Abstract

Background

   Tongxieyaofang (TXYF) was a traditional Chinese medicine (TCM) formula
   for the treatment of diarrhea with liver stagnation and spleen
   deficiency syndrome, but the potential targets and mechanisms have not
   been fully clarified. This study aims to explore the potential
   mechanisms of TXYF in alleviating diarrhea using network pharmacology
   and full-length 16S rRNA gene sequencing.

Methods

   Network pharmacology was applied to identify bioactive compounds and
   potential targets involved in the role of TXYF in alleviating diarrhea.
   Meanwhile, a model of diarrhea with liver stagnation and spleen
   deficiency syndrome was constructed by intragastric administration of
   Folium senna extract combined with restraint and tail pinch stress. The
   effect of TXYF on intestinal mucosal microbiota of diarrhea mice was
   analyzed by full-length 16S rRNA gene sequencing.

Results

   Network pharmacology analysis showed that kaempferol, wogonin,
   naringenin, and nobiletin were compounds associated with the efficacy
   of TXYF. TXYF may alleviate diarrhea via multiple BPs and pathways,
   including TNF signaling pathway, IL-17 signaling pathway, and Toll-like
   receptor signaling pathway, which are involved in TCM-gut
   microbiota-host interactions. Then, we found that TXYF administration
   reshaped the diversity and composition of the intestinal mucosal
   microbial community of diarrhea mice. Lactobacillus, primarily
   Lactobacillus johnsonii, was enriched by the administration of TXYF.
   After TXYF administration, the abundance of Lactobacillus, particularly
   Lactobacillus johnsonii, was enriched.

Conclusion

   Oral administration of TXYF may alleviate diarrhea through remodeling
   intestinal mucosal microbiota. Promoting the colonization of beneficial
   commensal bacteria in the intestinal mucosa through gut microbiota-host
   interactions may be a potential mechanism of TXYF in the treatment of
   diarrhea.

   Keywords: Chinese medicine, intestinal mucosal microbiota, diarrhea,
   Tongxieyaofang (TXYF), full-length 16S rRNA gene sequencing

1. Introduction

   Diarrhea is one of the most common symptoms in digestive. The incidence
   of diarrhea ranges from 2.2% to 73.8%, with variability depending on
   the definition applied ([27]Dionne and Mbuagbaw, 2023). Diarrhea
   remains a common issue in clinical practice, and finding effective
   administration strategies is still necessary. Tongxieyaofang (TXYF), a
   traditional Chinese medicine (TCM) formula, consists of Rhizoma
   Atractylodis Macrocephalae (BaiZhu, dried rootstocks of Atractylodes
   macrocephala Koidz.), Radix Paeoniae Alba (BaiShao, dried roots of
   Paeonia lactiflora Pall.), Pericarpium Citri Reticulatae (ChenPi, dried
   pericarps of Citrus reticulata Blanco and its Cultivars), and Radix
   Saposhnikoviae (FangFeng, dried roots of Saposhnikovia divaricata
   (Turcz.) Schischk.) in a ratio of 3:2:1.5:1. TXYF, initially recorded
   in “Danxi’s Experiential Therapy” during the 14th century, was a
   treatment with proven effects for diarrhea accompanied by abdominal
   pain resulting from liver stagnation and spleen deficiency. It was
   officially recommended as a fundamental prescription for patients
   afflicted with diarrhea associated with liver stagnation and spleen
   deficiency syndrome in the Expert Consensus on the TCM Diagnosis and
   Treatment for Diarrhea (2017). According to TCM theory, TXYF can soothe
   the liver and regulate Qi, strengthen the spleen, transform dampness,
   and stop diarrhea. Randomized controlled clinical studies have shown
   that TXYF, as a primary prescription, has significant efficacy in
   relieving loose stools, abdominal discomfort, and bloating in IBS-D
   ([28]Tang et al., 2019; [29]Liang et al., 2022), which provides
   clinical evidence for its effect on alleviating diarrhea and abdominal
   pain. Although the clinical efficacy is definite, the potential
   mechanism of TXYF in stopping diarrhea needs to be further clarified.

   Alterations in gut microbiota, also called dysbacteriosis, have been
   found in many gastrointestinal diseases with diarrhea as the main
   clinical manifestation ([30]Qiao et al., 2023; [31]Wu et al., 2023;
   [32]Zhou et al., 2023; [33]Luo et al., 2024). In previous work, we also
   observed significant differences in the intestinal cavity and
   intestinal mucosal microbiota between mice with diarrhea due to liver
   stagnation and spleen deficiency and control mice ([34]Yuan et al.,
   2020; [35]Zhang et al., 2020). Fecal microbiota transplantation of
   human fecal samples into germ-free mice can also construct human
   disease phenotype models ([36]Wang et al., 2022). Targeted manipulation
   of gut microbiota to treat diarrhea also has shown great potential and
   attracted wide attention ([37]Qiao et al., 2024; [38]Zhou et al.,
   2024). Under this background, researchers specializing in TCM uncovered
   that TCM can modulate the structure and metabolism of the gut
   microbiota to exert therapeutic effects ([39]Shao et al., 2020; [40]Zhu
   et al., 2023; [41]Wu et al., 2024). These discoveries provide a new
   scientific rationale for the application of TCM in modern society,
   particularly in gastrointestinal health and disease management. Delving
   into the effects of TCM on the gut microbiota can help reveal the
   modern scientific basis for its traditional efficacy and pave the way
   for the development of novel treatment approaches. Therefore, we
   designed to investigate if the effect of TXYF in alleviating diarrhea
   with liver stagnation and spleen deficiency syndrome is associated with
   gut microbiota regulation.

