Enteric fungi protect against intestinal ischemia-reperfusion injury via inhibiting the SAA1-GSDMD pathway

Abstract

Introduction: Prophylactic antifungal therapy has been widely used for critical patients, but it has failed to improve patient prognosis and has become a hot topic. This may be related to disruption of fungal homeostasis, but the mechanism of fungi action is not clear. As a common pathway in critical patients, intestinal ischemia-reperfusion (IIR) injury is fatal and regulated by gut microbiota. However, the exact role of enteric fungi in IIR injury remains unclear.

Objectives: This is a clinical study that aims to provide new perspectives in clarifying the underlying mechanism of IIR injury and propose potential strategies that could be relevant for the prevention and treatment of IIR injury in the near future.

Methods: ITS sequencing was performed to detect the changes in fungi before and after IIR injury. The composition of enteric fungi was altered by pretreatment with single-fungal strains, fluconazole and mannan, respectively. Intestinal morphology and function impairment were evaluated in the IIR injury mouse model. Intestinal epithelial MODE-K cells and macrophage RAW264.7 cells were cultured for in vitro tests.

Results: Fecal fungi diversity revealed the obvious alteration in IIR patients and mice, accompanied by intestinal epithelial barrier dysfunction. Fungal colonization and mannan supplementation could reverse intestinal morphology and function impairment that were exacerbated by fluconazole via inhibiting the expression of SAA1 from macrophages and decreasing pyroptosis of intestinal epithelial cells. Clodronate liposomes were used to deplete the number of macrophages, and it was demonstrated that the protective effect of mannan was dependent on macrophage involvement.

Conclusion: This finding firstly validates that enteric fungi play a crucial role in IIR injury. Preventive antifungal treatment should consider damaging fungal balance. This study provides a novel clue to clarify the role of enteric fungi in maintaining intestinal homeostasis.

Keywords: Antifungal therapy; Enteric fungi; Intestinal ischemia–reperfusion injury; Mannan; Pyroptosis.

Graphical abstract

Fig. 1

Characteristics of enteric fungi in IIR model. (A) The morphology and blood flow of the superior mesenteric artery (SMA) in IIR injury patients, before and 2 h after the start of hemodialysis by color Doppler ultrasound (n = 10). (B) Serum levels of intestinal barrier-related biomarkers (I-FABP, ZO-1, Citrulline) of IIR injury patients were detected by ELISA (n = 20). Wilcoxon matched-pairs signed-rank test was used to calculate the P-value. (C) Alpha diversity indices of genus level in patients. (D) Principal coordinate analysis (PCoA) using Bray–Curtis metric distance algorithms and unweighted UniFrac distance algorithms of beta diversity in patients. (E) The community of enteric fungi on the order and genus level in patients were analyzed with Wilcoxon signed-rank test to calculate the P-value. (F) LEfSe analysis was performed to explore the different species characteristics from phylum to genus levels in patients (LDA > 4). (G) CCA analysis at the genus level. Positive correlation between two indicators < 90°; Negative correlation between two indicators > 90°; No correlation between two indicators = 90°. (H) Correlation linear graph. The relative abundance of Candida and Saccharomyces was positively correlated with citrulline levels and negatively correlated with I-FABP and ZO-1 levels. (I) Alpha diversity indices at the genus level in mice model. (J) Beta diversity indices at the genus level in mice model (based on PCoA and NMDS). (K) LEfSe analysis was performed to explore the different species characteristics from phylum to genus levels between both groups (LDA > 2). (L) Random forests analysis at the genus level. (M) Correlation Heatmap graph of fungi and intestinal barrier-related biomarkers and tissue cytokines expression. Fig. 1 (A, C–H) (n = 10), Fig. 1 (B) (n = 20), Fig. 1 (I-M) (n = 5) *P < 0.05; **p < 0.01; ***p < 0.001.

Fig. 2

The colonization of single fungal strains alleviated IIR injury. (A) The experimental design of the present study. Briefly, 10⁸ CFU C. tropicalis or S. cerevisiae was colonized in the intestine by oral gavage before the IIR injury model was established. (B)C. tropicalis and S. cerevisiae abundance in mouse stool after single fungal strains colonization. (C) Serum FITC-dextran concentration in C. tropicalis and S. cerevisiae colonization groups. (D) Serum I-FABP concentration in different groups. (E) H&E staining and the Chiu’s score in different groups. Scale bar = 100 μm. (F) The expression of the ZO-1, occludin, and GSDMD proteins in different groups. (G) Cytokine levels of intestine tissue supernatant in different groups. (H) Fecal fungi abundance in germ-free mice treated with S. cerevisiae and C. tropicalis mixtures were identified by fungus plate culture and fence quantitative abundance test. (I) Serum FITC-dextran concentration in GF mice colonized with S. cerevisiae and C. tropicalis mixtures. (J) Serum I-FABP concentration in GF mice colonized with S. cerevisiae and C. trop mixtures. (K) H&E staining and the Chiu’s score in GF mice colonized with S. cerevisiae and C. tropicalis mixtures. Scale bar = 100 μm. The values were presented as the mean ± SD in Fig. 2(B–G) (n = 6), and Fig. 2 (H–K) (n = 5). *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control group.

