Specific pathogen-free male C3H/HeN (wild-type, WT), weighing between 20 and 25 g were obtained from the National Laboratory Breeding and Research Center (NLBRC, Taipei, Taiwan). C3H/HeJ (TLR4 mutant) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). C3H/HeJ mice have been demonstrated to have a missense mutation in the third exon of TLR4, yielding a nonfunctional TLR4 . All animal procedures were in compliance with regulations on animal used for experimental and other scientific purposes approved by the National Sun Yat-Sen University Animal Experiments Committee.
To evaluate the role of commensal microflora on thermal injury-induced intestinal barrier dysfunction, WT mice were fed with vehicle or oral antibiotics for 4 wks to deplete the intestinal commensals with or without LPS supplements in drinking water (10 μg/μl) at week 3. Wild type (WT) mice were randomly divided into four sham groups (control, LPS, antibiotics, antibiotics + LPS) (n = 6) and four burn groups (control, LPS, antibiotics, antibiotics + LPS) (n = 6 in each). The sham group was subjected to sham treatment and the burn groups were subjected to a 30 - 35% total body surface area (TBSA) burn injury. All animals received sterile saline (50 ml/kg i.p.) for fluid resuscitation right after burn or sham treatment. At 24 hr after burn, mesenteric lymph nodes were harvested for bacterial translocation assay. Also, the distribution of fluorescein isothiocyanate-dextran (FITC-dextran) across the lumen of small intestine in animals under anesthesia (ketamine and xylazine) was measured at 24 hr after injury to assess the intestinal permeability. In another experiment, the GSH level of the intestinal mucosa in animals with the same quantity and treatment was measured to assess the peroxidation produced after injury. Mid-ileum tissues were harvested for TLR4 immunohistochemical studies. In another experiment, the intestinal mucosa was harvested for NF-κB DNA-binding activity, TLR4 mRNA and protein expression assay at 8 hr after burn injury.
To evaluate the role of commensal microflora on thermal injury-induced neutrophil deposition and cytokines expression in lung, WT mice were randomly divided into four sham groups (n = 6) and four burn groups (n = 6 in each) as experiment 1. The animals were sacrificed at 8 hr after burn, and lung tissue was harvested for MPO activity. In another experiment, the lung tissue was harvested for the assay of TLR2, TLR4, and TNFα mRNA expression at 8 hr after injury.
Most major burn patients suffered from ileus and received combined antibiotics treatment to prevent sepsis [15, 16]. To evaluate the effect of antibiotics treatment with or without LPS supplement in the thermal injury-induced bacterial translocation, WT mice were randomly divided into one sham burn groups (n = 6) and three burn groups (burn, antibiotics, antibiotics + LPS) (n = 6 in each). The sham group was subjected to sham treatment and oral saline feeding. The burn group was subjected to burn treatment and oral saline feeding. The antibiotics group was subjected to oral antibiotics administration after burn. The antibiotics + LPS group was subjected to LPS supplements (10 μg/μl) in oral antibiotics administration after burn. At 48 hr after burn or sham burn, peritoneal cells as well as bone marrow cells were harvested for bacterial killing activity, mesenteric lymph nodes were harvested for bacterial translocation, and intestinal mucosa was harvested for TLR4 mRNA assay.
C3H/HeJ mice were randomly divided into three sham groups (n = 6) and three burn groups (n = 6 in each) as experiment 1. At 8 hr after thermal injury, lung was harvested for TLR4 mRNA assay, intestinal mucosa was harvested for TLR4 mRNA assay, and peritoneal cells were harvested from the abdominal cavity for bacterial killing activity and TLR4 as well as TNFα mRNA expression assay.
The thermal injury procedures were modified from those described by Walker et al. Briefly, animals were anesthetized intraperitoneally with ketamine (80 mg/kg) and xylazine (10 mg/kg), and a marked area of the shaved dorsal skin was exposed from a wooden template and immersed in 95°C water for 10 sec. This procedure produced a 30 - 35% TBSA burn of the mice. Total body surface area was calculated using murine-specific data  and average 40 to 48 cm2 for mice of the weight used. The burn injury caused 8% mortality within the first 4 hrs after burn. Nonsurviving animals were excluded from the subsequent study. The sham control animals were anesthetized, shaved and maintained in identical settings except that room temperature water was used for immersion.
