Introduction
Sepsis is a life-threatening organ dysfunction accompanied by systemic inflammation and immunosuppression as a consequence of the host response to microbial infections. Sepsis which carries high mortality and morbidity in the intensive care unit remains a major health burden [
1]. Therefore, there is a compelling need for developing novel sepsis therapies.
The pathogenesis of sepsis has been attributed, at least in part, to the loss of intestinal epithelial barrier. As the first line of defense, the gut intestinal epithelial barrier impedes the translocation of commensal bacteria from the gut lumen into the bloodstream. Accumulating evidence suggests that the intestinal barrier function is impaired during systemic inflammation as in sepsis. These include epithelial apoptosis, disruption of tight junctions leading to an increase in intestinal permeability [
2,
3]. The impaired gut barrier function may increase the risk of bacterial translocation from the gut lumen to the bloodstream, aggravating systemic inflammation. Clinically, bacterial translocation from the gut into the bloodstream has been demonstrated in patients with postoperative sepsis [
4]. An abnormal and severe derangement of intestinal permeability upon admission to an intensive care unit was found to predict subsequent development of multiple organ failure [
5]. However, the underlying mechanism of sepsis-associated gut barrier dysfunction remains elusive.
Cathelicidin represents one of the most important classes of antimicrobial peptides in mammals. It has bactericidal property, inhibits endotoxin-induced pyroptosis of leukocytes, suppresses the release of inflammatory mediators, and protects endothelial cells from apoptosis [
6,
7]. Cathelicidin can be induced by vitamin D3 (VD3), which has therapeutic properties outside of its classic functions related to bone and calcium homeostasis [
8,
9]. In particular, a growing body of evidence has shown the antibiotic-like properties of vitamin D [
10]. Thus, this natural compound may prove to be effective against sepsis, as an adjunct treatment modality. Previously, Chen and his colleagues suggested that VD3 exerts protective effects during infections by upregulating the expression of cathelicidin and beta-defensin 2 in phagocytes and epithelial cells [
11]. Another study found that systemic LL-37 (human cathelicidin) levels may be regulated by VD3 status [
12]. In our study, we aimed to investigate the role of murine cathelicidin-related antimicrobial peptide (mCRAMP), a rodent antimicrobial peptide analogous to human cathelicidin LL-37, in maintaining gut barrier function in sepsis and to explore the relationship between vitamin D3 status and cathelicidin production in CLP mice model.
Materials and methods
Animals
129/SVJ wild-type (
Cnlp+/+) and cathelicidin-knockout (
Cnlp−/−) mice were used. These mouse strains were generated as previously described [
13]. All animals were male and 8 to 10 weeks old. They were maintained in the Laboratory Animal Services Center of the Chinese University of Hong Kong at a controlled temperature of 25 °C ± 1 °C, relative humidity 55% ± 5%. A cycle of 12 h light/12 h dark was maintained prior to the experiments.
Cecal ligation and puncture
Polymicrobial sepsis was induced by cecal-ligation and puncture (CLP) [
14]. Under anesthesia with intra-peritoneal injection of ketamine (75 mg/kg) and xylazine (10 mg/kg), a 1-cm midline incision was made on the anterior abdomen. The cecum was exposed and ligated at 50% from the distal end. A through-and-through puncture was performed with a 22-gauge needle to induce sepsis. The cecum was then placed back into the peritoneal cavity. Sham-operated animals underwent abdominal incision and intestinal manipulation with neither ligation nor puncture. All animals were given 1 ml of normal saline by subcutaneous injection and placed on a warm towel immediately after the surgery. The survival rates and septic severity were recorded every 12 h until 7 days after the surgery. No antibiotic was given to the CLP-operated mice in order to assess the systemic inflammation after surgery [
15,
16] Concerning animal welfare, buprenorphine (0.01 mg/kg) was administered to the mice after surgery. Mice were given buprenorphine (0.01 mg/kg) daily until the end of the experiment if necessary.
VD3 prophylaxis
VD3 were purchased from Sigma Chemical Co. (St., Louis, MO). In the water control group, mice were pretreated with water by oral gavage at 48 h, 24 h, and 1 h before CLP. In the VD3 prophylaxis group, mice were pretreated with three doses of VD3 (50 μg/kg) by oral gavage at 48 h, 24 h, and 1 h before CLP. The doses of VD3 used in the present study were referred to as others [
17] .
Treatment with active VD3
1alpha, 25-dihydroxyvitamin D3 (1alpha, 25(OH)2VD3) were purchased from Cayman Chemical Co. (Ann Arbor, MI). Mice were treated with water or 1alpha, 25(OH)2VD3(50μg/kg) for 7 days after CLP by intraperitoneal injection.