   Network pharmacology, a novel model characterized by ‘multi-component
   and network target’ approaches in TCM research, has been introduced as
   an encouraging method to study the mechanisms of herbal formulas
   ([42]Yao et al., 2022; [43]Zhao et al., 2023). In this study, we aimed
   to identify the potential targets of TXYF alleviating diarrhea through
   network pharmacology and assess the effects of TXYF on the intestinal
   mucosa microbiota in mice with diarrhea with liver stagnation and
   spleen deficiency syndrome using full-length 16S rRNA gene sequencing.

2. Materials and methods

2.1. Animals and ethics statement

   Fifteen one-month-old male specific-pathogen-free Kunming mice,
   weighing (20 ± 2) g, were purchased from the Hunan SJA Laboratory
   Animal Co., Ltd (License No: SCXK (Xiang) 2016-0002). Mice were housed
   in the barrier environment of the Center for Experimental Animals of
   the Hunan University of Chinese Medicine with (23-25) °C, (47-53) %
   humidity, 12h light/dark cycle, with free access to water and food
   during adaptive feeding. Animal experiments were performed following
   the National Guidelines for Experimental Animal Welfare, approved by
   the Animal Ethics and Welfare Committee of Hunan University of Chinese
   Medicine.

2.2. Preparation of experimental decoctions

2.2.1. Folium senna extract

   All traditional Chinese herbs were purchased from the pharmacy of the
   First Affiliated Hospital of Hunan University of Chinese Medicine. The
   authenticated voucher specimens were deposited at School of Chinese
   Medicine, Hunan University of Chinese Medicine. Folium senna (Fanxieye,
   dried leaf of Cassia angustifolia Vahl or Cassia acutifolia Delile)
   samples (500 g) were cut into small pieces and soaked in boiled water
   (5000 mL) for 10 min. Then, the filtrates were evaporated to 500 mL
   with a thermostat water bath at 75°C to get the Folium senna extract
   with a crude drug concentration of 1 g/mL, which was stored at 4°C for
   model preparation later.

2.2.2. TXYF decoction

   The TXYF herbal materials (45 g), consisting of Rhizoma Atractylodis
   Macrocephalae (18 g), Radix Paeoniae Alba (12 g), Pericarpium Citri
   Reticulatae (9 g), and Radix Saposhnikoviae (6 g), were soaked in cold
   water (400 mL) for 30 minutes before boiled for 30 minutes. The
   filtrates were collected, and the herb residues were decocted again in
   the same way. Two times’ filtrates were mixed and evaporated to 100 mL
   to get the TXYF decoction with a crude drug concentration of 0.24 g/mL,
   which was stored at 4°C for drug administration later.

2.3. Process of animal experiment

   After three days of adaptive feeding in a barrier environment, mice
   were randomly divided into a control group (gtcm), a placebo group
   (gtmm), and a TXYF group (gttm), with 5 mice in each group. The
   repeated stress-related diarrhea model was established referring to our
   previous study ([44]Yuan et al., 2020; [45]Zhang et al., 2020).
   Briefly, mice in the placebo and TXYF groups were gavage with Folium
   Senna extract at a dose of 17.5 mL/kg/d after fasting for 12 hours. At
   the period of Folium Senna administration, each mouse in the placebo
   and TXYF groups was put in a 50 mL centrifuge tube separately to
   restrict its activities for 1 hour/day, with its tail nipped at the
   distal 1/3 using a binder clip at the same time. Mice in the control
   group were given the same dose of sterile water and free access to
   water, food, and activities during the period of modeling. After seven
   days of continuous modeling, mice in the placebo and TXYF groups were
   successfully induced symptoms of diarrhea with liver stagnation and
   spleen deficiency syndrome. Then, mice in the TXYF group were gavage
   with 0.35 mL TXYF decoction twice a day for three days. The control and
   placebo groups were given the same dose of sterile water at the same
   time.

   At the end of the TXYF treatment period, mice were sacrificed by
   cervical dislocation. The small intestinal mucosa samples of mice were
   collected referring to our previous study ([46]Shao et al., 2020;
   [47]Zhang et al., 2020). Briefly, the abdominal cavity of the mouse was
   opened at an ultra-clean workbench to take out the small intestine. To
   sample the mucosa conveniently, the intestinal wall was cut along the
   long axis. A sterile cover glass was used to scrape the intestinal
   mucosa tissue. All samples were collected and marked separately and
   stored at -80°C for further PacBio SMRT sequencing.

