Intrarectal Instillation of Clostridium difficile Toxin A Triggers Colonic Inflammation and Tissue Damage: Development of a Novel and Efficient Mouse Model of Clostridium difficile Toxin Exposure

Abstract

Clostridium difficile, a major cause of hospital-acquired diarrhea, triggers disease through the release of two toxins, toxin A (TcdA) and toxin B (TcdB). These toxins disrupt the cytoskeleton of the intestinal epithelial cell, increasing intestinal permeability and triggering the release of inflammatory mediators resulting in intestinal injury and inflammation. The most prevalent animal model to study TcdA/TcdB-induced intestinal injury involves injecting toxin into the lumen of a surgically generated “ileal loop.” This model is time-consuming and exhibits variability depending on the expertise of the surgeon. Furthermore, the target organ of C. difficile infection (CDI) in humans is the colon, not the ileum. In the current study, we describe a new model of CDI that involves intrarectal instillation of TcdA/TcdB into the mouse colon. The administration of TcdA/TcdB triggered colonic inflammation and neutrophil and macrophage infiltration as well as increased epithelial barrier permeability and intestinal epithelial cell death. The damage and inflammation triggered by TcdA/TcdB isolates from the VPI and 630 strains correlated with the concentration of TcdA and TcdB produced. TcdA/TcdB exposure increased the expression of a number of inflammatory mediators associated with human CDI, including interleukin-6 (IL-6), gamma interferon (IFN-γ), and IL-1β. Finally, we were able to demonstrate that TcdA was much more potent at inducing colonic injury than was TcdB but TcdB could act synergistically with TcdA to exacerbate injury. Taken together, our data indicate that the intrarectal murine model provides a robust and efficient system to examine the effects of TcdA/TcdB on the induction of inflammation and colonic tissue damage in the context of human CDI.

INTRODUCTION

Clostridium difficile is a Gram-positive, spore-forming, strictly anaerobic, toxin-producing (toxin A [TcdA], 308 kDa; toxin B [TcdB], 270 kDa) bacterium that is a major cause of nosocomial diarrhea. The emergence of the NAP1/027 strain, which is more virulent and exhibits increased resistance to antibiotics, has generated considerable concern. Recent hospital outbreaks and the increased occurrence of community-acquired C. difficile infections (CDI) have been associated with the NAP1/027 strain (33, 37). Altogether, the clinical data suggest that the incidence of CDI is on the rise and is no longer linked solely to the elderly in the hospital setting (26).

The clinical features of CDI are driven by TcdA and TcdB. Taylor and Bartlett (44) and Abrams et al. (1) were the first to describe the existence of clostridial toxins responsible for cell damage and gastrointestinal disease in the context of antibiotic treatment. Subsequently, Wren et al. were able to correlate the severity of antibiotic-associated diarrhea with the toxin production of C. difficile strains isolated from patients (50). TcdA and TcdB are monoglucosyltransferases that inhibit monomeric G-protein function, altering the cytoskeletal structure of cells. In the gastrointestinal tract, this leads to the disruption of the intestinal epithelial barrier, resulting in intestinal injury and the induction of mucosal inflammation (18). The resulting inflammatory response leads to the production of proinflammatory mediators, such as interleukin-1β (IL-1β) and CXCL-8/IL-8 (31, 41, 43). These events trigger the recruitment of inflammatory cells, propagating the collateral tissue damage that is apparent in patients with CDI (14, 15).

Studies assessing the effects of TcdA and TcdB on the intestinal epithelial barrier, enteric nerve function/survival, and activation of the innate immune system have used a variety of experimental methods. The predominant in vivo method is the “ileal loop” model, a surgery-based model that involves a laparotomy, ligation of the terminal ileum, injection of C. difficile toxin preparations proximal to the ligation, and closure of the surgical incisions (2, 5, 9, 13, 40). This procedure is time-consuming and involves the variables and risks associated with any small-animal surgical procedure. Importantly, it should be noted that the target organ of CDI in humans is the colon and not the ileum. More recently, Chen et al. developed a murine infection model for C. difficile (11). This procedure involves a 3-day treatment with an antibiotic cocktail in the drinking water followed by an intraperitoneal injection of clindamycin and subsequent administration of C. difficile spores via oral gavage. Animals are euthanized 4 days postinoculation and assessed for colonic tissue damage and inflammatory cell infiltrate (11). Although this model recapitulates many aspects of CDI, it requires extensive biosafety measures (e.g., special mouse housing to avoid spore dissemination) and highly trained animal technicians and is a multiday procedure.

Given the shortcomings of these experimental models and the need for a robust preclinical model to efficiently assess therapeutic intervention strategies for CDI, we sought to develop a new murine model of toxin-induced intestinal injury and inflammation. Here we describe and validate a novel mouse model of CDI that involves the intrarectal administration of C. difficile toxin(s) and report that TcdA, but not TcdB, is required to trigger colonic inflammation and tissue damage.

MATERIALS AND METHODS

TcdA/TcdB production.

