SEARCH

SEARCH BY CITATION

Keywords:

  • gnotobiotic IL-10-/- mice;
  • pancolitis;
  • commensal bacteria

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Background: Monoassociating gnotobiotic IL-10-deficient (−/−) mice with either nonpathogenic Enterococcus faecalis or a nonpathogenic Escherichia coli strain induces T-cell-mediated colitis with different kinetics and anatomical location (E. faecalis: late onset, distal colonic; E. coli: early onset, cecal). Hypothesis: E. faecalis and E. coli act in an additive manner to induce more aggressive colitis than disease induced by each bacterial species independently.

Methods: Germ-free (GF) inbred 129S6/SvEv IL-10−/− and wildtype (WT) mice inoculated with nonpathogenic E. faecalis and/or E. coli were killed 3–7 weeks later. Colonic segments were scored histologically for inflammation (0 to 4) or incubated in media overnight to measure spontaneous IL-12/IL-23p40 secretion. Bacterial species were quantified by serial dilution and plated on culture media. Mesenteric lymph node (MLN) CD4+ cells were stimulated with antigen-presenting cells pulsed with bacterial lysate (E. faecalis, E. coli, Bacteroides vulgatus) or KLH (unrelated antigen control). IFN-γ and IL-17 levels were measured in the supernatants.

Results: Dual-associated IL-10−/− (but not WT) mice developed mild-to-moderate pancolitis by 3 weeks that progressed to severe distal colonic-predominant pancolitis with reactive atypia and duodenal inflammation by 7 weeks. NF-κB was activated in the duodenum and colon in dual-associated IL-10−/− × NF-κBEGFP mice. The aggressiveness of intestinal inflammation and the degree of antigen-specific CD4+ cell activation were greater in dual- versus monoassociated IL-10−/− mice.

Conclusion: Two commensal bacteria that individually induce phenotypically distinct colitis in gnotobiotic IL-10−/− mice act additively to induce aggressive pancolitis and duodenal inflammation.

(Inflamm Bowel Dis 2007)

The IL-10-deficient (−/−) mouse is a well-characterized colitis model that develops intestinal inflammation when exposed to normal commensal bacteria, unlike its normal wildtype (WT) counterpart,1 suggesting that endogenously produced IL-10 prevents colitis. IL-10−/− mice maintained in sterile germ-free (GF) conditions remain free of intestinal inflammation and have no evidence of immune activation.2 However, these mice develop chronic progressive colitis (predominantly cecal) after colonization with specific pathogen-free (SPF) bacteria. This colitis is evident within a week after colonization and is severe by 3–5 weeks after initial colonization.2, 3 Of considerable importance, gnotobiotic IL-10−/− mice do not develop intestinal inflammation when monoassociated with Bacteroides vulgatus, which previously has been shown to selectively induce colitis in HLA-B27/β2 microglobulin TG rats4, 5 or Pseudomonas fluorescens, a bacterial species that expresses the I2 superantigen that has been implicated in Crohn's disease (CD).6, 7 We have previously shown that monoassociating IL-10−/− mice with either nonpathogenic strains of E. faecalis or E. coli induce variable disease phenotypes in IL-10−/− mice, as manifested by different kinetics and localization of colitis in E. faecalis- versus E. coli-monoassociated IL-10−/− mice.7E. faecalis induced a slower-onset (10–12 weeks post-inoculation), distal colon-predominant inflammation, which progressed to severe distal colitis and duodenal inflammation by 24 weeks. In contrast, E. coli-monoassociation induced earlier onset (3 weeks postinoculation) cecal-predominant inflammation that progressed to moderately severe to severe disease by 16 weeks. Both organisms induced bacterial species-antigen-specific interferon-γ (IFN-γ) production by mesenteric lymph node (MLN) CD4+ T cells from monoassociated mice. However, neither organism alone induced the level of aggressive disease that we observed in IL-10−/− mice colonized with the complex assortment of commensal bacterial species comprising SPF enteric flora. E. faecalis or E. coli are potentially important contributors within the complex SPF bacterial populations responsible for inducing colitis in this model. These 2 organisms may act additively or synergistically to induce intestinal inflammation, or they may work in conjunction with the other bacterial species found in the SPF environment to induce the rapid-onset, severe colitis that develops in IL-10−/− mice moved into SPF conditions.2, 3 Alternatively, they could have minor roles and the aggressive colitis in SPF IL-10−/− mice is due to other commensal bacterial species yet to be determined.

