Kaempferol Improves CCl4-Mediated Mouse LF

For investigating therapeutic effect of kaempferol on LF, CCl4 was intraperitoneally injected for inducing the LF mouse model, followed by treatment with colchicine and kaempferol. Colchicine, which is often used in clinical practice for its ability to alleviate hepatic fibrosis [14, 17, 33]. In this investigation, colchicine were used as a colchicine treated drug. Figure 1A shows the chemical structure of Kaempferol. From Fig. 1B, CCl4-induced mice had markedly improved weight loss compared to untreated control mice, whereas kaempferol and colchicine treated mice had notably improved weight loss. In addition, as can be seen in Fig. 1C-D, mouse liver of CCl4 group were rough and yellowish with obvious fiber deposition and significantly elevated liver coefficient. Nonetheless, following kaempferol (40 mg/kg) and colchicine administration, liver coefficient and liver deposition in the mice apparently decreased, while livers were reddish and close to normal. In addition, TBil, AST, and ALT levels within serum remarkably increased, whereas serum Alb and PT levels reduced after CCl4 injection. Collectively, serum markers (such as Alb, PT, AST, and ALT) were dose-dependently improved following kaempferol treatment in the mouse model induced by CCl4 (Fig. 1E-I).

Fig. 1

Kaempferol improves CCl4-induced mouse LF. A Kaempferol chemical structure. B Body weight alteration (%). C Typical images showing liver tissues. D Liver coefficient (%). E ALT, F AST, G TBil, H Alb, I PT, J HA, K PC III, and (L) LN levels in serum. M H&E and IHC staining to examine Collagen I and Collagen II. M H&E and IHC staining to examine Collagen I and α-SMA levels within liver sections (200×). Relative to NC, #P < 0.05; ##P < 0.01; ###P < 0.001; relative to CCl4, *P < 0.05, **P < 0.01, and ***P < 0.001

Serum HA, PC III, and LN levels have been frequently used for assessing the LF degree. In the case of LF, HA, PC III, and LN expression increases dramatically [34]. This work determined serum HA, PC III, and LN contents of model group following CCl4 treatment, and discovered that their levels were significantly increased (Fig. 1J-L). However, their levels dose-dependently decreased after treatment with kaempferol. In addition, histopathologic analysis of the liver showed that the control liver tissue was intact with uniform cytoplasmic distribution, and CCl4-mediated mice had inflammatory infiltrates and lipid vacuoles. According to immunohistochemistry, α-SMA and collagen I levels within liver tissue also increased significantly after CCl4 induction. However, kaempferol treatment ameliorated lipid vacuolization and hepatic tissue inflammation, decreased collagen and fibrous deposition, and down-regulated α-SMA and Collagen I expression (Fig. 1M). Consequently, kaempferol exerts the obvious efficacy in mice with CCl4-induced LF.

Kaemferol Regulated Mouse Intestinal Dysbacteriosis Induced by CCl4

For determining how kaempferol modulated intestinal flora, this study conducted an alpha-diversity analysis. As revealed by species abundance indices, Chao1, observed_species, shannon and Simpson indexes in kaempferol group remarkably increased relative to CCl4 group (p < 0.05, Table 1), indicating that the community richness and diversity were higher in kaempferol group. Secondly, we conducted beta-diversity analysis for exploring the similar and different community compositions among diverse taxa. PCoA analysis showed (Fig. 2A) that the NC group was farther away from the CCl4 group, and the difference in species composition was larger, suggesting significant changes in mouse fecal flora in CCl4-induced LF state. There was a certain degree of deviation from the CCl4 group and convergence towards normal group after administration, suggesting that kaempferol and colchicine enhanced fecal floral diversity in LF mice partially. Additionally, taxonomic compositions in every group at genus and phylum levels were examined. As revealed by bubble diagrams, every sample included Sumerlaeota, Bdellovibrionota, Verrucomicrobiota, Bacteroidota, Acidobacteriota, Cyanobacteria (Fig. 2B); besides, genus mostly consisted of Alloprevotella, Ruminococcus, Staphylococcus, Paramuribaculum, Muribaculum, Akkermansia, HT002, Enterorhabdus (Fig. 2C). Genus with highest levels in CCl4 group compared to the NC group was g__ Escherichia-Shigella, g__ GCA-900066575, g__ Oscillospiraceae_unclassified. Genus in the kaempferol group had ncreased abundance compared to the CCl4 group, including g__ Robinsonella, g__Prevotella, g__TM7x, g__Coriobacteriaceae_UCG-002. However, g__Caulobacter, g__Enterococcus, g__Oscillospiraceae_unclassified decreased in abundance. Thus, kaempferol elevated beneficial floral abundances (g__Robinsoniella, g__Prevotella), decreases harmful bacterial abundances, alters intestinal floral composition, and prevents tumorigenesis.