2.4. Network pharmacology analysis

   All ingredients in the four Chinese herbal medicines of TXYF were
   retrieved from the traditional Chinese medicine systems pharmacology
   (TCMSP) database ([48]Ru et al., 2014)
   ([49]https://tcmspw.com/tcmsp.php). Drug-likeness (DL) and oral
   bioavailability (OB) were two crucial properties of absorption,
   distribution, metabolism, and excretion (ADME)-related parameters of
   ingredients. To obtain the ideal potential bioactive compounds, we set
   the screening criteria of DL greater than or equal to 0.18 and OB
   greater than or equal to 30%. Protein targets related to every active
   ingredient of TXYF were obtained using TCMSP databases and converted to
   target names in Homo sapiens using UniProtKB ([50]UniProt, 2021)
   ([51]https://www.uniprot.org/).

   Human genes associated with diarrhea were identified from the GeneCards
   database ([52]Safran et al., 2010) ([53]https://www.genecards.org) with
   the keyword “diarrhea”. Then, the potential targets of TXYF in treating
   diarrhea were obtained by a comprehensive analysis of the direct
   related targets of TXYF and the diarrhea targets using the online tool
   Venny2.1 ([54]https://bioinfogp.cnb.csic.es/tools/venny/index.html).

   Potential targets for TXYF treat diarrhea were analyzed using the
   online database STRING11.0 ([55]https://string-db.org/) to construct a
   functional protein-protein interaction (PPI) network of multiple
   proteins ([56]Szklarczyk et al., 2019). To obtain appropriate
   confidence in identified protein interactions in the PPI network, we
   set the screening criteria of interaction scores greater than 0.9. To
   evaluate the role of each target in the PPI network, hub targets were
   identified using four different analysis methods (Degree, Closeness,
   MCC, and EcCentricity) in “cytoHubba” of cytoscape3.8.0 software
   ([57]Shannon et al., 2003). Then, active compounds related to hub
   targets were screened and visualized with a Sankey diagram to introduce
   the relationship between hub targets and corresponding bioactive
   ingredients.

   The “clusterProfiler” package in R software (v3.6.3) was employed to
   conduct GO and KEGG enrichment analysis.

2.5. Full-length 16S rRNA gene sequencing on the PacBio SMRT platform

   Total bacterial genomic DNA was extracted from intestinal mucosa
   samples using the E.Z.N.A.^® Bacterial DNA Kit (OMEGA, USA). The
   quality and quantity of the extracted genomic DNA were checked using a
   0.8% agarose gel electrophoresis and an ultraviolet spectrophotometer.

   For amplification of full-length 16S rRNA gene, the following primers
   were used: forward primer 27F (5’-AGAGTTTGATCMTGGMCTCAG-3’) and reverse
   primer 1492R (5’-ACCTTGTTACGACTT-3’). Polymerase chain reaction (PCR)
   amplification of 16S rRNA genes was conducted using the Q5^®
   High-Fidelity DNA Polymerase (New England BioLabs, USA). Amplified DNA
   products were checked using a 2% agarose gel electrophoresis and
   purified using Axygen^® AxyPrep DNA Gel Extraction Kit. The quantity of
   amplified DNA products was determined using Quant-iT PicoGreen dsDNA
   Assay Kit. Mock community PacBio library was constructed using PacBio
   SMRTbell Template Prep Kit 1.0. The prepared DNA library was sequenced
   on the PacBio Sequel platform by Shanghai Personal Biotechnology Co.,
   Ltd.

2.6. Bioinformatics analysis

   Quality control of raw sequencing reads was processed to generate
   high-quality sequences using quantitative insights into microbial
   ecology (QIIME, v1.8.0) software. The high-quality sequences were
   clustered into operational taxonomic units (OTUs) at a more than 97%
   similarity threshold by UCLUST. OTU taxonomic classification was
   carried out by BLAST searching the representative sequence of each OTU
   set against the Greengenes database. Bioinformatics analysis was
   performed based on a filtered OTUs table, in which OTUs containing less
   than 0.001% of total sequences across all samples were discarded. Alpha
   and beta diversity were introduced to distinguish the diversity within
   samples and between samples. For alpha diversity analysis, observed
   species, Chao1, and Shannon indexes were calculated in the QIIME
   program and visualized by GraphPad Prism 8. For beta diversity
   analysis, nonmetric multidimensional scaling (NMDS) analysis was
   performed based on unweighted and weighted UniFrac distances using R.
   The analysis of similarities (ANOSIM) based on unweighted and weighted
   UniFrac distances was carried out to compare different groups by QIIME.
   Linear discriminant analysis (LDA) effect size (LEfSe) ([58]Segata
   et al., 2011) was conducted to define microbial biomarkers of the three
   groups, which provided biological class explanations to establish
   statistical significance, biological consistency, and effect size
   estimation of predicted biomarkers. The threshold on the logarithmic
   LDA score for distinguishing features was set to 4.0. Predictive
   functional profiling of microbial communities was conducted using 16S
   rRNA data by Phylogenetic Investigation of Communities by
   Reconstruction of Unobserved States (PICRUSt) ([59]Langille et al.,
   2013) in the context of the Kyoto encyclopedia of genes and genomes
   (KEGG). The raw data will be made available by the authors, without
   undue reservation, to any qualified researcher.

2.7. Statistical analysis

   The data were expressed as mean ± standard deviation (SD) where
   applicable. SPSS 22.0 and GraphPad Prism 8.0 were used for statistical
   analysis. Because the data does not conform to the normal distribution,
   the Kruskal-Wallis H test and Mann-Whitney U test were used to evaluate
   the statistical significance of the differences between the three
   groups and the two groups respectively. Differences were considered
   statistically significant when p < 0.05. The threshold of p-value was
   determined with the original false discovery rate (FDR) method of
   Benjamini and Hochberg (BH) in multiple comparisons.