The goals of the current study were 1) to characterize the disease phenotype when GF IL-10−/− mice are simultaneously colonized with the previously studied nonpathogenic strains of E. faecalis and E. coli, and 2) to quantify immune responses in these dual-associated mice in comparison with WT dual-associated and IL-10−/− mice monoassociated with the same organisms. We hypothesize that dually associating gnotobiotic IL-10−/− mice with E. faecalis and E. coli will act additively or synergistically to potentiate the degree and alter the location of experimental colonic inflammation.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Mice and Bacterial Species

GF IL-10-deficient (−/−) and wildtype (WT) control (both on the inbred 129S6/SvEv background) mice were derived by hysterectomy at the Gnotobiotic Laboratory, University of Wisconsin, Madison. Generation of mice expressing enhanced green fluorescent protein (EGFP) under the control of an NF-κB-dependent promoter has been described.8 These NF-κBEGFP mice (C57BL/6 background) were crossed with IL-10−/− mice (129S6/SvEv background) or WT 129S6/SvEv mice and the 2 lines, designated IL-10−/− × NF-κBEGFP and WT × NF-κBEGFP, were subsequently rederived into GF conditions in the National Gnotobiotic Rodent Resource Center, Department of Laboratory Medicine, University of North Carolina at Chapel Hill (UNC-CH). Mice were monoassociated at 10–12 weeks of age with a human oral isolate of E. faecalis (strain OG1RF provided by Mark Huycke, MD)9 or a murine strain of E. coli that was isolated from WT mice raised in SPF conditions,7 or dually associated with both bacterial strains. Each bacterial strain was streaked on colistin-nalidixic acid (CNA) (E. faecalis) or MacConkey agar plates (E. coli) from glycerol stock solution and incubated for 24 hours. A single colony was picked and used to inoculate 10 mL of brain-heart infusion (BHI) media for each bacterial strain; the inoculated media was incubated for 18–24 hours. Mice were mono- or dual-associated with the bacterial inoculum at log-phase growth by rectal and oral swab. Bacterial mono- or dual-association and the absence of contamination by other bacterial species were confirmed by periodic aerobic and anaerobic culture of stool samples and at necropsy. Antigen-presenting cells (APC) were prepared from WT 129S6/SvEv mice (Taconic Laboratories, Germantown, NY) maintained under SPF conditions free of Helicobacter species. Mono- and dual-associated mice were maintained in the Gnotobiotic Core of the Center for Gastrointestinal Biology and Disease at North Carolina State University (NCSU), College of Veterinary Medicine, and the National Gnotobiotic Rodent Resource Center, (UNC). Animal use protocols were approved by the Institutional Animal Care and Use Committees (IACUC), NCSU and UNC.

Histological Scoring

Mice were killed at 3, 5, and 7 weeks after dual-association or at 3 weeks after monoassociation. Sections of duodenum, cecum, and distal colon were fixed in 10% neutral buffered formalin, embedded in paraffin, stained with hematoxylin and eosin (H&E), and scored in a blinded fashion by a single investigator using a validated scale as previously described.2, 7, 10

Colonic Tissue Fragment Cultures

Colonic tissue fragment cultures were prepared from the colon of each mouse as previously described2, 7 and 0.05 g/well were placed into 24-well plates (Costar 3524) and incubated in 1 mL of complete medium (RPMI 1640 medium supplemented with 5% fetal bovine serum, 50 μg/mL gentamicin, and 1% antibiotic/antimycotic; penicillin/streptomycin/fungizone: Invitrogen, Carlsbad, CA) for 20 hours at 37°C. Supernatants were collected and stored at −20°C prior to use for IL-12/IL-23p40 quantification by enzyme-linked immunosorbent assay (ELISA).

Bacterial Lysates

Lysates of bacteria were prepared from individual colonies of E. coli, E. faecalis, and a guinea pig B. vulgatus (obtained from Andrew Onderdonk, PhD (Harvard University)11 as previously described.7

Determination of Bacterial Colonization of Different Sections of the Intestinal Tract Using Luminal Content Culture

Bacteria were cultured from the cecum and the distal colon of representative IL-10−/− and WT mice euthanized 7 weeks after dual-association with E. faecalis plus E. coli. Cecal and colonic luminal contents were collected into preweighed sterile tubes and serially diluted (at 10−4, 10−6, and 10−8) in sterile phosphate-buffered saline (PBS). These diluted samples were plated on sheep blood agar containing CNA (E. faecalis) and MacConkey agar plates (E. coli) incubated at 37°C for 2 days under aerobic conditions, and levels of bacterial colonization were evaluated by counting colonies as described.7

EGFP Imaging in Dual-associated IL-10−/− × NF-κBEGFP and WT × NF-κBEGFP

Mice were killed 5 weeks after E. faecalis plus E. coli dual-association. Intact intestine (stomach to distal colon) was removed and imaged for EGFP expression using a charge-coupled device camera in a light-tight imaging box with a dual filtered light source and appropriate emission filters (LT-99D2 Illumatools, Lightools Research, Encinitas, CA). Intestinal tissues were then cut open longitudinally, washed briefly in PBS, and placed, lumen side facing the lens, on the stage of a Leica SP2 Upright Laser Scanning Confocal Microscope (Leica, Wetzlar, Germany). EGFP was excited at a wavelength of 495 nm and emission was detected with filters of the appropriate spectrum. The transmitted light channel allowed for special orientation in regards to the mucosal architecture. Image analysis was conducted using the Leica SP2 Laser Scanning Confocal Imaging Software.

