Western-type diet influences mortality from necrotising pancreatitis and demonstrates a central role for butyrate

Abstract

Objective: The gut microbiota are the main source of infections in necrotising pancreatitis. We investigated the effect of disruption of the intestinal microbiota by a Western-type diet on mortality and bacterial dissemination in necrotising pancreatitis and its reversal by butyrate supplementation.

Design: C57BL/6 mice were fed either standard chow or a Western-type diet for 4 weeks and were then subjected to taurocholate-induced necrotising pancreatitis. Blood and pancreas were collected for bacteriology and immune analysis. The cecum microbiota composition of mice was analysed using 16S rRNA gene amplicon sequencing and cecal content metabolites were analysed by targeted (ie, butyrate) and untargeted metabolomics. Prevention of necrotising pancreatitis in this model was compared between faecal microbiota transplantation (FMT) from healthy mice, antibiotic decontamination against Gram-negative bacteria and oral or systemic butyrate administration. Additionally, the faecal microbiota of patients with pancreatitis and healthy subjects were analysed.

Results: Mortality, systemic inflammation and bacterial dissemination were increased in mice fed Western diet and their gut microbiota were characterised by a loss of diversity, a bloom of Escherichia coli and an altered metabolic profile with butyrate depletion. While antibiotic decontamination decreased mortality, Gram-positive dissemination was increased. Both oral and systemic butyrate supplementation decreased mortality, bacterial dissemination, and reversed the microbiota alterations. Paradoxically, mortality and bacterial dissemination were increased with FMT administration. Finally, patients with acute pancreatitis demonstrated an increase in Proteobacteria and a decrease of butyrate producers compared with healthy subjects.

Conclusion: Butyrate depletion and its repletion appear to play a central role in disease progression towards necrotising pancreatitis.

Keywords: acute pancreatitis; antibiotics; bacterial infection; butyrate; diet.

Figure 1

Western diet increases mortality and inflammation in mice with ANP. (A) Mouse model of acute necrotising pancreatitis. (B) Serum amylase levels are significantly increased in mice administrated taurocholate in the pancreatic duct (ANP, n=9) compared with vehicle controls (saline, n=9) (p=0.001) or sham-operated mice (n=5) (p<0.001, Mann-Whitney test). (C–C′′) Microscopic (C, C′) and macroscopic (C′′) histological changes of pancreatic tissue of mice administered vehicle (C) or taurocholate (ANP, C′). Parenchymal necrosis (C′, arrow) and extrapancreatic fat necrosis in mice treated with taurocholate (C′′). (D) Histology score adapted from Schmidt’s criteria (p>0.05 by Mann-Whitney test). (E) Kaplan-Meijer curve of groups of mice that received a SD or WD diet prior to induction of necrotising pancreatitis shows increased mortality in the latter group (p<0.001 by log-rank test). (F and G) Systemic inflammation at 24 hours after taurocholate administration (ANP) measured by serum IL-6 (F) and serum TNFα (G). (F) n=5–10, p=0.036 by unpaired t-test). (G) n=5–6 per group, p=0.004 by unpaired t-test). *p<0.05; **p<0.01; ***p<0.001. ANP, acute necrotising pancreatitis; SD, standard chow diet; WD, Western-type diet.

Figure 2

Culture analysis of blood and pancreas in SD-fed and WD-fed ANP mice. Frequency of culture positivity in blood (A) and pancreas tissue (B). MacConkey media were used for Gram-negative and colistin/nalidixic acid (CNA) 5% sheep blood agar for Gram-positive bacteria, n=7–10 per group, p=0.002 blood Gram-positive, p=0.035 blood Gram-negative, p=0.015 for pancreas Gram-negative. Bacterial density (CFU) in blood (C) and pancreas (D) in mice at the lethal endpoint (moribund or 72 hours postoperatively). Significant differences were found for Gram-negative bacteria between WD-fed and SD-fed ANP mice (n=7–10 per group, p=0.013 in blood; p=0.003 in pancreas tissue by Mann-Whitney test). Incidence of identified cultured bacterial species in blood (E) and pancreatic tissue (F). *p<0.05; **p<0.01; ns, not statistically significant. ANP, acute necrotising pancreatitis; CFU, colony forming units; SD, standard chow diet; WD, Western-type diet.