3. Results

3.1. Network pharmacology analysis

3.1.1. Potential targets for TXYF treat diarrhea

   After filtered by the ADME threshold (OB ≥ 30% and DL ≥ 0.18) and
   removed the duplicates, a total of 32 bioactive ingredients in TXYF
   were screened by the TCMSP database ([60] Supplementary Table 1 ).
   Based on identified bioactive compounds in TXYF, a total of 145
   potential target genes responding to compounds of TXYF were retrieved
   using the TCMSP data server. An herb-compound-target (CT) network was
   constructed by Cytoscape 3.8.0 software, which included 181 nodes (four
   herb nodes, 32 bioactive compound nodes, and 145 compound-associated
   target nodes) and 431 compound-target corresponding connections ([61]
   Figure 1 ). According to the degree of the nodes in the CT network,
   kaempferol (degree = 62), wogonin (degree = 46), beta-sitosterol
   (degree = 39), naringenin (degree = 38), nobiletin (degree = 35),
   5-O-Methylvisamminol (degree = 25) were recognized as main compounds of
   TXYF.

Figure 1.

   [62]Figure 1
   [63]Open in a new tab

   The herb-compound-target (CT) network. The orange nodes stand for herbs
   in TXYF. The red nodes stand for bioactive compounds in TXYF. The blue
   nodes stand for targets responding to the efficacy of TXYF. The grey
   lines stand for corresponding connections between different category
   nodes. The size of each node indicates the number of connections. The
   larger the node, the more the connections.

   Targets related to diarrhea were obtained from the GeneCards database.
   Finally, 2 698 diarrhea-related targets (relevance score > median) were
   incorporated into the study. There were 94 intersection targets between
   targets responding to compounds of TXYF and targets related to
   diarrhea, which were recognized as potential targets for TXYF treat
   diarrhea.

3.1.2. GO and KEGG enrichment analysis

   The results showed that with a post-adjusted p-value threshold of 0.05,
   94 targets had enriched 1935 biological processes (BPs), 48 cellular
   components (CCs), and 131 molecular functions (MFs). The top 10
   enriched GO terms from each category were directly shown in
   [64]Figure 2A . The results showed that targets were enriched in BPs
   related to antibacterial and inflammation, including the response to
   lipopolysaccharide, response to molecule of bacterial origin, and
   response to oxidative stress. The CCs and MFs results indicated that
   most targets were localized in the membrane and related to steroid
   hormone receptor activity, nuclear receptor activity, ligand-activated
   transcription factor activity, and phosphatase binding. Meanwhile, the
   top 20 enriched KEGG pathways were visualized in [65]Figure 2B . The
   results showed that majority targets are associated with TNF signaling
   pathway, IL-17 signaling pathway, and Toll-like receptor signaling
   pathway. These findings suggest the complex pathological process of
   diarrhea and the therapeutic potential of TXYF as a treatment for
   diarrhea.

Figure 2.

   [66]Figure 2
   [67]Open in a new tab

   Enrichment analysis. (A) Bar plot for top 10 enriched GO terms of each
   sub-class. The length of each bar stands for enriched counts, and the
   color of each bar stands for adjusted p-value. The abscissa represents
   the enriched gene counts. (B) Dot plot for top 20 enriched KEGG pathway
   terms. The size of each node stands for enriched counts, and the color
   of each node stands for adjusted p-value. The abscissa represents the
   enriched gene ratio.

3.1.3. PPI network analysis and hub targets screening

   To investigate the relationships among the 94 intersectant targets, we
   constructed a PPI network using the STRING database and visualized it
   with Cytoscape software. After removing disconnected nodes, a PPI
   network including 82 nodes and 282 interactions were constructed with
   the highest confidence score of 0.9 as the threshold. Hub targets were
   identified using four different algorithms, Betweenness, Degree,
   Maximal clique centrality (MCC), and Closeness, in the cytoHubba app
   ([68] Figure 3A ). Then, core hub targets of the PPI network were
   obtained by intersecting hub targets (top 10) identified from four
   algorithms ([69] Figure 3B ). Therefore, five targets (AKT1, MAPK14,
   MAPK3, TP53, and MAPK1) were identified as the core hub targets of the
   PPI network. Four bioactive compounds (kaempferol, naringin, wogonin,
   nobiletin) linked to five core hub targets were then selected in
   reverse. They were also the main compounds in TXYF. The relationships
   between core hub targets and corresponding bioactive compounds were
   shown directly using a Sankey diagram ([70] Figure 3C ).

Figure 3.

   [71]Figure 3
   [72]Open in a new tab

   The PPI network analysis (A), the Venn diagram of core hub targets (B),
   and the Sankey diagram of core hub targets and corresponding compounds
   (C). In (A), (1) to (4) networks represent the analysis results based
   on four algorithms, Betweenness, Degree, MCC, and Closeness,
   respectively. The highlight red-to-yellow nodes are hub targets (top
   10). The blue nodes are also protein targets with the highest
   confidence (0.9) from the PPI network analysis. In (B), hub targets
   (top 10) identified from four algorithms were intersected to obtain the
   core hub targets for further analysis. In (C), the relationships
   between core hub targets obtained from the Venn diagram and
   corresponding bioactive compounds were shown directly. The up blocks
   stand for the core hub targets, and the underneath blocks stand for
   compounds, as follows: kaempferol, naringenin, wogonin, and nobiletin.

