ABSTRACT
Commensals limit disease caused by invading pathogens; however, the mechanisms and genes utilized by beneficial microbes to inhibit pathogenesis are poorly understood. The attaching and effacing mouse pathogen Citrobacter rodentium associates intimately with the intestinal epithelium, and infections result in acute colitis. C. rodentium is used to model the human pathogens enterohemorrhagic Escherichia coli and enteropathogenic E. coli. To confirm that Bacillus subtilis, a spore-forming bacterium found in the gut of mammals, could reduce C. rodentium-associated disease, mice received wild-type B. subtilis spores and 24 h later were infected by oral gavage with pathogenic C. rodentium. Disease was assessed by determining the extent of colonic epithelial hyperplasia, goblet cell loss, diarrhea, and pathogen colonization. Mice that received wild-type B. subtilis prior to enteric infection were protected from disease even though C. rodentium colonization was not inhibited. In contrast, espH and hag mutants, defective in exopolysaccharides and flagellum production, respectively, did not protect mice from C. rodentium-associated disease. A motAB mutant also failed to protect mice from disease, suggesting that B. subtilis-mediated protection requires functional flagella. By expanding our current mechanistic knowledge of bacterial protection, we can better utilize beneficial microbes to prevent intestinal disease caused by pathogenic bacteria, ultimately reducing human disease. Our data demonstrate that wild-type B. subtilis reduced disease caused by C. rodentium infection through a mechanism that required espH and functional flagella.
INTRODUCTION
The mammalian intestinal tract is colonized extensively by a complex microbiota. An estimated 500 to 1,000 microbial species are present within the gastrointestinal tract, and commensal microbes outnumber human cells 10-fold (37). Studies of germfree animals have demonstrated that normal, healthy gut development is dependent on the intestinal microbiota (15). Moreover, many studies suggest that alterations in host-microbe interactions may initiate or perpetuate intestinal diseases such as inflammatory bowel disease (7, 26, 30). Perturbing the commensal microbiota by administering antibiotics makes the host more susceptible to a variety of intestinal pathogens, including Salmonella enterica, Clostridium difficile, enterohemorrhagic Escherichia coli (EHEC), enteropathogenic E. coli (EPEC), and Citrobacter rodentium (11, 18). In contrast, administration of probiotic or beneficial microbes can prevent or ameliorate intestinal inflammation that results from chemical irritation or pathogen infection (8). Studies such as these demonstrate that the composition of the microbiota contributes to intestinal health; however, the bacterial and host mechanisms responsible for this beneficial effect remain largely unknown.
Several species of spore-forming bacteria are present within the mammalian gastrointestinal tract, including members of the genera Clostridia and Bacillus; however, few studies have examined the role of spore-forming bacteria in intestinal health. Most commensals cannot be cultured, and of the species that can be grown in the laboratory, many are not genetically tractable, making it challenging to determine the bacterial genes involved in intestinal immune regulation. Several Bacillus spp., including B. subtilis, are aerobic spore-forming bacteria present in the guts of humans and mice (10, 14). Bacillus spp. are complex organisms that exist as vegetative cells or metabolically inert spores or as part of a multicellular biofilm encased in a matrix. For decades, B. subtilis has been used as a model Gram-positive organism, and the genetics involved in flagellum synthesis, sporulation, and biofilm development have been elucidated. Since Bacillus spp. are present in the mammalian microbiota and B. subtilis is genetically tractable, B. subtilis is an attractive candidate to identify bacterial genes responsible for the immunomodulatory properties of beneficial microbes in a host infected with an enteric pathogen.
Diarrheal diseases caused by enteric pathogens such as EHEC and EPEC are a worldwide problem, affecting millions of children and adults annually, with few prophylactic and treatment modalities available. C. rodentium is genetically similar to the human attaching and effacing (A/E) pathogens EPEC and EHEC (3), and infections with these pathogens result in A/E lesions on the host epithelium. Because EPEC and EHEC colonize mice poorly, mice infected with C. rodentium are used as a model of A/E infections and acute colitis (3, 25). Following oral challenge of mice with 108 to 109 CFU, C. rodentium colonizes the cecal patch within a few hours and the colon 24 to 48 h postinfection; peak colonization typically is found 10 days postinfection (dpi) (35). C. rodentium infection causes acute colitis which is characterized by epithelial hyperplasia, loss of goblet cells, and mucosal infiltration by neutrophils and T cells (13, 21, 25, 31). Symptoms also include diarrhea. In most strains of immunocompetent mice, C. rodentium is not lethal and is cleared by both innate and adaptive immune responses 3 to 4 weeks postinoculation; pathogen clearance requires MyD88-dependent signaling and CD4+ T cells (12, 22, 33).