Figure 3

Compositional (16S rRNA gene) analysis of the intestinal microbiota of acute necrotising pancreatitis (ANP) mice. (A) Beta-diversity of cecal content (CE-C) and cecal tissue (CE-T) microbiota measured by weighted UniFrac. A significant difference in microbial composition was observed in Western-type diet (WD)+ANP cecum tissue versus WD+VEH (vehicle) and standard chow diet (SD)+ANP (4–7 mice per group, p=0.003 by Permutation Multivariate Analysis of Variance test). (B) Alpha diversity of CE-C and CE-T microbiota measured by Shannon index. Significant differences were observed in WD+ANP compared with SD+ANP (CE-C: p=0.009; CE-T: p=0.0043 by Mann-Whitney test). No significant differences were observed in WD+ANP versus WD+VEH. (C) Relative bacterial abundance at the phylum level. (D) Relative abundance at the genus level in CE-C and CE-T. (E) Relative abundance of Escherichia/Shigella genus in cecum mucosa (p=0.015 by Mann-Whitney test). *p<0.05; **p<0.01; ns, not statistically significant.

Figure 4

Phenotypic microarray and untargeted metabolomics analysis of intestinal microbiota of acute necrotising pancreatitis (ANP) mice. (A and B) Microbial phenotypic microarray (Biolog) analysis of cecum communities using GENIII plates varied by metabolic substrates. (A) Heatmap of phenotype microarray analysis. There was no statistical difference between standard chow diet (SD)-fed and Western-type diet (WD)-fed ANP mice (p=0.056). When stratified for survival (survivors vs non-survivors), there was a significant increase of metabolic activity in non-survivors (p=0.009 by multiple response permutation procedure). (B) Radial plot of significant carbohydrate substrates showing means with bars representing 95% CI (p<0.05 by Analysis of Variance). (C–E) Untargeted gas chromatography-mass spectrometry (GC-MS) metabolomics analysis of cecum content. (C) Heatmap of differential metabolites between four groups (p<0.05 by Analysis of Variance). (D) Significant test by sparse partial-least squares discriminant analysis (sPLS-DA) showing difference in metabolites of WD pancreatitis mice compared with SD pancreatitis mice (p=0.004) and WD vehicle control mice (p=0.002 by permutations significance test). (E) Univariate analysis shows differential cecal metabolites between SD +ANP and WD +ANP groups (p<0.1 by DESeq analysis). *p<0.05; **p<0.01; ***p<0.001.

Figure 5

Short-chain fatty acids in the cecum of ANP mice. (A–C) Targeted gas chromatography-mass spectrometry analysis of cecum content for butyrate (p<0.016) (A), acetate (p=0.040) (B) and propionate (p>0.05) (C), all by unpaired t-test. *p<0.05; ns, not statistically significant. ANP, acute necrotising pancreatitis; SD, standard chow diet; WD, Western-type diet.

Figure 6

Escherichia coli is a major contributor to mortality in Western-type diet (WD)+acute necrotising pancreatitis (ANP) mice. (A) Experimental design for gut depletion of E. coli. Gram-negative gut depletion (GNGD) was performed two times per day by oral gavage with non-absorbable antibiotics (neomycin 100 mg/L and polymyxin B 10 mg/L). (B) Kaplan-Meier survival curves demonstrating cumulative survival between treatment groups WD+ANP (n=14) and WD+GNGD+ ANP (n=6). The survival was significantly improved with GNGD (p=0.048, log-rank test). (C) Depletion of Escherichia/Shigella in cecum mucosal tissue in five out of six WD+GNGD+ ANP mice, based on 16S rRNA analysis (p=0.132 by Mann-Whitney). (C and D) Culture analysis demonstrates reduction of E.coli bacterial load in blood. Significant differences were found for incidence (n=6, 0=0.007 by Fisher exact test) and CFU (p=0.006 by Mann-Whitney test) of Gram-negative cultures, but not for Gram-positive. *p<0.05; **p<0.01; ns, not statistically significant.