Total Colonic RNA Isolation and Real-time RT-PCR Analysis of IL-12/IL-23p40 mRNA

RNA was extracted from colons using TRIzol (Invitrogen), and 1 μg of total RNA was reverse transcribed as previously described.12 Real-time reverse-transcriptase polymerase chain reaction (RT-PCR) was conducted in triplicate using the SybrGreen method and the ABI Prism 7700 sequence detection system (Applied Biosystems, Foster City, CA) as previously described.13 Relative induction of IL-12/IL-23p40 mRNA in IL-10−/− × NF-κBEGFP compared with WT × NF-κBEGFP mice was calculated after normalizing IL-12/IL-23p40 mRNA levels to 18S ribosomal RNA. Primers for IL-12/IL-23p40 mRNA and 18S ribosomal RNA were 5′-GGAAGCACGGCAGCAGAATA-3′ / 5′-aacttgagggagaagtaggaatgg-3′ (amplicon size: 179 bp) and 5′-CGCCGCTAGAGGTGAAATTCT-3′ / 5′-CATTCTTGGCAAATGCTTTCG-3′ (amplicon size: 66 bp), respectively. RT-PCR reaction conditions were: a single hold at 95°C for 10 minutes, followed by 40 cycles of denaturation (15 sec, 95°C), annealing and extension combined (60 sec, 60°C). The specificity of amplification was assured by melting curve analysis. RT-PCR analysis was conducted for total RNA from 3 independent animals per genotype.

CD4+ T-cell Isolation

CD4+ T cells were enriched from MLN cells harvested from IL-10−/− or WT mice by negative selection. B cells and CD8+ T cells were depleted using antibody-coated magnetic beads and cell sorting techniques according to the manufacturer's instructions (Miltenyi-Biotec, Auburn, CA) as previously described.7, 10 Over 95% of the enriched MLN cell populations expressed CD4 as determined by flow cytometry.

APC Preparation and CD4+ T cell Stimulation

Splenic APC were prepared from SPF 129S6/SvEv mice and pulsed overnight with 10 μg/mL of either E. faecalis, E. coli, or B. vulgatus lysate, or keyhole limpet hemocyanin (KLH: Pierce, Rockford, IL) as described previously.7, 10 CD4+ T cells (2 × 105 CD4+ T cells/well) were cocultured for 72 hours with antigen-pulsed APC (3 × 105 APC cell/well) in flat-bottom 96-well cell culture plates (Costar 3595), 0.2 mL per culture. Supernatants were collected after 3 days and stored at –20°C.

Cytokine Measurements

We used commercially available monoclonal antimouse IFN-γ and IL-12/IL-23p40 capture and detection reagents (BD Biosciences Pharmingen, San Diego, CA) in our validated ELISA protocols to measure amounts of IFN-γ secreted by stimulated CD4+ T cells and IL-12/IL-23p40 secreted constitutively by colon tissue fragment cultures.2, 7, 10 For detection of IL-12/IL-23p40, plates were coated with antimouse IL-12 p40/p70 (clone C15.6) and bound ligand was detected with biotin antimouse IL-12 p40/p70 (clone C17.8). For detection of IL-17, plates were coated with antimouse IL-17A, clone eBioTC11-18H10.1 and IL-17 was detected using biotin antimouse IL-17A, clone eBioTC11-8H4 (e-Bioscience, San Diego, CA). Cytokine levels were measured in triplicate supernatants and compared to standard curves generated using recombinant murine cytokines.

Flow Cytometry

MLN cells subpopulations (before and after negative selection) were evaluated as previously described7 on the FACScan (BD Biosciences, Mountain View, CA) using FITC anti-CD4, PE anti-CD8, or FITC anti-B220 (all from Invitrogen).

