ABSTRACT
Butyrivibrio fibrisolvens MDT-1, a butyrate-producing strain, was evaluated for use as a probiotic to prevent enterocolitis. Oral administration of the MDT-1 strain (109 CFU/dose) alleviated the symptoms of colitis (including body weight loss, diarrhea, bloody stool, organic disorder, and mucosal damage) that are induced in mice drinking water that contains 3.0% dextran sulfate sodium. In addition, myeloperoxidase (MPO) activity levels in colonic tissue were reduced, suggesting that MDT-1 mitigates bowel inflammation. The addition of MDT-1 culture supernatant inhibited the growth of nine clinical isolates of Campylobacter jejuni and Campylobacter coli that could potentially cause enterocolitis. Infection of mice with C. coli 11580-3, one of the isolates inhibited by MDT-1 in vitro, resulted in diarrhea, mucosal damage, increased MPO activity levels in colonic tissue, increased numbers of C. coli in the cecum, and decreased body weight gain. However, administration of MDT-1 to mice, prior to and during C. coli infection, reduced these effects. These results suggest that Campylobacter-induced enterocolitis can be alleviated by using B. fibrisolvens as a probiotic.
Since butyrate produced by intestinal microbiota has been implicated in protection against colon cancer and inflammatory bowel disease (IBD) (16), augmentation of butyrate production in the intestine is desirable for the maintenance of colonic health in humans and animals. We reported previously that oral administration of Butyrivibrio fibrisolvens MDT-1 (a butyrate-producing strain newly isolated from the rumen) to mice increased the rate of butyrate production by fecal microbiota and alleviated the formation of colorectal aberrant crypt foci (30). In this study, our aim was to investigate using the MDT-1 strain to alleviate the symptoms of enterocolitis, including IBD.
Enterocolitis occurs frequently in humans and animals, but the pathogenesis of IBDs, including ulcerative colitis and Crohn's disease, remains unknown. Although corticosteroids and sulfasalazine are currently used commonly for the treatment of patients with IBD, the utility of these agents is limited by their adverse effects (35). In the search for new therapies for IBD, it has been reported that oral administration of Lactobacillus plantarum (23), Streptococcus salivarius (39), Escherichia coli Nissle strain 1917 (25, 33), Bifidobacterium longum (15), Lactobacillus casei Shirota (26), VSL-3 (6), and some bioactive substances (19) partially suppressed IBD.
IBD can be experimentally induced by administering dextran sulfate sodium (DSS) orally in mice (11, 32), rats (37), hamsters (40), or guinea pigs (21). DSS can produce both acute and chronic ulcerative colitis, with features somewhat similar to the symptomatic and histological findings for colitis in humans (11, 32). The pathogenesis of DSS-induced colitis is presently unclear, but toxic effects on colonic epithelium (12, 32), alterations of luminal bacterial flora (32, 40), and activation of macrophage inflammatory responses (31, 32) have been suggested as possible mechanisms.
Polymorphonuclear neutrophils (PMNs) are the most abundant cell type in intestinal lesions in ulcerative colitis (38). Myeloperoxidase (MPO) is present mainly in PMNs, and MPO activity reflects the number of PMNs (8). Thus, MPO activity can be used as a quantitative index of intestinal inflammation (38).
In general, enterocolitis is often induced by pathogenic bacterial infection, and Campylobacter is now recognized as one of the major agents of enterocolitis in humans (7, 9) and animals (34). Many other pathogenic bacteria, including Escherichia coli O-157, are also known to induce enterocolitis (5). Because several strains of B. fibrisolvens have been reported to produce bacteriocins that are antibiotic against specific bacteria (22), the objective of the present study was to examine whether administration of B. fibrisolvens MDT-1 improves DSS-induced experimental colitis and Campylobacter-induced enterocolitis.
MATERIALS AND METHODS
Bacterial strain and culture conditions. B. fibrisolvens MDT-1 was obtained as described previously (30). Clostridium butyricum MIYAIRI 588, Campylobacter coli (15 strains; isolated from farm animals), and Campylobacter jejuni (5 strains; isolated from farm animals) were provided by Meiji Seika Kaisha, Ltd. (Tokyo, Japan).