C. difficile toxins TcdA and TcdB were produced as described previously (19, 36). Briefly, C. difficile strains (the VPI strain, ATCC 43255, designation VPI 10463, and the 630 strain, ribotype 012) were grown in brain heart infusion medium under anaerobic conditions. Dialysis tubing containing phosphate-buffered saline (PBS) was inoculated with an overnight culture and suspended in 750 ml of medium within a conical flask. The contents of the dialysis tubing were harvested at day 5 postinoculation by centrifugation (10,000 × g, 60 min). After centrifugation, the toxin-containing supernatant was removed, passed through a 0.22-μm filter to remove spores and then through a 100-kDa centrifugal spin filter (Chemicon, Millipore), and used as a source of TcdA/TcdB. This preparation was verified to be lipopolysaccharide free via high-pressure liquid chromatography (HPLC), and the purity was assessed by SDS-PAGE and Western blotting as published previously (19, 36). The concentration of TcdA and TcdB in TcdA/TcdB preparations was quantified by enzyme-linked immunosorbent assay (ELISA). Triplicates were coated with 100 μl of 10-fold-diluted supernatants overnight, washed with Tris-buffered saline (TBS), and blocked with 200 μl of TBS-Tween-skim milk (5%) for 2 h at 37°C. One hundred microliters of rabbit polyclonal anti-TcdA or anti-TcdB (1:500) was applied per well after washing 4 times with TBS-Tween. Plates were then incubated for 1 h at 37°C. Afterward, wells were washed 4 times with TBS-Tween, and 100 μl (1:1,000) of anti-rabbit-horseradish peroxidase (HRP) (Sigma-Aldrich, Oakville, Ontario, Canada; A4914) was applied to each well for 1 h at 37°C and then washed 4 times with TBS-Tween. The plates were then read at 410 nm using a Spectramax 340. TcdA and TcdB were purified from VPI TcdA/TcdB preparations by anion-exchange chromatography, and their activity was assessed using the Vero cell cytotoxicity assay as described previously (19).

Intrarectal instillation of C. difficile toxins.

All mice used in our studies were female and between 10 and 12 weeks of age. C57/BL6 mice (Charles River, Sherbrooke, Quebec, Canada) were used for the initial characterization of the model. ASC^−/− and wild-type littermates on the C57/BL6 background were bred in-house (36). The instillation of C. difficile toxins (TcdA/TcdB; purified TcdA and TcdB) was adapted from a previously published protocol to induce experimental colitis with the intrarectal instillation of 2,4,6-trinitrobenzene sulfonic acid (TNBS) (34). Briefly, a 5F infant feeding tube catheter with side ports (Mallinckrodt Inc., St. Louis, MO; catalogue no. 85771) was lubricated with water-soluble personal lubricant (Healthcare Lubricating Jelly, Toronto, Ontario, Canada) and inserted 2.5 cm (measured from the midway point between the 2 catheter side ports) up the colon. At this point, 100 μl of solution was slowly administered over 30 s while pressure was applied to the anal area to prevent leakage. Following the injection of the solution, the tube was slowly removed and the rectal pressure was maintained for a further 30 s. C. difficile toxins (TcdA/TcdB; purified TcdA and TcdB) were diluted in PBS to allow for a uniform 100 μl of solution to be injected. Treatment values indicated throughout the paper denote the total toxin administered in the 100-μl solution. Control animals were treated with 100 μl of PBS. All animal experiments were approved by the Health Sciences Animal Care Committee of the University of Calgary and conformed to the guidelines set forth by the Canadian Council for Animal Care. The health and welfare of each mouse were assessed at 30-min intervals following intrarectal toxin instillation. Given that death is not an acceptable experimental endpoint, based on the Canadian Council of Animal Care guidelines, mice were removed from the study and euthanized based on the following criteria: 1, animals that exhibited a hunched posture; 2, animals that remained isolated in their cages (i.e., failed to socialize with cage-mates); 3, animals that failed to exhibit any grooming behavior; 4, animals that did not interact with environmental enrichment; 5, animals that failed to respond to external environmental cues (i.e., did not quickly enter their nesting following a gentle finger poke to the abdomen or flank). Mice exhibiting two or more of the behaviors indicated in criteria 1 to 4 or exhibiting the behavior described in criterion 5 alone were deemed to be in distress and removed from the study based on humane grounds.

Assessment of cytokine levels.

Colonic segments were homogenized in lysis buffer (150 mM NaCl, 20 mM Tris, pH 7.5, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, and protease inhibitor cocktail [Complete Minitab; Roche]). Tissue cytokine levels were assessed using a Luminex XMap assay according to the manufacturer's instructions (Luminex Corp., Toronto, Ontario, Canada). Further assessment of tissue IL-1β was performed using a commercial enzyme-linked immunosorbent assay (ELISA) kit (BD Biosciences, Mississauga, Ontario, Canada) according to the manufacturer's instructions.

Tissue MPO assay.

Tissue myeloperoxidase (MPO) activity was determined as described previously (20). MPO activity was measured in units per milligram of tissue, where 1 unit of MPO was defined as the amount needed to degrade 1 μmol of H₂O₂ per minute at room temperature.

In vivo intestinal permeability assay.