Our study has confirmed that a specific beneficial bacterium, B. subtilis, if given orally to mice 24 h prior to exposure to pathogenic C. rodentium (9), greatly decreases intestinal disease even though the pathogen colonizes successfully. We further identified two B. subtilis mutants, espH and hag mutant strains, defective in exopolysaccharide and flagellum synthesis, respectively, that did not protect mice from C. rodentium-associated disease. When the two nonprotective mutants were administered together or the mice received a motAB mutant with nonfunctional flagella, the protective effect of wild-type B. subtilis was not restored. These novel findings suggest that espH and functional flagella must be present in the same bacterium in order to limit disease that results from infection with an enteric pathogen. This study increases our understanding of the bacterial genes required to limit intestinal disease.
MATERIALS AND METHODS
Mice.All animal experiments were performed according to protocols approved by the Institutional Animal Care and Usage Committee at Loyola University Medical Center, Maywood, IL. C57BL/6 founders were from Charles River Laboratory (Wilmington, MA), and mice utilized for experiments (4 to 6 weeks of age) were bred at Loyola University Chicago, Chicago, IL. Sterile standard chow and tap water were given to mice ad libitum.
Bacterium and spore preparation.B. subtilis wild-type 3610, DS76 (espH mutant), DS1143 (hag mutant), and DS222 (motAB mutant) strains were kindly provided by Daniel Kearns (Indiana University, Bloomington) and have been described previously (19). B. subtilis spores were germinated via exhaustion as described previously (9). Briefly, a 16-h B. subtilis culture was diluted 1:100 in Difco sporulation medium (DSM) and incubated for 48 h at 37°C, with shaking at 250 rpm. Bacteria were pelleted (at 10,000 × g for 10 min at 4°C), washed in ice-cold sterile water, and heat shocked at 68°C for 30 min to kill any remaining vegetative cells. Spores were washed three times with ice-cold water and stored at 4°C until being administered to mice. Spores were counted by culturing serial dilutions, and this count was verified using a Petroff-Hausser counting chamber. On the day of administration, B. subtilis spores were washed with ice-cold sterile water, resuspended in 100 μl of phosphate-buffered saline (PBS), and given to mice via oral gavage. For infection studies, C. rodentium ATCC 51459 was cultured for 16 h in Luria-Bertani (LB) medium and washed once in PBS and an infectious dose of 5 × 108 CFU was resuspended in 100 μl of sterile PBS for administration to mice by oral gavage.
Study design.B. subtilis spores (109 in 100 μl PBS) were orally administered to specific-pathogen-free C57BL/6 mice 24 h prior to infection with 5 × 108 CFU of C. rodentium by oral gavage. Age- and gender-matched mice were utilized for each experiment. To assess disease, mice were euthanized 10 days postinfection and tissues were collected. Day 10 postinfection was chosen because at this time the pathogen is well established and colitis is evident (23). To assess if mice had diarrhea, feces were examined and scored as described previously (27). Briefly, normal feces are hard, dry pellets and are scored as 1. Mice with mild to severe diarrhea have slightly soft stool (scored as 2), very soft stool (scored as 3), or unformed stool (scored as 4). Colons were collected and processed for histological analysis.
Histology.Distal colons from euthanized mice were collected and fixed overnight in 10% formalin-buffered phosphate, then dehydrated through a gradient of alcohol, cleared with xylene, and infiltrated with paraffin. After processing and infiltration, the tissues were sectioned longitudinally at 4-μm intervals and stained with hematoxylin and eosin (H&E). Epithelial hyperplasia and goblet cell loss in the distal colon were evaluated as described previously (36). Briefly, images of each colon were taken using a Leica DM IRB fluorescence microscope equipped with a MagnaFire charge-coupled device camera. Heights of five well-oriented crypts per mouse were measured from 2 to 3 regions. Epithelial cells are shed during C. rodentium infection, and intact crypts were difficult to find in diseased mice.