Assessment of sepsis morbidity
Septic morbidity was evaluated by Murine Sepsis Severity (MSS) score. Briefly, a score was assigned based on appearance, level of consciousness, activity, response to the stimulus, eyes, respiratory rate, and respiration quality.
Biochemical analyses
Serum alanine transaminase (ALT) and aspartate aminotransferase (AST) levels were determined using Vet Test Chemistry Analyzer (IDEXX) according to the manufacturer’s instructions. Serum vitamin D levels were measured using the vitamin D ELISA kit (#501050, Cayman).
Reverse transcription-quantitative PCR
Total RNA was extracted from ileal tissues by RNAiso Plus reagent according to the commercial protocol (TaKaRa, Japan). For each specimen, a total of 500 ng RNA was reverse-transcribed into cDNA using PrimeScript RT reagent (TaKaRa, Japan). Quantitative real-time PCR was performed with Quantstudio 12 K Flex Real-time PCR system (Life Technologies, Thermo Fisher Scientific, MA, USA) using primers targeting
Muc1,
Muc2,
Muc3,
Muc4,
Cnlp, and
β-actin [
18‐
21].
Histology and immunofluorescence
Harvested ileal tissues were washed briefly in cold phosphate-buffered saline and fixed in Carnoy’s solution (60% ethanol, 30% chloroform, and 10% glacial acetic acid) at 4 °C for 4 h. Fixed tissues were stored in 80% ethanol at 4 °C before tissue processing. Processed sections were stained with Alcian-blue followed by periodic acid Schiff reaction. Expression of cathelicidin was detected in a series of ileal specimens harvested in the acute phase of sepsis. For immunofluorescence, dewaxed and rehydrated slides of murine ileal sections were blocked with 10% bovine serum immunofluorescence buffer (0.1% bovine serum albumin, 0.2% Triton X-100, 0.5% TWEEN 20 in phosphate-buffered saline) and then incubated with mouse mCRAMP (Santa Cruz,1:200) antibodies overnight at 4 °C followed by Alexa Fluor anti-mouse 546 secondary antibodies (1:2000). 4′,6-diamidino-2-phenylindole (DAPI) was used for DNA counterstain. Fluorescent images were captured using a confocal microscope (Leica).
Apoptosis assay
Apoptosis was assessed by an in situ cell death detection kit (Roche Applied Science) and confirmed by immunoblotting using antibodies targeting caspase-3 and cleaved caspase-3.
Intestinal permeability assay and tight junction proteins
Mice were gavaged with 4 kD fluorescein isothiocyanate (FITC)-dextran (500 mg/kg) at 21 h after CLP or sham surgery. After 3 h, blood was collected and the intensity of FITC determined by fluorometry. The expression of tight junction proteins, claudin-1, and occluding was evaluated by immunoblotting.
Profiling of ileal transcriptome
Total RNA was extracted from ileal tissues at 24 h after CLP or sham surgery using RNAiso Plus (TaKaRa, Shiga, Japan). The poly-A RNA was purified and used for library construction. The sample libraries were sequenced with the Illumina HiSeq 2000 sequencing system (Illumina, San Diego, CA, USA). Clean reads were aligned to
Mus musculus primary DNA index files (release-94). Transcripts were then assembled by Cufflinks [
22]. Differentially expressed genes (DEGs) between
Cnlp+/+ CLP and
Cnlp+/+ Sham mice, as well as
Cnlp−/− CLP and
Cnlp−/− Sham mice were identified using edgeR packages. Short Time-series Expression Miner (STEM) software was adopted for the identification of co-expression gene clusters among four groups of mice. The co-expression pattern of particular gene clusters was confirmed and visualized by Pheatmap R package. Pathway analysis was performed with enrichr R package and visualized by ggplot2. Protein–protein interaction network was generated in STRING. The interaction between genes was defined according to “experiments,” “databases,” and “co-expression.” The network topology was analyzed with the “NetworkAnalyzer” plugin in cystoscope.
Intestinal epithelial cell isolation
The small intestine was prepared by cutting the gut about 1 cm downstream from the stomach and 1 cm upstream from the cecum. Forceps were used to remove Peyer’s patches and the attached mesenteric fat carefully. The small intestine was then placed into a 50 mL conical tube containing 30 mL of CMF HBSS (Hank’s balanced salt solution with phenol red, Ca2
+, and Mg2
+-free) with 5% FBS and 2 mM EDTA and shook at 250 rpm for 20 min at 37 °C in order to remove epithelial cells and intraepithelial lymphocytes. The intestine was rapidly minced and incubated in 20 mL of pre-warmed collagenase solution (1.5 mg/mL of collagenase VIII and 40 μg/mL of DNase I in CMF HBSS/FBS) with a shaking frequency of 200 rpm for 20 min at 37 °C for digestion [
23].