Table1 Alpha diversity analysis reflects species richness, diversity in each groupFig. 2

Kaempferol regulates intestinal microbial richness, diversity and intestinal floral community structure within CCl4-induced mice. A Chao1 index. B Observed_species indexes. C Shannon and (D) Simpson indexes. E Principal coordinate analysis (PcoA) analysis. Microbial community bar plots based on (F) phylum and (G) genus. H Venn plots of kaempferol vs CCl4 versus colchicine vs CCl4 after taking intersections of the differential flora

When kaempferol and colchicine were used, Lachnospiraceae_NK4A136_group, Enterorhabdus and Alistipes had markedly elevated abundances (p < 0.05) in model in comparison with control group. Nonetheless, Turicibacter, Clostridium_sensu_stricto_1 and Romboutsia apparently decreased (p < 0.05). The intestinal flora in which kaempferol reversed the imbalance was derived by taking the intersection of the differential flora of kaempferol vs CCl4 and colchicine vs CCl4, e.g. g__Robinsoniella, g__Prevotella_9, g__Erysipelotrichaceae_UCG- 003, g__Dorea, g__Eubacterium]_nodatum_group, g__Coriobacteriaceae_UCG-002 (Fig. 2D).

Kaemferol Attenuates Liver Fibrosis Depending on Intestinal Flora

For verifying that protection of kaempferol on CCl4-induced mice was regulated via intestinal flora, microbiota of kaempferol-treated CCl4-induced mice were transplanted into recipient CCl4-induced mice by daily gavage. CCl4 recipient mice receiving the intestinal flora of the kaempferol group of donor mice exhibited significantly attenuated CCl4-induced hepatic fibrosis-related symptoms compared to CCl4 recipient mice receiving the intestinal flora of the CCl4 donor mice. Compared with CCl4 receptor mice receiving intestinal flora from the CCl4 donor mice, the CCl4 receptor mice receiving intestinal flora from donor mice of the kaempferol treatment group also exhibited apparently alleviated CCl4-induced LF-related symptoms. Relative to CCl4-FMT mice, kaempferol-FMT mice increased the body weight of mice (Fig. 3A) and resulted in significantly lower liver deposition and liver coefficient, with reddish, near-normal livers (Fig. 3B). Compared with CCl4-FMT group, kaempferol-FMT group down-regulated HA, PC III and LN expression in serum (Fig. 3C-E). HE staining results suggested that lipid vacuoles and inflammatory infiltration appeared in CCl4-FMT group, however, kaempferol-FMT group restored liver tissue integrity and made the cytoplasmic distribution homogeneous. According to immunohistochemical analysis, Collagen I and α-SMA levels remarkably increased within liver tissue in CCl4-FMT group. However, kaempferol treatment ameliorated lipid vacuolization and liver tissue inflammation, decreased collagen/fiber deposition, and down-regulatedα-SMA and Collagen I (Fig. 3F). Consequently, fecal microbiota in kaempferol-exposed CCl4 donor mice had an ameliorative effect on cirrhosis similar to that of the kaempferol-exposed group. On the contrary, microbiota of CCl4-induced mice did not reverse cirrhosis symptoms in CCl4 recipients. To further investigate whether protection of kaempferol against CCl4-exposed mice was dependent on the intestinal microbiota, we administered antibodies to deplete the intestinal microbiota in CCl4-induced mice, and then examined liver fibrosis-related indexes using HE staining and immunohistochemical experiments. Consequently, intestinal flora mediated the protection of kaempferol within CCl4-induced mice. In short, intestinal flora transferred the amelioration induced by kaempferol administration.