Colonization by C. rodentium.C. rodentium colonization was assessed by collecting fresh fecal samples 7, 10, 14, 21, and 28 days post-pathogen infection. Feces were homogenized in 500 μl of sterile 20% glycerol in PBS, and serial dilutions were cultured on selective MacConkey agar plates. Plates were incubated for 16 h at 37°C, and only colonies that displayed the characteristic pink center surrounded by a white rim (C. rodentium) were counted. The level of colonization was calculated and expressed as the number of CFU per gram of feces.
B. subtilis colonization.B. subtilis colonization was assessed by collecting fresh fecal samples 1, 2, 3, 4, 6, 8, and 11 days postinoculation. Feces were homogenized in 500 μl of sterile 20% glycerol in PBS, and serial dilutions were cultured on selective LB medium plates with chloramphenicol (5 μg/ml) to determine the total number of B. subtilis CFU. Vegetative cells were killed by heating at 68°C for 30 min, and serial dilutions were plated onto LB medium plates to quantify the spores. Plates were incubated for 16 h at 37°C; the resulting colonies are represented as the number of CFU per gram of feces.
Germination assay.B. subtilis spores (106 spores/ml) were incubated at 37°C in LB medium to initiate germination. Aliquots were removed at 0, 15, 30, and 60 min. Untreated aliquots were serially diluted on LB medium plates to determine the total number of bacteria present. The remaining aliquots were heated at 68°C for 30 min to kill vegetative cells. The heated samples were serially diluted on LB medium plates to quantify spores.
Biofilm assay.Overnight cultures were diluted to an optical density at 600 nm (OD600) of 0.1, and 20 μl of this dilution was used to inoculate a well containing 2 ml of LB medium. A glass coverslip was positioned diagonally with respect to the bottom of the well. The plates were incubated at 37°C. After 24 h, the coverslip containing the pellicle (biofilm at the liquid-air interface) was removed from the well, washed three times with water, stained with 0.2% crystal violet for 15 min, and washed three times. The crystal violet was redissolved in ethanol, and the OD450 was determined. A coverslip treated with medium alone was utilized to determine the retention of crystal violet in the absence of bacteria. The OD450 for this sample was subtracted as background from the corresponding values for all other samples.
Motility assay.Overnight B. subtilis cultures were diluted to an OD600 of 0.1, and 0.5 μl was used to inoculate a motility plate (containing LB medium with 0.3% agar). Plates were incubated at 37°C for 8 h, and then colony diameters were measured.
Statistical analysis.All experiments were performed a minimum of three times, and results were analyzed using Student's t test. Error bars denote standard errors of the mean (SEM). Differences were considered statistically significant if P was <0.05.
RESULTS
Effect of wild-type B. subtilis pretreatment on C. rodentium-infected mice.To establish if wild-type B. subtilis could protect mice from disease induced by pathogenic C. rodentium, mice were given 109 wild-type B. subtilis spores 24 h prior to infection with 5 × 108 CFU C. rodentium by oral gavage. Ten days postinfection, mice (n = 8 to 12/condition) were assessed for signs of disease, which included colonic epithelial cell hyperplasia, goblet cell loss, and diarrhea. In colonic sections, we observed that the crypt height in C. rodentium-infected mice increased roughly 40% compared to that in the control PBS- and B. subtilis-treated mice (Fig. 1A to C and E). In contrast, mice pretreated with wild-type B. subtilis prior to pathogen infection were protected from C. rodentium-associated colonic epithelial cell hyperplasia and were not statistically different from the PBS controls (Fig. 1D and E). Goblet cells were readily observed in colonic tissues from uninfected and B. subtilis-pretreated mice (Fig. 1A, D, and F), whereas very few goblet cells were observed in C. rodentium-infected mice (Fig. 1C and F), consistent with the results of previous studies (2).