Flow cytometry
After blocking Fc receptors with anti-mouse CD16/CD32 (BD Biosciences), small intestinal epithelial cells were stained with anti-mouse Ly-6G (BioLegend), anti-mouse F4/80 (BD Biosciences), anti-mouse CD86 (BD Biosciences), anti-mouse CD206 (BD Biosciences), and anti-mouse CD45 (BD Biosciences). The stained cells were analyzed on a FACSCalibur flow cytometer (BD Biosciences). The data were analyzed using FlowJo Software (FlowJo, Ashland, OR). Neutrophils were defined as Ly6G+ cells and macrophages as F4/80+ cells and M1 macrophages as F4/80+ CD86+ and M2 macrophages as F4/80+ CD206+. Lymphocytes were defined as CD45+ cells.
Statistical analysis
Multiple group comparisons were performed by two-way ANOVA or non-parametric Kruskal-Wallis followed by the Tukey’s t test. Mortality was compared by Kaplan-Meier survival curves and analyzed by the log-rank test. P values less than 0.05 were considered statistically significant.
Discussion
Cathelicidin is one of the immunomodulatory proteins involved in the pathogenesis of sepsis [
29]. Clinical studies have shown that human cathelicidin was 50% lower in critically ill patients with severe sepsis compared to non-septic patients and was further downregulated in septic shock [
30]. Stratification of critically ill patients by different levels of plasma cathelicidin revealed that those with less than 116 ng/mL at admission had four-fold increased risk for 90-day mortality as compared to those with cathelicidin > 238 ng/mL, after controlling for confounders, and also more likely to develop sepsis during the same hospital stay [
31]. These suggest that cathelicidin has an important role in sepsis.
In this study, we demonstrated that upon induction of sepsis by CLP in mice, the expression of cathelicidin was increased by four-fold. The increased expression of this peptide was more prominent in the first 4 h upon sepsis induction, indicating that cathelicidin is involved in the acute phase of sepsis. Instead of a sequential occurrence of hyper-inflammation or immunosuppression [
32], recent studies suggested a paradigm shift in sepsis pathogenesis in which both processes persist over the course of the disease, leading to persistent inflammation and catabolism syndrome [
32]. Given the anti-inflammatory properties of cathelicidin, its gradual decrease at a late stage after CLP in mice may explain why wild-type mice would die at a later stage. The protective role of cathelicidin was also confirmed in survival analysis between
Cnlp wild-type and knockout group. Consistently, human cathelicidin protects rats against sepsis after bacterial challenge [
33] and the increased expression of cathelicidin in adipocytes surrounding the colon limits the release of bacteria from mice with experimental colitis [
34]. Nevertheless, contradictory evidence also exists in the literature. Severino et al. reported that wild-type C57BL/6 mice succumbed more rapidly to CLP compared with cathelicidin-deficient mice [
35]. The discrepancies between this report and our study might arise from the genetic backgrounds of mice (129/SVJ and C57BL/6, respectively). In this regard, mice of different genetic backgrounds could exhibit divergent antimicrobial activity [
36].
Along with the change of cathelicidin expression as revealed by real-time PCR and immunostaining, there were signs of intestinal barrier dysfunction including heightened permeability to fluorescein dextran, reduced mucin production, lowered tight junction protein expression, and increased apoptotic activity. The bacterial load in the blood also became higher after the induction of sepsis. These conditions were further exaggerated in the cathelicidin-knockout mice, whose survival period was significantly shortened after CLP. These confirmed the protective role of cathelicidin in preserving gut barrier function in sepsis.
Mucins are structural components of mucus, which lines the gastrointestinal mucosa, and are important in preventing harmful microbes from entering the bloodstream [
37]. The expression of various mucin genes differs upon encountering microbial challenges. Of note,
Muc1 is increased considerably after infection [
37], a finding that is in agreement with our observation that
Muc1 and
Muc2 genes were upregulated after induction of experimental sepsis. The magnitude of expression was reduced after knocking out cathelicidin. Although the mechanism of cathelicidin in control of mucin production remains unclear, the administration of exogenous cathelicidin to rats has been shown to increase the thickness of the mucus layer in the intestine [
38].
Apoptosis and tight junction alterations are important mechanisms through which intestinal microbes invade the hosts [
39]. In our study, we observed higher activity of apoptosis after CLP. This was further exaggerated after knocking out cathelicidin, an antimicrobial peptide that inhibits kidney cell apoptosis by reducing the endoplasmic reticulum stress [
40]. The disruption of gut barrier integrity may partially explain the higher bacterial load seen in the cathelicidin-knockout group.