Depletion of gut commensal microflora and reconstitution of commensal-depleted animals with TLR ligands
Commensal bacterial products have been known to engage TLRs and confer protection against dextran sulfate sodium (DSS)-induced intestinal epithelial injury . Animals were provided ampicillin (A; 1 g/L; Sigma), vancomycin (V; 500 mg/L; Abott Labs), neomycin sulfate (N: 1 g/L; Pharmacia/Upjohn), and metronidazole (M; 1 g/L; Sidmack Labs) in drinking water for four weeks. Previously, a four-week oral administration of vancomycin, neomycin, metronidazole, and ampicillin with the same dose described above in mice has been proved to deplete all detectable commensals . Previously, this oral antibiotics protocol has no significant effect on nutrition and systemic effect [13, 18]. To those animals receiving LPS, drinking water was supplemented with 10 μg/μl of purified E. coli 026:B6 LPS (Sigma) at week 3 and continued in drinking water for the duration of sham treatment or thermal injury. LPS, a membrane constituent of gram-negative bacteria, was the best-studied TLR ligand and was recognized by TLR4 and MD-2, a molecule associated with the extracellular domain of TLR4 .
Quantification of intestinal permeability
The assay of intestinal permeability was modified from the method described by Otamiri et al. . A 5-cm segment of the jejunum and proximal ileum was dissected with the beginning at 5 cm distal to the ligament of Treitz with well protected superior mesenteric vessels. The bilateral end of the isolated intestine was clamped with rubber bands to prevent the leakage of FITC-dextran. 200 μl of 0.1 M phosphate buffer saline (pH 7.2) containing 25 mg of FITC-dextran (MW 4,400, Sigma) was injected into the lumen. After 30 min, blood sample (100 μl) was taken by a puncture of the portal vein and immediately diluted with 1.9 ml of 50 mM Tris (pH 10.3) containing 150 mM NaCl. The diluted plasma was centrifuged at 4°C, 3,000 g for 7 min and the supernatant was analyzed for FITC-dextran concentration with a fluorescence spectrophotometer (Hitachi, F-2000) at the excitation wavelength of 480 nm and the emission wavelength of 520 nm. Standard curves for calculating the FITC-dextran concentration in the samples were obtained by diluting various amounts of FITC-dextran in a pool of mice plasma, then diluted and centrifuged in the same manner as the samples before measurement.
Determination of glutathione (GSH) level
The intestinal mucosa glutathione level was quantitated by the fluorescence probe o-phthalaldeyde (sigma) which can react with GSH and has high quantum yield. Mix 1.89ml of 50 mM potassium phosphate buffer (pH8.0) with 10ul of supernatant obtained and add 100 μl of 1mg/ml o-phthalaldeyhyde (freshly prepared in absolute methanol). The samples were incubated at room temperature for 15 min and fluorescence was measured at an excitation wavelength of 350 nm and an emission wavelength of 420 nm. The data was expressed as GSH content (mM).
Bacterial translocation to MLN
The collected mesenteric lymph nodes were weighed and homogenized in 500 μl of sterile saline. Aliquots of the homogenate from each tissue were plated onto TSB (Tryptic Soy Broth) agar plates (DIFCO, Detroit, Michigan, USA). The plates were examined after aerobic incubation at 37°C for 24 hr to determine whether commensal depletion with or without LPS altered burn-induced translocation. Representative colonies were expressed as colony forming unit per gram of organ tissue (CFU/g tissue).
Determination of lung myeloperoxidase activity
Lung content of myeloperoxidase (MPO) was determined to assess the degree of pulmonary neutrophil infiltration . Mice were anesthetized and the thorax was opened with median sternotomy. The bilateral lungs and heart were harvested together and the pulmonary vasculature was cleared of blood by gentle injection of 10 ml sterile saline into the right ventricle. The lungs were then blotted dry of surface blood and weighed.
Lung tissues was placed in 50 mM potassium phosphate buffer (pH 6.0) with 0.5% hexadecyltrimethylammonium bromide and homogenized. The homogenate was sonicated on ice and centrifuged for 30 min at 3,000 g, 4°C. An aliquot (0.1 ml) of supernatant was added to 2.9 ml of 50 mM potassium phosphate buffer (pH 6.0) containing 0.167 mg/ml of O-dianisidine and 0.0005% hydrogen peroxide . The rate of change in absorbance at 460 nm was measured over 3 min. One unit of MPO activity was defined as the amount of enzyme that reduces 1 μmole of peroxide per min and the data were expressed as units per gram of lung tissue (Units/g tissue).