Fig. 3

Prophylactic antifungal therapy aggravated IIR injury in mice. (A) The experimental design. Mice were pretreated with fluconazole for 2 weeks, and the IIR model procedure was subsequently performed (ischemia 30 min, reperfusion 6 h). (B) Quantitative fungal abundance (ITS, C. tropicalis and S. cerevisiae) was determined by qPCR of fungal rDNA and expressed as copy numbers (copies/g). (C) Serum fluconazole concentration in mice detected by ultra-high performance liquid chromatography. (D) Serum alanine aminotransferase (ALT) level. (E) Intestinal permeability test determined by FITC-dextran transepithelial permeability assay. (F) Serum I-FABP concentration detected by ELISA. (G) The histopathological damage of IIR injury was estimated by H&E staining and the Chiu’s score. Scale bar = 100 μm. (H) Cytokine levels of intestine tissue supernatant were determined using the LEGENDplex™ assays. (I) The ZO-1, occludin, and GSDMD protein expressions were analyzed by WB. (J) Serum FITC-dextran concentration in GF mice. (K) Serum I-FABP concentration in GF mice. (L) H&E staining and the Chiu’s score in GF mice. Scale bar = 100 μm. The values were presented as the mean ± SD in Fig. 3(B–I) (n = 6) and Fig. 3 (J–L) (n = 5). *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control group.

Fig. 4

Enteric fungal clearance did not exacerbate GSDMD cleavage in GSDMD^-/- mice. (A) The experimental design of the present study. GSDMD ^-/- mice were pretreated with fluconazole for 2 weeks, and the IIR model procedure was subsequently performed (ischemia 30 min, reperfusion 6 h). (B) GSDMD protein expression in GSDMD^-/- mice. (C) Quantitative microbiological abundance (16S, ITS, C. tropicalis, and S. cerevisiae) of GSDMD^-/- mice compared with the WT mice. (D) Serum FITC-dextran concentration in GSDMD^-/- mice. (E) Serum I-FABP concentration in GSDMD^-/- mice. (F) The expressions of ZO-1 and occludin protein in GSDMD^-/- mice. (G) H&E staining and the Chiu’s score in GSDMD^-/- mice. Scale bar = 100 μm. (H) The levels of cytokines in the supernatant of GSDMD^-/- mice intestine tissue. The values were presented as the mean ± SD in Fig. 4 (B) (n = 2) and Fig. 4 (C–H) (n = 5–8). *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control group. WT, wild-type, GSDMD^-/-, GSDMD-knockout mice.

Fig. 5

Mannan replicated the protective effects of enteric fungi on IIR injury. (A) Mice were pretreated with mannan for 2 weeks, and the IIR model procedure was subsequently performed (ischemia 30 min, reperfusion 6 h). (B) Serum FITC-dextran concentration in mice pretreated with mannan. (C) Serum I-FABP concentration. (D) Cytokine levels of intestine tissue supernatant were determined LEGENDplex™ assays. (E) H&E staining and the Chiu’s score. Scale bar = 100 μm. (F) The ZO-1, occludin, and GSDMD protein expressions were analyzed by WB. (G) Serum FITC-dextran concentration in GF mice pretreated with mannan. (H) Serum I-FABP concentration in GF mice pretreated with mannan. (I) H&E staining and the Chiu’s score in GF mice pretreated with mannan. The values were presented as the mean ± SD in Fig. 5(B–I) (n = 5–6). *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control group.

Fig. 6

Mannan’s protective role in IIR injury depends on macrophage involvement. (A) Molecular structure diagram of mannan attached to the cy3 fluorescent marker. (B) Co-localization of mannan and macrophages in the ileum, detected by immunofluorescence staining. Scale bar = 50 μm and 10 μm, respectively. In vitro, Cy3-mannan was added to RAW264.7 cells, which showed that mannan entered RAW264.7 cells after 6 h of incubation. (C) The number of ileum F4/80⁺CD45.2⁺ macrophages in different groups were analyzed by flow cytometry in mice. (D) Serum FITC-dextran concentration in mice. (E) Serum I-FABP concentration. (F) H&E staining and the Chiu’s score. Scale bar = 100 μm. (G) The expressions of ZO-1, occludin, and GSDMD protein were analyzed by western blotting. (H) mRNA expression of GSDMD, IL1β, IL18, and SAA1 in ileum tissue after establishing the IIR mouse model. The values were presented as the mean ± SD in Fig. 6 (B-H) (n = 4). *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control group.

Fig. 7

Single fungal strains and mannan both alleviated GSDMD cleavage induced by reducing SAA1 expression. (A) Volcano plot of differentially expressed genes in intestinal tissue treated with C. trop detected by RNA-Seq. The x-axis of the volcano plot shows the log2 transformed gene expression fold change. The y-axis shows the negative log10 transformed adjusted q-values of the differential expression. The upregulated genes are represented by red dots. The downregulated genes are represented by blue dots. SAA1 mRNA expression was detected by qPCR. (B) Volcano plot of differentially expressed genes in the mannan group detected by RNA-Seq. The upregulated genes are represented by red dots. The downregulated genes are represented by green dots. SAA1 mRNA expression in the mannan group was detected by qPCR. (C) SAA1 protein expression in different groups was detected by immunohistochemistry assay. Scale bar = 20 μm. (D) Co-localization of SAA1 protein and macrophages marked with F4/80. Scale bar = 50 μm and 20 μm, respectively. (E) SAA1 mRNA expression in RAW264.7 cells was detected by qPCR. RAW264.7 cells were pretreated with mannan(1 mg/mL) for 6 h, and then move to the medium containing LPS(1 µg/mL) to incubate 2 h. (F) SAA1 protein expression in RAW264.7 macrophage was detected by WB. (G) GSDMD protein expression in MODE-K cell treated with active SAA1 protein in the OGD/R model. (H) mRNA expression of Occludin, ZO1, GSDMD, NLRP3, IL1β and IL18 in MODE-K cell treatment with SAA1 detected by qPCR in the OGD/R model. The values were presented as the mean ± SD in Fig. 7 (A-H) (n = 3–6). *p < 0.05, **p < 0.01, ***p < 0.001 compared with the control group. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)