Effect of wild-type B. subtilis on C. rodentium-associated disease 10 days postinfection as demonstrated by histological analysis of colonic tissues. (A to D) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received PBS only (A), wild-type (wt) B. subtilis only (B), C. rodentium (Cr) only (C), or wild-type B. subtilis spores prior to C. rodentium infection (D). Orange arrows indicate goblet cells. (E and F) Colonic crypt heights (E) and goblet cell measurements (F). Results are averages from three independent experiments, and a total of 6 to 12 mice were assessed for each group. P values were calculated by Student's t test.
To determine if B. subtilis could prevent other symptoms of C. rodentium-associated disease such as diarrhea, we examined feces from mice. C. rodentium-infected mice had mild to moderate diarrhea, and only one mouse treated with B. subtilis prior to infection had soft stool (Fig. 2). In contrast, mice treated with PBS or wild-type B. subtilis spores alone did not have soft stool or diarrhea (Fig. 2). These data suggest that B. subtilis reduced C. rodentium-induced diarrhea.
Effect of wild-type B. subtilis on C. rodentium-associated disease 10 days postinfection as demonstrated by diarrhea scores. Results are averages from three independent experiments; a total of 8 to 12 mice were assessed for each group. P values were calculated by Student's t test.
C. rodentium colonization at 7, 10, 14, 21, and 28 days postinfection was assessed to determine if treatment with wild-type B. subtilis prevented colonization by, or enhanced clearance of, the pathogen. Serial dilutions of homogenized feces cultured on MacConkey agar plates showed, as reported previously, that mice were highly colonized with C. rodentium 7 days postinfection and that the pathogen was cleared 21 to 28 days postinfection (13, 35). At every time point, mice that received B. subtilis prior to infection had numbers of C. rodentium CFU similar to those in mice exposed to only C. rodentium (Fig. 3). No C. rodentium was detected in the feces of either group 28 days postinfection. These data suggest that B. subtilis did not delay colonization by, or enhance clearance of, C. rodentium.
C. rodentium colonization was assessed by culturing serial dilutions of fecal homogenates. N.D., none detected (the level was less than 103 CFU/g feces). ●, values for mice receiving C. rodentium only; ■, values for mice receiving wild-type B. subtilis prior to C. rodentium infection. Results are averages from three independent experiments, and a total of 8 to 10 mice were assessed for each group.
From these data, we conclude that wild-type B. subtilis attenuated disease associated with C. rodentium infection since mice pretreated with wild-type B. subtilis prior to pathogen infection had normal colonic crypt heights and no goblet cell loss. Additionally, wild-type B. subtilis prevented diarrhea in the treated mice and it did not mediate protection by reducing C. rodentium colonization.
Genes required for B. subtilis-mediated protection.Since gastrointestinal biofilms contribute to intestinal health (24, 32) and because an earlier study demonstrated that wild-type Lactobacillus crispatus, but not a mutant strain that lacked the capacity to initiate biofilms, protects mice from dextran sulfate-induced colitis (5), we hypothesized that the capacity of B. subtilis to adhere and form a biofilm contributed to the protective effect of B. subtilis. To determine if B. subtilis biofilms are required for protection, we treated mice with an espH or tasA B. subtilis strain defective in biofilm formation prior to C. rodentium infection. As a negative control, we pretreated mice with a hag mutant which forms biofilms but cannot make flagella (19). Mice were pretreated with each mutant prior to pathogen infection, and disease was assessed 10 days postinfection with C. rodentium by examining the colon and feces. As predicted, the espH biofilm mutant did not protect mice from disease, as demonstrated by colonic epithelial hyperplasia (Fig. 4B and E), goblet cell loss (Fig. 4F), and diarrhea (Fig. 4G). Surprisingly, administration of the other biofilm mutant, the tasA strain, provided protection and suppressed C. rodentium-associated disease (data not shown). Also unexpectedly, mice pretreated with the hag strain, the flagellum mutant, prior to pathogen infection had colonic hyperplasia (Fig. 4D and E), goblet cell loss (Fig. 4F), and diarrhea (Fig. 4G). As expected, administration of a B. subtilis strain alone did not induce disease, indicating that the dosage of B. subtilis administered was not detrimental (Fig. 1 to 4). From these data, we conclude that the espH and hag B. subtilis mutants do not protect mice from C. rodentium-induced disease.