Figure 7

Faecal microbial transplant (FMT) in mice following acute necrotising pancreatitis (ANP) exacerbates bacterial dissemination and worsens survival. (A) Experimental design. Mice fed a Western-type diet (WD) for 2 weeks received either FMT (from healthy mice), sterile faecal filtrate (SFF) or vehicle (VEH) by oral gavage at 0, 24 and 48 hours after ANP induction. (B) Kaplan-Meijer survival curves demonstrating significantly attenuated survival with WD +ANP+ FMT compared with control mice that received SFF (WD +ANP+ SFF) (p=0.047 by log-rank test). (C) Total colony forming units (CFU) were increased in the WD +ANP+ FMT group (p=0.024 by Mann-Whitney test). (D) Incidence of Gram-positive and Gram-negative culture positivity in the pancreas. No significant differences were observed (Fisher exact test). (E) Incidence of cultured bacterial species in pancreatic tissue. *p<0.05.

Figure 8

Oral and systemic butyrate supplementation reduces mortality and bacterial dissemination in Western-type diet (WD)+acute necrotising pancreatitis (ANP) mice. (A) Butyrate was supplemented either preoperatively in the drinking water (BUTDW, 100 mM) for 4 weeks or systemically with intraperitoneal injections (BUTIP, 40 mM, 500 µL) at 0, 7 and 14 hours postoperatively. (B) Kaplan-Meijer survival curves demonstrating significantly reduced mortality with both oral (WD+ANP+BUTDW, n=9, p=0.002) and systemic (WD+ANP+BUTIP, n=14, p=0.019) butyrate supplementation. Survival curves are analysed using log-rank test. (C) Serum endotoxin levels are significantly attenuated by BUTDW (n=5, p=0.003 by Mann-Whitney) and BUTIP (n=5, p=0.012 by unpaired t-test) (D, E) Culture analysis of blood (D) and pancreas (E). p=0.016 of Gram-negative bacteria in BUTIP versus vehicle (VEH) and p=0.028 in BUTDW versus VEH in blood (D). p=0.012 of Gram-negative bacteria in BUTIP versus VEH, and p=0.088 in BUTIP versus VEH in pancreas (E). (F) Expression of tight junction genes indicates increased expression of CLDN1 (n=5, p=0.016) and CHD1 (p=0.032) in the cecum tissue of BUTDW mice versus the untreated control group (WD+ANP). (G) 16S rRNA analysis of relative abundance. Relative abundance at the phylum level demonstrating depletion of Proteobacteria in the BUTDW group to the level of WD+VEH (p=0.008 vs WD+ANP in cecum lumen). Depletion of Proteobacteria was also observed in WD+ANP+BUTIP mice but did not reach statistical significance. *p<0.05; **p<0.01; ns, not statistically significant.

Figure 9

16S rRNA analysis of human samples. (A) Relative abundance of microbiota at the phylum level in faeces of 15 healthy volunteers (HV) and 36 patients with acute pancreatitis (AP) including 26 patients with mild AP (MAP) and nine patients with severe AP (SAP). (B) Proteobacteria was significantly increased in AP compared with HV (p=0.0002). No differences were observed between MAP and SAP (p=0.224, both by Mann-Whitney test). (C) Relative abundance of microbiota at the genus level. Blautia, Akkermansia and Escherichia/Shigella were among the dominating species in MAP while Escherichia/Shigella, Enterococcus and Streptococcus—in patients with SAP. In two patients with SAP, a near monoculture community was observed at the genus level (Enterococcus and Streptococcus) as indicated by arrows. (D) Relative abundance of Escherichia/Shigella was significantly increased in AP compared with HV (p<0.001). No differences between MAP and SAP (p=0.540, both by Mann-Whitney test). (E) Relative abundance of Streptococcus was significantly increased in AP compared with HV (p=0.003), and in SAP compared with MAP (p=0.047, both by Mann-Whitney test). (F) Relative abundance of butyrate producers was significantly decreased in patients with AP as compared with HV (p<0.001). When stratified for severity, no statistical difference was found between MAP and SAP (p=0.100, both by unpaired t-test). *p<0.05; **p<0.01; ***p<0.001, ns, not statistically significant.