B. fibrisolvens and C. butyricum were grown in B. fibrisolvens medium, which consisted of a basal medium (14) supplemented with short-chain fatty acids, vitamins, and trace minerals (13). Culture conditions and procedures were previously described (14). Cell growth was estimated by changes in optical density at 600 nm. Cells were harvested at an optical density at 600 nm of 1.3 to 1.5 (late log growth) and cooled immediately on ice. Viable cells were counted using Hungate's roll tube method in a medium containing 2% agar (29). Viability was scored as the mean number of colonies from counts of five separate roll tubes for each sample.
Mice.Four-week-old male Jcl:ICR mice (CLEA Japan, Tokyo, Japan) were used for all experiments. Mice were housed in plastic cages with wire tops and permitted free access to a commercial diet (CE-2; CLEA Japan, Tokyo, Japan) and water. The diet consisted of a mixture of 249 g crude protein, 46 g crude fat, 37 g crude fiber, 67 g ash, and 514 g nitrogen-free extracts, with an energetic content of 829 kJ/kg feed. All of the animal experiments in this study were approved by the Institutional Animal Care and Use Committee of Meiji University (IACUC-05-0007).
Oral administration of B. fibrisolvens and C. butyricum to mice. B. fibrisolvens MDT-1 and C. butyricum grown to late log stage were collected by centrifugation (5,000 × g, 10 min, 4°C) and washed with 0.8% NaCl. Washed cells resuspended in 0.8% NaCl (1013 CFU/liter) were administered to mice by gavage at a level of 0.1 ml or 109 CFU · mouse−1 · dose−1 (or day−1).
Animal experiment 1 (DSS-induced colitis).The protocol of this experiment is outlined in Fig. 1. Ten mice were used for each group. Using a stomach tube, mice in group A (control) were given 0.1 ml of 0.8% NaCl solution once a day for 3 days, followed by 7 days of drinking water. Group B mice were given 0.1 ml of 0.8% NaCl solution once a day for 3 days, followed by 7 days of drinking water containing 3% DSS (molecular weight, 40,000; MP Biomedicals, Germany) (32). Live MDT-1 and C. butyricum (109 CFU/dose) were administered once a day for 3 days to group C and D mice, respectively, followed by 7 days of drinking water containing 3% DSS.
Protocols for animal experiments 1 (A) and 2 (B). Group A, addition of 0.8% NaCl solution (N) for 10 days (control); group B, addition of 0.8% NaCl for 3 days and then DSS (D); group C, addition of B. fibrisolvens MDT-1 (B) for 3 days and then DSS; group D, addition of C. butyricum (Cb) for 3 days and then DSS; group E, addition of 0.8% NaCl for 10 days (control); group F, addition of 0.8% NaCl for 10 days and C. coli (Cc) on day 0; group G, addition of B. fibrisolvens MDT-1 for 10 days and C. coli on day 0. A, autopsy.
Daily clinical assessment of DSS-treated animals included measurement of the volume of drinking water consumed and body weight, evaluation of stool consistency, and the absence or presence of blood in the feces by ocular inspection. After 10 days, mice were sacrificed by cervical dislocation under ether anesthesia for autopsy. For each mouse, the liver and spleen were weighed separately, and the colon was removed. After exclusion of the cecum, colon length and weight were measured and tissue samples were then prepared for histological (five mice in each group) and MPO activity (the other five mice in each group) evaluations.
Disease activity index.Disease activity was quantified using clinical scores previously established by Hartmann et al. (18) that assess body weight loss, stool consistency, and rectal bleeding. For body weight, no loss was counted as 0 points, 1 to 5% loss as 1 point, 5 to 10% loss as 2 points, 10 to 20% loss as 3 points, and >20% loss as 4 points. For stool consistency, no points were given for well-formed pellets, 2 points were given for pasty and semiformed stools that did not stick to the anus, and 4 points were given for liquid stools that did stick to the anus. Bleeding was scored as 0 points for no blood observed, 2 points for a positive finding of blood, and 4 points for gross bleeding. The mean of these three scores represented the clinical score, with a theoretical range of 0 (healthy) to 4 (maximal severity of disease).
Histology.Histological examination was performed on samples of distal colon coming from 5 mice in each experimental group of 10 animals. Samples were fixed in 10% formalin before being embedded in paraffin. After deparaffinization, samples were cut into 5-μm-thick sections and stained with hematoxylin and eosin. The mucosal damage was evaluated independently by two investigators blinded to the study groups and quantified using mucosal damage scores previously established by Oda (28). These scores derived from the two investigators were averaged. The following three parameters were used: surface epithelial loss, crypt destruction, and inflammatory cell infiltration into the mucosa. A score of 0 to 4 was assigned to each of the three parameters according to the extent and severity of the change: 0, no change; 1, localized and mild; 2, localized and moderate; 3, extensive and moderate; and 4, extensive and severe. The sum of the scores from the three parameters was defined as the mucosal damage score for each animal. In addition, toluidine blue staining (pH 2.5) was also performed to evaluate sulfated polysaccharide in the mucosa (21).