Fig. 3

Kaempferol alleviated inflammation and liver fibrosis in CCl4-induced mice depending on intestinal flora. A Body weight. B Typical pictures for liver samples. C HA, D PC III, and (E) LN contents in serum. F H&E and IHC staining conducted to examine Collagen I and α-SMA. F H&E and IHC staining conducted to examine Collagen I and α-SMA levels within liver sections (200×). G Principal coordinate analysis (PcoA) analysis. H Microbial community bar plot drawn based on genus

PCoA analysis and Adonis analysis showed (Fig. 3G, Table 2) that the kaempferol-FMT group was more distant from the CCl4-FMT group, with greater differences in species composition. Taxonomic compositions at the genus level were examined. According to the histogram (Fig. 3H), the kaempferol-FMT group up-regulated g__Rikenellacea… compared to the CCl4-FMT group..ae_RC9_gut_group, g__Monoglobus, g__Alloprevotella, g__Bacteroides, g__Incertae_Sedis, g__Prevotellaceae_UCG-001, and downregulated g__ Eubacterium]….m]_siraeum_group, g__Robinsoniella, g__Candidatus_Arthromitus.

Table 2 Adonis analysis of all groupsKaemferol Altered the Metabolic Profile of CCl4-Induced Mice

Metabolites are important communication mediators between the intestinal microbiota of LF mice and the transcriptional levels and physiological status of the host liver monocyte population.PCA analysis showed (Fig. 4A) that the NC group was farther away from the CCl4 group and had a greater difference in the composition of metabolites, which suggests that the differential metabolites of the mice in the state of LF induced by CCl4 were significantly changed. There was a certain degree of deviation from the CCl4 group and convergence toward normal group after administration, suggesting that kaempferol and colchicine enhanced serum differential metabolite diversity in LF mice partially. Clustering heatmap exhibits metabolite relative levels of each group, and the heatmap suggests that the metabolite expression of kaempferol differs greatly from that of the CCl4-treated group and converged to the metabolite expression of the NC group (Fig. 4B-C). Metabolic differences between kaempferol-treated and CCl4-treated groups were investigated by metabolomics. In the volcano plot (Fig. 4D), 101 metabolites had up-regulation, whereas 56 metabolites showed down-regulation, such as the differential metabolites S-Adenosylmethionine, 6-Methyl-2- thiouracil, 4,4'-Sulfonylbisphenol, (4E,10E)-13,21-Dihydroxy-8,14,19- Trimethoxy—4,10,12,16—tetramethyl -3,20,22 -triox…, 5(S)-HpETE, and Choline were upregulated. According to KEGG pathway enrichment, those 15 most significant pathways regulated by kaempferol were Arachidonic acid metabolism, Bile secretion, Carbon metabolism, ABC transporters, etc. (Fig. 4E).

Fig. 4

Kaemferol changes metabolic profiles in CCl4-induced mice. A PCA. B Cluster heatmap analysis of metabolites across different groups. C Differential metabolite cluster heatmap analysis comparing kaempferol and CCl4. D Volcano plot illustrating differential metabolites between kaempferol and CCl4. metabolites between kaempferol and CCl4. E KEGG enrichment analysis for kaempferol versus CCl4. F Volcano plot depicting differential metabolites in kaempferol. metabolites in kaempferol-FMT compared to CCl4-FMT. G KEGG analysis on differential metabolites between kaempferol-FMT and CCl4-FMT. H Intersection of the differential metabolites from kaempferol-M versus CCl4 and those from kaempferol-FMT versus CCl4-FMT. I The intersection of metabolic pathways differing between kaempferol-M vs CCl4 and kaempferol-FMT vs CC14-FMT

