Cathelicidin preserves intestinal barrier function in polymicrobial sepsis

Abstract

Objectives: The intestinal epithelium compartmentalizes the sterile bloodstream and the commensal bacteria in the gut. Accumulating evidence suggests that this barrier is impaired in sepsis, aggravating systemic inflammation. Previous studies reported that cathelicidin is differentially expressed in various tissues in sepsis. However, its role in sepsis-induced intestinal barrier dysfunction has not been investigated.

Design: To examine the role of cathelicidin in polymicrobial sepsis, cathelicidin wild-(Cnlp^+/+) and knockout (Cnlp^-/-) mice underwent cecal-ligation and puncture (CLP) followed by the assessment of septic mortality and morbidity as well as histological, biochemical, immunological, and transcriptomic analyses in the ileal tissues. We also evaluated the prophylactic and therapeutic efficacies of vitamin D3 (an inducer of endogenous cathelicidin) in the CLP-induced murine polymicrobial sepsis model.

Results: The ileal expression of cathelicidin was increased by three-fold after CLP, peaking at 4 h. Knockout of Cnlp significantly increased 7-day mortality and was associated with a higher murine sepsis score. Alcian-blue staining revealed a reduced number of mucin-positive goblet cells, accompanied by reduced mucin expression. Increased number of apoptotic cells and cleavage of caspase-3 were observed. Cnlp deletion increased intestinal permeability to 4kD fluorescein-labeled dextran and reduced the expression of tight junction proteins claudin-1 and occludin. Notably, circulating bacterial DNA load increased more than two-fold. Transcriptome analysis revealed upregulation of cytokine/inflammatory pathway. Depletion of Cnlp induced more M1 macrophages and neutrophils compared with the wild-type mice after CLP. Mice pre-treated with cholecalciferol (an inactive form of vitamin D3) or treated with 1alpha, 25-dihydroxyvitamin D3 (an active form of VD3) had decreased 7-day mortality and significantly less severe symptoms. Intriguingly, the administration of cholecalciferol after CLP led to worsened 7-day mortality and the associated symptoms.

Conclusions: Endogenous cathelicidin promotes intestinal barrier integrity accompanied by modulating the infiltration of neutrophils and macrophages in polymicrobial sepsis. Our data suggested that 1alpha, 25-dihydroxyvitamin D3 but not cholecalciferol is a potential therapeutic agent for treating sepsis.

Keywords: Antimicrobial peptide; Bacterial translocation; LL-37; Sepsis.

Fig. 1

Murine cathelicidin-related antimicrobial peptide (mCRAMP) is upregulated after cecal-ligation and puncture (CLP) induced sepsis in wild-type (WT) mice (n = 6 per group) from which total RNA and protein were collected from distal ileum over a period of time for a real-time PCR and b immunofluorescence for mCRAMP. Genetic knockout (KO) of Cnlp led to f reduced survival and e higher sepsis severity score (n = 18 for WT mice; n = 17 for KO mice). FITC dextran 4 kD was orally gavaged at 21 h after CLP with serum harvested after 3 h. Genetic KO of Cnlp led to c increased serum concentration of FITC-labeled dextran 4 kD (FD-4) and d increased bacterial DNA upon experimental sepsis. Error bars denote standard error of the mean. *P < 0.05; ***P < 0.001; ****P < 0.0001

Fig. 2

Effects of cecal-ligation and puncture (CLP) or sham surgery (Sham) on acid mucin in distal ileum of cathelicidin wild-type (Cnlp^+/+) or knockout (Cnlp^−/−) mice (n = 6 per group) at 24 h after CLP as determined by quantitative real-time PCR and a Alcian blue perioidic acid Schiff reaction, respectively. b The number of acid-mucin-producing globlet cells per villus and c the expression of mucin genes MUC1 and MUC2 were compared. Error bars represent standard error of the mean. *P < 0.05; **P < 0.01

Fig. 3

Deletion of Cnlp reduced tight junction of intestinal epithelial cells. The protein levels of a, b occludin and claudin-1 were detected in cathelicidin-knockout (Cnlp^−/−) mice (n = 6) compared to wild-type mice (Cnlp^+/+) (n = 6) after CLP-induced sepsis by immunoblotting. Error bars denote standard error of the mean. ***P < 0.001; ****P < 0.0001