3.2. TXYF administration altered the community diversity in intestinal
mucosal microbiota in diarrhea mice

   By high-throughput sequencing, 50005, 44322, and 42813 filtered
   high-quality sequences were obtained from samples of the control,
   placebo, and TXYF groups respectively, with the length concentrated at
   1500 bp. A Venn diagram of OTU numbers showed that the gut microbiota
   community was changed with the treatment of TXYF ([73] Figure 4A ). The
   shared OTUs among the three groups were 188 (11.39%). The rarefaction
   curve demonstrated that the current sequencing depth was adequate to
   detect the microbial diversity in these community samples ([74]
   Figure 4B ).

Figure 4.

   [75]Figure 4
   [76]Open in a new tab

   (A) Venn diagram of OTUs (at distance 0.03). (B) Rarefaction curve
   based on observed species. (C) Box plots of α-diversity index of OTUs
   number (1), Chao1 (2), and Simpson (3). (D) NMDS of OTU-based
   unweighted (1) and weighted (2) UniFrac distances. gtcm, the control
   group; gtmm, the placebo group; gttm, the TXYF group.

   Diversity analysis was performed at two levels – within the sample and
   across samples. Alpha diversity indicators, such as observed species,
   Chao1 and Shannon indexes, were applied to reflect richness and
   evenness of community within a single sample ([77] Figure 4C ). As the
   results suggested, the placebo group has the highest community richness
   among the three groups. Compared with the placebo group, community
   richness and diversity decreased in the TXYF group and reached the same
   level as the control group. However, there was no statistical
   differences among the three groups (p > 0.05). Nonmetric
   multidimensional scaling (NMDS) of OTU-based weighted and unweighted
   UniFrac distance was conducted to present the similarities and
   differences in community composition across different samples and
   groups ([78] Figure 4D ). NMDS highlighted a significant segregation of
   the gut microbiota according to the unweighted UniFrac distance
   (ANOSIM, R = 0.1724, p = 0.040; Adonis, R2 = 0.18382, p = 0.033),
   suggesting that TXYF administration was a potential factor in
   alterating community composition.

3.3. TXYF administration changed the community composition in intestinal
mucosal microbiota in diarrhea mice

   To investigate whether there were representative taxa with differences,
   which may contribute to differences in diversity, we analyzed the
   community structure and composition of the three groups at the
   taxonomic level from phylum to species. OTUs were annotated to 17
   phyla, 35 classes, 61 orders, 116 families, 199 genus, and 375 species.
   The proportion of unclassified OTUs was 4.06% (67/1650), suggesting
   that the full-length 16S sequencing technology has a high resolution in
   cluster analysis. In [79]Figure 5A , the numbers of taxa at class and
   family levels in the TXYF group were significantly less than the
   control group (p < 0.05). In detail, compared with the placebo group,
   class Acidimicrobiia and 13 family taxa were newly colonized in the
   TXYF group, while seven class taxa and 21 family taxa were missing from
   the TXYF group.

Figure 5.

   [80]Figure 5
   [81]Open in a new tab

   (A) Histogram of the number of microbial species of each sample at
   different taxonomic levels. (B) Radar graph of the relative abundance
   of dominant phylum in the intestinal mucosa. (C) Box plots of relative
   abundance of Bacteroidetes (1) and F/B ratio (2). (D, E) Stack diagram
   of relative abundance of genus (D) and species (E) in the intestinal
   mucosa. *p < 0.05, **p < 0.01, Kruskal-Wallis H test. gtcm, the control
   group; gtmm, the placebo group; gttm, the TXYF group.

   Firmicutes (52.09%), Proteobacteria (38.21%), Bacteroidetes (6.41%),
   and Actinobacteria (2.69%), accounting for more than 98%, were the
   dominant communities in the intestinal mucosa of mice in the TXYF group
   ([82] Figure 5B ). Moreover, compared with the control group, the
   relative abundances of Bacteroidetes in both placebo and TXYF groups
   were significantly decreased, which resulted in a significant increase
   in the Firmicutes to Bacteroidetes (F/B) ratios in both groups (p <
   0.05) ([83] Figure 5C ).

   At the genus rank, dominant populations varied among groups ([84]
   Figure 5D ). Specifically, Staphylococcus (29.64%), Pelomonas (12.91%),
   Ralstonia (10.69%), Candidatus Arthromitus (10.41%), Lactobacillus
   (8.57%), and Vibrionimonas (5.18%), whose relative abundance was
   greater than 5%, were the six enriched genera in the TXYF group. They
   accounted for 77.4% of the total in the TXYF group. Only four genera
   had relative abundance greater than 5% in the placebo group, namely
   Pelomonas(22.59%), Staphylococcus(19.05%), Candidatus
   Arthromitus(18.21%), and Ralstonia(13.88%). They accounted for 73.73%
   of the total in the placebo group. The relative abundances of
   Pelomonas(27.39%), Ralstonia(17.74%), Candidatus Arthromitus(13.91%),
   Vibrionimonas(7.06%), Ochrobactrum(5.65%), Barnesiella(5.41%), and
   Bacteroides(5.28%)were greater than 5%, accounting for 82.44% of the
   total in the control group.