To assess intestinal permeability, we measured the movement of fluorescein isothiocyanate (FITC)-dextran from the lumen of the gastrointestinal tract into the plasma. All mice were gavaged with FITC-dextran (Sigma-Aldrich, Oakville, Ontario, Canada; molecular mass, 3 kDa; 60 mg/100 g body weight) 1 h prior to the intrarectal instillation of TcdA/TcdB or PBS vehicle control. After the 2- or 4-hour exposure period, whole blood was obtained by retro-orbital bleeding at the time of euthanasia. Plasma was isolated, and FITC-dextran measurements were performed in triplicate in a fluorometric plate reader at 488 nm. Data are expressed as total fluorescence units in 100 μl of plasma.

Histological assessment.

The severity of TcdA/TcdB-induced colitis was scored histologically using two different parameters on coded, hematoxylin-and-eosin-stained slides in a blinded fashion by a board-certified pathologist. A histological scoring system was used to assess pathological tissue architectural changes: 0, normal; 1, vacuolation/blebbing; 2, loss of epithelium; 3, complete loss of crypt architecture. An inflammation score was used to assess the severity of the inflammatory response: 0, normal; 1, increased number of inflammatory cells in lamina propria; 2, increased number of inflammatory cells in submucosa; 3, dense inflammatory cell mass, but not transmural in nature; 4, transmural inflammation. Furthermore, an assessment of each section was performed to determine an estimate of the percentage of the colonic tissue section exhibiting architectural changes (% architecture change) or exhibiting increased inflammatory cell infiltrate (% tissue inflammation).

Immunohistochemistry.

Colonic tissue was removed upon necropsy, placed into Zamponi's fixative, and held at 4°C overnight. Following sucrose cryoprotection, tissue samples were embedded in OCT cryoembedding medium and subsequently sectioned. Immunohistochemical analysis was performed on frozen sections to assess the infiltration of F4/80-positive macrophages (AbD Serotec, Raleigh, NC; catalogue no. MCA 497B) and CD3-positive T cells (BD Pharmingen, Mississauga, Ontario, Canada; catalogue no. 550275). The total number of F4/80- and CD3-positive cells was quantified at 40× by counting the number of positive cells in the region from the muscularis mucosa to the lumen in 10 individual 250-μm spans per section.

TUNEL staining.

To assess the extent of apoptotic cell death induced by intrarectal instillation of TcdA/TcdB, we performed terminal deoxynucleotidyltransferase dUTP-mediated nick end labeling (TUNEL) staining on paraffin-embedded tissue sections using a commercially available kit (DeadEnd fluorometric TUNEL system; Promega, Madison, WI). Paraffin-embedded tissues were dewaxed, hydrated, and incubated with 20 μg/ml proteinase K to strip proteins from the nuclei. Fragmented DNA was then identified by the incorporation of fluorescein-12-dUTP during an incubation step with terminal deoxynucleotidyltransferase at 37°C for 1 h. Sections were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) and mounted using FluorSave reagent (Calbiochem, Gibbstown, NJ). Additionally, sections were incubated with 10 units/ml of DNase I (Promega) prior to the labeling step in order to artificially fragment the DNA and provide a positive control for the assay. Dried sections were visualized using the DeltaVision RT confocal microscope (Applied Precision, Issaquah, WA) with filters for DAPI and FITC.

Statistical analysis.

Parametric data are expressed as means ± standard errors of the means. Nonparametric data are expressed as means. One-way and two-way analysis of variance (ANOVA) tests were used for multiple comparison analyses on parametric data followed by Tukey's multiple comparison test to determine statistical differences. Multivariate analysis of covariance (MANCOVA) tests were used for multiple comparison analyses on nonparametric data followed by Dunn's multiple comparison test to determine statistical differences. Kaplan-Meier survival analysis was performed to assess the sensitivity of mice to TcdA/TcdB treatment. For this analysis, mice removed from the study and euthanized according to the criteria set forth under “Intrarectal instillation of C. difficile toxins” were considered to have not survived. The log-rank Mantel-Cox test was used to determine if the survival curves of the various groups were statistically different.

RESULTS

Intrarectal administration of TcdA/TcdB triggers colonic inflammation.

To determine the sensitivity of the colon to TcdA/TcdB, we performed a dose-response study. Increasing doses of TcdA/TcdB (25, 50, 75, and 100 μg TcdA/TcdB, each administered in 100 μl of PBS; 4-h total exposure) were instilled into the colon of C57/BL6 mice. We observed a dose-dependent increase in tissue myeloperoxidase (MPO) content (an index of neutrophil infiltration) (Fig. 1A) following TcdA/TcdB treatment. Furthermore, at doses greater than 50 μg, mice exhibited poor survival. Kaplan-Meier survival analysis revealed a statistically significant reduction in the survival of mice treated with 75 and 100 μg of TcdA/TcdB compared to all other doses. As depicted in Fig. 1B, a number of mice were removed from the study and euthanized according to the criteria set out in Materials and Methods (i.e., 100% of mice treated with the 100-μg dose at 3 h and 50% of the mice treated with the 75-μg dose at 4 h). Mice treated with the 50-μg dose displayed some decreased mobility and reduced social behavior but continued to exhibit grooming behavior and interact with nesting materials. Mice treated with 25 μg did not exhibit any changes in behavior indicative of distress.

Fig 1.