Statistical Analysis

Histologic scores were analyzed with the Mann–Whitney test (SAS, Cary, NC) by statisticians in the Biostatistics Core of the Center for Gastrointestinal Biology and Disease. The paired Student's T-test was used for all other data. Statistical significance was defined as P < 0.05 for comparisons indicated.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Kinetics of Histologic Disease Progression in E. faecalis Plus E. coli Dual-associated IL-10−/− Mice

We investigated whether dual colonization with E. faecalis plus E. coli induced colitis in gnotobiotic GF inbred 129S6/SvEv IL-10−/− and WT mice. The degree of histologic colitis in the distal colon and cecum was evaluated 3, 5, or 7 weeks after inoculation with both E. faecalis and E. coli (Fig. 1). The dual-associated IL-10−/− mice developed mild cecal inflammation and distal colitis within 3 weeks that was moderate in both regions by 5 weeks. By 7 weeks, the dual-associated IL-10−/− mice had marked pancolitis that was most severe in the distal colon (distal colon: 3.6 ± 0.3; cecum: 2.9 ± 0.4). WT 129S6/SvEv mice, similarly colonized for 7 weeks, had no histological evidence of colitis (distal colon: 0.4 ± 0.1 cecum: 0.7 ± 0.2). The degree of inflammation was increased in dual-associated IL-10−/− compared to E. coli- or E. faecalis-monoassociated IL-10−/− mice (as shown in Table 1). At 3 weeks of bacterial colonization, there was no histologic evidence of inflammation in E. faecalis-monoassociated IL-10−/− mice and only mild cecal-predominant inflammation in E. coli-monoassociated IL-10−/− mice. Cecal and colonic inflammation in the dual-associated IL-10−/− mice was significantly greater than the inflammation noted in the monoassociated IL-10−/− mice. Furthermore, these IL-10−/− dual-associated mice developed pancolitis similar to IL-10−/− mice transferred into SPF conditions.2 These results show that bacterial inoculation with both organisms induces disease that is more aggressive than with each bacterial species individually, as characterized by earlier disease onset and wider anatomical distribution of colitis.

thumbnail image

Figure 1. Kinetics of histologic disease progression in E. faecalis plus E. coli dual-associated IL-10−/− or wildtype (WT) mice. Blinded histological scores (mean score ± SD) of distal colon and cecum are shown from IL-10−/− dual-associated mice at different timepoints (n = 5 mice/timepoint) after initial bacterial inoculation and from WT mice (n = 4) studied 7 weeks after inoculation are shown. *P < 0.02 versus the histological score for both distal colon and cecum from WT mice.

Download figure to PowerPoint

Table 1. Inflammation and Cytokine Production in E. faecalis Plus E. coli Dual- or Monoassociated Mice
a. Mouse Type (wks Colonized)Bacterial ColonizationIL-12/IL-23p40 (ng/50 mg)Distal Colon ScoreCecum Score
  • Results from IL-10-/- E. faecalis plus E. coli dual- (n = 5) or monoassociated mice (E. faecalis n = 5; E. coli n = 9) colonized for 3 weeks are shown; WT dual-associated mice were colonized for 7 weeks (n = 4), correlating with the latest time point studied in IL-10-/- mice. IFN-γ and IL-17 were measured in supernatants of MLN CD4+ T cells collected 3 days after stimulation with APC pulsed with the bacterial lysate indicated per column. IL-12/IL-23 p40 was measured in colon tissue fragment culture supernatants.

  • a

    P < 0.01 versus WT controls.

  • b

    P < 0.02 versus IL-10-/- E. faecalis plus E. coli.

  • c

    P < 0.005 versus IL-10-/- E. faecalis plus E. coli.

IL-10-/- (3 wks)E. faecalis + E. coli1.8 ± 0.1a1.6 ± 0.5a1.9 ± 0.3a
WT (7 wks)E. faecalis + E. coli0.6 ± 0.20.4 ± 0.10.7 ± 0.2
IL-10-/- (3 wks)E. coli0.5 ± 0.06c1.0 ± 0.1c1.4 ± 0.4b
IL-10-/- (3 wks)E. faecalis0.3 ± 0.03c0.7 ± 0.3b0.8 ± 0.5b
b. Mouse TypeBacterial ColonizationIFN-γ (ng/mL) E. faecalis-Pulsed APCIFN- γ (ng/mL) E. coli-Pulsed APCIL-17 (ng/mL) E. faecalis-Pulsed APCIL-17 (ng/mL) E. coli-Pulsed APC
IL-10 -/- (3 wks)E. faecalis + E. coli16.6 ± 3.2a106.6 ± 9.3a1.9 ± 0.4a4.7 ± 0.6a
WT (7 wks)E. faecalis + E. coli1.2 ± 0.211.5 ± 1.80.1 ± 0.020.3 ± 0.04
IL-10 -/- (3 wks)E. coli0.2 ± 0.01c7.6 ± 1.0c0.07 ± 0.02b1.5 ± 0.1b
IL-10 -/- (3 wks)E. faecalis3.5 ± 0.4b1.0 ± 0.3c7.3 ± 0.4c0.9 ± 0.4c