MPO activity of colonic tissue.MPO activity was assayed according to Tian et al. (38). Briefly, the most distal 6-cm segment of the colon was mashed with mortar and pestle in 1% (wt/wt) HTAB (hexadecyltrimethylammonium bromide; Wako, Osaka, Japan) dissolved in 20 mM potassium phosphate buffer (pH 6.0, 50 mg of tissue/ml). The mashed tissue was sonicated, and the extract was then cleared by centrifugation (20,000 × g, 15 min, 4°C). The supernatant (0.01 ml) was mixed in a microplate well with 0.2 ml of 20 mM potassium phosphate buffer (pH 6.0) containing 20 mM guaiacol and 0.05% (wt/wt) hydrogen peroxide. Microplates were left at room temperature for 20 min, and then each reaction was terminated with 0.05 ml of 0.8 N sulfuric acid. The absorbance at 460 nm was measured with a multilabel counter (Arvo 1420; Wallac, Turku, Finland), and the increase in absorbance relative to that of a blank with no extract was used to assess MPO activity.
Inhibitory effect of MDT-1 on the growth of Campylobacter.The supernatant of MDT-1 cultures grown to late log stage was collected by centrifugation (20,000 × g, 10 min, 4°C) and filtered aseptically through a membrane filter (pore size, 0.22 μm). Brain/heart infusion agar (Becton Dickinson, New Jersey) containing 5% lysed horse blood (Nippon Bio-Test Laboratories, Tokyo, Japan) was supplemented to a level of 10% with either this supernatant or the plain growth medium used for B. fibrisolvens (as a control). The agar plates were then inoculated with individual strains of Campylobacter coli (15 strains) or Campylobacter jejuni (5 strains) and incubated at 42°C under microaerophilic conditions (5% O2, 10% CO2, and 85% N2) generated with CampyPak Plus (Becton Dickinson, New Jersey) in a GasPak jar (Becton Dickinson, New Jersey). Growth inhibition was judged by the significant reduction in colonies formed after 48 h of incubation.
Animal experiment 2 (effect of MDT-1 on infection with Campylobacter). C. coli 11580-3 was grown in a microaerophilic environment for 24 h at 42°C on brain/heart infusion agar supplemented with 5% lysed horse blood. Bacterial cells collected by scraping were resuspended in 0.8% NaCl solution (107 CFU/ml).
The protocol of this experiment is outlined in Fig. 1. As in experiment 1, 10 mice were used for each group, and 0.1 ml of 0.8% NaCl solution was given to group E mice (control) once a day throughout the experimental period (10 days). Group F mice were similarly given 0.8% NaCl solution, but C. coli 11580-3 (107 CFU) was administered on the fourth day (one dose). For group G mice, MDT-1 (109 CFU/dose) was administered once a day for 10 days, and C. coli 11580-3 (107 CFU) was administered on the fourth day. Stool consistency was scored during the experimental period. All mice were sacrificed 7 days after C. coli inoculation for autopsy. The colon was removed, and cecal contents were harvested by extrusion. The colonic tissue samples were then prepared for histological (five mice in each group) and MPO activity (the other five mice in each group) evaluations. The cecal contents were diluted with 0.8% NaCl solution (1:1), and 100-μl aliquots were plated on the Campylobacter agar with 5% antimicrobics and 10% sheep blood (Becton Dickinson, New Jersey). C. coli colonies were counted after 48 h of incubation in a 42°C microaerophilic environment.
Statistical analysis.Data were analyzed by one-way analysis of variance, and Tukey's test was used when the F test was significant. The Kruskal-Wallis test and Dunn's procedure as a multiple-comparison procedure were also performed to compare the means of nonparametric or abnormally distributed data. Differences with P values of <0.05 were considered significant. Data are expressed as means ± standard errors (SE). All statistical analyses were performed with the SigmaStat statistical analysis system (Jandel Scientific, San Rafael, CA).
RESULTS
Effect of MDT-1 on DSS-induced colitis.None of the mice in group A (control) exhibited any of the clinical signs of DSS-induced colitis, i.e., body weight loss, diarrhea, and bloody stool (Fig. 2).