Intrarectal instillation of C. difficile toxins TcdA and TcdB leads to a dose-dependent increase in colonic inflammation and disease severity in C57/BL6 mice. (A) Colonic myeloperoxidase (MPO) assessed at 4 h after instillation of TcdA/TcdB (25 to 100 μg each in 100 μl of PBS). n = 5/group/concentration. *, P < 0.05 compared to PBS control treatment (100 μl of PBS alone); #, P < 0.05 compared to 25- and 50-μg groups; Φ, no significant difference between marked groups. (B) Kaplan-Meier analysis assessing the effect of increasing concentrations of TcdA/TcdB on the survival of C57/BL6 mice after 4 h of exposure. P < 0.05. (C) The effects of intrarectal instillation of 50 μg of TcdA/TcdB over a period of 10 h. Colonic myeloperoxidase (MPO) assessed at 2, 4, and 10 h post-intrarectal toxin instillation. n = 6/group/time point. *, P < 0.05 compared to PBS control. (D) Kaplan-Meier survival analysis between TcdA/TcdB- and PBS-treated groups. Note that the health and welfare of each mouse were assessed at 30-min intervals following intrarectal toxin instillation. Given that death is not an acceptable experimental endpoint based on the Canadian Council of Animal Care guidelines, mice removed from the study and euthanized based on criteria indicated in Materials and Methods were considered to have not survived.

Next, we sought to assess the time dependence of the inflammatory responses. In this case, a significant increase in colonic MPO content was apparent 2 h after intrarectal administration of 50 μg of TcdA/TcdB (Fig. 1C). The animals in our study tolerated the 50-μg dose of TcdA/TcdB at the 2- and 4-hour time points, displaying some reduced mobility. However, approximately 20% of the toxin-treated mice had to be removed from the study prior to the 10-hour endpoint (Fig. 1D).

Intrarectal administration of TcdA/TcdB leads to increases in colonic tissue damage and intestinal permeability.

C. difficile toxins cause extensive damage to intestinal epithelial cells, triggering the infiltration of neutrophils, an effect that contributes to the formation of the pseudomembranes which are pathognomonic of CDI (22). To determine whether the inflammatory response initiated by the intrarectal instillation of TcdA/TcdB would parallel that observed in the tissue of patients with CDI, we performed histological analysis of colonic sections isolated from toxin-treated mice (Fig. 2A to C). Specimens from TcdA/TcdB-treated mice revealed immune cell infiltration, beginning as early as 2 h post-toxin treatment (Fig. 2B). At the 4-hour time point, we observed enhanced neutrophil infiltration into the mucosa and the luminal regions (Fig. 2C), a hallmark of the inflammatory response in colonic tissue sampled from CDI patients (22). This inflammatory response was associated with considerable tissue damage, including the sloughing off of epithelial cells, submucosal edema, and crypt shortening (Fig. 2C). Quantitative assessment of these sections revealed architectural changes and tissue inflammation to be present in a significantly greater percentage of the overall tissue cross-section at the 4-hour time point compared with the PBS-treated control and 2-hour time point (Fig. 2D and E). Furthermore, the tissue fluid content, an indication of edema, was significantly increased at the 4-hour time point (data not shown).

Fig 2.

Intrarectal instillation of C. difficile toxins TcdA and TcdB (50 μg in 100 μl PBS) triggers colonic tissue damage, intestinal inflammation, and increased intestinal permeability in a time-dependent fashion. (A to C) Representative colonic sections were stained with hematoxylin and eosin from control (A), 2-hour (B), and 4-hour (C) toxin-treated groups. (D and E) Histological assessment of colonic sections stained with hematoxylin and eosin. Sections were given an overall score for tissue architecture based on criteria specified in Materials and Methods (D), and the percentage of each section exhibiting architectural changes and inflammatory cell infiltrate was assessed (E). n = 6/group/time point. *, P < 0.05 compared to control (100-μl PBS installation) and the 2-hour group. Note that colonic sections were assessed in a blinded fashion by a board-certified pathologist. (F) Intrarectal toxin instillation increases intestinal permeability as assessed by the migration of FITC-dextran into the serum. Mice were given a gastric gavage of FITC-dextran (15 mg/mouse in 100 μl of PBS) 1 h prior to intrarectal instillation of toxin. Mean fluorescence was assessed in the serum in duplicate for each mouse. n = 6/group/time point. *, P < 0.05 compared to control mice (100-μl PBS instillation). (G) Colonic MPO assessed at 2 and 4 h post-intrarectal toxin instillation. n = 6/group/time point. *, P < 0.05 compared to control mice (100-μl PBS instillation).

During an infection with C. difficile, the release of TcdA and TcdB into the luminal space and the subsequent entry of the toxins into the intestinal epithelial cell are one of the earliest pathogenic events. The internalization of TcdA/TcdB into an intestinal epithelial cell leads to changes in the function of the cytoskeleton, resulting in cell rounding and a reduction in the ability of the epithelium to form a tight barrier (18). Thus, we hypothesized that intrarectal administration of TcdA/TcdB would lead to an increase in intestinal permeability, an effect that could play a role in the extensive tissue inflammation and collateral tissue damage observed in CDI. To test this, mice were provided FITC-dextran by oral gavage 1 h prior to toxin exposure. TcdA/TcdB treatment led to a significant increase in the flux of FITC-dextran from the lumen into the blood (Fig. 2F), indicating increased intestinal barrier permeability. This effect was paralleled by the colonic MPO values assessed at 2 and 4 hours post-toxin instillation (Fig. 2G).

