1 Praxiology Change of Mice with Stomach Perfusion of Folium sennae extracts
During gastric administration with sterile saline, the mice in pfck group had dense and glossy back hair with bright eyes; a diet water quantity, stools and mental state were normal. Diarrhea symptoms appeared in the afternoon of the first day of Folium sennae extracts given by gavage in pfm group mice, and feces began to become thinner and softer. Feces of pfm group mice became thinner and perianal contamination on the second day of modeling. Mice in pfm group began to show hairy back, sleepiness, arched back and dry back; and diarrhea symptoms in the afternoon were more serious than those in the morning of the third day of modeling. On the 8th day, it was found that the stool in pfck group was water-like after dissecting the mice, while in pfm group, the stool was granular, moist, dark, normal in size and shape. There was no significant difference in body weight change and body weight change rate between pfm group and pfck group [(9.97+4.48) g vs (10.26+5.00) g, p-value = 0.06; (29.23+9.87) % vs (28.74+10.86) %, p-value = 0.06].
2 Intestinal Microbial Diversity in Diarrhea Mice Caused by Folium sennae extracts
2.1 Sample sequences and operational taxonomic units
The number of 16S rRNA PCR sample sequences is positively correlated with the number of microbial populations, and the intensity of sample sequences indicates the number of DNA fragments, hence the size of a certain microbial population. There were a total of 475,973 effective sequences and 439,518 high quality sequences obtained by sequencing mostly between 150 bp and 300 bp in this study for subsequent analysis.
According to sequence homogeneity, Qiime soft was used in clustering 97 % similarity DNA sequences into a same Operational Taxonomic Unit (OTU). As shown in Figure 1, 288 and 202 OTUs were uniquely identified from two groups, respectively; and the number of OTUs coincided between the two groups was 749. The result suggested that the OTU of bacterial gene had decreased after diarrhea caused by Folium sennae.
2.2 Bacterial diversity in intestinal contents of diarrhea mice caused by Folium sennae
In order to illustrate the diversity of intestinal microbiota in the two group of mice, we calculated Alpha diversity and Beta diversity [18]. The Alpha diversity index includes Chao1 (showing the richness of bacteria) and Shannon (showing the diversity of bacteria): the higher the Alpha diversity, the better the richness of the bacterial species, the more uniform the number of intestinal bacteria, and the more stable the microbiota. The results showed that both the Chao1 index curve (Figure 2A1) and Shannon index curve (Figure 2B1) of the sample had entered the plateau and reached saturation, which suggested that the amount of sequencing data of our study is large enough to reflect the vast majority of microbial species information in the sample. As shown in Figure 2, within all the range of gene sequences pfm group mice showed lower Chao1 index and Shannon index. Overall, Chao1 index (Figure 2A2) and Shannon index (Figure 2B2) decreased with no statistically significant after Folium sennae intervention, and Shannon index decreased with statistical significance (p-value < 0.05). It suggested that Folium sennae could affect intestinal microbiota diversity in mice.
Beta diversity is a comparative analysis of microbial community composition of different samples. PCA and NMDS analysis was conducted to measure differences in communities of bacteria at genus level to illustrate the Beta diversity. PCA of log2-transformed normalized abundance for OTUs with normalized total abundance. Unifrac analysis was carried out to obtain the distance matrix of the differences between samples by using the evolutionary information of species and the abundance information of species. The distance matrix information between the samples was analyzed by NMDS. As shown in Figure 3, with PCA (Figure 3A) or NMDS (Figure 3B) analysis of OTU, OTUs of pfm group were mainly concentrated in the right quadrant. Thus it can be inferred that the intestinal microbial diversity of normal mice could be changed by Folium sennae.
3 Active Ingredients of Folium Sennae Regulate the Key Protein CYP3A4 and Influence Tryptophan Metabolic Pathways Associated with Diarrhea
3.1 Filtration of active ingredients and potential target genes associated with diarrhea of Folium sennae
Total 53 chemical constituents of Folium sennae were retrieved from TCMSP. All compounds were subjected to Active Perl 5.26 screening, and a total of 10 active compounds had OB≥ 30 % and DL≥ 0.18. Therefore, the selected 10 compounds (Table 1) from Folium sennae were subjected to further analysis. After removing the redundancy, total of 370 potential targets (Table S1) were obtained from the 10 active ingredients. Genecard database was used to establish a new diarrhea-related gene database, which was docked with the selected 370 target genes of Folium sennae. Finally, 51 potential target genes of Folium sennae-Diarrhea were filtered out (Table S2).
3.2 Compounds-Targets-Disease (C-T-D) network construction and analysis
After removing the redundancy lacking target protein gene information, we further eliminated compounds whose target genes did not intersect with the diarrhea-related gene database. A total of 4 compounds were incorporated into the C-T-D network construction, including rhein, kaempferol, sitosterol and stigmasterol. As shown in Figure 4, the 4 active compounds and related 51 target genes constructed the network schematic diagram. Totally, this C-T-D network is composed of 57 nodes (4 active compounds and 51 potential targets) and 333 edges. In this picture, the edges indicated an association between the active ingredients, target genes and diarrhea. Degree values indicated the intensity of the interaction among the ingredients of Folium sennae, target genes and diarrhea, particularly in the C-T-D network kaempferol was the potential active ingredient that could regulate the CYP3A4 protein target gene, which had an effect on diarrhea.