To verify that the differential metabolite regulation by kaempferol on CCl4-induced mice was mediated by the intestinal flora, the faecal colony-transplanted kaempferol-FMT and CCl4-FMT groups were subjected to non-targeted metabolomics assays. Based on the results of the PCoA analyses (Fig. 4A), difference in kaempferol-FMT compared with CCl4-FMT groups had a certain degree of deviation and converged toward the normal group, suggesting that the differential metabolite regulation of CCl4-induced mice by kaempferol was mediated by the intestinal flora. In the volcano plot (Fig. 4F), 122 metabolites showed up-regulation, whereas 82 metabolites had down-regulation. Based on KEGG pathway enrichment, the top 15 pathways regulated by kaempferol-FMT compared with CCl4-FMT were Glycine, serine and threonine metabolism, Bile secretion, Neuroactive ligand-receptor interaction, and others (Fig. 4G). Taking the intersection of the differential metabolites of kaempferol-M vs CCl4 and kaempferol-FMT vs CCl4-FMT yielded 13 identical differential metabolites, as shown in the Wayne diagram (Fig. 4H). Taking the intersection of the differential metabolic pathways of kaempferol-M vs CCl4 vs kaempferol-FMT vs CCl4-FMT yielded 36 identical differential metabolic pathways, as shown in the Wayne diagram (Fig. 4I). It was concluded that kaempferol enriches the above metabolites and metabolic pathways depending on gut microbe.

Kaemferol-Mediated Enrichment of g__Robinsoniella, g__Erysipelotrichaceae_UCG-003 and g__Coriobacteriaceae_UCG-002 Attenuates Liver Fibrosis via Regulating Lipid Metabolism

Alterations of intestinal floral structure may impact the host metabolic phenotype. Consequently, we used Spearman correlation analyses for exploring correlations after kaempferol administration. Spearman correlation analysis of 19 key metabolites abolished through kaempferol and 8 gut microorganisms showed that:1,9-Nonanedithiol, 1-(4-Fluorophenyl) -5-oxo-3-pyrrolidinecarboxylic acid, 5(S)-HpETE, 5-[(2 -Phenylethyl)amino]-1,3,4-thiadiazole-2(3H)-thione were mostly significantly correlated with intestinal bacteria (Fig. 5A). Additionally, Hyodeoxycholic acid, Chenodeoxycholic acid, Deoxycholic acid and g_Betaproteobacteria_unclassified showed strong positive correlation.5-[(2- Phenylethyl)amino]-1,3,4-thiadiazole-2(3H)-thione, all-trans -5,6-Epoxyretinoic acid, Dihydroergocristine, LPI(20:4), Dimethylbenzyl_carbinyl_hexanoate, LPI(18:0) and g__ Robinsoniella were strongly positively correlated.Histidine, 13,15-Dihydroxy-9-methyl-10-oxabicyclo [10.4.0]hexadeca-1(12),13,15-trien-11-one, Grandiflorenic acid, 1,9- Nonanedithiol, 1-(4-Fluorophenyl)-5-oxo-3-pyrrolidinecarboxylic acid, 5(S)-HpETE and g__Coriobacteriaceae_UCG-002 were strongly positively correlated.1,9-Nonanedithiol, 1 -(4-Fluorophenyl)-5-oxo-3-pyrrolidinecarboxylic acid, 5(S) -HpETE and g__Erysipelotrichaceae_UCG-003 are strongly positively correlated. histidine, 5-Methylcytidine, all-trans -5,6-Epoxyretinoic acid, LPI (18:0) was strongly and negatively related to g__Eubacterium]_nodatum_group. 3-Hydroxybenzoic acid. 17alpha-Hydroxypregnenolone showed a strong negative correlation with g__Robinsoniella.