We have published previously that C. difficile toxins trigger intestinal inflammation and tissue damage through the activation of an ASC-containing inflammasome (36). In the current study, deletion of ASC led to a significant reduction in the inflammation and colonic tissue damage observed following intrarectal administration of TcdA/TcdB (Fig. 3G). Furthermore, ASC^−/− mice exhibited an attenuated IL-1β response following intrarectal administration of TcdA/TcdB (Fig. 3G) (36).

Fig 3.

Genetic deletion of ASC renders mice resistant to colonic tissue damage and inflammation following intrarectal instillation of TcdA/TcdB (50 μg in 100 μl PBS; 4 h). (A to C) Representative colonic sections stained with hematoxylin and eosin from wild-type (WT) mice treated with PBS (A), WT mice treated with TcdA/TcdB (B), and ASC^−/− mice treated with TcdA/TcdB (C). Histological assessment of colonic sections stained with hematoxylin and eosin. n = 6/group/time point. (D) Colonic sections were scored for tissue architecture based on criteria specified in Materials and Methods. (E) The percentage of each section exhibiting architectural changes and inflammatory cell infiltrate was assessed. n = 6/group/time point. (F) Colonic MPO assessed at 4 h post-intrarectal toxin instillation. n = 6/group. (G) Assessment of colonic IL-1β levels 4 h post-intrarectal instillation of TcdA/TcdB. n = 6/group. *, P < 0.05 compared to WT mice treated with PBS (100-μl PBS instillation); #, P < 0.05 compared to WT mice treated with TcdA/TcdB.

TcdA and TcdB trigger cell death when administered into the colon.

In addition to triggering changes in the permeability of intestinal epithelial monolayers, C. difficile toxins cause cell detachment and cell death (6, 16, 31). Indeed, extensive colonic tissue death is a hallmark of severe CDI and an indication of poor prognosis (3, 12, 32). To determine if intrarectal instillation of TcdA/TcdB provoked intestinal epithelial cell death, we performed TUNEL staining on colonic Swiss-roll preparations (Fig. 4). PBS-treated control tissues exhibited negligible levels of TUNEL-positive cells (Fig. 4A). However, TcdA/TcdB triggered extensive cell death when administered into the colon, an effect that could be observed as early as 2 h following toxin administration (Fig. 4B). After 2 h, TUNEL-positive cells were mainly localized to the lamina propria adjacent to colonic crypts, suggesting nonepithelial cell death. At the 4-hour time point, TUNEL-positive cells were noted in the luminal region of stained sections associated with debris (Fig. 4C), indicative of epithelial cells that have been shed from the mucosa, a pathological feature reminiscent of the pseudomembranes observed in patients with CDI.

Fig 4.

Intrarectal instillation of TcdA/TcdB (50 μg in 100 μl PBS) causes extensive cell death as indicated by TUNEL positivity in colonic sections. Colonic Swiss-roll preparations stained with DAPI and TUNEL from control (A) (100 μl of PBS), 2 h of TcdA/TcdA exposure (B) (50 μg in 100 μl of PBS), or 4 h of TcdA/TcdB exposure (C) (50 μg in 100 μl of PBS).

Intrarectal administration of TcdA/TcdB increases the expression of inflammatory mediators.

We, and others, have published reports indicating the importance of the innate immune system in the initial response to C. difficile toxins (19, 27, 36, 41, 49). Furthermore, clinical reports indicate that innate inflammatory mediators, such as IL-1β and IL-8, are elevated in tissue isolated from CDI patients (43). To characterize the inflammatory response following intrarectal administration of TcdA/TcdB, we performed a bead-based cytokine array on colonic tissue protein extracts isolated from mice treated with PBS or toxin solution (50 μg) for 4 h. As outlined in Table 1, a large number of cytokines and chemokines were significantly induced following TcdA/TcdB exposure. As reported previously, toxin exposure led to a significant increase in colonic IL-1β and KC (CXCL1) (27, 36, 41, 43, 49), the murine ortholog of human IL-8. Furthermore, the chemokines monocyte chemoattractant protein 1 (MCP-1) (CCL2), MIP-1α (CCL3), MIP-1β (CCL4), and IP-10 (CXCL10) were also significantly elevated following TcdA/TcdB exposure. In addition to mediators associated with the innate immune response, T-helper type 1 (gamma interferon [IFN-γ], tumor necrosis factor alpha [TNF-α], IL-2, and IL-12)-, T-helper type 2 (IL-5 and IL-13)-, and T-helper type 17 (IL-17)-related cytokines were significantly elevated following toxin exposure.

Table 1.