Effects of the oral administration of DSS, B. fibrisolvens MDT-1, and C. butyricum on the body weight (A), stool consistency score (B), fecal blood score (C), and disease activity index (DAI) (D) of mice. (○) Group A, DSS (−)/bacteria (−). (□) Group B, DSS (+)/bacteria (−). (•) Group C, DSS (+)/MDT-1 (+). (▪) Group D, DSS (+)/C. butyricum (+). + or − in parentheses indicates the presence or absence, respectively, of DSS, B. fibrisolvens MDT-1, and C. butyricum (bacteria). Data are expressed as means ± SE (n = 10). Different letters (a to c) at each time point indicate a significant difference between groups (P < 0.05).
No significant differences in the volume of drinking water between the four groups were observed during the experimental period. Mice treated with DSS demonstrated a significant decrease in body weight (group B) (Fig. 2A), which was reduced by administration of MDT-1 (group C) but not C. butyricum (group D). Diarrhea and bloody stool induced by treatment with DSS (group B) (Fig. 2B and C) were also alleviated significantly in mice administered MDT-1 (group C), but mice administered C. butyricum did not show a similar effect (group D). The presence of MDT-1 promoted a reduction in overall disease activity, as quantified by the disease activity index score (Fig. 2D). Furthermore, apparent colonic inflammation that was macroscopically observable in group B mice was less severe in group C mice (Fig. 3).
Effect of the oral administration of DSS, B. fibrisolvens MDT-1, and C. butyricum on the histological damage score in mice. Bars indicate SE (n = 5). Different letters (a to c) indicate a significant difference between groups (P < 0.05).
The ratio of spleen to body weight, which increased greatly with DSS treatment alone, diminished significantly in mice administered MDT-1 (P < 0.05) (Table 1). Despite the fact that there were no significant changes in cecum or colorectum weight with DSS in either the absence or the presence of MDT-1 (Table 1), colorectum length was markedly reduced in DSS-treated mice. As a result, there was a corresponding increase in the ratio of colorectum weight to unit length, i.e., the colorectum was abnormally shortened and thickened. However, these DSS effects were diminished in mice administered MDT-1.
Effect of the oral administration of DSS, B. fibrisolvens MDT-1, and C. butyricum on the sizes of organs of micea
MPO activity in colonic tissue.MPO activity (per wet tissue weight) in colonic tissue increased greatly in mice treated with DSS, suggesting that DSS promotes an increase in neutrophil density that signals an inflammatory response. Administration of MDT-1 reduced the increase in MPO activity (Fig. 4), suggesting that MDT-1 can help alleviate DSS-induced inflammation. Again, C. butyricum had no alleviating effect.
Effect of the oral administration of DSS, B. fibrisolvens MDT-1, and C. butyricum on the MPO activity in colonic tissue. *, change in absorbance/g of intestinal tissue. Bars indicate SE (n = 5). Different letters (a to c) indicate a significant difference between groups (P < 0.05).
Prevention of infection with Campylobacter spp. by MDT-1.Among the 15 strains of C. coli and 5 strains of C. jejuni examined, the addition of MDT-1 culture supernatant inhibited growth in 6 strains of C. coli and 3 strains of C. jejuni (data not shown).
Mice infected with one of the strains inhibited by MDT-1, C. coli 11580-3, lost body weight, and their cecal contents contained viable C. coli cells (Table 2), although no C. coli cells were detected in control mice. Administration of MDT-1 to mice alleviated the body weight loss and decreased the number of C. coli cells in the cecal contents to approximately 5% of that in nontreated, infected mice. Diarrhea induced by infection of C. coli (group F) was also alleviated significantly in mice administered MDT-1 before and during C. coli infection (group G) (Fig. 5).
Effect of infection with C. coli 11580-3 and the oral administration of B. fibrisolvens MDT-1 on the stool consistency score in mice. (○) Group E. (□) Group F. (•) Group G. Data are expressed as means ± SE (n = 10). Different letters (a to c) at each time point indicate a significant difference between groups (P < 0.05).
Effect of the administration of B. fibrisolvens MDT-1 on infection with C. coli 11580-3 in micea
MPO activity in colonic tissue increased greatly in mice infected with C. coli (Fig. 6). Administration of MDT-1 reduced the increase in MPO activity. Colonic tissue obtained from C. coli-infected mice shows acute inflammatory changes. The acute lesions were characterized by mild, focal, and nonulcerative infiltrates of neutrophils localized predominantly to the intercriptal lamina propria. Mice given MDT-1 showed smaller acute inflammatory changes induced by Campylobacter infection (data not shown).