   At the species level, enriched species with a relative abundance
   greater than 5% in the TXYF group were Pelomonas saccharophila (12.91%
   vs 22.44% in gtmm), Staphylococcus lentus (12% vs 10.98% in gtmm),
   Staphylococcus sciuri (10.4% vs 1.31% in gtmm), Ralstonia insidiosa
   (10.4% vs 13.59% in gtmm), Candidatus Arthromitus sp. SFB-mouse (10.27%
   vs 18.04% in gtmm), Lactobacillus johnsonii (7.3% vs 0.57% in gtmm),
   Staphylococcus saprophyticus (5.85% vs 4.84% in gtmm), Vibrionimonas
   magnilacihabitans (5.18% vs 4.41% in gtmm) ([85] Figure 5E ).

3.4. Microbial biomarkers associated with the administration of TXYF

   To further identify the biomarkers associated with TXYF administration,
   we used LEfSe analysis to analyze microbial abundances at taxonomic
   levels from phylum to species. A total of 23 specific microbial
   biomarkers were identified across the three groups with linear
   discriminant analysis (LDA) score greater than 4 and p-value less than
   0.05 ([86] Figure 6A ). Taxa including Gammaproteobacteria, Bacillales,
   Lactobacillales, Corynebacteriales, Staphylococcaceae, Aerococcaceae,
   Staphylococcus, Aerococcus, and Aerococcus_viridans were significantly
   enriched in the TXYF group, which could be recognized as biomarkers
   associated with the treatment of TXYF.

Figure 6.

   [87]Figure 6
   [88]Open in a new tab

   The potential biomarkers associated with the administration of TXYF.
   (A) Identified taxonomic features of significant differences across
   three groups by LEfSe (LDA score ≥ 4 and p-value ≤ 0.05). (B)
   Identified discriminative species between the TXYF (gttm) and placebo
   (gtmm) groups by Metastats analysis.

   In addition, the Metastats analysis based on the number of sequences
   (i.e., absolute abundance) was performed to compare the differences in
   composition at species level between the TXYF and placebo groups. As
   shown in [89]Figure 6B , there were significant increases in six
   species, such as Staphylococcus sciuri, Proteus vulgaris, Proteus
   mirabilis, Lactobacillus intestinalis, Stenotrophomonas maltophilia,
   and Lutispora thermophila, and decreases in seven species, including
   Porphyromonas gingivalis, Barnesiella intestinihominis, Ochrobactrum
   lupini, Methylobacterium jeotgali, Anaeroplasma abactoclasticum,
   Marvinbryantia formatexigens, and Clostridium aldenense, in the TXYF
   group compared with the placebo group.

3.5. TXYF administration regulated the microbial function

   To predict the potential functions of microbial communities and assess
   the differences in microbial functions across groups, the BIOM file of
   OTU abundances was matched with the reference genomes of the KEGG
   database. The 16S sequencing dataset of 15 samples resulted in a
   functional dataset containing 5734 KOs. Further statistical analysis
   showed that 48 KOs were significantly altered between the TXYF and
   placebo groups ([90] Figure 7A ). Administration of TXYF significantly
   increased relative abundances of KOs in pathways such as two-component
   system, quorum sensing, biosynthesis, and metabolism of cofactor,
   butanoate, and tryptophan ([91] Figure 7B ). But in the meantime, TXYF
   reduced the relative abundance of KOs in pathways such as bacterial
   secretion system, secondary bile acid biosynthesis, and citrate cycle
   ([92] Figure 7C ).

Figure 7.

   [93]Figure 7
   [94]Open in a new tab

   Function prediction based on the KEGG database by PICRUSt. (A) Heatmap
   for the relative abundance of discriminative KOs between the TXYF
   (gttm) and placebo (gtmm) groups. (B) The main KEGG pathway of the KOs
   with significantly increased expression in the TXYF group. (C) The main
   KEGG pathway of the KOs with significantly decreased expression in the
   TXYF group.

4. Discussion

   According to the theory of TCM, TXYF was widely used to treat diarrhea
   due to liver stagnation and spleen deficiency syndrome in clinical
   practices. A meta-analysis of randomized trials also indicated that
   TXYF as a treatment for IBS-D was more effective and safer in improving
   the clinical effectiveness rate and decreasing the adverse events than
   conventional medication such as probiotics, pinaverium bromide,
   trimebutine, and Oryzanol ([95]Zhou et al., 2019b). The clinical
   efficacy of TXYF is positive, but the mechanism of TXYF relieving
   diarrhea is indefinite. Folium senna combined with restraint stress is
   an effective method of establishing the mice model of liver stagnation
   and spleen deficiency syndrome ([96]Yuan et al., 2020; [97]Zhang
   et al., 2020). In this study, we used network pharmacology to identify
   potential active compounds and targets for TXYF to alleviate diarrhea
   and focused on the effects of TXYF on the intestinal mucosal microbiota
   of diarrhea model mice using PacBio SMRT sequencing technology.

