Changes in colon gene expression associated with increased colon inflammation in interleukin-10 gene-deficient mice inoculated with Enterococcus species

Abstract

Background: Inappropriate responses to normal intestinal bacteria may be involved in the development of Inflammatory Bowel Diseases (IBD, e.g. Crohn's Disease (CD), Ulcerative Colitis (UC)) and variations in the host genome may mediate this process. IL-10 gene-deficient (Il10-/-) mice develop CD-like colitis mainly in the colon, in part due to inappropriate responses to normal intestinal bacteria including Enterococcus strains, and have therefore been used as an animal model of CD. Comprehensive characterization of changes in cecum gene expression levels associated with inflammation in the Il10-/- mouse model has recently been reported. Our aim was to characterize changes in colonic gene expression levels in Il10-/- and C57BL/6J (C57; control) mice resulting from oral bacterial inoculation with 12 Enterococcus faecalis and faecium (EF) strains isolated from calves or poultry, complex intestinal flora (CIF) collected from healthy control mice, or a mixture of the two (EF.CIF). We investigated two hypotheses: (1) that oral inoculation of Il10-/- mice would result in greater and more consistent intestinal inflammation than that observed in Il10-/- mice not receiving this inoculation, and (2) that this inflammation would be associated with changes in colon gene expression levels similar to those previously observed in human studies, and these mice would therefore be an appropriate model for human CD.

Results: At 12 weeks of age, total RNA extracted from intact colon was hybridized to Agilent 44 k mouse arrays. Differentially expressed genes were identified using linear models for microarray analysis (Bioconductor), and these genes were clustered using GeneSpring GX and Ingenuity Pathways Analysis software. Intestinal inflammation was increased in Il10-/- mice as a result of inoculation, with the strongest effect being in the EF and EF.CIF groups. Genes differentially expressed in Il10-/- mice as a result of EF or EF.CIF inoculation were associated with the following pathways: inflammatory disease (111 genes differentially expressed), immune response (209 genes), antigen presentation (11 genes, particularly major histocompatability complex Class II), fatty acid metabolism (30 genes) and detoxification (31 genes).

Conclusions: Our results suggest that colonic inflammation in Il10-/- mice inoculated with solutions containing Enterococcus strains is associated with gene expression changes similar to those of human IBD, specifically CD, and that with the EF.CIF inoculum in particular this is an appropriate model to investigate food-gene interactions relevant to human CD.

Figure 1

Summary of the colon histological injury score (HIS) in Il10^-/- and C57 mice in response to various treatments. Data represent mean ± SD for 5 mice fed the AIN-76A diet, and housed in specific pathogen free (SPF) or conventional (C) conditions, or housed in conventional conditions and orally inoculated with 12 Enterococcus strains (EF), with conventional intestinal flora from normal C57 mice (CIF), or a 1:1 combination (EF·CIF). Asterisks denote Il10^-/- mice with values significantly different (*P < 0.05, ***P < 0.01) from C57 mice in the same treatment group, while the "†" symbol denotes Il10^-/- mice with values significantly (P < 0.05) different from Il10^-/- mice under SPF conditions for the same histological parameter. One Il10^-/- mouse in the CIF treatment group died during the course of the trial. An autopsy was unable to unequivocally establish the cause of death.

Figure 2

Hierarchical clustering of 20 microarray slides (representing RNA pooled from either two or three mice per slide) and 6,521 probe sets. A heat map of 6,521 transcripts and 2 dendograms that group gene probes (left) and microarray slides (top) together is shown. This represents all differentially expressed probes for the comparison Il10^-/- vs. C57 mice for any of the five treatment groups (SPF, C, EF, CIF, EF·CIF). Each line is a probe, and each column is an array slide. Expression signal intensities are shown in red and blue, indicating high and low expression, respectively. Arrays for Il10^-/- mice are represented in yellow and C57 mice in blue in the bar above the heat map.

Figure 3

Generation of a biological network of genes of the most significant Canonical Pathways for the EF·CIF (Il10-/- vs. C57) comparison. The network was generated by IPA using all molecules from significantly affected Canonical Pathways (Fatty Acid Metabolism; LPS/IL-1 mediated inhibition of RXR function; Tryptophan Metabolism; β-Alanine Metabolism; Valine, Leucine and Isoleucine Degradation; Antigen Presentation Pathway; Interferon Signaling; IL-10 Signaling; Fatty Acid Elongation in Mitochondria). Connections were applied based on known interactions between these genes within the Ingenuity Pathways Knowledge Base. Central genes and their direct interactions were identified and were supported by published information. *Genes that are detected 2 or more times on the array. Genes or gene products are represented as nodes, and the biological relationship between two nodes is represented as a line (i.e. an edge). All edges are supported by at least 1 reference from the literature. Red and green colored nodes indicate genes with up- and down-regulated expression, respectively. The intensity of the colors specifies the degree of up- or down-regulation. Greater intensity represents a higher level of differential expression. Nodes and edges are displayed with various shapes and labels that present the functional class of genes and the nature of the relationships between the nodes, as shown in legend below the figure.

Figure 4

Generation of a biological network of genes of the most significant Canonical Pathways for the Il10-/- (EF·CIF vs. C) comparison. The network was generated by IPA using all molecules from significantly affected Canonical Pathways (Oxidative Phosphorylation; Antigen Presentation Pathway; IL-10 Signaling; Interferon Signaling; LPS/IL-1 mediated inhibition of RXR function; Circadian Rhythm Signaling). Connections were applied based on known interactions between these genes within the Ingenuity Pathways Knowledge Base. Central genes and their direct interactions were identified and were supported by published information. *Genes that are detected 2 or more times on the array. Genes or gene products are represented as nodes, and the biological relationship between two nodes is represented as a line (i.e. an edge). All edges are supported by at least 1 reference from the literature. Red and green coloured nodes indicate genes with up- and down-regulated expression, respectively. The intensity of the colours specifies the degree of up- or down-regulation. Greater intensity represents a higher level of differential expression. Nodes and edges are displayed with various shapes and labels that present the functional class of genes and the nature of the relationships between the nodes, as shown in the legend to Figure 3.

Figure 5

Real time quantitative PCR validation of gene expression results from microarray analysis. Real time quantitative PCR was performed to determine the relative expression of Ifng (a), Sult1a1 (b), Cyp2c40 (c), Ces2 (d) and Ncf4 (e), five genes which showed differential expression (as determined using microarray analysis) in Il10^-/- mice as a result of bacterial inoculation. Results for the differentially expressed genes were normalized against the geometric mean of CANX and HPRT. An asterisk indicates a significant difference in gene expression for the comparison of interest.

Figure 6

Overall study design. Both Il10^-/- and C57 mice were fed AIN-76A diet from the time of arrival (34 ± 3 days of age), and were randomly divided into five treatment groups with five animals per group. One group of mice was housed in SPF conditions, a second group was maintained under conventional conditions (C), while the remaining groups were kept in conventional conditions and orally inoculated (200 μl) with solutions containing either the 12 strains of Enterococcus listed in Table 1 (EF), conventional intestinal flora (CIF) derived from healthy age-matched C57 mice raised under conventional conditions, or a combination of the two (EF·CIF). Inoculation was performed at approximately 5.5 weeks of age, and tissue sampling at 12 weeks of age.