Detailed Material and Methods
Mice
C57/BL6 (wild type, WT) and congenic B6.129P2-Nos2tm1Lau/J (iNOS KO,Jackson Laboratory, Bar Harbor, ME, U.S.A.) mice were randomized to caecal ligation and perforation (CLP) or sham operation, in accordance with the standard operating procedures of the Department of Comparative Medicine, University of Toronto, Canada.
Caecal Ligation and Perforation (CLP)
Our objective was to create a model of haemodynamically stable sepsis.Based on pilot data (n= 22WT), perforating the cecum twice with a 22G needle resulted in mild hypotension and low 48 h mortality (<10%). Mice were anaesthetised (0.1-0.5% inhaled isoflurane in 100% oxygen). The peritoneal cavity opened via a 7mm midline incision. The caecum was identified, ligated at its midpoint, and punctured twice. Afterexpressing a small amount of fecal material it was returned to the abdomen, and the abdomen was closed. Buprenorphine 1g (0.05-0.1 mg/kg) in 0.5ml sterile 0.9% saline was administered subcutaneously in the scruff of the neck, at the end of the procedure. Sham operated animals underwent identical procedure except that the caecum was mobilizedonly (not tied and not punctured). Fluid and analgesia were as above.Mice were returned to their cages with free access to food and water and 0.5 ml 0.9% saline was administered subcutaneously in the scruff of the neck every 8 hours.
Hemodynamic Measurements
48 hours after surgery, mice were anaesthetised (intraperitoneal 100mg/kg ketamine and 10 mg/kg xylazine). Mice were placed on a warming pad (37°C) under a dissecting microscope. The right common carotid artery was cannulated (Millar Mikro-Tip pressure transducing catheter: 1.4F sensor, 2F catheter; HoustonTX). In vivo pressure tracings from the aorta and left ventricle were recorded (SonoLAB software; Sonometrics Corp., LondonOntarioCanada) and analysed using Cardiosoft and Origin 6.0 (Sonometrics Corp., and Microcal Software, NorthamptonMA). The heart was removed, emptied of blood, and snap frozen.
Biochemical Measurements
Blood was collected from the carotid artery at the conclusion of the experiment for chemistry analysis, serum creatinine, serum creatinine, aspartate aminotransferase [AST] and alkaline phosphatase [ALP](Vita-Tech Canada Inc, Markham, Ontario, Canada).
Histology
Whole hearts from 3 animals/group were stored in 4% formalin and sent for routine staining with Hematoxylin and Eosin (H & E). H&E sections (10 per animal) were examined by a single investigator blinded to the treatment status of each animal. The degree of myocardial injury was assessed using an adapted arbitrary myocardial injury scoring system (14), as follows: grade 0, no lesions; grade 1, focal areas of myocardial edema; grade 2, focal lesions extending over a wider area of myocardial edema associated with cellular gaps and myocardial fiber disruption, ; grade 3, confluent lesions of myocardial edema, focal areas of necrosis, and cellular infiltration; and grade 4, confluent lesions throughout the heart, gross cellular necrosis, cellular infiltration, fiber disruption and mural thrombi.
RNA Isolation and Microarray
Total RNA from whole hearts was isolated using TRIzol Reagent (Life Technologies, Rockville, MD and Invitrogen, Burlington, Ontaio, Canada), and purified using RNAeasy kit (Qiagen, Mississauga, Canada and Qiagen, Chatsworth, CA) as per manufacturers specifications.Total RNA from 3 WT sham operated and 6 WT CLP animals was hybridized to MOE430A (22,000 expressed sequences tags [ESTs]). RNA from 3 iNOS KO sham operated and 3 iNOS KO CLP animals was hybridized to MOE430A2.0 (45,000 ESTs includes all probes from MOE430A), as per manufacturers specifications. Complete array data set and experimental protocol was submitted to the NationalCenter for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) according to MIAME standard (GSE 9667).