Effects of espH and hag spores on C. rodentium-associated disease 10 days postinfection. (A to D) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received espH B. subtilis spores (A), espH B. subtilis spores prior to C. rodentium infection (B), hag B. subtilis spores (C), and hag B. subtilis spores prior to C. rodentium infection (D). H&E-stained sections from PBS-treated and C. rodentium-infected mice are shown in Fig. 1. (E to G) Colonic crypt height measurements (E), goblet cell measurements (F), and diarrhea scores (G). Results are averages from three independent experiments, and a total of 6 to 12 mice were assessed for each group. ∗, P < 0.05 (Student's t test).
We tested if the inability of the espH and hag mutants to induce a protective response resulted from colonization defects by culturing fecal homogenates and found no differences in the number of CFU between wild-type and mutant strains 4 days after B. subtilis administration (Fig. 5A). Spores were the dominant form recovered from the fecal samples for all strains (Fig. 5A). After 4 days, B. subtilis CFU numbers were below the level of detection (102 CFU/g feces). At later time points (days 6 to 11 postinoculation), we recovered a few spores from all animals, although not at every time point, suggesting that all strains of B. subtilis persisted in the gastrointestinal tract at very low levels throughout the course of our experiment.
Assessment of B. subtilis colonization, germination, motility, and biofilm development. (A) B. subtilis colonization was assessed by culturing serial dilutions of fecal homogenates. Results are averages from two independent experiments, and a total of 5 or 6 mice were assessed. (B) Germination of B. subtilis spores in LB medium. (C) B. subtilis motility in soft-agar plates. (D) Biofilm formation by B. subtilis strains. Germination, motility, and biofilm results are averages from three independent experiments.
We also assessed germination, biofilm formation, and motility for each strain to determine if any unexpected phenotypes occurred (e.g., was the flagellum mutant also impaired in biofilm formation?). First, we examined the germination capacities of all strains, since a reduced capacity to germinate in either the espH or hag mutant could explain why the espH and hag strains were not protective. More than 99.5% of spores germinated within 1 h, and no differences were observed at any of the earlier time points assessed (Fig. 5B). The majority of B. subtilis bacteria in feces were recovered as spores (Fig. 5A), making it difficult to assess if the mutants differ from the wild type in the ability to germinate in vivo. However, since espH and hag are expressed only by vegetative cells, it is difficult to envision how these genes could alter the germination of spores in vivo. Collectively, our data suggest that all strains are able to germinate successfully.
Additionally, we performed a motility assay with all strains to ensure that the espH mutant was as motile as the wild type. Both the wild type and the espH mutant were able to swim and form a large colony in soft-agar plates (Fig. 5C). In contrast, the hag mutant, which does not produce flagella and does not swim, did not form a large colony (Fig. 5C). These data suggest that at least in vitro the espH mutant is motile. Finally, we assessed if the hag mutant was comparable to the wild type in its capacity to synthesize matrix exopolysaccharides and form a biofilm. Equal quantities of crystal violet were retained by the wild type and the hag mutant (Fig. 5D), indicating that these strains are capable of biofilm formation. As expected, little to no crystal violet was retained by the biofilm-deficient espH mutant (Fig. 5D). In summary, we did not observe any unexpected differences in germination, biofilm formation, or motility for any of the three strains. We suggest that lack of protection by espH and hag mutants was due not to defects in colonization or germination, but instead to the absence of flagella and exopolysaccharides.
Effect of coadministration of espH and hag spores to C. rodentium-infected mice.Since neither the espH nor the hag mutant was sufficient for protection from an enteric pathogen, we tested whether administration of the two mutants together could suppress C. rodentium-associated disease. We hypothesized that if two immunomodulatory signals are required, then administration of both mutant spores could restore the protective effect of wild-type B. subtilis. To test this possibility, we introduced equal numbers of espH and hag spores into mice 24 h prior to C. rodentium infection. Ten days postinfection, mice were assessed for disease by examination of colonic and fecal samples. Coadministration of mutant spores did not restore the protective effect of wild-type B. subtilis since the pretreated mice had colonic epithelial hyperplasia, goblet cell loss, and diarrhea (Fig. 6). These data indicate that espH and hag must be expressed by the same bacterium for B. subtilis to mediate its protective effect.