Luminex bead-based analysis of inflammatory mediators isolated from colonic tissue of control (PBS-treated) or intrarectal TcdA/TcdB-treated mice (50 mg in 100 ml PBS for 4 h)^a

Analyte	Mean concn ± SEM (pg/ml)		P (t test)
	Control	TcdA/TcdB
Chemokine
CCL2/MCP-1	22.6 ± 1.7	6,650.7 ± 350.9	<0.005
CCL3/MIP-1α	4.4 ± 0.6	181.0 ± 24.1	<0.005
CCL4/MIP-1β	10.3 ± 4.7	74.5 ± 7.6	<0.005
CCL5/RANTES	39.4 ± 6.7	57.4 ± 7.9	NS
CXCL1/KC	235.5 ± 35.5	22,543.0 ± 324.2	<0.005
CXCL6/LIX	BDL	804.7 ± 113.1	<0.005
CXCL9/MIG	98.3 ± 24.4	1,139.1 ± 84.5	<0.005
CXCL10/IP-10	91.7 ± 14.4	2,547.2 ± 227.3	<0.005
Cytokine
IFN-γ	1.4 ± 0.5	46.6 ± 4.8	<0.005
IL-1α	30.7 ± 13.5	98.5 ± 5.6	<0.05
IL-1β	2.5 ± 0.8	555.2 ± 108.2	<0.005
IL-2	2.03 ± 0.3	4.8 ± 0.2	<0.005
IL-3	0.5 ± 0.3	0.8 ± 0.1	NS
IL-4	BDL	0.14 ± 0.09	NS
IL-5	0.7 ± 0.1	93.3 ± 19.7	<0.05
IL-6	4.6 ± 0.7	9,937.4 ± 338.2	<0.005
IL-7	12.7 ± 1.2	13.1 ± 1.1	NS
IL-9	16.1 ± 2.8	17.5 ± 0.8	NS
IL-10	1.4 ± 0.3	4.7 ± 0.5	<0.005
IL-12p70	4.3 ± 1.2	45.4 ± 2.1	<0.005
IL-13	21.1 ± 4.5	80.8 ± 3.5	<0.005
IL-17	0.26 ± 0.05	155.7 ± 19.0	<0.005
LIF	4.9 ± 0.5	313.4 ± 28.8	<0.005
TNF-α	0.3 ± 0.1	15.2 ± 2.8	<0.005
Growth factor
G-CSF	10.1 ± 0.8	6,201.8 ± 279.1	<0.005
M-CSF	7.3 ± 0.5	20.7 ± 0.6	<0.005
VEGF	10.4 ± 1.0	21.7 ± 2.0	<0.005

^an = 6/group. BDL, below detection limit; NS, not significant; G-CSF, granulocyte colony-stimulating factor; M-CSF, macrophage colony-stimulating factor.

TcdA and TcdB trigger the infiltration of macrophages into the colonic tissue.

It is apparent from our data, and the reports from others, that the innate immune system plays a key role in the initial inflammatory response to C. difficile toxins (19, 27, 36, 41, 49). Clinical findings suggest that innate immune cells and their respective mediators are elevated in samples of colonic tissue isolated from CDI patients (22). In our model, neutrophils play a key role in the inflammatory response to TcdA/TcdB, as evidenced by the increased colonic tissue MPO level. Our analysis of the inflammatory mediators following TcdA/TcdB exposure revealed increased levels of a numbers of chemokines known to drive macrophage chemotaxis (Table 1). To determine if macrophages were a component of the inflammatory response in the colon following intrarectal administration of TcdA/TcdB, we performed immunohistochemistry on colonic sections. Colonic sections stained for F4/80, a marker for macrophages, revealed a time-dependent increase in the number of cells in the colonic mucosa and in regions beyond the muscularis mucosa following administration of toxin (Fig. 5).

Fig 5.

Intrarectal instillation of TcdA/TcdB leads to a time-dependent increase in macrophages in colonic tissue. (A to C) Immunohistochemical analysis of F4/80-positive macrophages in colonic sections of mice treated with TcdA/TcdB (50 μg) for 2 and 4 h compared to PBS control. (D) Quantification of F4/80 macrophages expressed as cells per 250 μm of muscularis mucosa. n = 6/group/time point. *, P < 0.05 compared to control (100-μl PBS instillation); #, P < 0.05 compared to control and 2-h group.

Intrarectal administration of TcdA/TcdB from different strains triggers colonic inflammation, an effect associated with TcdA and TcdB concentrations.

To determine whether TcdA/TcdB preparations from different C. difficile strains could elicit similar inflammatory responses and associated colonic tissue, we compared the responses to toxin preparations from the VPI and 630 strains. Instillation of TcdA/TcdB produced by the 630 strain (50 μg for 4 h) resulted in an intermediate response compared to that of the control (100 μl PBS) and the VPI strain (50 μg for 4 h). Histological assessment revealed that TcdA/TcdB derived from the 630 strain increased architectural damage and colonic inflammation to levels greater than those in the control group, but the VPI group still exhibited the most severe response (Fig. 6A to C; summarized in Fig. 6D and E). The trends observed in our histological assessment were recapitulated by our biochemical assay that revealed that the TcdA/TcdB produced by the VPI strain led to the greatest increase in colonic tissue MPO, followed by TcdA/TcdB preparations produced by the 630 strain (Fig. 6F). To determine if differences in the responses to the TcdA/TcdB preparations were due to differences in the concentrations of TcdA and TcdB produced by each strain, we performed ELISAs. Indeed, the concentrations of TcdA and TcdB in the TcdA/TcdB preparations derived from cultures of the VPI strain were significantly greater than those observed in preparations derived from the 630 strain (Fig. 6G).