It has been reported that cathelicidin improves septic mice survival by inhibiting pyroptosis of macrophages and preventing exaggerated inflammatory responses [
41]. Consistent with this finding, our transcriptome analysis of ileal tissues revealed that expression of inflammatory genes (
Grb2,
Rela,
Jun) were shown as the most popular hub genes (interaction degree large than 20) in the upregulated gene cluster. An increased intestinal inflammatory response has been shown to be associated with gut barrier dysfunction in rodents [
42]. Collectively, these suggested that cathelicidin depletion would exaggerate pro-inflammatory response, which was also verified by the KEGG and Reactome pathway analyses. Further mechanistic studies will be needed to determine if cathelicidin controls pro-inflammatory response via
Grb2,
Rela, and
Jun.
It was demonstrated that human cathelicidin synergistically enhanced the endogenous inflammatory mediator interleukin-1β and chemokines such as macrophage chemoattractant proteins in human peripheral blood mononuclear cells [
43]. M1 macrophages can rapidly kill pathogens to help the primary host defense, which mainly play a role in pro-inflammation, and M2 macrophages routinely repair and maintain tissue integrity, which serve an anti-inflammatory function [
44]. In our study, we observed a dramatic M1-to-M2 shift in the small intestine after CLP and depletion of cathelicidin tended to induce more M1 but not M2 macrophages compared with the wild-type mice after CLP. So 24 h after CLP, the immune state of mice seems immunosuppressive with macrophages polarizing to an M2 phenotype. Given that human cathelicidin directs macrophages differentiation toward proinflammatory macrophages [
45], depletion of cathelicidin may lower the pro-inflammation response in the immune environment during CLP. Apart from modulating the function of macrophages, cathelicidin can induce the migration of neutrophils and eosinophils by the formyl-peptide receptor, FPR2 [
46]. In our study, CLP induced infiltration of neutrophils into small intestine and depletion of cathelicidin exaggerated neutrophil infiltration compared with the wild-type mice after CLP. Apart from CLP-induced infiltration of macrophages and neutrophils, we examined the infiltration of lymphocytes into the small intestine. Results demonstrated that there is no significant difference between CLP groups and sham groups. In line with our study, two clinical studies reported that there were no significant differences in T cell and B cell populations between septic patients and the corresponding control group [
47,
48]. Collectively, these suggested that CLP would induce more infiltration of macrophages and neutrophils into the small intestine. Cathelicidin depletion would exaggerate pro-inflammatory response, which was associated with elevated production of neutrophils and M1 macrophages.
Parekh and colleagues analyzed the patient data of 61 patients with sepsis and utilized the CLP model, demonstrating that sepsis and severe sepsis are associated with vitamin D deficiency, which in turn is associated with more severe sepsis [
49]. Accumulating evidence suggests that VD3 exerts protective effects during infections by up-regulating the expression of cathelicidin and beta-defensin 2 in phagocytes and epithelial cells [
11]. In our study, we observed that VD3-pretreated mice had better survival after CLP and these mice also recovered faster with a better MSS score. Along with increasing of mucin1 expression, there were signs of upregulation of cathelicidin with VD3 pretreatment as revealed by real-time PCR and immunostaining. The bacterial load in blood became lower in mice after induction of cathelicidin with VD3. These confirmed that VD3 could up-regulate the cathelicidin and protect against sepsis.
Furthermore, we assessed the therapeutic use of the active form and the inactive form of VD3 in our CLP model. We observed that administration of calcitriol (an active form of VD3) but not cholecalciferol (an inactive form of VD3) after the onset of sepsis led to a better survival outcome in CLP mice. In line with recent publications, high-dose VD3 (cholecalciferol, inactive form of VD3) did not improve the survival outcomes of critically ill patients in terms of 90-day mortality [
50]. Since hepatic cytochrome P450 (CYPs) play an essential role in the conversion of VD3 into 25-hydroxyVD3 together with additional evidence showing that hepatic CYPs dysfunctions are linked to sepsis [
51‐
53], we further examined the functions of the liver after the onset of sepsis. Our results demonstrated that CLP induced hepatic damage and the associated downregulations of hepatic CYPs at mRNA level, resulting in decreased serum intermediate and active VD3. Fortunately, the administration of calcitriol (an active form of VD3) can bypass hepatic biotransformation of cholecalciferol into 25-hydroxyVD3 mediated by CYP system, directly entering the circulatory system and exerting the beneficial effects. Taken together, we confirmed that the active form of VD3 but not the inactive form of VD3 is a therapeutic drug in our CLP model. Noticeably, the latter worsened 7-day mortality and the associated symptoms in CLP-operated mice, the mechanism of which remains unclear.
This study has potential limitations. First of all, the sample size in our survival analyses is relatively small (n = 8–11). Moreover, only male mice were used for behavioral studies, considering the lesser influence of sex hormones on male mice during the estrous cycle. Our results may not be directly applicable to females. Last but not the least, no antibiotics were given to the CLP-operated mice for all experiments, which might undermine the direct extrapolation of our research findings into the clinical settings.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.