Fig. 5

Kaempferol-mediated enrichment of g__Robinsoniella, g__Erysipelotrichaceae_UCG-003 and g__Coriobacteriaceae_UCG-002 attenuates liver fibrosis via regulating lipid metabolism. A Correlation analysis of differential intestinal microbiota and serum metabolites in kaempferol versus CCl4. B Correlation analysis of intestinal microbiota and serum metabolites in kaempferol versus colchicine. C Venn diagram illustrating the intersection of differential intestinal microbiota. D Venn diagram depicting the intersection of differential metabolites

To find the gut flora and serum metabolites specifically regulated by kaempferol, the data of differential gut flora and differential serum metabolome of Kaempferol vs Colchicine were analyzed (Fig. 5B) and intersecting Wayne plots of kaempferol vs CCl4 vs Kaempferol vs Colchicine were done. Differential intestinal flora intersecting Wayne plots showed that kaempferol specifically regulated five intestinal flora (Fig. 5C), namely g__Robinsoniella, g__Erysipelotrichaceae_UCG-003, g__Dorea, g__Eubacterium]_nodatum_ group, g__Coriobacteriaceae_UCG-002, Differential metabolite intersection Wayne plots showed that kaempferol specifically regulates 12 serum metabolites (Fig. 5D), namely 5-Methylcytidine, all-trans-5,6-Epoxyretinoic acid, the LPI (18:0), LPI (20:4), 13,15-Dihydroxy-9-methyl-10-oxabicyclo[10.4.0]hexadeca-1(12),13,15-trien-11-one, 5-[(2-Phenylethyl)amino]-1,3,4 -thiadiazole-2(3H)-thione, 1-(4-Fluorophenyl)-5-oxo-3- pyrrolidinecarboxylic acid, Dihydroergocristine, Grandiflorenic acid, 1,9-Nonanedithiol, 5(S)-HpETE, Dimethylbenzyl_carbinyl_hexanoate, which are mainly lipids and lipid-like molecules.

The above results suggest that, in a limited number of samples, these five bacteria can serve as specific bacteria for distinguishing between healthy and LF mice. kaempferol is critical for treating LF treatment through decreasing the bacterial genera g__Dorea, g__Eubacterium]_nodatum_group and increasing the bacterial genus g__Robinsoniella, g__ Erysipelotrichaceae_UCG-003, g__Coriobacteriaceae_UCG-002, which is important for the treatment of LF and we will further verify in subsequent experiments.

Kaemferol Downregulates the Th17/IL-17 Signaling Pathway in PDGF-Induced LX2 Cells

For exploring cellular mechanisms of protection against chronic hepatitis induced by kaempferol, we analyzed hepatic stellate cell gene expression profiles by RNA-Seq. Based on Heatmap clustering analysis, kaempferol treatment regulated a variety of gene levels (P < 0.01, Fig. 6A), and the volcano showed the genes regulated by kaempferol, including 59 with up-regulation whereas 35 with down-regulation (P < 0.05, Fig. 6B), with statistically significant differences between the groups. According to KEGG analysis, kaempferol remarkably down-regulated various inflammatory pathways such as Th17 cell differentiation, Toll like receptor, Jak stat, and IL-17 pathways (Fig. 6C). In line with GO annotation, kaempferol could effectively regulate a variety of immune-related functions, such as innate immune response, complement activation, cytokine stimulus cellular response (Fig. 6D). Interestingly it has been extensively documented that production of the pro-inflammatory cytokine interleukin 17 (IL-17) family can be triggered by commensal microorganisms (bacteria and fungi) and is important for the maintenance of intestinal homeostasis by regulating mucosal immunity and gut barrier integrity [34, 35]. IL-17 signalling promotes gut barrier immunity by regulating microorganisms but also drives tumour growth and inflammatory progression [36]. Gut dysbiosis induced by systemic or intestinal epithelial deletion of IL-17RA has been documented to induce pancreatic and brain tumour growth due to overdevelopment of Th17, as well as B cells circulating to distant tumours [37]. Given the role of IL-17-IL-17RA in the maintenance of intestinal homeostasis, it has been documented that intestinal IL-17-IL-17RA signalling and its microbial regulation can influence immune regulation and ultimately tumour growth [38]. IL-10 signalling in intestinal macrophages drives the anti-inflammatory Th17 cell phenotype. Gut microbiota-specific Th17 cells have regulatory properties and suppress effector T cells via c-MAF and IL-10 [39, 40].