Fig 6.

Intrarectal installation of TcdA/TcdB derived from different C. difficile strains results in various levels of intestinal inflammation and tissue damage, an effect that correlates with TcdA and TcdB content. Representative colonic sections were stained with hematoxylin and eosin from control (A), TcdA/TcdB from the VPI strain (50 μg; 4 h) (B), and TcdA/TcdB from the 630 strain (50 μg; 4 h) (C). Histological assessment of colonic sections stained with hematoxylin and eosin. (D) Colonic sections were scored for tissue architecture based on criteria specified in Materials and Methods. (E) The percentage of each section exhibiting architectural changes and inflammatory cell infiltrate was assessed. Note that colonic sections were assessed in a blinded fashion by a board-certified pathologist. (F) Colonic MPO assessed at 4 h post-intrarectal toxin instillation. n = 6/group/time point. *, P < 0.05 compared to control (100-μl PBS installation); **, P < 0.05 compared to the 630 strain. (G) TcdA and TcdB content of TcdA/TcdB preparations from VPI and 630 strains measured by ELISA. *, P < 0.05 compared to the 630 strain.

Intrarectal administration of TcdA, but not TcdB, triggers colonic inflammation and tissue damage.

The roles of TcdA and TcdB in the pathogenesis of CDI have been subject to extensive study. Seminal studies by Triadafilopoulos et al. reported that purified TcdA, but not TcdB, triggered fluid accumulation and inflammation in the ileum (45, 46). However, a recent study by Lyras et al. revealed that mutant C. difficile strains lacking detectable TcdA production could still cause disease in a hamster infection model (29). In our intrarectal model, administration of purified TcdA (5 μg and 10 μg) resulted in the recruitment of inflammatory cells to the colonic mucosa (Fig. 7A to C, summarized in Fig. 7F and G), severe mucosal tissue damage, and an increase in tissue MPO (Fig. 7I), effects that exhibited dose dependence. Interestingly, administration of TcdB (5 μg and 10 μg) had no effect on colonic mucosal inflammation or tissue damage (Fig. 7A, D, and E, summarized in Fig. 7G and H). However, administration of purified TcdA and TcdB together at the lower concentration (5 μg each) triggered an inflammatory response greater than that observed when mice were administered TcdA (5 μg) alone (Fig. 7G and I). These data suggest that TcdA and TcdB may act synergistically to trigger colonic inflammation and tissue damage during C. difficile infection.

Fig 7.

Intrarectal instillation of purified TcdA, but not TcdB, triggers colonic tissue damage and intestinal inflammation. (A to F) Representative colonic sections were stained with hematoxylin and eosin from the following groups: control (A), 5 μg TcdA for 4 h (B), 10 μg TcdA for 4 h (C), 5 μg TcdB for 4 h (D), 10 μg TcdB for 4 h (E), and 5 μg TcdA plus 5 μg TcdB for 4 h (F). Histological assessment of colonic sections stained with hematoxylin and eosin. (G) Colonic sections were scored for tissue architecture based on criteria specified in Materials and Methods. (H) The percentage of each section exhibiting architectural changes and inflammatory cell infiltrate was assessed. Note that colonic sections were assessed in a blinded fashion by a board-certified pathologist. (I) Colonic MPO assessed at 4 h post-intrarectal toxin instillation. n = 6/group/time point. *, P < 0.05 compared to control (100-μl PBS installation) and TcdB groups (5 and 10 μg); **, P < 0.05 compared to TcdA (5 μg).

DISCUSSION

Our initial aim was to develop a mouse model that exhibits the hallmarks of CDI (i.e., a robust inflammatory response, disruption of the colonic epithelial barrier, and tissue damage) that could be used as a more expedient assessment tool to examine therapeutic strategies. We envision this model to be advantageous in preclinical validation studies for therapeutic agents designed to target the pathophysiological attributes of TcdA and/or TcdB through inhibiting their enzymatic function, preventing their binding and entry into intestinal epithelial cells, or inhibiting the inflammatory response that may result from the breakdown of the intestinal epithelial barrier.

The ileal loop surgical model is currently the gold standard for assessing the pathogenic role of TcdA and TcdB in CDI and evaluating therapeutic modalities that specifically target TcdA/TcdB actions. This model involves the surgical generation of a terminal ileal ligation, termed the ileal loop, and administration of TcdA and/or TcdB into the lumen of the small intestine. Ileal loops have been used to assess toxin-induced inflammation and intestinal tissue damage in a number of species, including rabbits, rats, and mice (4, 8, 24, 25, 38, 39, 49). We, and others, have used this model in the mouse to characterize the inflammatory response and intestinal tissue damage triggered by TcdA and TcdB (13, 19, 36, 49). However, the time required to perform the ligation procedure and the potential issues associated with small-animal surgery limit the use of the ileal loop model in large-scale mouse studies designed to assess the efficacy of therapeutic anti-TcdA/TcdB modalities. The novel toxin exposure model described here was based, in part, on the well-established protocol to induce experimental colitis in mice through the intrarectal installation of 2,4,6-trinitrobenzene sulfonic acid (TNBS) (34). Intrarectal instillation of TcdA/TcdB into the distal colon of the mouse triggered a robust inflammatory response associated with neutrophil and macrophage infiltration, increased intestinal permeability, and colonic tissue damage, all hallmark features of human CDI (21).