Effect of infection with C. coli 11580-3 and the oral administration of B. fibrisolvens MDT-1 on the MPO activity in colonic tissue. *, change in absorbance/g of intestinal tissue. Bars indicate SE (n = 5). Different letters (a and b) indicate a significant difference between groups (P < 0.05).
DISCUSSION
Administration of MDT-1 to mice alleviated the body weight loss, diarrhea, bloody stool, mucosal damage, enlargement of spleen, and abnormal shortening and thickening of colorectum caused by DSS treatment (Fig. 2 to 4; Table 1). This alleviation is consistent with macroscopic observations that colonic inflammation improved in the presence of MDT-1. Overall, these results are consistent with the hypothesis that colitis can be mitigated by introducing MDT-1 to gut flora.
The decrease in intestinal tissue MPO activity observed in mice administered MDT-1 suggests that MDT-1 can improve DSS-induced colitis. Because secretion of tumor necrosis factor alpha and interleukin-1β by PMNs increases in active ulcerative colitis and infectious colitis, these molecules are likely to be important contributors to the initiation and perpetuation of mucosa inflammation (8).
The production of butyrate, which was previously reported to have an anti-inflammatory effect (30), is increased in the intestine following MDT-1 administration (2). Although it is presently unclear whether butyrate production promotes the alleviating effect of MDT-1 on colitis, enhanced butyrate production in the gut may, at least in part, explain the effect. Administration of C. butyricum or its fermentation products has also been reported to improve DSS-induced colitis in rats, and it has been suggested that this effect may be due to increased butyrate production (3, 4). However, since C. butyricum MIYAIRI 588 had little effect on the symptoms of DSS-induced colitis, the alleviating effect of MDT-1 in our experiments may not be due to butyrate alone.
Some strains of resident intestinal bacteria, such as Bacteroides vulgatus, Clostridium difficile, and E. coli, are thought to be potential pathogens associated with IBD (23). Campylobacter is also known to cause enterocolitis (27). Inhibition of the growth of nine strains of Campylobacter by the MDT-1 culture supernatant (Table 2) suggests that MDT-1 causes extracellular release of an antibacterial substance(s). Several strains of B. fibrisolvens are known to produce bacteriocins with strain-specific, but not species-specific, antibacterial activities (22). Although it is presently unknown whether MDT-1 cells produce antibacterial substances, it is possible that the alleviation of colitis seen with MDT-1 may be due to inhibition of pathogenic bacterial growth.
Administration of MDT-1 alleviated the overall symptoms of C. coli 11580-3 infection in mice (Table 2; Fig. 5 and 6). Prevention of body weight loss may have been due to the anti-inflammatory effect of butyrate as well as to the inhibition of C. coli growth by MDT-1 in the intestine. Because the inhibition of C. coli by B. fibrisolvens is likely to be strain specific, there may be alternative strains of B. fibrisolvens that could be administered to humans and animals to inhibit the growth of other C. coli strains or possibly species of Campylobacter, which would provide a treatment for Campylobacter-induced enterocolitis.
It has been reported that humans and pet animals are sometimes infected with Campylobacter via animal products, especially poultry (1, 17, 20, 24). In Japan, antibiotics have been used to prevent colonization of Campylobacter in chickens; however, the use of antibiotics can result in the development of antibiotic-resistant bacteria. Therefore, alternative strategies to reduce the infection of Campylobacter in chickens, such as a B. fibrisolvens probiotic, are needed (10).
In conclusion, oral administration of B. fibrisolvens MDT-1 to mice alleviated DSS-induced colitis and Campylobacter-induced enterocolitis. Although it is conceivable that an MDT-1-produced antibacterial substance(s), as well as butyrate, exerts these effects, the underlying mechanisms have yet to be elucidated.
ACKNOWLEDGMENTS
This study was supported in part by a grant-in-aid for scientific research (no. 15580240 and 16780190) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT) and by the “Collaboration with Venture Companies” Project for Private Universities, with a matching fund subsidy from MEXT for 2001 to 2005.
FOOTNOTES
- Received 22 July 2006.
- Returned for modification 17 August 2006.
- Accepted 12 September 2006.
- Copyright © 2006 American Society for Microbiology