Representative histologic sections from dual-associated IL-10−/− mice evaluated 7 weeks after bacterial inoculation (Fig. 2a–c) show that severe inflammation develops throughout the entire colon and duodenum. Inflammation was manifested in various regions of the colon by crypt hyperplasia and abscesses, infiltration of predominantly mononuclear cells with focal neutrophils, and goblet cell depletion, similar to the histologic features of SPF IL-10−/− mice.2 In addition, duodenal infiltration was present in the dual-associated mice after 7 weeks of colonization. Control WT dual-associated mice did not develop inflammation in any region of the intestinal tract even after 7 weeks of bacterial monoassociation (Fig. 2d–f). All representative histologic sections are shown at the same magnification (20×; note similar muscularis mucosae thicknesses in all sections) from IL-10−/− and WT dually associated mice.

thumbnail image

Figure 2. Histological findings in the intestinal tract of E. faecalis plus E. coli dual-associated IL-10−/− or WT mice. H&E-stained sections at 20× magnification from a representative dual-associated IL-10−/− (a) distal colon, (b) cecum, or (c) duodenum or WT (d) distal colon, (e) cecum, or (f) duodenum obtained 7 weeks after colonization with E. coli plus E. faecalis. Progressive inflammation, manifested by marked crypt hyperplasia, goblet cell depletion, and lamina propria cellular infiltration, is seen in the IL-10−/− dual-associated mice but not in WT control tissue. Mild cellular infiltration seen in WT mice is representative of normal physiological changes, not colitis.

Download figure to PowerPoint

E. faecalis Plus E. coli Dual-association Induces Duodenal as Well as Colonic NF-κB-Activation and IL-12/IL-23p40 mRNA Expression in IL-10−/− But Not WT Mice

NF-κB is a transcription factor critically important for the induction of a wide variety of genes. We used IL-10−/− × NF-κBEGFP and control WT × NF-κBEGFP transgenic mice to assess the activation of NF-κB during the course of colitis that develops after E. faecalis plus E. coli dual-association. These mice express EGFP under the NF-κB-promoter, leading to EGFP accumulation in cells that activate NF-κB, and thus provide a means to track NF-κB-activation in real-time by EGFP imaging.8, 14 IL-10−/− × NF-κBEGFP and WT × NF-κBEGFP transgenic mice (n = 3/group) were colonized with E. faecalis / E. coli. When the intestinal tract was imaged macroscopically 5 weeks after dual-association, IL-10−/− but not WT mice clearly demonstrated EGFP expression (Fig. 3a–c). Thus, E. faecalis plus E. coli dual-association induces NF-κB-activation in the colon and duodenum of IL-10−/− but not WT mice. To evaluate transcriptional activity following NF-κB activation, we isolated total RNA from colonic sections and measured mRNA expression of the NF-κB dependent gene IL-12/IL-23p40 by real-time RT-PCR. Importantly, colonic tissue isolated from the dual-associated IL-10−/− × NF-κBEGFP mice demonstrated a 9.1 ± 4.3-fold increase in IL-12/IL-23p40 mRNA over WT × NF-κBEGFP animals (P < 0.05). Thus, E. coli / E. faecalis dual-association induces duodenal as well as colonic NF-κB-activation, leading to increased IL-12/IL-23p40 mRNA expression in IL-10−/− mice.

thumbnail image

Figure 3. Expression of EGFP in the intestinal tract of E. faecalis plus E. coli dual-associated IL-10−/− × NF-κBEGFP or WT × NF-κBEGFP mice. Mice were dual-associated for 5 weeks. Images of (a) the colon and (b) the stomach and small intestine from a representative mouse of each type are shown. EGFP expression in a Peyer's patch from a representative IL-10−/− mouse is shown in (c) low (upper panel) and high (lower panel) magnifications by confocal microscopy.

Download figure to PowerPoint

Spontaneous Intestinal IL-12/IL-23p40 Secretion Progressively Increases in Dual-associated IL-10−/− Mice

We addressed whether mucosal immune responses accompanied the presence of histological and clinical inflammation. Statistically significant increases in proinflammatory IL-12/IL-23 p40 production measured from colonic explant cultures were evident even at the earliest timepoint studied (3 weeks after colonization) when compared to noninflamed WT dual-associated mice. Furthermore, spontaneous colonic IL-12/IL-23p40 secretion increased in proportion to the duration of bacterial colonization (Fig. 4), paralleling the degree of histological inflammation.

thumbnail image

Figure 4. Colonic IL-12/IL23p40 produced in E. faecalis plus E. coli dual-associated IL-10−/− and WT mice. Spontaneous IL-12/IL-23 p40 secretion was quantified from colonic explant culture supernatants incubated for 20–24 hours. IL-10−/− mice, n = 5 mice/timepoint, WT mice, n = 4 (*P < 0.005 versus WT).