TcdA and TcdB alter the cytoskeleton of intestinal epithelial cells, leading to the disruption of the tight junctions and a subsequent increase in epithelial permeability (17, 18, 31). In the clinical scenario, colonoscopic assessment of CDI patients often reveals extensive tissue damage, pseudomembranes, and even signs of ischemia and necrosis (22). Histological assessment reveals significant epithelial cell death driven by both apoptotic and necrotic mechanisms (22). Similarly, in our model, we observed evidence of intestinal barrier dysfunction as early as 2 h following the instillation of TcdA/TcdB, as revealed by increased movement of luminal FITC-dextran into the circulation. This occurrence was also associated with a significant increase in the colonic MPO levels at the same time point, indicating the influx of granulocytes, likely neutrophils, into the tissue. We also observed a parallel time-dependent increase in the number of F4/80-positive macrophages stained in colonic sections. Although the histological features of the colonic sections did not indicate significant structural alterations that might contribute to the increased permeability at 2 h, the appearance of TUNEL-positive cells localized to the lamina propria adjacent to colonic crypts at this time point suggests that the introduction of TcdA/TcdB triggered cell death. At this early time point, TcdA/TcdB could be penetrating the epithelial barrier and causing cell death of immune cells, most notably macrophages (30, 42). To our knowledge, the appearance of subepithelial cell death has not been observed in vivo using other CDI models; thus, our intrarectal instillation model could be also used to study immune cell cytotoxicity following TcdA/TcdB exposure (10). Interestingly, the appearance of TUNEL-positive cells being shed from the mucosa and infiltrating neutrophils and debris in the luminal region parallels the early events thought to lead to the development of pseudomembranes, a hallmark of CDI (21, 47).

The inflammatory response to C. difficile toxins is thought to involve the activation of the innate immune system (24, 27, 36, 49). Indeed, innate inflammatory mediators such as IL-1β and IL-8 are found to be elevated in the tissue of patients with CDI (43). Furthermore, studies have correlated fecal IL-1β content with disease severity in CDI patients (43). In a previous study using colonic mucosal biopsy specimens from healthy control patients, we reported that TcdA/TcdB exposure could trigger the release of IL-1β in an inflammasome-dependent fashion (36). Following intrarectal instillation of TcdA/TcdB, we observed a significant elevation in a number of different inflammatory mediators, including IL-1β, IL-6, and IFN-γ, all of which have been reported to be associated with CDI (23, 35, 43, 48). Additionally, we observed a significant increase in the expression of KC, the mouse ortholog of human IL-8. Taken together, these data suggest that the inflammatory response observed in our model is highly representative of that observed in the tissue of CDI patients.

Finally, we examined whether purified TcdA or TcdB could individually trigger colonic inflammation and tissue damage or whether both toxins were required. It has yet to be fully determined whether TcdA or TcdB is responsible for the colonic damage and inflammation observed in human CDI. Previous studies have assessed the activity of TcdA and TcdB in the ileum, despite the colonic involvement in human CDI. In these studies, purified TcdA, but not TcdB, triggered small intestinal inflammation when administered intragastrically (28). Furthermore, administration of TcdA, but not TcdB, triggered fluid accumulation, immune cell infiltration, and tissue damage in an ileal loop (45, 46). In contrast to these studies, Lyras et al. reported that TcdA-deficient TcdB-expressing strains could still trigger disease in the hamster infection model (29). Given that the major site of colonization, inflammation, and damage in the C. difficile-infected hamster is the cecum (7), translation of these data to human CDI, which exhibits robust involvement of the descending colon, is uncertain. In our study, we observed that TcdA, but not TcdB, triggered colonic inflammation and tissue damage when administered intrarectally. However, the low dose of TcdB (5 μg) administered along with the low dose of TcdA (5 μg) synergized to trigger an inflammatory response and tissue damage similar to that observed when mice were treated with the high dose of TcdA alone (10 μg). Interestingly, Lyerly et al. (28) described a similar effect in their initial characterization of purified TcdA and TcdB, albeit in the small intestine. Taken together, these data suggest that TcdA and TcdB may act together to mediate colonic inflammation and tissue damage in CDI.

The use of reliable animal models that recapitulate key pathological features of human disease is paramount for development of therapeutic strategies for use in the clinical setting. Our intrarectal toxin instillation model displays key features of the inflammatory response and colonic tissue damage observed in CDI, including colonic tissue damage, intense neutrophilic and monocytic infiltration, and enhanced cytokine production (22). We feel that those studying the pathogenic effects of TcdA and TcdB on the host will find great value in the intrarectal instillation procedure, due to the robustness of the model, its efficient nature, and its ability to recapitulate the key feature of CDI. Furthermore, we expect that this in vivo model will be important for the development and evaluation of therapeutic strategies designed to target the toxigenic effects of TcdA and TcdB in the pathogenesis of CDI.