Fig. 4

Increased apoptosis was detected in cathelicidin-knockout (Cnlp^−/−) mice (n = 6) compared to wild-type mice (Cnlp^+/+) (n = 6 per group) after CLP-induced sepsis as demonstrated by a, b TUNEL staining and c, d immunoblotting for cleaved caspase-3. All specimens were collected at 24 h after CLP or Sham surgery. Error bars denote standard error of the mean. **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 5

Ileal transcriptomes of septic cathelicidin wild-type and knockout mice. a Heatmap was generated using genes in cluster 8 and cluster 16. The transcriptome datasets from the septic and non-septic wild-type mice but not the cathelicidin knockout mice were published in Inflamm Res. 2019; 68 [9]:723–726. b Protein-protein interaction network was constructed in STRING using the source of “experiments,” “databases,” and “co-expression” and visualized by cytoscape. Nodes in round shape have a degree large than 22. Several inflammation-related genes were highlighted as hub genes according to the topology analysis. c, d Top 30 KEGG and Reactome pathways were plotted. A series of inflammation related pathways were enriched by both sources

Fig. 6

Deletion of endogenous cathelicidin increased neutrophils and macrophages into small intestine. Effects of CLP or sham surgery on the relative proportion of neutrophils and macrophages in small intestine of cathelicidin wild-type (Cnlp^+/+) or knockout (Cnlp^−/−) mice (n = 6 per group) at 24 h were determined by flow cytometry. a Neutrophils were defined as Ly6G⁺ cells and b macrophages as F4/80⁺ cells and c M1 macrophages as F4/80⁺ CD86⁺ and d M2 macrophages as F4/80⁺ CD206⁺. Error bars denote standard error of the mean. **P < 0.01; ****P < 0.0001

Fig. 7

CLP-induced polymicrobial sepsis did not influence the lymphocyte count in the ileum (n = 5 for the sham-operated group; n = 7 for the CLP group). Flow cytometric analysis of lymphocytes isolated from small intestines of sham-operated or CLP cathelicidin wild type (Cnlp^+/+) was performed. Cells were labeled with anti-mouse CD45⁺ lineage surface markers

Fig. 8

Effect of VD3 on peritonitis-induced polymicrobial sepsis. a All mice (n = 6 per group) underwent CLP surgery pretreated with water or VD3 by gavage at 48 h, 24 h, and 1 h before CLP. Mice pretreated with VD3 had b decreased 7-day mortality (n = 10 per group) and c significantly lower MSS score (n = 10 per group) and d lower level of fluorescein dextran entering the bloodstream (n = 6 per group). Mucin production increased in mice pretreated with VD3. The expression levels of e Muc1 and f Muc2 were detected in mice pretreated with VD3 compared to mice pretreated with water after CLP-induced sepsis by immunoblotting. Error bars denote standard error of the mean. **P < 0.01; ***P < 0.001; ****P < 0.0001

Fig. 9

Pretreatment with VD3 up-regulated the ileal expression of cathelicidin in sepsis. Effect of VD3 on expression of mCRAMP in distal ileum of wild-type mice at 24 h after CLP were determined by immunofluorescence and (a) real-time quantitative PCR (b). n = 6 per group, error bars denote standard error of the mean. c, d Prophylactic efficacy of VD3 required cathelicidin in CLP-induced polymicrobial sepsis. Cathelicidin-knockout mice (Cnlp^−/−) pretreated with water or VD3 by oral gavage at 48 h, 24 h, and 1 h before CLP (n = 5 per group). There was no significant difference between water control and VD3 group in terms of c 7-day mortality as well as d MSS score in Cnlp^−/− mice. **P < 0.01; ****P < 0.0001

Fig. 10

Effects of inactive and active VD3 after the onset of CLP-induced polymicrobial sepsis. a All mice underwent CLP surgery and were administered with water or inactive VD3 (50 μg/kg) for one time by oral gavage immediately after CLP (n = 8 per group). Mice treated with inactive VD3 even had a higher 7-day mortality and b higher day-1 MSS score. CLP-induced polymicrobial sepsis resulted in hepatic damage. CLP-operated mice demonstrated increased c AST and ALT levels, d decreased CYP2R1 and CYP27A1 mRNA levels, and e reduced serum VD3 levels (intermediate plus active forms) (n = 4–5 per group). For active VD3 treatment, all mice underwent CLP surgery and were administered with water or active VD3 (50 μg/kg) for 7 days by intraperitoneal injection (n = 11 per group). Mice treated with active VD3 had better outcomes in terms of f 7-day mortality, g day 2 and day 3 MSS score as well as d higher serum vitamin D3 levels (intermediate plus active forms).*P < 0.05; **P < 0.01; ***P < 0.001