Download figure to PowerPoint

Bacterial Antigen-specific CD4+ T cell Responses in E. faecalis plus E. coli Dual-associated Mice

To address whether dual-associated mice exhibited bacterial antigen-specific responses to E. faecalis and/or E. coli, we cultured MLN-derived CD4+ T cells (isolated from IL-10−/− and WT dual-associated mice) with APC that had been pulsed in vitro with either E. faecalis or E. coli lysates or with KLH or B. vulgatus lysates (as unrelated protein and bacterial antigen controls, respectively). IL-10−/− CD4+ MLN T cells produced high levels of IFN-γ when stimulated with either E. faecalis or E. coli lysate-pulsed APC, but not with KLH or B. vulgatus-pulsed APC (Fig. 5a). Interestingly, IFN-γ secretion by CD4+ T cells isolated from dual-associated mice stimulated with E. faecalis-pulsed APC increased slightly with time after bacterial colonization (3 weeks: 16.6 ± 3.2 ng/mL; 7 weeks: 24.3 ± 1.1 ng/mL). In contrast, IFN-γ secretion by CD4+ T cells stimulated with E. coli-pulsed APC decreased with time after dual-association (3 weeks: 106.6 ± 9.4 ng/mL; 7 weeks: 48.2 ± 6.3). In addition, the overall IFN-γ levels produced in cultures of CD4+ T cells stimulated by E. coli-pulsed APC were higher at all timepoints relative to E. faecalis-pulsed APC values. We also evaluated IL-17, a cytokine thought to be crucial for the development of intestinal inflammation.15 As shown in Figure 5b, IL-17 production by CD4+ T cells from dual-associated mice did not directly correlate with IFN-γ levels. Instead, IL-17 secretion in response to E. faecalis-pulsed APC increased with time after dual-association much more dramatically than did IFN-γ. IL-17 production in response to E. coli-pulsed APC also increased between 3 and 7 weeks after dual-association. Furthermore, the amounts of IL-17 secretion induced by E. coli and E. faecalis-pulsed APC were comparable at 5 and 7 weeks after dual association, in contrast to the greater ability of E. coli lysates to induce IFN-γ relative to E. faecalis lysates.

thumbnail image

Figure 5. Bacterial antigen-specific CD4+ T-cell responses in MLN of E. faecalis plus E. coli dual-associated mice. CD4+ T cells prepared from mesenteric lymph node (MLN) cells of dual-associated IL-10−/− mice (n = 5 mice per timepoint) or WT (n = 4) were cocultured with KLH or bacterial lysate (E. faecalis; E. coli; B. vulgatus)-pulsed antigen presenting cells prepared from T-cell-depleted splenocytes. (a) IFN-γ and (b) IL-17 were measured from cell culture supernatants collected at 72 hours.

Download figure to PowerPoint

Direct Comparison Among Mice Monoassociated with E. faecalis or E. coli or Dual-associated with Both Bacterial Species

Next, we compared mono- versus dual-associated IL-10−/− mice. We chose to evaluate these mice after 3 weeks of colonization, based on previous studies showing that IL-10−/− mice monoassociated with E. faecalis do not develop histologic evidence of intestinal inflammation during the first 4–5 weeks of colonization, while E. coli-monoassociated IL-10−/− mice develop varying degrees of cecal-predominant disease that is detectable histologically after 3 weeks.7 The results shown in Table 1 compare histologic inflammatory scores for the cecum and distal colon and spontaneous (colonic IL-12/IL-23p40) and stimulated (IFN-γ and IL-17) proinflammatory cytokine secretion from IL-10−/− dual- or monoassociated mice colonized for 3 weeks with WT dual-associated mice colonized for 7 weeks. As shown in Table 1a, amounts of IL-12/IL-23p40 in colonic tissue fragments from mice colonized with both bacteria were higher than observed in colonic fragment cultures from monoassociated mice. Likewise, inflammation of the distal colon and cecum was also significantly higher in tissue from dual-associated mice compared to mice monoassociated with either E. faecalis or E. coli. The picture for cytokine production by CD4+ T cells is complex (Table 1b). CD4+ cells from dual-associated mice produced more IFN-γ but less IL-17 in response to E. faecalis-pulsed APC compared to cells from E. faecalis-monoassociated mice. The in vitro IFN-γ response to E. coli-pulsed APC was very high in cultures of MLN CD4+ T cells from dual-associated mice and IL-17 was also produced in higher quantities by the cells from these mice compared to cells from E. coli-monoassociated but not from E. faecalis-monoassociated mice. The results included in Table 1 for E. coli-monoassociated mice have been previously published,7 with the exception of IL-17, a cytokine we did not evaluate in earlier studies.

Luminal Bacterial Concentrations Do Not Affect the Degree of Intestinal Inflammation

We quantified luminal bacterial concentrations using fecal samples obtained from different intestinal regions of the dual-associated mice to determine whether or not the 2 different bacteria were able to colonize equally. Luminal bacterial load was quantified by culturing serially diluted weighed luminal contents from the cecum and distal colon obtained at the time of necropsy from mice colonized for 7 weeks. No statistically significant differences in bacterial growth of E. faecalis or E. coli were noted by these standard culture techniques between the luminal contents of the distal colon versus cecum of dual-associated IL-10−/− and when compared to WT mice (Table 2).