3.3 PPI network construction and analysis
We constructed functional protein association PPI network to screen target protein genes that play a key role in diarrhea caused by Folium sennae. 49 target genes (2 disconnected nodes of genes, F7 and SCN5A, were hid) associated with active ingredients and diarrhea were imported into the STRING database for PPI network construction and analysis. There are 49 interacting targets in the network, resulting in 278 edges representing the interaction between proteins. As shown in Figure 5, the 49 protein genes were clustered into three groups according to the interaction relationship, showing three different node colors; in the PPI network, the line color indicates 7 different type of interaction evidence: gene neighborhood , gene fusions, gene co-occurrence, co-expression, protein homology, texting and experimented.
According to the protein relationship of PPI network, the protein relationship is further analyzed by the number of protein links. Combining the disease score of genes (the higher the score, the more evidence is available for disease regulation), and the disease score originates in the diarrhea related gene database (Table S2), the key proteins are selected. As shown in Figure 6, the color of protein dots represents the counting of PPI network protein connections. The horizontal ordinate represents the disease score of genes, and there are three protein genes include AKT1, TNF and CYP3A4 with disease score greater than 15. The higher the disease score represents the stronger the evidence of diarrhea related protein genes, and AKT1, TNF and CYP3A4 are the main target protein genes of diarrhea caused by Folium sennae.
3.4 Kyoto encyclopedia of genes and genomes (KEGG) metabolic signaling pathway analysis
Intestinal microbiota can synthesize vitamins and amino acids necessary for human growth and development through bacterial metabolism, and it participates in the metabolism of sugars and proteins [11]. Both metabolism of substance and energy in human body by intestinal microbial community and metabolites of intestinal microbiota community play an important role. In order to effectively connect with the microbiota function, we choose the perspective of metabolic pathways to explore the mechanism of diarrhea caused by Folium sennae and intestinal microbiota disorder. To determine the relevant metabolic signaling pathways involved in diarrhea effect of Folium sennae and intestinal microbiota, we conducted pathway enrichment analysis using KEGG metabolic pathways. A total of 51 targets obtained 80 KEGG signaling pathways, and 57 channels were significantly enriched (p-value < 0.05). As shown in Figure 7, among the 80 KEGG signaling pathways there were 8 KEGG metabolic signaling pathways of gene enrichment. The color of the bars in the graph was decided the p-values, and there were statistically significant the first 6 metabolic signaling pathways (p-value < 0.05). The abscissa indicates the count of proteins enriched in the pathway. The pathway enriched more protein, the more evidence suggests that this metabolic pathway is the main mechanism of diarrhea caused by Folium sennae. Drug metabolism - cytochrome P450 (CYP450) and metabolism of xenobiotics by CYP450 are the top 2 metabolic pathways of diarrhea caused by Folium sennae with count 7 and 6 respectively. In addition, the results of this study suggested that arginine and proline, tyrosine and tryptophan metabolism are the main pathways of diarrhea caused by Folium sennae.
Previous studies have suggested that the amino acid tryptophan played a key role in the agonistic binding of an inducer of CYP450 by activation of the constitutive androstane receptor and confirmed the structural impact of mutations of tryptophan on CYP450 [19, 20]. Therefore, we hypothesized that diarrhea caused by Folium sennae could affect tryptophan metabolism by diversity disorder of intestinal microbiota. Based on the sequencing results of microbial taxonomy, we validated the hypothesis by simulating intestinal microbial metabolic activities with VHM.
4 Diarrhea Caused by Folium sennae Affects Intestinal Bacterial Characteristic and Tryptophan Metabolism
4.1 Diarrhea caused by Folium sennae extracts affects intestinal bacterial characteristic in mice
According to the OTU annotation results, the abundance level and composition ratio of each sample in different taxonomic levels had been obtained, which reflects the community structure of pfck group and pfm group in different taxonomic levels. By comparing the community structure of the two groups in different taxonomic levels, the effects of Folium sennae on the community structure of intestinal microbiota in mice were analyzed. In this study, abundance of bacteria of OTU in genus level of single sample in pfck group and pfm group was obtained. Based on the information of microbiota and its abundance, the structure map of microbiota community could reflect comprehensively the distribution of microbiota and abundance in samples, and the microbiota with higher abundance (top 20) could be found (Figure 8). The picture had reflected the overall difference in the intestinal bacterial characteristic of the two groups of mice.
Lactobacillus had high abundance in pfck group (57 %) and pfm group (69 %) on phylum level. And as shown in Figure 9, comparing the OTU abundance of the two groups, there were 33 bacterial microbiota communities (p-value < 0.05) with significant difference in OTU abundance between the two groups. Comparing the OTU abundance of the two groups, there were 10 bacterial microbiota communities with OTU abundance greater than 0.1 % and q-value less than 0.05 on genus level. Paraprevotella, Streptococcus, Epulopiscium, Sutterella and Mycoplasma increased significantly in the genus level of intestinal microbiota in mice with Folium sennae intervention (q-value < 0.05); the abundance of Adlercreutzia, Lactobacillus, Dehalobacterium, Dorea and Oscillospira were reduced in the pfm group (q-value < 0.05).
4.2 VMH simulation of tryptophan metabolism in intestinal microbiota
Bacteria communities labeled by OTU which its abundance changed with the intervention of diarrhea caused by Folium sennae were selected to simulate tryptophan metabolism of intestinal bacteria. The intestinal metabolic function of the 10 microbiota from the above results was simulated, 7 of which were related to tryptophan metabolism through the synthesis of functional enzymes (Table. 2). In theory, all the microflora except Paraprevotella interfered by Folium sennae extracts were influence EX_trp_L(e) and TRPt2r to complete the metabolism of tryptophan, and L - Tryptophan was produced by the intestinal microbiota enzymatic reaction and completes amino tryptophan metabolism. We therefore concluded that diarrhea caused by Folium sennae influenced mainly tryptophan metabolism of intestinal microbiota.