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

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.

PE-10 tubing (Fisher Scientific, Waltham, MA) was inserted into the common bile duct. The cannula was temporarily fixated with an 8-0 prolene suture, and a micro-clamp was placed at the proximal hepatic duct. The pancreas was retrograde infused for 10 min with either 50 ul saline (control mice) or 4% sodium taurocholate hydrate (pancreatitis mice) with 1% methyl-blue (Sigma-Aldrich, Saint-Louis, MO).
The cannula, micro-clamp and ligature were removed and the puncture hole closed with a single 8-0 prolene suture. The abdominal wall and skin were closed with, respectively with 6-0 vicryl and 5-0 nylon interrupted sutures. Post-operatively, injections of bupinorphrine were given every 12 hours for at least 48 hours. All procedures were performed under sterile conditions according to local guidelines and policies.
BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)

Bacterial dissemination
Blood and organ tissues (pancreas, liver, spleen, mesenteric lymph nodes and lung) were weighted, homogenized in saline and 10-fold serial dilutions were plated on selective media for gram-positive (Colombia CNA with 5% sheep blood) and Gram-negative bacteria (Difco™ MacConkey Agar, Becton Dickinson, Franklin Lakes, NJ). Colony forming units were counted after 24 hours of incubation at 37 degrees Celsius and normalized by volume (blood) or weight (tissue). Positive blood and pancreas tissues culture were sent to the Clinical Microbiology laboratory for identification using a Vitek MS (bioMérieux, Inc., Durham, NC). Isolates were cultured in liquid TSB medium and stored in 10% glycerol at -80 degrees Celsius.

Histology
Pancreatic tissues at the pancreatic head were fixated in formalin for 24 hours following sacrifice. The slides were stained with an H&E staining and scored for parenchymal necrosis and inflammation according to criteria of Schmidt by an experienced pathologist. 1

Serum laboratory markers
Amylase was detected in serum using the UCSC Clinical Laboratory via routine methods. Serum interleukin-6 (IL-6), interleukin-1b (IL-1b) and Tumor necrosis factor alpha (TNFα) were measured using

Bacterial community Phenotype Microarray analysis
The metabolic activity of the intestinal bacterial community was determined as previously described. 2 Briefly, cecal contents were homogenized in saline, centrifuged at 100G for 3 min and the supernatant  The samples were analyzed using a high-resolution accurate mass 7200B QTOF GC-MS Agilent, Santa Clara, CA) with a mass range 35-800, acquisition rate 6 spectra/s. Samples were randomized in batches of 5, and injected at a 25:1 split ratio. A temperature program of 70°C for 5 min was used, ramped at 5°C/min until 320°C, and held for 3 min. A blanc sample containing the solvent (pyridine) was run between each sample to clean the column of any residue. The chromatograms were processed (peakpicking, definition of pseudospectra and annotation) using R (R Core Team, 2014) with the "runGC" function of the metaMS package. 5 The resulting pseudospectra were annotated using the commercial Fiehn GC/MS Metabolomics RTL Library 6 . Features with matching factors ≥60% were positively identified.

16S rRNA gene sequencing of human fecal samples
DNA was extracted from fecal material using a repeated bead beating protocol 7 (method 5). DNA was purified using Maxwell RSC Whole Blood DNA Kit. 16S rRNA gene amplicons were generated using a single step PCR protocol targeting the V3-V4 region. 8 PCR products were purified using Ampure XP beads Forward and reverse reads were truncated to 240 and 210 bases respectively and merged using USEARCH. 9 Merged reads that did not pass the Illumina chastity filter, had an expected error rate higher than 2, or were shorter than 380 bases were filtered. Amplified Sequence Variants (ASVs) were inferred for each sample individually with a minimum abundance of 4 reads. 10 Unfiltered reads were than mapped against the collective ASV set to determine the abundances. Taxonomy was assigned using the RDP classifier 11 and SILVA 12 16S ribosomal database V132.

RNA isolation and RT-qPCR
Cecum tissues used for RNA isolation were collected in RNAlater and stored at -80 degrees Celsius until analysis. 20-30 mg of tissue was homogenized using a Precellys 24 tissue homogenizer (Bertin instruments, Montigny-le-Bretonneux, France). Total RNA was isolated using the NucleoSpin kit Ct values were calculated using LinRegPCR software. 13 Relative gene expression was performed using the delta-delta Ct method with b2-microglobulin (B2M) as reference. 14