   Network pharmacology was an efficient approach to elucidate the
   intrinsic multi-component, multi-target, and multi-pathway features of
   Chinese medicine prescriptions. In this study, we obtained 32
   compounds, including kaempferol, wogonin, naringenin, nobiletin,
   paeoniflorin, and Atractylenolide IIIt as active fractions of TXYF. The
   intestinal epithelium barrier consists of a single layer of enterocytes
   and tight junctions (TJs). A leaky barrier may contribute to
   hyper-intestinal permeability, which will lead to intestinal
   dysfunction and systemic diseases ([98]Camilleri et al., 2017).
   Malfunction of TJ can also trigger the development of infection,
   inflammation, obesity, and even autoimmune diseases ([99]Wen et al.,
   2020). Chen et al ([100]Chen et al., 2020). reported that kaempferol
   supplementation up-regulated the expression of tight junction proteins,
   butyrate receptors, and transporters in the intestinal mucosa in acute
   alcoholic liver injury mice. The beneficial action of kaempferol on
   LPS-induced inflammation and barrier dysfunction in a coculture model
   of intestinal epithelial cells and intestinal microvascular endothelial
   cells was associated with inhibiting the NF-κB signaling pathway
   activation ([101]Bian et al., 2020). In addition, studies suggested
   that supplementations of TCM bioactive compounds such as wogonin,
   naringenin, nobiletin, and paeoniflorin could suppress inflammatory
   response and maintain the intestinal barrier function ([102]Azuma
   et al., 2013; [103]Wang et al., 2015; [104]Wu et al., 2019; [105]Wen
   et al., 2020). Atractylodes III was also considered to have an
   inhibitory effect on mRNA and protein expressions of pro-inflammatory
   (IL-1β, TNF-α, IL-6) and anti-inflammatory cytokines ([106]Zhou et al.,
   2019a).

   Furthermore, the anti-inflammatory effect of TCM bioactive compounds
   was related to gut microbiota and its metabolites. It has been reported
   that the anti-arthritis activities of kaempferol were linked to
   functions of reversion the perturbation of metabolites involved in
   energy production and the tryptophan, fatty acid, and secondary bile
   acid metabolisms in the gut contents ([107]Aa et al., 2020).
   Paeoniflorin has active impacts in decreasing the serum contents of
   corticosterone, corticotropin-releasing hormone, and
   adrenocorticotropic hormone and increasing serotonin and
   5-Hydroxyindoleacetic acid in the prefrontal cortex and hippocampus in
   post-traumatic stress disorder mice ([108]Qiu et al., 2018). Moreover,
   the anti-inflammatory and immunomodulatory actions of paeoniflorin were
   microbiota-dependent and related to indole-3-lactate and epithelial
   autophagy ([109]Fan et al., 2020). Therefore, we speculate that the
   bioactive compounds of TXYF may interact with gut microbiota after oral
   administration and play roles in alleviating diarrhea by maintaining
   the intestinal epithelial barrier and regulating intestinal microbes
   and their metabolites.

   GO and KEGG pathway enrichment analysis further verified the above
   inference. The effect of TXYF in alleviating diarrhea may be related to
   multiple BPs and pathways, such as the inflammatory responses and the
   immune systems, which are mediated by the TNF signaling pathway, the
   IL-17 signaling pathway, and the Toll-like receptor (TLR) signaling
   pathway. Specifically, TNF signaling has an active impact on
   maintaining epithelial cell tight junctions and intestinal barriers. It
   was stated that loss of intestinal barrier function mediated by TNF
   signaling was highly relevant to the inflammatory pathophysiology
   observed in Crohn’s disease and celiac disease ([110]Kolodziej et al.,
   2011). TLR signaling pathway was proven to play an essential role in
   maintaining tolerance to commensal microbiota and regulating specific
   immune responses against pathogens, which is vital for the homeostasis
   of the intestine and activation of innate immunity ([111]Takeda and
   Akira, 2004). For example, Yi et al ([112]Yi et al., 2021). found that
   dietary anethole supplementation can alleviate intestinal barrier
   disruption and intestinal inflammation induced by enterotoxigenic
   Escherichia coli through modification of TLR signaling and intestinal
   microbiota. However, further studies are needed to confirm the exact
   role of these pathways in the process of TXYF relieving diarrhea.

   Clear evidence from full-length 16S rRNA gene sequencing indicates that
   TXYF has a direct effect on the intestinal mucosal microbiota of
   diarrhea model mice. From a macro perspective, although no significant
   difference was observed in bacterial richness and evenness across
   groups according to alpha diversity analysis, beta diversity analysis
   based on unweighted UniFrac distance suggested a statistical difference
   in microbial composition among the three groups. We also found that the
   Chao1 and Simpson indices of the TXYF group were in the middle level of
   the three groups. These findings indicate that Folium senna combined
   with restraint stress might perturb the biodiversity in the intestinal
   mucosa by increasing bacterial richness, which is consistent with early
   studies ([113]Zhang et al., 2020, [114]2021). TXYF might be a driving
   force in regulating the intestinal mucosal microbiota of diarrhea mice
   to a healthy state.