Microarray Analysis
Probe based analysis using Robust Microarray Analysis (RMA, and quantile data normalization (per gene and per chip) was performed in MeV 4.0 of TM4 (Microarray software suite: filters were applied to Log2 data, based on normalized raw intensity values (excluded ESTs with raw intensity values 50 and absent in overhalf [8 out of 15] of the experimental conditions). Summarization was performed by median polish. For analysis we included 15,525 ESTs that passed quality filters. Multi-class analysis was performed in SAM (significance analysis of microarray)(15). Functional predictions were performed with Ingenuity (Ingenuity Systems, Inc.Redwood City, CA).
Gene Set Enrichment Analysis
Gene Set Enrichment Analysis (GSEA, was used to detect coordinated expression within treated samples of a priori-defined groups of genes (16-20). Gene sets are available from Molecular Signatures DataBase (MolSigDB, Briefly, GSEA calculates an enrichment score (ES) that reflects the degree to which a gene set is overrepresented at the extremes (top or bottom) of the entire ranked list of microarray data – where genes are ranked according to the expression difference (signal/noise ratio) between exposure to CLP-induced myocardial depression or not. The ES is calculated by walking down the list, increasing a running-sum statistic when it encounters a gene that is in the gene set and decreasing it when it encounters genes that are not. The magnitude of the increment depends on the correlation of the gene with the phenotype (i.e. myocardial depression ‘yes’ or ‘no’). The enrichment score is the maximum deviation from zero encountered in the random walk; it corresponds to a weighted Kolmogorov–Smirnov-like statistic.
The software then estimates the statistical significance (nominal P value) of the ES by using an empirical phenotype-based permutation test procedure that preserves the complex correlation structure of the gene expression data. Specifically, the software permutes the phenotype labels and recomputes the ES of the gene set for the permuted data, which generates a null distribution for the ES. The empirical, nominal P value of the observed ES is then calculated relative to this null distribution. Importantly, the permutation of class labels preserves gene-gene correlations and, thus, provides a more biologically robust assessment of significance than would be obtained by permuting genes. Finally, when an entire database of gene sets is evaluated, GSEA adjust the estimated significance level to account for multiple hypothesis testing. GSEA first normalizes the ES for each gene set to account for the size of the set, yielding a normalized enrichment score (NES). It then controls the proportion of false positives by calculating the false discovery rate (FDR) corresponding to each NES. The FDR is the estimated probability that a set with a given NES represents a false positive finding; it is computed by comparing the tails of the observed and null distributions for the NES. A total of 22,232 probes were imported into GSEA. GSEA was run according to default parameters: collapses each probe set into a single gene vector (identified by its HUGO gene symbol), permutation number = 1000, and permutation type = “gene-sets”. Calculation of the false discovery rate (FDR) was used to correct for multiple comparisons and gene set sizes (21). Sepsis induced transcriptional profiles were analyzed separately for each strain.
Real-Time Quantitative Reverse Transcription–Polymerase Chain Reaction
RT-PCR and qRT-PCR was preformed as previously described (22) using RNA extracted from hearts not used for microarray (biological replicates).
Western Blotting
Tissue lysates,from animals not used for microarray, were used for Western blotting as previously described (23).
Statistical Analysis
Data were analyzed using SPSS v11.0 (SPSS Inc, ChicagoIL). Mean values for continuous variables were compared using Student’s 2-tailed t test for 2 independent samples. Mortality rates were compared using Fisher’s exact test. All clustering (of physiological and microarray data) was performed using Euclidean distance. Linkage method was “complete clustering” in JMP Statistical Discovery Software ( Power and sample size calculations were performed in v2.1.30 ( Results for each genotype were analyzed separately then compared. Treated/control ratios were tested for deviation from unity by calculation of confidence limits. Mean values were compared by 1-way ANOVA followed by a Dunnett test when appropriate, and by Student’s t test where appropriate. P<0.05 was accepted as statistically significant.