Fig. 6

RNA sequencing analysis reveals the effects of kaempferol on PDGF-induced LX2 cells. A Heat map of genes in different groups. B Volcano plot of differentially expressed genes in Kaemferol versus NC group. C KEGG enrichment analysis. D GO enrichment analysis

Single-Cell Sequencing-Based Exploration of the Mechanism by which Kaempferol Improves Prognosis in Mice with Liver Fibrosis

To explore the single-cell profile in liver landscape of mice with liver fibrosis after kaempferol modulation, single-cell transcriptional sequencing was conducted in liver samples of CCl4 group and kaempferol groups. After quality control, 39,440 cells (18,788 in the CCl4 group and 20,652 in kaempferol-treated) were analyzed to further identify cell types, as shown in Fig. 7. Thus, therefore, 12 cell types (Fig. 7A) were identified according to classical cell markers (Fig. 7B), such as B Cells (n = 1386, with expression of Cd79a, Cd79b), conventional dendritic cells (n = 221, with expression of Cd80, Cd83), Endothelial Cells (n = 3794, Clec4g, Ptprb), Erythroid Cells (n = 761, with expression of Alas2Bpgm), Hepatocytes (n = 936, expressing Car3, Alb), Kupfer cells (n = 1478, Clec4f, C1qa), liver capsular macrophages (n = 8653, expressing S100a4), Macrophages (n = 1203, expressing Cd74), Myeloid cells (n = 12,582, expressing Serpine1, Ltc4s), Neutrophils (n = 6036, Cxcr2 S100a8), Plasmacytoid Dendritic (n = 904, with expression of Siglech, Ccr9), T & NK Cells (n = 1486, with expression of Cd3d, Cd3g, Nkg7). As documented [36], endothelial cells are related to the migration of endothelial cells, hepatocytes are associated with organic acid catabolic metabolic processes, hematopoietic stem cells are involved in organization of extracellular matrix, and T/B cells are related to cellular differentiation, whereas kaempferol treatment resulted in significant changes in 12 cell abundances within the liver (Fig. 7C-D). Specifically, the relative percentages of Myeloid cells, T & NK Cells, Kupfer cells, and Hepatocytes significantly elevated following kaempferol exposure, whereas conventional dendritic cells, Endothelial Cells. Erythroid Cells, Neutrophils and Plasmacytoid Dendritic declined. Additionally, kaempferol exposure significantly changed gene expression patterns (genes with up- and down-regulation, Fig. 7E) of these 12 cells. As suggested by GO and KEGG analyses on DEGs, kaempferol mainly regulated Carbon metabolism, Complement and coagulation cascades, Biosynthesis of amino acids and other pathways (Fig. 7F-G).

Fig. 7

Kaempferol enhances the single-cell RNA sequencing profiles of CCl4-induced murine liver tissue. A UMAP analysis for cell trace visualization. B Violin plot illustrating the expression levels of marker genes across different cell types. C Comparative analysis of cluster distributions between the CCl4 and kaempferol-treated groups. D Proportional representation of each cell type in the livers of mice from both CCl4 and kaempferol groups. E Analysis of up-regulated and down-regulated gene proportions following kaempferol treatment in each group. F Gene Ontology enrichment analysis for differentially expressed genes. G KEGG pathway enrichment analysis for differentially expressed genes

