Animal Procedure and Sample Collection
The animal experiment was carried out in accordance with the guidelines and regulations of the Animal Care and Ethical Committee of the Sichuan Agricultural University and was in compliance with the ARRIVE guidelines. The ethical approval for the animal procedure was granted by the Animal Care and Use Committee of the Sichuan Agricultural University (SCAUAC-20200051).
After acclimatisation for a week, 14 male weaned pigs aged 28 d (Duroc × Landrace × Yorkshire) were randomly allocated into two groups, the group challenged by enterotoxigenic Escherichia coli K88 (the ETEC group, n = 7) and the unchallenged group (the CON group, n = 7). The inoculum for the ETEC group is Luria Broth containing ETEC (100 mL, 1×109 CFU/mL, serotype O149: K91: K88ac; China Veterinary Culture Collection Centre, Wuhan, China), and the CON group received 100 mL sterilised Luria broth. The two groups were reared on the same diet and water ad libitum, kept in separate rooms to avoid cross contamination at controlled temperature (28-30˚C). The ingredient composition and nutrient levels of the diet are presented in Table S1.
Self-defined diarrhoea scores (1, normal; 2, pasty; 3, semi-liquid; 4, watery) were recorded at 0, 4, 8, 16, 20, 24, 30 and 36 hs after the ETEC challenge. All pigs were euthanized with an intramuscular injection of 15 mg/kg body weight of pentobarbital sodium under anaesthesia 48 h after the ETEC challenge. Two sections of ileal tissue (2 cm in length) were collected from each pig with one stored in 4% paraformaldehyde solution for histology and the other snap-frozen and stored at -80 °C for proteomic and transcription analyses.
Intestinal Morphology
Ileal tissues were cross sectioned and stained with the periodic acid Schiff method (PAS staining). Ten villi and crypts of each section were measured (Image ProPlus 6.0, company, city, country) under a microscope and the villi-crypt ratio (VCR) was calculated.
Proteomics of the Ileal Tissues
The ileal samples were processed according to the enhanced filter-aided sample preparation (eFASP) protocol 48. Briefly, frozen tissue samples were homogenised using a TissueLyser II (Qiagen, Gaithersburg, MD, USA) in a lysis buffer containing 4% sodium dodecyl sulphate, 0.2% deoxycholic acid, 50 mmol/L dithiothreitol and 100 mmol/L ammonium bicarbonate (pH 8.0). The lysate was incubated at 95˚C for 5 min and centrifuged (16,000 × g, 4˚C, 20min). Protein concentration of the obtained supernatant was determined by a BCA protein quantification kit (Thermo Scientific, Waltham, MA, USA). Supernatant containing 100 µg protein was transferred onto a centrifugal filter (Amicon Ultra, 10 kDa, Millipore, Darmstadt, Germany) and washed twice by mixing with an exchange buffer (8 mol/L urea, 0.2% deoxycholic acid, 100 mmol/L ammonium bicarbonate, pH 8.0) followed by a centrifugation (14,000 × g, 15 min). The protein was reduced by Tris(2-carboxyethyl)phosphine (TCEP, 0.01 mol/L, 1:50 [v/v]), alkylated with iodoacetamide and digested by trypsin (1 µg/100 µg protein, Promega, Madison, WI, USA). Tryptic peptides were recovered and purified by phase extraction using ethyl acetate acidified by formic acid (1%, v/v). Vacuum-dried tryptic peptides were resuspended in 2% acetonitrile with 0.1% formic acid and applied onto a Dionex RSLC UPLC System (Thermo Scientific) coupled to an Orbitrap Fusion Lumos Mass Spectrometer (Thermo Scientific). One µg of peptide was injected onto a 2 cm C18 material-trapping column and separated on an analytical column (Acclaim PepMap100, 75 µm ID, 15 cm, 100 Å, Thermo Scientific) with both columns kept at 40˚C. The peptides were eluted at a stable flow rate of 300 nL/min with a linear gradient from a solution of 2.4% acetonitrile and 0.1% formic acid to a solution of 78% acetonitrile and 0.1% formic acid in 150 min. Mass spectrometric data were obtained in the positive ionization mode in a data-dependent acquisition (DDA) mode. The mass spectra are available at the ProteomeXchange Consortium (proteomexchange.org) with the data set identifier PXD028066.