   Then, we further concentrated on the changes in specific taxa related
   to the administration of TXYF. We recognized some positive taxa that
   were associated with the protective effects of TXYF. In this study, the
   relative abundance of Lactobacillus increased significantly in the TXYF
   group. Lactobacillus, a commensal inhabitant of animal and human
   gastrointestinal tracts, is considered friendly bacteria with a wide
   range of applications from probiotics to therapeutic bacteria in both
   humans and animals ([115]O’Flaherty et al., 2020). In our study,
   Lactobacillus johnsonii and Lactobacillus acidophilus, which are
   important members of the Lactobacillus acidophilus complex (LAC)
   ([116]Ramachandran et al., 2013), were enriched with the administration
   of TXYF. L. johnsonii completely lacked genes encoding biosynthetic
   pathways to produce essential nutrients, such as amino acids, purine
   nucleotides, and cofactors. Thus, it has high nutritional requirements
   for the intestinal environment, and its development and function depend
   on the host or other intestinal microorganisms to maintain
   ([117]Pridmore et al., 2004). Co-occurrence network analysis showed
   that L. johnsonii was positively correlated with L. intestinalis,
   suggesting a mutualism relationship between L. johnsonii and L.
   intestinalis. It has been established that L. johnsonii L531 could
   alleviate Salmonella infantis-associated enteritis and promote
   intestinal secretory IgA production via regulation of NLRC4 and NLRP3
   inflammasomes, NF-κB signaling pathway activation and suppression of
   mitochondrial damage ([118]Xia et al., 2020). Oral pre-administration
   of L. acidophilus played a protective role in dextran sodium
   sulfate-induced colitis by targeting the intestinal TJ barrier that
   involves the TLR-2 receptor complex ([119]Al-Sadi et al., 2021). Based
   on the effect of inducing the expression of µ-opioid and cannabinoid
   receptors in intestinal epithelial cells and mediating analgesic
   functions in the gut, L. acidophilus was recognized as a potential
   therapeutic approach for abdominal pain and irritable bowel syndrome
   ([120]Rousseaux et al., 2007). However, further studies are warranted
   to investigate the exact beneficial mechanism of L. johnsonii and L.
   acidophilus in diarrhea mice with liver stagnation and spleen
   deficiency syndrome.

5. Conclusion

   In this study, network pharmacology and full-length 16S rRNA gene
   sequencing were used to investigate the active compounds and targets of
   TXYF in alleviating diarrhea and the effect of TXYF on the intestinal
   mucosal microbiota of diarrhea model mice. Kaempferol, wogonin,
   naringenin, and nobiletin are compounds associated with the efficacy of
   TXYF in relieving diarrhea. TXYF may alleviate diarrhea via multiple
   BPs and pathways, including TNF signaling pathway, IL-17 signaling
   pathway, and Toll-like receptor signaling pathway, which are involved
   in TCM-gut microbiota-host interactions. TXYF was directly verified to
   remodel the diversity and composition of the intestinal mucosal
   microbiota in mice with diarrhea associated with live stagnation and
   spleen deficiency syndrome. Lactobacillus, primarily Lactobacillus
   johnsonii, was enriched by the administration of TXYF and were
   identified as being involved in the therapeutic role of TXYF in
   alleviating diarrhea. In conclusion, oral administration of TXYF may
   alleviate diarrhea through remodeling intestinal mucosal microbiota.
   Promoting the colonization of beneficial commensal bacteria in the
   intestinal mucosa through gut microbiota-host interactions may be a
   potential mechanism of TXYF in the treatment of diarrhea.

Funding Statement

   The author(s) declare financial support was received for the research,
   authorship, and/or publication of this article. This work was supported
   by the Scientific Research Foundation of Hunan Provincial Education
   Department (No. 23B1054) and the Start-up Fund for Doctoral Research of
   Hunan University of Medicine (No. 202304).

Data availability statement

   The datasets presented in this study can be found in online
   repositories. The names of the repository/repositories and accession
   number(s) can be found below: [121]https://www.ncbi.nlm.nih.gov/,
   PRJNA861669.

Ethics statement

   The animal study was approved by Animal Ethics and Welfare Committee of
   Hunan University of Chinese Medicine. The study was conducted in
   accordance with the local legislation and institutional requirements.

Author contributions

   HS: Funding acquisition, Conceptualization, Investigation, Formal
   analysis, Writing – original draft. LW: Investigation, Visualization,
   Writing – review & editing. HZ: Supervision, Writing – review &
   editing.

Conflict of interest

   Author HZ was employed by Kangpu Pharmaceutical Co., Ltd.

   The remaining authors declare that the research was conducted in the
   absence of any commercial or financial relationships that could be
   construed as a potential conflict of interest.

Generative AI statement

   The authors declare that no Generative AI was used in the creation of
   this manuscript.

Publisher’s note

   All claims expressed in this article are solely those of the authors
   and do not necessarily represent those of their affiliated
   organizations, or those of the publisher, the editors and the
   reviewers. Any product that may be evaluated in this article, or claim
   that may be made by its manufacturer, is not guaranteed or endorsed by
   the publisher.

Supplementary material

   The Supplementary Material for this article can be found online at:
   [122]https://www.frontiersin.org/articles/10.3389/fcimb.2024.1502373/fu
   ll#supplementary-material
   Supplementary Table 1

   The information of obtained bioactive compounds of TXYF.
   [123]Table1.xlsx^ (10.6KB, xlsx)

References