Table 2. Bacterial Growth in the Distal Colon and Cecum of E. faecalis Plus E. coli Dual-associated IL-10-/- and WT Mice
Mouse TypeE. faecalis Distal ColonE. faecalis CecumE. coli Distal ColonE. coli Cecum
  1. Results from IL-10-/- and WT control dual-associated mice colonized for 7 weeks are shown (n = 4-5 mice per group). Bacterial counts (108 cfu per gm stool) were quantified from serial dilutions of cecal and distal colonic fecal contents obtained at necropsy and plated on CNA and MacConkey agar plates. No significant differences were noted between bacterial counts in cecal versus distal colonic contents or between samples obtained from IL-10-/- versus WT control mice.

IL-10-/-24.0 ± 11.915.6 ± 18.718.2 ± 11.523.8 ± 28.9
WT17.8 ± 8.110.0 ± 14.219.8 ± 5.618.3 ± 13.8

IL-10−/− Dual-associated Mice Can Develop Reactive Atypia of the Distal Colon

Escalation of disease progression and aggressiveness was noted in dual-associated IL-10−/− mice when compared to IL-10−/− mice monoassociated with either E. faecalis or E. coli. We previously reported that 2 of 12 IL-10−/− mice monoassociated with E. faecalis for greater than 30 weeks develop features of severe reactive atypia in the distal colon associated with underlying extensive active and chronic inflammation, but no dysplastic changes or invasive carcinoma; E. coli-monoassociated IL-10−/− mice do not develop any epithelial atypia even at late timepoints of bacterial colonization.7E. faecalis-monoassociated mice did not develop epithelial changes prior to 30 weeks. However, 3 out of 5 IL-10−/− mice colonized with both E. faecalis plus E. coli for 7 weeks developed reactive atypia of the distal colonic epithelium associated with crypt architectural distortion and transmural inflammation. These changes were characterized by palisading epithelial cells and increased nuclear density, with accompanying adenomatous changes of the crypts (Fig. 6).

thumbnail image

Figure 6. Reactive atypia of the distal colon of E. faecalis plus E. coli dual-associated IL-10−/− mice. Reactive atypia of the distal colonic epithelium, with crypt architectural distortion and transmural inflammation characterized by palisading epithelial cells and dense nuclei is noted in IL-10−/− dual-associated mice colonized for 7 weeks (section is shown at 20× magnification).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Our results demonstrate the ability of 2 common nonpathogenic commensal bacterial species that individually induce different phenotypes of enterocolitis in a single genetically defined host,7, 13 together potentiate the aggressiveness of intestinal inflammation. These data support and expand the results of our previous studies by showing that inoculating IL-10−/− mice with both organisms leads to more rapid progression of severe intestinal inflammation with wider regional distribution of disease. Proinflammatory cytokine-driven immune activation, manifested by spontaneous colonic IL-12/23p40 secretion and IFN-γ production from MLN CD4+ T cells stimulated with E. faecalis and E. coli lysate pulsed–APC, was significantly higher in dual-associated IL-10−/− compared with E. faecalis- and E. coli-monoassociated mice at an early phase of bacterial colonization (3 weeks). Interestingly, E. faecalis pulsed APC activated a lower IL-17 response in MLN CD4+ T cells from dual-associated compared to monoassociated mice at the earliest timepoint we evaluated. However, IL-17 production in response to components of both bacteria accompanies the subsequent development of intestinal inflammation evaluated at 5 and 7 weeks after dual association. These findings highlight the complex nature of bacterial-driven activation of immune responses with both Th1 and Th17 profiles in the IL-10−/− model of colonic inflammation.

Previously, investigators have studied the differential ability of several nonpathogenic bacterial species to variably induce or ameliorate disease in genetically susceptible rodent inflammatory bowel disease (IBD) models. IL-2 gene-deficient (IL-2−/−) mice develop severe colitis and progress to early death when raised in conventional housing16 but much milder, slow-onset, and nonlethal intestinal inflammation when raised in GF conditions.17 Waidmann et al18 showed that gnotobiotic IL-2−/− mice monoassociated with a nonpathogenic E. coli strain developed colitis, whereas IL-2−/− mice either monoassociated with a nonpathogenic B. vulgatus strain or dually associated with E. coli plus B. vulgatus did not develop colonic inflammation. These results indicate that B. vulgatus ameliorates colitis in IL-2−/− mice when co-inoculated with a colitogenic E. coli strain. This finding further emphasizes the importance of interactions between different components of the normal commensal microbiota in maintaining the tenuous balance between the induction, maintenance, and amelioration of intestinal inflammation.