Kaempferol Treatment Improves Hepatic Lymphocyte Immune Function

Lymphocytes exhibited obvious DEG following kaempferol administration, consisting of 890 with up-regulation whereas 1292 with down-regulation (Fig. 8A), therefore, lymphocytes were focused in later analyses. Lymphocytes are important for liver immunomodulation. This work described hepatic lymphocyte subtypes, such as T cells, B cells, and NK cells, according to critical genetic markers. From Fig. 8B, T cells are classified as 4 subtypes, CD4_naive, CD4_Tem, CD4_Treg, and CD8_T, while B cells are categorized into B_naive, and B_plasms. Violin plots exhibit subtype-specific gene levels in each subtype (Fig. 8C), where C-X-C motif chemokines (CXCLs), also called the macrophage inflammatory protein, exerts the main effect on chemokine leukocytes in the inflammatory region. Kaempferol administration increased the CXCL family expression within hepatic lymphocytes, indicating enhanced immunomodulation. In liver disease, immune cells are the alarm system through promoting cytokine release and immune cell recruitment into the liver tissue to exert its protection. We found that, kaempferol administration remarkably enhanced CD4 + T and CD8 + T cellular proportions and decreased mature B cell proportion in comparison with controls, revealing that kaempferol activates immune cells for protecting the body through inducing inflammatory reactions to regulate inflammation. In addition, GO annotation was carried out based on DEGs across cell types. From Fig. 8D, those up-regulated pathways of kaempferol group were mainly related to the activation of immune cells, adaptive immune response according to immunoreceptor somatic recombination of immunoglobulin superfamily structural domains, conforming to prior results in liver injury [41, 42]. With regard to down-regulated DEGs, GO annotation showed the effect of kaempferol on inhibiting myeloid differentiation, negative regulation of immune system processes, liver inflammation and apoptosis (Fig. 8E). In summary, hepatic lymphocytes after kaempferol treatment exhibited enhanced immune function.

Fig. 8

kaempferol improved liver lymphocyte immune function after treatment. A Volcano plot showing up- and down-regulated differential genes (DEGs) between the Kaempferol and CCl4 groups. B UMAP showing the distribution of samples corresponding to the identified cell subtypes. C Violin plots illustrating subtype-specific gene expression for each subtype. D Bubble plots showing GO enrichment analysis of up-regulated DEGs across the subtypes. E Bubble plots showing GO enrichment analysis of down-regulated DEGs across the isoforms

Correlation Analysis of Liver Monocyte Population Transcription Factors with Metabolome and Gut Flora

To explore the correlation between the components of the enterohepatic axis, differential transcription factors were correlated with differential intestinal flora. The correlation analysis heatmap (Fig. 8F) showed that g__Eubacterium]_nodatum_group, g__Ruminococcus]_gnavus_group, g__Dorea, and g__Betaproteobacteria_unclassified were associated with the hepatic transcription factors Fosl1(+), Mef2c(-), Cebpe(+), Fosb(+), Sox17(-), Mbd4(+), and Nfib(+) are strongly positively correlated with the differential intestinal flora g__Robinsoniella, g__Prevotella_9, g__ Citrobacter, g__Erysipelotrichaceae_UCG-003, g__Coriobacteriaceae_UCG-002 were strongly positively correlated with the liver transcription factors Zfp553(-), Nfe2( +), Esrrg(-), Snai1(+), Sox17(-), Mbd4(+), Nfib(+), Hoxb2(+), Twist1(-), Maf(+), Pbx1(+) were strongly negatively correlated. Differential transcription factors were correlated with differential serum metabolites, and the correlation analysis heatmap (Fig. 8G) showed that 3-Hydroxybenzoic acid, 17alpha-Hydroxypregnenolone, Chenodeoxycholic acid. Deoxycholic acid, Hyodeoxycholic acid and liver transcription factors Sox17(-), Hoxb2(+), Nfya(-), Klf5(-), Cebpd(+), Gm14295(-) were strongly positively correlated, while the differential serum metabolites all-trans-5,6-Epoxyretinoic acid, 5-[(2- Phenylethyl)amino]-1,3,4-thiadiazole-2(3H)-thione, 5-Methylcytidine. Histidine, 13,15-Dihydroxy-9-methyl-10-oxabicyclo[10.4.0]hexadeca -1(12),13,15-trien-11-one, N,N-Diethyl-2- aminoethanol was strongly negatively correlated with the liver transcription factors Klf5(-), Cebpd(+), Gm14295(-), Fosl1(+), Mef2c(-), Cebpe(+), Fosb(+). The correlation analysis of the stages of the enterohepatic axis could further confirm that kaempferol treats liver fibrosis through the enterohepatic axis.