Polymerase chain reaction (PCR) and quantification of PCR product
Total RNA was isolated from cells using TRIZOL reagent (Invitrogen, Life Technologies) as described previously . Reverse transcription-generated cDNA encoding TLR2, TLR4, and TNFα genes were amplified using PCR. Sets of primers were designed according to those genes documented in GenBank. The sequences are 5'-AGTGGGTCAAGGAACAGAAGCA-3' (sense) and 5'-CTTTACCAGCTCATTTCTCACC-3' (antisense) for TLR4, 5'-TCTGGGCAGTCTTGAACATTT-3' (sense) and 5'-AGAGTCAGGTGATGGATGTCG-3' (antisense) for TLR2, 5'-CAGCCTCTTCTCATTCCTGCTTGTG-3' (sense) and 5'-CTGGAAGACTCCTCCCAGGTATAT-3' (antisense) for TNFα, and 5' GTGGGCCGCTCTAGGCACCA3' (sense) and 5' CGGTTGGCCTTAGGGTTCAG3' (antisense) for β-actin gene as a control.
Bacterial killing activity of peritoneal cells and bone marrow cells
The peritoneal cavity was washed with 5 ml PBS containing 0.1% BSA and 10 mM EDTA. The peritoneal cells were collected and resuspended in HBSS as 106 cells/ml. Bone marrow cells were harvested from bilateral femoral and tibial bone marrow. Red cells depletion was performed using erythrolysis. After 5 min of preincubation, the cell suspension was incubated with E. coli (108/ml) at 37°C for 1 h with shaking. The cells were removed as the pellet after centrifugation at 200 × g for 10 min, and E. coli number in the supernatant was counted [24, 25].
Protein levels of TLR4 of intestinal mucosa were measured by Western immunoblotting. Homogenized samples (50 μg of protein each) were subjected to 12.5% SDS-PAGE under reducing conditions. Proteins were transferred onto PVDF membranes (Millipore) by using a Semi-Dry Electrophoretic system (Bio-Rad). The TLR4 was identified by rabbit monoclonal antibody (Cell Signaling Technology, Inc.). The membranes were incubated with the secondary antibody (Biotinylated anti-rabbit IgG) (Perkin-Elmer Life Science, Boston, USA) for 1 hr at room temperature. Blots were developed by the ECL Western blotting detection reagents (Perkin-Elmer).
The following antibodies were used for immunohistochemical stains: Rabbit monoclonal immunoglobulin G to TLR4 (Cell Signaling Technology, Inc.), biotinylated secondary antibodies and peroxidase-conjugated streptavidin (Dako). The TLR4 antibody with a 1:300 dilution was used. Sections of the paraffin fixed mice lung (4 μm thickness) were deparaffinized with xylene and graded ethanol. For antigen retrieval in the TLR4 staining, sections were soaked in a citrate buffer containing NP-40 (pH 6.0, Sigma) and heated in a microwave oven (600W) for 10 min. Endogenous peroxidase was blocked with 2% hydrogen peroxide in 70% methanol for 10 min at room temperature. Sections were incubated with primary antibodies for 2 hr, biotinylated secondary antibodies for 20 min at room temperature, and then subsequently processed by the avidin-biotin peroxidase complex method with 3-amino-9-ethylcarbazole (AEC) as the chromogen. Sections were lightly counterstained with Mayer's hematoxylin and viewed under a light microscope. Negative control sections were also incubated, but without primary antibody.
Electrophoretic mobility shift assay for NF-κB and AP-1
Nuclear extracts were prepared as described . Intestinal mucosa were harvested in hypotonic buffer and pelleted by centrifugation. The pellets were suspended in nuclear extract buffer. After 15 min on ice the suspensions were centrifuged and the supernatants were transferred to new tubes. The Bandshift kit (Promega Corp. Madison, WI) was used according to the manufacturer's instructions. Consensus and control oligonucleotides (Santa Cruz Biotechnology Inc.) were labeled by polynucleotides sequences included the AP-1 consensus (5' to 3') (CGCTTGATGACTTGGCCGGAA) or the NF-κB consensus(5' to 3') (AGTTGAGGGGAC-TTTCCCAGGC) (1.75pmol/μl). After the oligonucleotide was radiolabeled, 5 μg of nuclear protein was incubated with 2 μg of poly(dI-dC) and 5,000-10,000 cpm of γ[32P]-ATP-labeled oligonucleotides. After 30 min at room temperature, the samples were analyzed on a 4% polyacrylamide gel. The gel was dried and visualized by autoradiography.
Values are expressed as means SEs. Intergroup comparisons were made using one-way ANOVA followed by Bonferroni correction. Statistical analysis was performed on Prism software (GraphPad). Data were expressed as mean ± standard deviation of the mean in all figures, and p < 0.05 is considered to be statistical significance. The bacterial count, MPO activity, FITC, glutathione level, and bacterial translocation between groups were assessed with one-way analysis of variance (ANOVA), followed by Scheffe's F test.