It is possible that differences in disease progression are, in part, due to the number of bacteria present or the degree of mucosally adherent bacteria within a specific region of the intestinal tract. The bacterial concentrations quantified from the distal colon and cecum by routine culture and PCR did not explain the regional differences in histologic inflammation in mice monoassociated with E. faecalis or E. coli in our previous study of monoassociated IL-10−/− mice.7 These results are further supported by our current results, which show no significant differences in bacterial concentrations of E. faecalis versus E. coli in different colonic regions. Waidmann et al18 also showed that the anti-colitogenic effect of B. vulgatus on E. coli in IL-2−/− mice is not due to a reduction in E. coli numbers within the intestinal tract, suggesting that potentiating and/or ameliorating effects of different commensal organisms are due to their differential abilities to breach the mucosal barrier and/or to activate the resident macrophages and dendritic cells and to induce antigen-specific immune responses rather than altering luminal bacterial composition.

There are parallels between these observations in rodent colitis models and human IBD.19 Duchmann et al20, 21 have shown that peripheral blood CD4+ TCRαβ+ T cell clones from IBD patients respond to several commensal bacteria, including E. coli. In addition, invasive E. coli and Enterococci have been recovered from inflamed intestinal tissue immediately adjacent to ulcers and fistulae of patients with CD.22 Moreover, enteroadherent/invasive E. coli strains were found in greater frequency in the ileal mucosa of patients with postoperative recurrence of disease compared to either CD patients with chronic ileal lesions or healthy controls.23 Other investigators have isolated different microorganisms and their associated proteins from CD patients, including P. fluorescens,6 and documented serologic responses to defined bacterial antigens in the majority of CD and ulcerative colitis (UC) patients.24 Swidsinski et al25, 26 demonstrated mucosal association of coliforms and Bacteroides species in patients with CD, UC, and nonspecific intestinal inflammation.

The question arises whether different disease phenotypes observed in IBD patients may be driven in part by various combinations of distinct commensal bacteria that act in an additive manner. Several published reports show that patients with more severe disease phenotypes (i.e., intestinal perforating disease, fibrostenosis, small intestinal resection) had a high frequency of antibody responses to bacterial and fungal antigens.27, 28 In addition, patients with antibody responses to multiple microbial antigens appeared to have an increased likelihood of complicated small bowel disease phenotypes. These observations suggest that additive effects of immune responses to antigens from various commensal microorganisms intensify intestinal inflammation and complications of CD.

The clinical implications of understanding the relationship between bacterial antigen-driven immune responses in genetically susceptible individuals extend beyond the diagnostic arena. Antibiotics with different narrow spectra of bactericidal activity in colitic SPF IL-10−/− mice preferentially treat different colonic regions,29 suggesting the ability to tailor antimicrobial therapy in IBD patients, depending on which bacterial species are present. Metronidazole, which has predominant anaerobic antimicrobial activity, either alone or in combination with ciprofloxacin, is more active in colonic versus ileal CD,30, 31 suggesting that differential bacterial colonization in the colon versus the small intestine is at least partially responsible for different disease phenotypes. In pilot studies, Mow et al32 showed that CD patients with ileal +/− right-sided colonic disease whose sera contained antibodies reactive with both E. coli outer membrane porin C (OmpC) and I2 had the highest disease remission rate when treated with metronidazole + ciprofloxacin + budesonide compared to patients with antibody reactivity to either OmpC or I2 or to neither bacterial protein. Although not statistically significant, these findings are intriguing and warrant future study.

The results of our study demonstrate that 2 common nonpathogenic commensal organisms that individually induce phenotypically distinct colitis in the same genetically susceptible host act synergistically to induce aggressive, early-onset distal colonic and cecal inflammation accompanied by amplified bacterial antigen-specific immune responses. The clinical and pathogenic implications of this observation include the likelihood that various commensal enteric bacterial species interact to induce and perpetuate mucosal immune responses that accompany intestinal inflammation. Elucidating which specific components of the complex luminal microbiota drive pathogenic versus protective immune responses can guide therapeutic manipulation of the complex bacterial milieu of the distal intestine.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors thank Maureen Bower and Jerilyn Shaw of the National Gnotobiotic Rodent Resource Center (Department of Laboratory Medicine, UNC-CH) and Donna Kronstadt (Gnotobiotic Core, Center for Gastrointestinal Biology and Disease, College of Veterinary Medicine, NCSU) for maintaining the gnotobiotic mice; Joseph Galanko, Biostatistics Core, Center for Gastrointestinal Biology and Disease, UNC-CH; Desmond McDonnell and Bayley Crane (NCSU) for technical assistance; and the Center for Gastrointestinal Biology and Disease for providing administrative support.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES