prevotellaceae UCG-001, Lachnospiraceae, Ruminococcus 1 and Escherichia-Shigella. In brief, these obvious changes in the gut microbiota at different taxonomic levels under the intervention of CH confirmed the regulatory effect of CH on the gut microbiota.

SCFAs, as important metabolites of the gut microbiota, have a significant impact on intestinal homeostasis and immune regulation. The results of principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) analyses (Figures 9A, B) suggested that the distribution of the CH group was close to that of the SHAM group, indicating the similarity of the composition of SCFAs between the SHAM and CH groups. Further analysis revealed (Figures 9C, D) that the levels of acetate, propionate, butyrate and valerate as well as total SCFAs were significantly decreased in the SAP group compared to the SHAM group (P < 0.05), whereas the intervention of CH restored them almost to the level of the SHAM group (P<0.05).

Spearman correlation analysis was used to determine the correlation between gut microbiota, SCFAs and disease phenotypes in SAP rats. Correlation analysis of SCFAs and phenotypes (Figure 9E) showed that the level of SCFAs was negatively correlated with pancreatic injury (amylase and lipase) and pathological damage (pancreatic and colonic pathological scores) and positively correlated with intestinal barrier proteins (ZO-1 and occludin). In addition, correlation analysis of gut microbiota with SCFAs showed that the relative abundance of Enterobacter, Enterococcus, and Escherichia-Shigella were negatively correlated with the level of SCFAs, and Lachnospiraceae NK4A136 group, Ruminococcus 1, and Prevotellaceae UCG-001 were positively correlated with the level of SCFAs (Figure 9F). Next, correlation analysis of gut microbiota and phenotypes showed that Enterobacter, Enterococcus and Escherichia-Shigella were positively correlated with pancreatic injury (amylase and lipase) and pathological damage (pancreatic and colonic pathological scores) and negatively correlated with intestinal barrier proteins (ZO-1 and occludin), while Lachnospiraceae NK4A136 group, Ruminococcus 1 and Prevotellaceae UCG-001 showed the exact opposite trend (Figure 9F). Finally, the associations between gut microbiota, SCFAs and disease phenotypes in SAP rats after CH intervention were visualized using correlation network plots (Figure 9G). These results suggested that CH treatment increased the relative abundance of SCFAs-producing bacteria in the rats and that SCFAs alleviated SAP by maintaining intestinal mucosal barrier function.

The original contributions presented in the study are publicly available. This data can be found here: https://www.ncbi.nlm.nih.gov/sra/PRJNA1220697↗, PRJNA1220697.