Protein annotation and quantification were carried out using MaxQuant (version 1.5.2.8) 49 against the Uniprot database (Sus scrofa, UP000008227, last modified 2021-01-29). Detection of at least two unique peptides per protein and protein being present in at least 50% of the samples in each group were required. Protein abundance data were normalized and two-based logarithm transformed using the Perseus software (version 1.6.5.0) 50 before data analysis.
RT-qPCR of Ileal Genes
Transcription levels of selected genes in the ileum were tested by quantitative RT-qPCR using predesigned primers (Table S3). Briefly, tissue RNA was extracted using Trizol Reagent (TaKaRa Biotechnology, Dalian, China) according to the manufacturer’s instructions. The reverse transcription was performed using a cDNA reverse transcription kit (Vazyme Biotechnology, Nanjing, China) and the obtained cDNA was amplified with a SYBR green kit (Vazyme Biotechnology, Nanjing, China) on an ABI-7900HT Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). The transcription levels of target genes were normalised to the housekeeping gene, ACTB, and analysed using the 2 −∆∆Ct method 51.
Full-Length 16S Sequencing of the ileal mucosal microbiome
The full-length 16S sequencing of the ileal mucosal microbiome was performed as previously described 52. Briefly, the ileal mucus was scraped and the bacterial genomic DNA was extracted with a DNA Stool Mini-Kit (QIAGEN, Hilden, Germany). The DNA concentration was estimated on a NanoDrop spectrophotometer (Thermo Scientific). The 16S rRNA gene was amplified by PCR with specific primers. The DNA libraries were constructed on the amplicons with the SMRT Bell technology on a PacBio RS II sequencer (Pacific Biosciences, Menlo Park, CA, USA) 53. The raw reads were processed using Lima software (version 1.11.0) to obtain circular consensus sequence (CCS) reads. Further sample sorting, trimming, clustering of OTUs were conducted with the USEARCH software 54.
Data Analysis
All data analysis were conducted in R 55 integrated with R Studio 56. Body weight, diarrhoea score, intestinal morphology and transcription levels of ileal genes were analysed by student’s t test or Wilcoxon sum-rank test.
The proteomic data were firstly analysed with a semi multivariate method, the NSC analysis to locate proteins associated with the ETEC challenge using the package pamr in R 57. The mount of shrinkage value was determined using a 5-fold cross validation, and the NSC probability analysis with a probability cut-off of 90%. Proteins selected with differentiating power by NSC were further verified by Student’s t test or Chi-squared test. A protein with a p value < 0.05 and absolute effect size > 0.8 was regarded as significantly different between the treatment groups.
For the mucosal microbiomic analysis, OTUs present in at least 50% of all samples or higher than 75% of all counts were selected and a total of 111 were selected out. OTU information and abundance were aligned with treatment groups for data analysis in R. Alpha-diversity shown as Shannon index was calculated and compared between the treatment groups. Beta-diversities based on the unifrac and weighted unifrac distances were tested by permutation test and presented in PCoA plots. Wilcoxon sum-rank test was used to analyse taxon abundance between the treatment groups.
Correlation Analysis of The Microbiomic and Proteomic Data
Correlation between the mucosal microbiomic and proteomic data was firstly conducted by a multivariate correlation analysis, rCCA, with the R package mixOmics 58. The regularising λs for the two datasets were obtained by permutation. Any correlation with the correlation product > 1.0 was selected and were further verified with Spearman correlation analysis. Only the correlations with p < 0.05 and correlation coefficient (ρ) > 0.8 were presented.