Effect of coadministration of espH and hag spores on C. rodentium-associated disease 10 days postinfection. (A and B) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received espH and hag B. subtilis spores only (A) or espH and hag B. subtilis spores prior to C. rodentium infection (B). H&E-stained sections from PBS-treated and C. rodentium-infected mice are shown in Fig. 1. (C to E) Colonic crypt height measurements (C), goblet cell measurements (D), and diarrhea scores (E). Results are averages from three independent experiments; a total of 6 to 8 mice were assessed for each group. ∗, P < 0.05 (Student's t test).
Effect of administration of motAB spores to C. rodentium-infected mice.Since flagella are important for bacterial localization and are also recognized by several host receptors, such as Toll-like receptor 5 (TLR5) and IPAF, we hypothesized that the inability of the hag mutant to protect mice could be due to a lack of motility or a failure to produce ligands to modulate host signaling pathways. If protection by B. subtilis requires localization to a particular niche, we hypothesized that a motAB mutant, which produces structurally intact flagella but lacks the stator proteins required for turning of the flagella (16), would not be protective, since the motAB mutant still produces the ligands required for the host receptors TLR5 and IPAF but is not motile (16). We found that mice that received the motAB strain prior to infection with C. rodentium were not protected from disease and had colonic cell hyperplasia, goblet cell loss, and diarrhea (Fig. 7). We conclude that B. subtilis utilizes functional flagella to be protective.
Effect of motAB spores on C. rodentium-associated disease 10 days postinfection. (A and B) Sections of fixed colons were observed at a magnification of ×100 after H&E staining. Images are representative of mice that received motAB B. subtilis spores only (A) or motAB B. subtilis spores prior to C. rodentium infection (B). H&E-stained sections from PBS-treated and C. rodentium-infected mice are shown in Fig. 1. (C to E) Colonic crypt height measurements (C), goblet cell measurements (D), and diarrhea scores (E). Results are averages from three independent experiments; a total of 6 to 8 mice were assessed for each group. ∗, P < 0.05 (Student's t test).
DISCUSSION
Commensal bacteria contribute to mammalian health in a variety of ways, and multiple studies have established a relationship between microbial composition and intestinal health. Further, several groups have established that administration of beneficial microbes attenuates inflammation induced by chemical irritants and intestinal pathogens. Here, we show that a beneficial microbe, B. subtilis, protects from C. rodentium-associated disease and the protective effect is mediated by genes involved in exopolysaccharide and flagellum production. These results confirm and extend those of D'Arienzo et al. (9), who previously demonstrated that multiple doses of B. subtilis PY79 reduce C. rodentium-associated disease.
The finding that the administration of both espH and hag spores did not protect mice from disease suggests that both genes must be expressed by the same bacterium to generate a protective response. Additionally, since the motAB mutant, with nonfunctional flagella, did not protect mice from disease, we suggest that flagella are required for at least the localization of B. subtilis to a particular intestinal niche. We further suggest that, once B. subtilis is in the correct niche, espH and perhaps flagella are required for host immunomodulation. Consistent with our model, previous studies demonstrated that the exopolysaccharide polysaccharide A from Bacteroides fragilis suppresses chemically induced intestinal inflammation by expanding anti-inflammatory T regulatory cell populations in the mesenteric lymph nodes and lamina propria (28, 29). These studies suggest that bacterial polysaccharides, particularly those produced by beneficial microbes, have profound roles in maintaining intestinal homeostasis and limiting inflammation induced upon infection with enteric pathogens. Previous studies also demonstrated that administration of flagella alone is sufficient to elicit an anti-inflammatory response and protect mice from the intestinal pathogens Clostridium difficile and Enterococcus faecium (18, 20). B. subtilis flagella, in addition to localizing bacteria to the correct niche, may suppress intestinal inflammation directly. Future experiments are needed to distinguish which of these possibilities is occurring.
Microbe-associated molecular patterns (MAMPs) are conserved bacterial motifs which are recognized by host pattern recognition receptors (PRRs) such as TLRs, Nod-like receptors (NLRs), and C-lectins (7, 34). Well-studied bacterial MAMPs include flagella and lipopolysaccharide (LPS). Previously, it was shown that suppression of intestinal disease by bacterial flagella or exopolysaccharides, such as polysaccharide A, requires TLR5 or TLR2, respectively (18, 28). Based on these data, we predict that B. subtilis mediates its protective effects via TLR5 and/or TLR2.
Although we hypothesize that B. subtilis induces a protective immune response that limits C. rodentium-associated disease, B. subtilis may instead prevent disease by interfering with C. rodentium pathogenesis. Bacterial pathogenesis can be reduced by altering pathogen colonization or suppressing production of virulence factors. The data from this study demonstrate that wild-type B. subtilis did not mediate its protective effect by preventing colonization by or enhancing clearance of C. rodentium, since similar pathogen titers were observed in the presence of B. subtilis. However, colonization by C. rodentium was assessed by culturing fecal homogenates, and this technique does not assess the quantities of C. rodentium bacteria bound to intestinal mucus or the host epithelium. C. rodentium is an A/E pathogen that binds directly to the host epithelium and injects virulence factors into host cells to cause disease (3). In B. subtilis-treated mice, C. rodentium may colonize the mucus layer rather than directly attaching to the host epithelium, and thus B. subtilis may limit disease by altering the site of C. rodentium colonization rather than suppressing C. rodentium growth. Alternatively, B. subtilis may suppress disease by inhibiting the locus for enterocyte effacement (LEE) operons, where many C. rodentium virulence factors are encoded (13, 23). These possibilities could be tested in future experiments.
For initial studies to confirm that B. subtilis could prevent disease associated with C. rodentium, we utilized spores instead of vegetative cells, since spores remain viable after transiting through the stomach, allowing more viable bacteria to be administered to the host. Interestingly, others have shown that spore-forming Clostridia spp. induce transforming growth factor β (TGF-β) production and an environment that promotes the production of colonic T regulatory cells (1), which can suppress inflammation. Conversely, uncultivable spore-forming segmented filamentous bacteria (SFB) promote the development of inflammatory Th17 cells in the small intestine and mice colonized with SFB produce higher quantities of antimicrobial defensins and are more resistant to C. rodentium infection (4, 17). These studies indicate that spore-forming bacteria may be important regulators of intestinal homeostasis and that select spore-forming bacteria could prevent disease due to infection with an enteric pathogen. Bacterial spores begin to germinate in the small intestine, and vegetative cells are present in the small and large intestines (6). Previous studies did not determine if vegetative cells or spores are required for a protective effect. The results of this study suggest that B. subtilis spores germinate and that vegetative cells, which express espH and functional flagella, are needed for a protective effect during C. rodentium infection.
Our study demonstrates that wild-type B. subtilis reduced disease associated with C. rodentium infection, including epithelial cell hyperplasia, goblet cell loss, and diarrhea, even though pathogen colonization was not reduced. This protective effect required espH and functional flagella since B. subtilis mutants lacking either of these features did not prevent disease. Further, when the espH and hag mutants were administered together prior to infection with C. rodentium, the protective effect was not restored. These findings identify novel genes that are required to prevent disease due to an enteric pathogen. This study increases our molecular understanding of how beneficial microbes and mammalian hosts cooperatively interact to prevent detrimental inflammation.
ACKNOWLEDGMENTS
We thank Daniel Kearns for all B. subtilis strains utilized in this study. In addition, we thank members of the Knight lab for many thoughtful discussions. We thank Stephen M. Small for technical assistance.
This work was supported by the National Institutes of Health (grant NIAID 50260 to K.L.K.).
FOOTNOTES
- Received 22 August 2011.
- Returned for modification 24 September 2011.
- Accepted 22 November 2011.
- Accepted manuscript posted online 5 December 2011.
- Copyright © 2012, American Society for Microbiology. All Rights Reserved.
